Alarm Buzzer above 2000 RPM, possible Malfunctioning Over Revolution Warning Alarm?
 

Starting a dedicated thread as I was digging around the Coolant Level Sensor Alarm issue, and that issue appears to be unrelated.


Continued from this thread:



Almost disappointingly, I don’t see anything that would make me super suspicious of the “Body CPU” ? In terms of corrosion or caps failing is an obvious way?

These photos are as removed uncleaned and untouched.

There is a little white residue on the back side of the through hole board, almost undetectable in person, the photo captures it very well. It brushed off effortlessly.

The “Body CPU” is the CPU #2 just to be sure we are talking about the same thing?

Attachment 746136

Attachment 746137

Attachment 746138

Attachment 746139

Attachment 746140

Attachment 746141

Attachment 746142

Based on the above CPU2 photos, pull the instrument cluster?


And specifically the tach module?

>Take a look at the solder joints along each of the 22 pins on IC1

Can you expand on "IC1" is this the IC on the Tach module with 22 pins? Looking for some reference to IC1 and 22 pins, coming up short ATM for conformation this is explicitly what you are referencing.

As luck would have it, I have Parts of an extra cluster… and a tach module if it would help troubleshooting?



To be clear this is Not the tach from the car that is having issues with a rogue buzzer , simply spare parts sitting in a box that has been tucked away for many Many years.



Assuming this Tach module could be the cause, I assume I could simply swap it out to check if the alarm buzzer continues to activates at above 2k RPM?



Would that be satisfactory way to test without damaging my pretty and clean unmolested cluster and tach?

Extra tach Parts for reference

Attachment 746132

Attachment 746133

Attachment 746134

Attachment 746135

rotarypower101,

I didn't see anything wrong with the Body CPU, and yes...I meant to say CPU#2. I use both terms interchangeable so my apologies for any confusion.

Would you take a few more photos of CPU#2? This time, could you please show a closer angle on the component legs? This will help me identify any potential problems with capacitor leakage. Additionally, the angle of the photo should try to show a profile of the capacitors. This will help identify and capacitor swelling, especially from their tops.

The white powdery stuff (I did see traces of it in your photos) could potentially be a problem. That could be some corrosion. Or it could be some form of sea salt? You can brush down the back side of CPU#2 with isopropyl alcohol then let it dry. It doesn't need to be soaked but brush on a small amount along the board then check for that white residue. I would also wipe out the plastic cavity and cover where CPU#2 sits with isopropyl alcohol. This would also clean up any residual salts and/or mung.

As it stands, I don't see any issues with CPU#2. But I want to make sure it's ok before it gets tucked away. The photos I requested would remove any doubts.

Onto the tachometer board: Yes, the IC1 mentioned in my post is the IC1 on the tach board. It has 22 total pins, 11 per side. Based upon your photos of the spare tach board, I don't see anything unusual. If the solder on any of those pins to IC1 were dull-grey, frosted, or cracked then that is a sign of a cold solder joint. A cold solder joint means that the electrical bond is broken and must be repaired. This is done by "re-flowing" the solder. I typically recommend replacing the old solder with fresh solder to ensure a positive electrical and mechanical bond.

All of the electronics on the tach board are there to drive the motor and tach needle. That means the signal driver (and over-rev buzzer) originates from the speedo board. For future reference, if your tach needle starts to bounce around then refer back to that Bouncing Tach thread. The main point of failure are the solder joints to IC1 (on the tach board). Simply reflowing the solder usually fixes a bouncing tach needle.

Since you have a spare tach available, you could install it into the cluster then test it. However, I doubt that is the problem. As stated above, the speedo board drives the tach. Therefore, we should take a closer look at the speedo board for failed items. There are a few threads that discuss it. One thread really dives into the weeds here: Troubleshooting the FD Speedometer-Odometer-Tachometer In that thread, you can see a lot of close ups of solid state components (resistors, capacitors, etc...).

In order to get better visibility of the speedo board, you will need to desolder the speedo face. How are your solder skills? Before removing the speedo face, take as many close up photos of the speedo board at various angles. Post them here or in the troubleshooting the FD Speedo thread (linked above) and I will share my insight.

Does that help?

Hi, thank you!



Some other details if it helps clarify, to hopefully give or build a better picture of the state of the vehicle.



In the PNW, we don’t use winter salt around here often/typically, we are on the “Coast” but far enough away that it seems to be a complete non factor. Probably due to wet weather and great drainage we don't get a build up even for the small quantities that are present.



Generally speaking the car is very clean internally very rarely do I ever see any indication of degradation of the electronics or any components showing corrosion or associated issues. New engine harness, I can’t even think of any other electrical issues that come to mind.

Are these included photos sufficient? May I describe the parts if the photos lack the detail you are requesting?

The white powder material had a localized horizontal element to it (across the board), CPU2 has a vertical orientation, so my assumption is this may be OK as if it were the electrolyte would it release vertically and typically on the cap side?

The back side of the board looks shinny on all the joints, I see nothing I would even remotely call corrosion.

The front of the board gives no definite indication of residue or corrosion, most joints have a nice shinny appearance. There are some joints that have a "grainy slightly dull grey appearance" it doesn't appear to be corrosion, but there is not a glossy sheen to the solder, and occur near the larger caps.

The caps to my Untrained eye look parallel on the sides, no bulges on the sides or top, a set of calipers used to determine parallelism shows very little deviation from a visual inspection.

I cleaned the board and case with 91% Isopropyl Alcohol.

I can take any additional photos requested if helpful, please ask if you have Any question about any obscured component you are curious about.

I can reflow any joints that are requested or advantageous to do so to help minimize any questions or doubts moving forward.

Visually, I would assume this CPU2 is sound, without another FD to swap a known working part, it would be hard to suspect it just visually ATM IMO.
Are there any PNW people still into FDs? I know of no enthusiasts left unfortunately, makes it very hard to get any local help and support... :( If anyone knows of FD focused people still around, and specifically PFC Tuning, I could really use some help!

​​​​​​​Apologies for the dark lines, the lighting seems to be giving some kind of artifact...
Case photo included if helpful.

Attachment 746121

​​​​​​​Attachment 746122

Attachment 746123

Attachment 746124

Attachment 746125

Attachment 746126

Attachment 746127

Attachment 746128
Case for any evidence of electrolyte?

​​​​​​​Attachment 746129
​​​​​​​

The Extra speedo board I have is terrible looking on the top, not much more to say there, it is awful in many places! Back is fine, but the top has green corrosion in too many places to list and just a messy build with lots of residue flux just left which I think is pretty typical from other photos I have seen.



To confirm, go ahead and pull the instrument cluster?

Once I do this inspect the Tach first as noted above? But this is likely not the issue assuming it is in better condition than this extra tach?

And any specific advice in advance about pulling the cluster?

>How are your solder skills?
I have quality tools, and know how to use them, I can solder anything you put in front of me down to the smallest SMD.

Photos of extra speedo board:

Attachment 746116

Attachment 746117

Attachment 746118

Attachment 746119

Attachment 746120

I am a little concerned about the "fuzz" on the legs of C1, C2, and C3. That could be some electrolyte leakage but I am not sure. Could you describe it or take a closer pic of them? Additionally, what is the capacitance and voltage rating of C2?

Use this thread as a guide: FD01 Body CPU CPU#2 Components List

That link also has some sample photos of bad capacitors.

That white powdery stuff could have been from condensation caused by seasonal heat cycling. Maybe? Since it is cleaned, then I'd say there is nothing more to worry about.

I just noticed one last thing. Take a look at the transistors inside the blue squares. You will see a green deposit on the center leg of the transistor. This is caused by leaked electrolyte from a capacitor or two. Clean up those legs with isopropyl alcohol then take more pictures of them.


https://cimg6.ibsrv.net/gimg/www.rx7...153893d7e4.jpg
Examine the legs of the transistors inside these 2 blue boxes.


I am of the opinion that C2 and C3 were previously replaced. And some electrolyte from the original capacitors may not have been fully cleaned up. Is this the 1st time you dug into the Body CPU? I also don't like the fuzzy legs on C1, C2, and C3. They could be suspect!

I see that your spare speedo board has lived a rough life! I also noted the socket for IC3. This board had its fare share of playtime and experimentation!

Copy all on your solder skills. That's a bonus!

Take a look at C2 on the sample speedo board. It looks like the top of it is bulged up. Now take a look at C3. See how C3 has a squared top? That bulged top is an indication that C2 is well beyond its life. Note, a bad capacitor doesn't have to swell up or out. That only shows a sign of physical damage. Electrically, it may have failed long before that.

Additionally, do not lay over capacitors, as shown in the sample. They should stand upright and have some clearance between the eyelet of the circuit board and the base of the component.

Before you pull the cluster and inspect the speedo board, let's put our minds at ease with CPU#2 first. For planing purposes, I would follow this train of thought:
1. remove the cluster (just got done figuring out how to fix the FD odometer has excellent instructions)
2. remove and inspect the tach (do not take off the tach face and needle) for poor solder joints
3. remove and inspect the speedo board for bad components.

I can't offer up help in the PacNW for tuning advice. Perhaps others may have recommendations.

Does that sound like a plan?

Absolutely, Thank you for taking the time and effort to Help, SO refreshing to be back here compared to several other platforms forums I frequent!

Here is a small video to show the Explicit feedback I get. It seems pretty obvious it is RPM based... but its always nice to leave a little room for doubt

How I missed it I am not sure, the speedo is seemingly in sync with the tach…



After startup ~22second mark, the speedo seems to respond to a signal from the tach after a brief amount of time…

https://www.youtube.com/watch?v=ZyHD...ature=youtu.be

PFC Commander for a more consistent/accurate RPM VS alarm buzzer

https://www.youtube.com/watch?v=q-7K...ature=youtu.be

Thank you for sharing the video. It does help fill in the picture better!

I noted that your oil level light was on during the recording. Did you disconnect it? I believe the Body CPU is a red herring. So clean it then close it up. We are done for now. The speedo following the tach is a tell-tale sign that your speedo board requires repair.

Go ahead and follow the instructions to remove the cluster and let's jump down the rabbit hole!

Let's talk about ground wires...have you done any work inside the engine bay? The primary ground point for the speedo is on the ground wire to the engine. It is ground point #3. BTW, you will need the Wiring Diagram Manual (WDM) because we will reference Diagram C-1a and C-1b a lot!

I keep a copy of my FSM (with WDM) at my computer because ya never know when its needed to answer a question.

BTW, the Body Electrical Troubleshooting Manual (BEM) had you test points in this circuit. That was part of the Flowcharts 23-24 that you posted in the original thread.

You need to verify that ground wire is properly grounded at the engine and to the chassis. If that ground wire is off or damaged then the instrument cluster will be harder to troubleshoot. Believe me, this ground wire makes ALL the difference because the speedo board uses it separately from the other ground wires in the instrument cluster.

While we explore the speedometer board, we are only going to visually inspect it. First inspect it with the speedo face on. I have a strange suspicion that the speedo face will need to be removed. But let's not get ahead of ourselves, ok?

Well spotted!!!
There is definitely more than I noticed on my ..."careful inspection" !


>what is the capacitance and voltage rating of C2?
10μF, 16V


The biggest thing to note after cleaning is that component Q12 corrosion seems to have eroded the center pin to very poor cross sectional area.

Attachment 746112


>"fuzz" on the legs of C1, C2, and C3. That could be some electrolyte leakage but I am not sure. Could you describe it or take a closer pic of them?
There is a very slight dull grey hue to C1 C2 C3 on the legs

Attachment 746113

Attachment 746114

Attachment 746115

Any other components to look at closer...or at all..., can't believe I missed those green fuzzballs, I was really looking at it hard and didn't see them...even when blowing up those photos...

>I noted that your oil level light was on during the recording

I have checked my oil level, it is fresh oil and right to the topped up mark.
I believe it is a axillary oil level system I have added and tied into the OEM system. I run 2 stroke oil in an aux tank to OMP.

>I believe the Body CPU is a red herring.
Thoughts on component Q12 from the post above ?

I will start working on the cluster removal.

>Let's talk about ground wires...have you done any work inside the engine bay?
Yes, anything specific that would be good to comment or detail?


I have a new fresh engine harness, and I believe a good multi point aux grounding to the block chassis and other important points.
Harness ground points are fresh and clean and well grounded, but it has been a few years since I have been in there.

>you will need the Wiring Diagram Manual (WDM) because we will reference Diagram C-1a and C-1b a lot!
_Z_Wiring_Diagram.pdf at the ready

> The primary ground point for the speedo is on the ground wire to the engine. It is ground point #3.
I will need to familiarize myself with this.

>I keep a copy of my FSM (with WDM) at my computer because ya never know when its needed to answer a question.
I have a hard copy and multiple monitors.

>BTW, the Body Electrical Troubleshooting Manual (BEM) had you test points in this circuit. That was part of the Flowcharts 23-24 that you posted in the original thread.
I am very poor at troubleshooting comprehension...I misinterpret and often struggle.
I tried to check it as outlined, but can do it again if you have any explicit instruction or advice how to do it more thoroughly, and with more purpose.

>You need to verify that ground wire is properly grounded at the engine and to the chassis. If that ground wire is off or damaged then the instrument cluster will be harder to troubleshoot. Believe me, this ground wire makes ALL the difference because the speedo board uses it separately from the other ground wires in the instrument cluster.
Will look into this in the next couple of minutes.

>While we explore the speedometer board, we are only going to visually inspect it. First inspect it with the speedo face on. I have a strange suspicion that the speedo face will need to be removed. But let's not get ahead of ourselves, ok?
Have the time will tear into as soon as search for and validate the "ground point #3."

For God's sake man, don't poke at that transistor's center leg! Don't even wobble it! It is fine so long as the metal is intact.

As a preventative measure, I would replace C1, C2, and C3. Yes, C2 is supposed to be 10uF (microfarads) @ 16V. Is that stamped on the body of C2? For now, the Body CPU is working so let's allow that sleeping dog to lye.

I do have 2 "oh by the ways" on the CPU#2:
1. The black capacitors are definitely replacements because I see the "Nichicon" manufacturer stamp. The originals were not stamped with OEMs. The majority of caps should be brown, like C1. There are some exceptions but that's a topic for another day.

2. Take a look at this spot:

https://cimg6.ibsrv.net/gimg/www.rx7...436ccafa96.jpg

Clean up that black spot on the one leg of that resistor. It may be a shadow effect but it could be old electrolyte reacting with the resistor's leg. Then inspect it for any breaks. Don't poke at it though!

No worries about spotting the mung built up on the transistor legs. Your eye was not trained to spot it. Luckily, you know someone who can spot it! ;)

No need to post the Diagrams here. Just keep them for reference. If you'd like, make a hard copy and use some color pencils (or crayons) to trace specific segments of the wire diagram.

Red can indicate a bad result, green can indicate a good result, blue for tracing a path, etc...

> It is fine so long as the metal is intact.

Good idea to plate it in solder until the component arrives for replacement?

>As a preventative measure, I would replace C1, C2, and C3.

Will do, I am always dealing with "disposable" items for components so... eBay and other low cost alternatives.
Any good places to get a known quality component where they don't ding you on shipping? or just do a digikey part and not fret?

> I do have 2 "oh by the ways" on the CPU#2:
1. The black capacitors are definitely replacements because I see the "Nichicon" manufacturer stamp. The originals were not stamped with OEMs. The majority of caps should be brown, like C1. There are some exceptions but that's a topic for another day.

It is a 93 touring if that has any bearing or interest on the topic.

>2. Take a look at this spot:

I believe that is a shadow, there are 4 restores that appear to be all of the same value in that area, and they all look identical.

>No worries about spotting the mung built up on the transistor legs. Your eye was not trained to spot it. Luckily, you know someone who can spot it!

Thank you! the Rotary community is unquestionably a better group than is typical found. I assume its the niche hardware and passion, but I have So much trouble elsewhere.

To the point people will tease you and discourage trying to repair a piece of hardware, had Seadoo PWC ECU that I doggedly pursued a fix on, and it was rough! Got it in the end from guess and check and got lucky..couldn't very well break it worse... there were some factors that made it worth keeping the original for that a aftermarket just couldn't equal, and OEM replacements were prohibitive.

Does "nichicon" brand suffer from capacitor plague?

I didn't see any obvious hits from searching those terms.

Its irregular to have that brand on those boards?

The case for CPU2 I don't believe has been opened before, unless someone knew how to open it without leaving any telltale apprentice marks.

I redid the C2-35 tests in the light for CPU2 after cleaning the board of the issues found above.



*I do Not see “B+” ~12Vdc battery voltage on Step 2 pin 1L (BR)



I am seeing ~0.2Vdc



The other day and today I saw 12VDC on pin 1C B/Y

And ~4Vdc which is inside the 3-6Vdc range stated on 2B Y/L

I did not do step 3,4 because they are inside the cluster, and it was impending I would have that open soon. And my assumption was that was the same connection tested at CPU2.

So it looks like it may be the CPU2…

Taking a lunch break and going to go try it again after reseating all the plugs and checking again. Its possible I made a mistake testing.

*Took a small ride reseated the plugs and tested again, see the same thing on pin 1L (BR) , I think I need to locate a new CPU2...

How much might that cost to source?

In the meantime while I wait for parts, I added pigtail to the buzzer...
Attachment 746107

The Nichicon caps were not the original components. There is nothing wrong with them, I use them in my repairs. So you should feel confident that the replacement caps should last for another 20 years. The original brown caps are susceptible to failure and leakage.

Still, there is always the possibility of failure. However, that possibility is normally low.

Do not mess with that transistor! Seriously, don't add solder to its legs. The potential to cause more harm than good is considerably great. Not to mention, that transistor is no longer made. Therefore, you have to hit the secondary market.

As I mentioned before, the replacement of C1, C2, and C3 on CPU#2 would be a good preventative measure. The greyish fuzz and green fuzz on surrounding components was not a confidence builder.

My "Components Only" thread provides part numbers for the majority of speedometer board components. They can be found between the DigiKey and Mouser wharehouses.

Thanks for double checking the shadow I spotted on CPU#2.

Thanks for the compliments about the forum decorum. We have some very knowledgeable members who don't mind sharing that knowledge with others. Each of us have a little niche specialty - whether it's engine management, suspension, or electrical/electronics.

The pigtail for the buzzer should not be necessary, unless you want a louder buzzer! I suppose that is one way to temporarily disable the buzzer. But keep in mind, that buzzer is multi-functional!

Help me understand your testing procedures you double checked, please.

What did you follow - Flowchart #24? If so, did then this should be your troubleshooting steps:

Step 1.a: Turn the ignition switch to the ON position (Do not start the car - the engine should be OFF).
Step 1.b: Using a DMM, measure voltage at Pin 1C (Blue/Yellow wire) of the CPU#2 connector (the connector must be plugged into the CPU).

Hint: Insert a narrow paper clip or a sewing needle into the backside of the connector at Pin 1C. Pin 1C is located on the top row, 2nd pin from the right, as you look at the connector. You want it to contact the metal lead. Do NOT puncture or cut into the wire's insulation. Then touch the red lead to the paper clip and the black (ground) lead to chassis ground. The chassis ground must be any bare metal that connects to the chassis of the car. Do not use a painted surface. There is plenty of structural and unpainted metal under the dash that you can use as a chassis ground.

Step 1.c: The result should be +12v (B+). This step verifies that battery voltage is applied to the Body CPU. If it is not there, then check the METER fuse or potentially troubleshoot the wires between the Body CPU and the ignition switch.

Step 2.a: Measure the voltage at Pin 1L (Brown wire) of the CPU#2 connector (the connector must be plugged into the CPU, the ignition switch is ON, and the engine OFF).

Hint: Just like Step 1.b, insert a jumper pin behind Pin 1L of the Body CPU connector. Pin 1L is located on the bottom row, 6th pin from the left. Connect the negative (black) lead to chassis ground.

Step 2.b: The result should be +12v (B+). This step verifies that voltage is coming from the instrument cluster (the speedo board) when the ignition is placed in the ON position and the engine is OFF. It will read 0v during all other times. You may have to turn off the ignition then turn it back to the ON position (I don't recall if there is a time delay on that circuit and there is no mention of it in the FSM).

If this is good then you should move to Step 3. If it is bad, then there is either a wiring issue or a problem with your speedo board.

We already know you have an issue with your speedo board because the speedo follows the tach.

In Step 3 of Flowchart #24, you are supposed to verify that battery voltage is leaving Pin 3B. This pin is the Buzzer output from your speedo board to your buzzer in the Body CPU.

The remainder of the steps verify the inputs to the instrument cluster (speedo board).

Does that make better sense now?

I'd also like you to verify your ECU is properly connected AND grounded. A poorly grounded PFC may cause the speedo to follow the tach.

>The pigtail for the buzzer

Simply added that because I am of the mind it is there for a very good reason also. Which is exactly why I would like it fully operational if it can be fixed.

I can still technically use the car in its current state, assuming it were necessary, and disable on the fly. Only because I can was that done.





>Step 1.a: Turn the ignition switch to the ON position (Do not start the car - the engine should be OFF).

Step 1.b: Using a DMM, measure voltage at Pin 1C (Blue/Yellow wire) of the CPU#2 connector (the connector must be plugged into the CPU).


I used my Fluke 17B+ both from a chassis ground point in the first test as well as stretched directly from the negative terminal on the battery after I got a fail on pin 1L in step 2.

Pin 1C I have ~12Vdc on ignition ON



>Step 2.a: Measure the voltage at Pin 1L (Brown wire) of the CPU#2 connector (the connector must be plugged into the CPU, the ignition switch is ON, and the engine OFF).



Pin 1L currently I have ~0.2Vdc with ignition on immediately after testing 1C for B+

I have tried to do this Flowchart No.24 C2-35 Step 2 test multiple times now (4 separate times trying to reseat the plugs hoping something changed). I hastily made a mistake looking for a quick fix the first time, or something changed when cleaned and CPU#2 was reinserted.



I think I am Failing Step 2


May I ask if you have thoughts, would an issue with the speedo board cause a possible erroneous reading on 1L, where CPU#2 would not be at fault?



>Does that make better sense now?

Becoming more clear.



>I'd also like you to verify your ECU is properly connected AND grounded. A poorly grounded PFC may cause the speedo to follow the tach.



Question is this a ground on the back side of what I think was a fan ”update box” IIRC?

I don’t see any ground continuity here at the loop terminal when the PFC is Not bolted down (which I think most are not? as they never came with any threaded holes to facilitate) which was surprising!… Assuming that is what it was supposed to be or is it? I believe that double black wire with a loose vinyl covering and loop terminal goes to the main harness? Is this loop terminal "feeding" a ground signal t other PFC or "Supplying" a ground elsewhere?

And this is the correct placement for it?

If there is anything important I can expand or elaborate on further that you are not satisfied with please ask :)


Attachment 746090









Some measurements if applicable.



With my PFC Not bolted to the frame I see ~0.8 Mega Ohm to battery ground



With PFC bolted down as pictured I see 0.2 ohm





I added a temporary direct ground from the battery to the PFC, and I saw no feedback change to the above videos on the cluster.





At ~5500 RPM the “sync” between the Speedo and the Tach seems to be lost, then recovers the sync momentarily after RPM dips below 5500RPM


Is there any other checks or tests I can do before pulling that cluster out that would be wise to check or inspect? Even on a hunch?


Attachment 746091

Attachment 746092

Also, mostly unrelated to the topic, but since I was looking at the PFC, you don’t happen to understand what this harness was, and how it fits into the picture?



I have asked, and I don’t think I ever got a satisfactory answer.



What is it?



Why was it on my original engine harness?



And has it been rendered obsolete by some update to the new harness?



I believe I was told to exclude it…Its been a long time…

As well as this stray lonely single female spade down by the PFC? Perhaps some OEM feature not used by the consumer?

Attachment 746088

Attachment 746089

Let's see if I can address your questions and offer more feedback.

1. I need to confirm - Was your engine running during the voltage tests at CPU#2 Pin 1L?

2. Since Flowchart #24 Step 2 is currently failing, then you will have to remove the instrument cluster. Then follow Steps 3 through 5.

3. I want to add an intermediate step in between Step 3 and 4: verify the wiring between CPU#2 and the instrument cluster.
3.a: After the instrument cluster is removed, measure the resistance between Connector C1-01-3, Pin 3B and CPU#2 Connector J4-01 Pin 1L.
3.b.: Disconnect Connector J4-01 (CPU#2 connector) then connect a jumper wire on Pin 1L to ground. Read the hint below.

Hint: Use the chassis to help measure the resistance between these two pins. There is no need to extend wires between these 2 points. This is achieved by using a jumper wire on J4-01 Pin 1L. Connect that jumper wire to the chassis.

3.c: Connect your DMM red lead to C1-01-3 Pin 3B and the black lead to the chassis close to this connector.
3.d: Measure the resistance. If it is a short (next to 0 ohms) then the wire is good. If it reads anything other than a short then a problem exists along that wire's path.

4. Once the intermediate step is completed, then follow through with Steps 3 through 5 of the BEM Flowchart #24.

5. To answer your question (as mentioned in my last post), your Speedo has a problem. It was confirmed with your video. It could be one, two, or more problems. That is why your test results fail at Step 2 of the Flowchart #24. Right now, we don't know if the speedo is actually bad or if something else is causing the speedo to be bad. Basically, the flowchart will help us isolate the problem to the speedo, ECU, or the wiring in between them. If we blindly throw parts (or test others) at the problem we do 2 things: waste resources (time, energy, money, etc) and we don't learn about the failed item. By using a logical approach, we learn. Does that make sense? Sorry for the soap box speech there.

Additionally, the flowchart assumes a stock ECU setup. It doesn't know (and it doesn't care) that you have another ECU installed. That is why it is VERY important ensure an aftermarket ECU will still send signals to all the right places in the car - it's more than just a control device for the engine. I'll address the ECU and photos next.

6. In the first photo with the red arrow, that is a ground connection. It's hard to tell where it goes but it looks like that could be Ground Point 6. Ground Point 6 supplies ground to the 2nd half of the instrument cluster. According to Diagram C-1b, if you remove that ground then you should lose your high beam warning light. The other warning should still work because there are alternates paths to ground via Ground Points 3 and 1. Moreover, Diagram C-1b has a second black wire tied to it, which goes to the Right Hand Turn Signal light bulb (and splits off as a ground plane for the instrument cluster). Comparing your photo to the diagram leads me to believe that wire is Ground Point 6.

6.a: That ground wire can be tested by placing one DMM test lead on the head of that bolt and the second DMM test lead to a different chassis location, such as a different bolt within reach.

6.b: Since this is a chassis ground point, it could be used to provide a ground for other aftermarket systems. Ground is ground (with a few exceptions) so if someone uses Ground Point 6 as a ground for an aftermarket device then it shouldn't make any difference to that device. However, it is recommended to use a separate ground for an ECU because other devices could very easily interfere with it or damage it. I believe the PFC still uses the OEM ground point. I run a stock ECU so I cannot confirm where it gets a ground from.

7. It is unnecessary to run a separate ground wire from the ECU to the battery negative (ground) terminal. Unless you suspect that connection to be bad. Otherwise, continue to use the ground wires and chassis ground as provided. It helps cut down on extra work! :)

8. Could you please explain this result?

"With my PFC Not bolted to the frame I see ~0.8 Mega Ohm to battery ground"

"With PFC bolted down as pictured I see 0.2 ohm"

8.a: Where did you take this measurement from?

8.b: Hypothetically speaking, if you have the PFC plugged into the harness but remove its ground connection from the chassis then you should see a jump in resistance between a ground (pin on the plugged-in connector) and chassis ground. The increased resistance indicates that the path to ground is no longer direct. It is taking a long trip and other electrical items are sharing that path. That would be similar to drinking water from a cup using a silly straw instead of using a straight straw. In the end, you still get the water, but the water travels a lot longer in the silly straw. Does that make sense?

8.c: I believe the PFC uses its mounting bolts and its metal shell as a ground to the chassis. Again, I am speculating here.

9. To answer your question: "Is there any other checks or tests I can do before pulling that cluster out that would be wise to check or inspect? Even on a hunch?" You will have to remove and inspect the instrument cluster and speedo!

9.a: To help make removal of the cluster and hood easier, you can loosen the steering column mounting bolts. This will allow a bit more clearance without the need to pull the steering wheel from the column.

9.b: Do not remove the bolts without supporting the steering wheel column in some fashion. These bolts also have a torque specification that is found in the FSM. Please tighten these bolts to the proper torque spec!

10. Why are the green, black, and red wires on your PFC so tight (2nd photo)? If possible, please give them some slack. Otherwise the tension applied on them may lead to issues down the road - like wire chaffing somewhere along their path.

11. I do not know what that wire harness segment is from the photo from your next post. You could post that photo in a "Wire Harness ID thread" and someone should be able to give you an answer.

12. I don't remember the purpose of that black spade connector. I am not an ECU kinda guy so maybe other members could chime in. Ask the all-powerful, wise, and magnificent @DaleClark if he could opine. He should, since I just tagged him! Thanks in advance for the assist, Dale!

OK, this turned out to take much longer to compose and looks like a novel! Sorry for the extra long format.

Thank you

No, ignition on, and engine off

Sounds good, done

For those interested, I used Dales instrument cluster removal procedure:https://www.rx7club.com/3rd-generati...3/#post8765957

From 3B to 1L I see 0.2 ohmThe terminals on the plugs look clean and shinny

With Connector J4-01 disconnected still, If I ground pin 1L out to the chassis and measure resistance at 3B I get 0.8 ohm

*if it is known, is there a reason why when you initially connect a probe that the resistance reads higher and then slowly "hunts" decreasing to a stable point?

An aside, is there any unassociated maintenance or inspections I can do while the cluster is out?

Step 3: I seem to have another problem.

With the cluster removed and ignition ON

Step 3:

I get 0.132Vdc with ignition on pin 3B on the white cluster plug.

C2-35 Step 3 calls for pin 3B to be B+ Battery voltage

Since CPU#2 seems to have not passed Step2, is this a possible reason for an erroneous reading?

It seemed "strange" to me the PFC is not grounded by this wire. My guess is this wire is meant to feed a ground to some place rather than provide a ground to the ECU.

If I am not mistaken "Most people" just shove their PFC into the existing space that the original ECU had to live, many times isolated with velcro or double stick tape with no intentional metal to metal contact.
My assumption was that without being bolted in, the loop terminal would have provided a direct ground to the PFC case.
Admittedly I do not fully understand these things well, but the feedback I have gotten is that the grounding on this car specifically is "Poor" and results in "ground loop interference" issues? (which is why I tried to increase these and make sure they are well facilitated in critical areas)
I thought the Case to the PFC would have been grounded based on how others mount this device with no direct/intentional metal to metal contact.


This measurement was taken from the chassis to the ECU bracket. As I was curious if the ECU was grounded well if not bolted in like mine.


That is the crux of my curiosity.
The PFC, or at least every one I have seen, does Not come with any facilitation to bolt the case to the stock brackets! IIRC it is generally isolated with the included mounting method, tape or velcro.


This is a AUX plug for the stock MAP sensor used as a secondary reference IIRC, again its been a Very long time...
the only reason it appears taut is because the PFC is pulled out to make a check on C2-36 Step5, and the wire is the shortest/limiting factor. once reinstalled it has ample slack.
Thank you for checking! :)



[/QUOTE]

Believe me I appreciate detail!
I try to limit my verboseness and descriptiveness, as it tends to not always help, but feel I need those tools to get the details out.

So if I sound brief or incomplete anywhere, it's me trying too hard to limit the wordiness where I think appropriate. Or I will tend to have a wall of text...

Some preliminary photos of the cluster speedo board :(

Attachment 746077

Attachment 746078

Attachment 746079

rotarypower101,

Thanks for digging into the instrument cluster and taking some preliminary photos of the speedo.

Please desolder the speedo face because we need to take a deep dive into this board! There are several problems that we need to address.

I'll try to briefly explain with these edited photos:

https://cimg0.ibsrv.net/gimg/www.rx7...43e00ffa3c.jpg
These red boxes are very BAD! Bad as in they look fried. I suspect the board may be physically charred in certain spots.

In the above photo, the components in these red boxes are either damaged from leaked electrolyte or suffered a large voltage spike and are dead. The cause could be many things - a current spike from somewhere else or a current spike that resulted from leaked electrolyte across these eyelets. The latter would be a short circuit condition. This may be the reason why your initial measurements passed but subsequent measurements fail. The circuit board looks like it took a catastrophic failure in between the measurements. It is doubtful that this damage was caused by taking measurements (if you had that question in mind).

I ask that you desolder the speedo face and before doing anything else, take photos of these specific areas from various angles. Get in close and use the highest quality setting on your camera. Do not resize them. These photos will serve as the "before" set. Next, take several q-tips and isopropyl alcohol to clean around these components. Do your best to remove as much mung as possible without damaging the components. From this angle, be VERY careful around the legs of TR5. One of these legs appears to be broken or partially severed. Additionally, the red box to the left of TR5 has an exposed section of the PCB. That is normal but what isn't normal is the scorch marks that surround it.

After these areas are cleaned to the best of your ability, take a second set of photos. Let's call this the "cleaned" set. Once done, please post the "before" and "cleaned" sets. We must determine whether or not the board could be saved. Burned components are easier to replace but a damaged PCB is a headache! Based on these initial photos, the board may need professional help! To prepare you for a possible solution, you may just need to replace the speedo board because of its extensive damage. However, let's not throw the baby out with the bath water just yet. We still need to verify the extent of the damage. Afterwards, we can make the call - fix it or replace it. These boards are getting harder to find and they are expensive.


https://cimg6.ibsrv.net/gimg/www.rx7...079e56df5c.jpg
The capacitor, C16 inside the red box (left) does not look good. Note the brown ring at its base and the swelled top.

In the 2nd photo, C16 looks like it failed. The brown ring at its base and swelled top are signs of a bad cap. I would recommend replacing it along with C17 (directly in front of it) if we go in the direction of repairing the speedo board. Use the same approach of taking photos with these locations as well. Take a look at R25 - it appears to have one eyelet that is blackened/charred. Use the same cleaning method described above. You should get better photos of this area once the speedo face is removed.

https://cimg7.ibsrv.net/gimg/www.rx7...e38d40f74e.jpg
These 3 capacitors are the ones that fail the most. They are C3, C4, and C6. They appear to have swelled tops.

In the 3rd photo, capacitors C3, C4, and C6 are the most susceptible to failure. Typically C3 spews out electrolyte and it runs down the board. Because the board sits in a vertical position, this acid just runs down the length of the board and eats everything in its path. This leaked electrolyte could be the cause of the short circuit and all of that charring.

What does your tachometer look like? Take a photo of its circuit board. But please, only show the top side (component side) of the board. Do not remove the tach face & needle.

Regarding the PFC, I am not an expert in its use, theory of operation, or installation. Our forum has a dedicated section for it, look in the Engine Management section for the PFC. Other gurus like @DaleClark would be more qualified to answer how to properly ground your PFC. I also know the PFC forum has a FAQ. It should cover proper grounding of the PFC.

Oh, in case you ask, we will save the "after" photos when the board is repaired (or maybe replaced).

P.S. If you cannot locate a USDM speedo then consider getting a JDM speedo. DaleClark has a JDM to USDM speedo conversion thread that has helped many other members. If you do need a new speedo then give a call to Ray Crowe first. He should be able to source one for you.

To me that’s a Very bad position to be in…



It would be unfortunate to repair everything, only to have what ever caused this to repeat this failure!



Based on the look of it, what ever happened fried these components with heat, I think…



Is that a common thing on a low voltage DC PCB to short from small traces of capacitor electrolyte? I don’t see a lot of evidence of damage from corrosion.



C3 is noticeably bulging, and is in a very good position to leak over the most damaged area.



Have you seen anything remotely similar to this before from a leaky cap? A short rather than corrosion?



I am certainly NO expert on the topic, but I don’t see much corrosion or high levels of effected areas from electrolyte if I were to describe it currently. Admittedly from very limited angles.



Several of the diodes seem to have a dark “black scorched” residue away from the epicenter of all the skid marks. D2, D3, seem to have obvious surface scorch material on the leads, even D4 completely on the other side of the board behind the digital odometer has a similar darkened char reside on the leads.



Simply as an observation that feels noteworthy.



I can say fairly confidently, given the proper advice I can fix nearly anything on a PCB! Seen and fixed much worse and spend a good deal of time working on components like this …



The technical expertise and background is my known Weak area…I know enough to be dangerous, with guidance I feel very confident I can build and fix nearly anything in front of me.



Saying that, I can accept when an object is too far gone, and will defer to advice and experience on that topic.







C16 C17 I noticed the same thing, I assume this is electrolyte creeping up the shrink? I noticed this on that extra board I have also.



Pulling the speedo and tach now

I don’t see anything obvious on the back side of the Tach board by peaking between the face and the board from a shallow oblique angle..

Attachment 746072

Attachment 746073

Attachment 746074

Attachment 746075

Attachment 746076

Copy all. Tach looks good but will double check on my computer. Using the GT app to read & reply.

Take some pics of the backside of the board, with attention to these key areas. Then post. What are the other marks you mentioned on the backside?

1 more thing: please continue with the plan to clean those char spots.

If you take a look at my troubleshooting thread, there are examples of burn & char marks. There should be other photos that document electrolyte leakage. It will look similar to charing because there is an exothermic reaction between the acid, solder mask, wire leads & traces.

What is the GT app? Searched and didn't get anything on "GT app forum" I assume it is some forum portal specific to mobile apps?

Working on the Speedo board now, just removed the actuator.

The other marks are just general comments about noticeable areas that seem different from others, and seem to be obvious and noticeable details seemingly worthy of mentioning as POI.

GT app = GarageTalk for mobile devices.

You can keep track of multiple InternetBrands car forums with one app.

I can focus in on anything else of interest and post those also if it is not clear enough?

Trying to post an album, please comment if it doesn't work properly.

https://imgur.com/a/EipvaZe


Anything of interest that I can capture better before cleaning the board?

Hopefully I did enough for you to see what you need to see?

Does this work better in a dedicated Album or images posted here? It appears there is a limit of 12 photos per post.

If anything looks like it needs more work or improvements, please don't hesitate to point it out.

Cleaned Photos

rotarypower101,

OK, I have some good news and even better news.

The good news: That was electrolyte that you cleaned from the board and the other components do not appear to be damaged. The even better news: the PCB is not damaged! Even more better news: the board can be repaired!

Thanks for those up close photos. I would prefer to have the photos posted in the thread. So that I may directly comment on them and so other members could evaluate them in the future. Links to 3rd party photo hosting sites are great for the short term but do not last past a few years. When photos are posted on the RX7Club, they are not going anywhere and members could see them without having to use a separate site or have to stop reading the current thread (bouncing back and forth).

Based upon the photos you posted, I would recommend a full capacitor replacement. Just the solid state components. Use my thread for parts here: Speedometer Parts Thread. Post 76 in that thread has the updated list of components with part numbers.

The reason for the full capacitor replacement: Install fresh capacitors that replace the one that failed (especially C3) and the others that will fail shortly after. This is "cheap" insurance that the board will not fail for the remainder of its life.

Could you post some close up photos of just C3? It is a perfect example of capacitor swelling from the top and bottom. Those photos will help others identify a bad cap.

I had no idea it could look like a proper scorched PCB, I was completely convinced that was a thermal meltdown, and it is right in front of my eyes, the little lifted bubbled surfaces following plausible trace paths between what appears to be associated components, the remote test point all by itself over there blackened and bubbled…

Some Notes while cleaning:

ZD3 looks discolored compared to my spare board or to the other zeners near it.

ZD4 ZD5 ZD7 had a deep thick coating that required what seemed like excessive mechanical abrasion to remove.

Concerned about the top pad on ZD4, if I were to guess, the plating on top of the PCB is gone, and all that remains it the through hole material.



Does electrolyte attack/target copper specifically more virulently than other typical metallic materials used in electronics?


https://cimg4.ibsrv.net/gimg/www.rx7...afed834c8.jpeg
https://cimg5.ibsrv.net/gimg/www.rx7...3f8f80a9d.jpeg
https://cimg0.ibsrv.net/gimg/www.rx7...341147ce0.jpeg
https://cimg1.ibsrv.net/gimg/www.rx7...e7d0f8c70.jpeg
https://cimg5.ibsrv.net/gimg/www.rx7...8bf45947c.jpeg
https://cimg8.ibsrv.net/gimg/www.rx7...9b0fb38d6.jpeg
https://cimg0.ibsrv.net/gimg/www.rx7...6dcef1654.jpeg
https://cimg5.ibsrv.net/gimg/www.rx7...c2b6f0071.jpeg

Two components really look like they took a beating - DA1 and TR7. That is because they were the 1st components in the wake of flowing electrolyte. Aren't you glad that you cleaned up the board with alcohol? I have seen that before, where leaked acid appears to look like charring and scorch marks.

I saw the eyelets on the Zener diode pack, ZD4, ZD5, and ZD7. They look fine. The anode (positive side) of ZD5 and ZD7 are tied to the ground plane (on the back side of the board). If you are worried about the solder joint on ZD4 then measure resistance between the positive leg of ZD4 on the top side and the bottom pad of R6 (SMD component) on the bottom side of the PCB. This will test the wire trace between the top side of the board to the bottom side of the boar without damaging the eyelet. If it measures a short (should be 0.2 ohms) then the wire trace is good. If it measures an open then the trace is broken. BTW, the anode (positive leg) of ZD4 is the side that is opposite of the black line on the diode. The black line on a diode always points to the negative material, sometimes called the cathode.

If you are worried about an exposed eyelet around ZD4 then you could dab it with a clear coat of nail polish. That will act as a topical touch-up sealant. The one thing you must remember about doing that: you must remove the clear coat with nail polish remover before you do any solder work to ZD4. Otherwise, I wouldn't worry too much about it.

I see the discoloration you mentioned on ZD3. It should be fine. Testing a Zener diode is tricky. It normally involves desoldering it then inserting the Zener into a test circuit. I don't recommend doing that. Just leave it in place. You can use a crude test and compare it to my measurement as follows:
1. Use the diode function on your DMM.
2. Place the red lead on the anode (positive side) of ZD3 and the black lead on the cathode (negative side) of ZD3.
3. The DMM should measure about 813 ohms. This is known as a "forward bias" of the diode.
4. Swap the lead so that the red lead is on the cathode (negative side) of ZD3 and the black lead is on the anode (positive side) of ZD3.
5. The DMM should measure infinity, or "1 " or "O/L". This is known as a "reverse bias" of the diode.

Note: You may hear one beep when the diode is forward biased. That is normal, dependent upon the multimeter. Additionally, a Zener diode operates in a different way from a normal diode. Basically, it's a one-way valve with unique operating zones.

I'd also like you to apply a second (or third, etc) coat of isopropyl alcohol to those affected areas again. This time, use an acid brush or a soft-wide bristle model paint brush. Pay attention to the legs and underside of those components. This will further neutralize any remaining electrolyte. Pay particular attention to DA1 and TR7 but don't apply a lot of pressure. You want to use a dabbing motion to allow the bristles to do the work. Then let it air dry.

To answer your question, the electrolyte will eat thru just about everything in its path when left alone. I cannot quantify how fast it eats through copper or other metals because I don't know the pH level of the acid. However, I saw the result of this acid eat away the leg of one transistor because it was allowed to sit for an undisclosed amount of time (let's say 1yr+). It ate the leg at a very unusual spot - the bottom edge of the transistor. I speculated the acid pooled at that "low" point because of gravity and the "wicking" property of the metal leg. It was barely visible and I spotted it well after I made several repair attempts on that board.

Thanks for the photos of C3. It doesn't look that bad, minus the slight bulge in the top lid.

Here are some photos from your cleaned photo album that have some of my observations:

https://cimg7.ibsrv.net/gimg/www.rx7...31cbbc4c1b.jpg
What capacitor is on the left? Is that C6? That one looks like it "gave up the ghost". Notice the swelling/deformation along the base. Note the removal of plating on the legs of DA1.



https://cimg9.ibsrv.net/gimg/www.rx7...497790c35a.jpg
Again, the capacitor on the left looks like C6. DA1 and TR7 is front and center. Use an acid brush with isopropyl alcohol along the legs and bottom edge of these components. I see a green tinge on the bottom of TR7 and DA1.



https://cimg0.ibsrv.net/gimg/www.rx7...ef6c3761e1.jpg
This looks like a good example of a failed capacitor. It should be C6. Note the abnormal blister along its bottom edge.



https://cimg1.ibsrv.net/gimg/www.rx7...7a733018cc.jpg
Again, this should be C6 that appears to have a blister around the narrow ring. You can also see the acid ate away the plating on 3 legs (might be DA1).



https://cimg2.ibsrv.net/gimg/www.rx7...d5db2cd464.jpg
A different shot of C6. IC1 is in the foreground. Note the warping of the bottom section of C6.

YES, when I first saw that I was certain it was a overloaded circuit.
Perhaps from a failed component that would be beyond my ability to track down. I doubt there is high current in that area, but I have seen that, and it was remarkably similar in likeness.

Measures 0.2 ohm, so all good it appears.

Only a concern if there was a trace leading to it that was unseen. I doubt this is a multilayer board with near 80s era consumer level tech... but it was worth noting.

Using a Fluke 17B+:

If I measure "ohms of resistance" on ZD3 I get 13.42 kilo ohms my malfunctioning board and 13.32 kilo ohms on my spare board in both directions.

On ZD3 when I use my "diode mode" black lead to negative cathode and red lead to positive anode I get 0.714v
red lead to negative cathode and black lead to positive anode I get 1.185v
And a very similar result with my extra board.

Am I perhaps misunderstanding anything here? With the voltage /ohm as well as the resistance discrepancy that is seemingly closely repeated on both boards?


Will do!

Thank you for taking time to help a random stranger!

C4 I believe, to the middle left is DA1 to the rear middle left TR7 to the right DA2, and on the far right C6 I believe

I will take another stab at it, I actually cleaned that one again after being able to see it in the photos. (I have one of those little USB/WiFI microscopes, I find it useless... you don't happen to use one you can recommend? )

She's ripe!
I have both my malfunctioning board and extra board I can stack and see the side by side difference, to my untrained eye the "don't look that bad by themselves" but having a direct comparison makes it apparent for the ones that are not so pronounced.


She has a little foopa going on there...

I assume the game plan is to simply replace all caps, and hope for the best? Are there other places in this car susceptible to this issue to worry abut also? Similar to CPU2? Any other body modules that are critical to look into?

I am asking for advice on capacitor brands and lines to purchase on a ask electronics reddit post, would you have any advice? I don't foresee this type of failure if repaired being the ultimate demise of this car, but I have absolutely no intentions of parting with it, and I have at least ~30 years it took for this to cause a issue with it left...

If there are trustable brands, I would love to try and do the best I can, unless its just all a crap shoot?

Its a shame you can't just throw tantalums at it, I don't suppose they make a through hole package in the required voltage and farad capacities that would be suitable? Or not cause some unforeseen issue as a substitute?

Furthermore I don't suppose they typically design for this phenomena? placing caps at the bottom of a vertical board rather than the top? Perhaps isolating caps and employing an air gap on vertical boards?

Has capacitor technology improved at all for electrolytics where this is less of a problem? Or is 30 years just asking too much of a device of that design?

Good to hear that the solder joint to ZD4 is good. Based upon your measurements of ZD3, it sounds like it checks out fine. Dependent upon the multimeter, the diode function would either display voltage or resistance. The important thing to note is when a diode is forward biased the value should be small. When the diode is reversed biased, the value should be considerably larger. I have a cheap-o Harbor Freight DMM that displays resistance for the diode function. A proper DMM, like your Fluke 17B+, would display the bias voltage.

Upon further inspection of those photos, the damaged capacitor is C6. See how close that cap is to IC1? The black component behind that capacitor is DA2. It could only be C6. Ha! Foopa (fupa, actually).

Regarding the capacitor replacement, the list of part numbers on my parts thread will provide many more years of use. They are either made by Panasonic or Nichicon and are excellent replacements. The speedo board requires the use of electrolytic caps because the circuits they connect require a specific polarity.

As you said, the game plan is to replace the capacitors. That should bring your speedo board back to life. There is a chance that you may have to re-examine the board further. We can only do so much from a visual inspection.

Remember, auto manufacturers need to make a profit from every vehicle sale. They design a product based upon many compromises. It has to last for a specific time and they use parts from the lowest bidder. I don't believe the Mazda engineers thought about product longevity, such as capacitor lifecycles. Basically, I wouldn't burn up too many brain cells over it.

I will get those parts selected, then post back when they are installed.


Thank you very much!

Hoping that will fix the issue.

Any point in checking the existing hardware by swaping the caps from the extra board to the malfunctioning one just to validate it works?

Glad to lend a hand.

Be advised, we may have to do more troubleshooting because all we did was perform a visual inspection. If everything goes well then we will celebrate!

Cannibalizing consumable items from one board to the next is a bad idea. Please do not replace bad caps with more bad caps (or potentially bad caps). I would suggest getting a few spare caps in case you booger something up.

You seem rather confident behind a soldering iron. If you would like a refresher or learn about some tips & tricks then check out my General Soldering Tips & Tricks thread. Or you could offer some advice of your own! I have not worked with SMD components before - not with the proper tools at least. So maybe you could add value to that thread?

May I ask about C15?



C15: 0.015uF, 50V, DigiKey P/N: 445-181133-ND, measured .016uF on LCR Meter.



Is this a recommended replacement as well? Due to age or preventative maintenance?



Or simply for completeness of the parts list and documentation?





Other than the electrolytic capacitors are there some other components that would be “wise” to order, hedging bets they may be needed as a common or typical failure?







In response to any soldering, I would like to help where I am able.

But I am not a formally trained operator by any measure…



Simply been using one since I was able to run a screw driver and tear anything apart out of curiosity and mod…



So while I feel supremely confident and have very wide breadth of experience in any consumer grade device, I don’t have any authority to assert what is “correct” or “right”. But I would certainly supply any advice upon request based on my experiences specifically.



Soldering is a strange thing, I feel from my perspective people are afraid of it, or screwing something up more than lack the capacity to do it well.



I have a ME background, so maybe its a small bit of understanding small details like insulating oxides, heat transfer, and other small details that gets me past the basic hurdles even in the worst scenarios… but I have been doing it since I was a little kid with poor tools, its almost fun with good tools.

C15 doesn't need to be replaced unless there is evidence that would support its replacement - such as burned up wire traces. I didn't notice anything unusual about it on your board.

You may have read earlier in that parts thread that C15 has a code stamped on it. It took me a while to decode it. I wasn't 100% sure that it was properly decoded because there are many standards to capacitor codes and my research was purely hypothetical. After a while, I used an LCR meter that has the ability to test inductors (L), capacitors (C), and resistors (R), hence LCR. The meter confirmed my hypothesis with a small margin of error. Now, the original format of that capacitor is no longer made, so I chose the closest type of capacitor based upon its datasheet specifications.

I also want to address something that concerns me. I found your reddit post asking others the same question you asked me about the quality of components. It happens to be a copy & paste of those components listed on the Speedo Components list thread.

The offending thread in question: Reddit Capacitor Replacement Post

My concerns are:
A. The exact question you posed to that forum was answered in our forum. It was specifically answered by the originator of that thread.
B. Give credit where credit is due. You could have cited one or 2 components and provide those reddit users with a background.
C. Replacement part numbers, both manufacturer and warehouse specific numbers, were provided to make it easy for others to purchase or do their own research.
D. Do you know how much time and energy was spent researching those components?
E. Why seek outside confirmation? Especially when you can hear it from the horse's mouth (who also did the research).
F. Have you read that parts list thread? It is more than just a "use this part number" thread. It documents the journey such as C15's measurements and invites other members to contribute to the parts list.

By posting to reddit, you are asking a different group of people to do the research for you. Is that really necessary? Are we conducting a double-blind study or something? In all, do you not trust the information and research presented? Would you feel more comfortable with another member helping you repair your speedo board?

That was not the intention at all.

I can edit and cite where that list comes from if that is at all a issue, or completely remove it if requested. I was not thinking that far ahead, just a random quick thought I could ask and maybe learn about the topic of capacitor brands. Of which I know nearly nothing other than to match ratings.

I posted that list from a cut and pasted list I was using as an organizational tool while putting together an order.

My intuition was the response to that type of post would be “Depends on the specific type of components needed in question” so it made sense at the time to include it.

I was not asking for any substitutes, simply hoping I could learn if there were brands more likely to offer reliability as a generality for electrolytic capacitors.

That was the intention anyway…

Maybe I misinterpret that threads wealth of information from inexperience, but I didn’t see anything that helped me understand if there were more reliable brands to consider.

If it helps, I apologize if I presented it poorly, and am more than happy to correct or remove anything you feel is incorrect to post. Just searching for information that is ambiguous at best to me.

rotarypower101,

Thank you for the explanation. The need to edit posts outside of this forum is not necessary. In the future, please cite where your information originates so others (either here or elsewhere) could better understand your question (and perspective). I don't mind an argument to use Brand X over Brand Y. Please, do some independent research then present an argument for discussion.

My background is in electronics. I am a retired Navy Aviation Electronics Technician with 26 years of experience. I troubleshot and repaired many circuits from various devices, such as radio, radar, navigation, aerial photographic equipment, and weapons systems in Navy aircraft - mostly F-14 Tomcats and F/A-18 Hornets. I am very familiar with burned up circuit boards, solid state components, and their repair/replacement. I share this experience with our members when it comes to electrical/electronic repair. When identifying suitable replacement parts for the speedometer, I and a few other members spent several hours researching datasheets and specifications for several components. We looked at component polarity (for the caps, specifically) temperature ranges, tolerance levels, part availability, and even the physical dimensions for replacements. I measured some components with a caliper just to be certain the replacements would fit and orient themselves on the board.

Therefore, recommending Brand X over Brand Y for component replacements would be a complex answer. One manufacturer would have the right criteria while another would not. Generally speaking, the Nichicon or Panasonic electrolytic capacitors typically met the criteria for use in the speedo board. Does this satisfies your curiosity? I applaud your enthusiasm for knowledge. May we continue with your board repair?

How did you fare when ordering replacement parts for the speedo?

This is by far the best use of this forum I've ever seen! Calm, collected, informative, deliberate.

That helps bring into focus that the selection has been curated much more carefully than I had perceived, thank you for taking the effort to detail that.

Well, I think..., received my components late yesterday evening, and so its time to move forward! :)

I have just removed all electrolytic capacitors, tried to clean and prep the top surface pads with fresh solder as the capacitors top pads even after removal nearly all had fairly obvious signs of what I assume is/was a resilient oxidation on the top surface of the solder that did not easy remove , and tried to clean up some of the trouble spots as well...

Attachment 746022


Each site seems to be clearly marked for polarity, save for one, C9

On C9, BP refers to the component being “Bi Polar”?
Polarity does not matter on C9?
Despite the component as received with Long leg (positive) is agnostic on site C9?

Are there any accommodations I should make before placing components into their new homes?

May I get any instruction you feel is helpful when placing these components, in terms of positioning, height above PCB or any other typical considerations that would be good to adhere to? Perhaps any not obvious pitfalls with placing capacitors to avoid?

For other Novices that may find this in the future, it took about 20 minutes to remove caps, with...multiple background distractions.

These are only my personal observations, they may not be strictly recommended…

C9 is probably the hardest to remove for the combination of small component, and a broad wide possibly a “ground plane” as it draws in the heat. It may be advantageous to simply cut the legs long and work with the legs individually rather than the whole component if you are having difficulty. I did manage to get it out in 1 piece, but it was a consideration I had.

C16,C17 are only moderately difficult because of small size, use light prying pressure if required. But be Careful not to add pressure if not completely molten, Do Not lift a pad or trace! I don’t believe it will be required to nip legs on these IMO.

Generally speaking they all "walked out" for me by applying alternating heat to the pads.


I used the tools pictured.

Solder wick, flux, solder, a solder sucker, side cutters, xacto, small flat head screwdrivers and a pair of needle nose pliers to grasp legs rather than the components when removing. (specifically because the legs of the caps appear to be “crimped” to create mechanical friction inside the through holes)

Attachment 746018



The solder iron I used Both a thick wedge and a fine tip on Both a Weller WES51 and a WESD51, again maybe not strictly recommended, but I found success at:

Attachment 746019



~500º for all but the broad traces, which I boosted to ~600º and kept the application as brief as was possible.

Attachment 746020



I also uses a DIY helping hand and PCB clamp to grasp the board

Attachment 746021

The Novice advice I would give is, don’t be afraid of it! But be careful, you can cause yourself a lot of grief if you are impatient…

If anyone that desires or wants advice from a non professional perspective, please feel free to PM me or post here.

I am glad that we can continue with this repair!

That was a good idea to place painter's tape over the LCD display. Just make sure the adhesive isn't too tacky. It may remove the paint on the black housing. I typically stick the tape along my pant leg twice then apply it to an object. That usually reduces the adhesive property of the tape enough to still stick but not enough to cause problems if left alone for a while.

Yes, you are correct. C9 is a bipolar cap so it doesn't matter how it is oriented in the circuit. As for the height of the replacement parts, I don't have a hard and fast measurement. I normally reference adjacent components and roughly eyeball it from the removed component.

Refer back to my general soldering tips & tricks thread. It cites how to put double bends in the legs.

I offer this helpful hint when installing replacement components: Bend one leg of the component after it is placed in the eyelets. It doesn't have to be a 45* or 90* kink. Just put enough of a bend on the leg so it will remain in place. Then solder the opposite leg. Once the opposite leg is soldered, then straighten out the bend and solder the remaining one. This should keep the component from moving wildly. After both legs are soldered in place, cut the tails off at the peak of the solder joint.

Also remember to use flux and apply solder from the backside of the board. The flux will allow the solder to quickly flow through the eyelet and ensure a good electrical and mechanical bond. A good solder joint should look like a mountain peak on both sides of the eyelet. This circuit board really requires solder to be on both sides of the eyelet because some components need contact across the entire eyelet (some eyelets have solder runs on both sides of the board). Lastly, do not bend the components after they are soldered onto the board. They need to stand straight up and the bend puts undue stress on the component. For example, IC2 looks like it has a serious bend to it. If you try to bend it back then it may cause a break at the base of the device (not at the base of the solder joint).

One final thought on soldering the legs into position: Use the double-bend approach to passing each leg through the eyelet. This will ensure that the solder fills the eyelet and surrounds the leg evenly.

Does that help?