Megasquirt Forum Area is for discussing Megasquirt EMS

Megasquirt Random loss of all power. Log attached

Thread Tools
 
Search this Thread
 
Old 08-19-15, 10:47 PM
  #1  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
Random loss of all power. Log attached

I've had this happen ~4 times ever. Once was when I first got the car back on the road earlier this year, once on the drag strip a few weeks ago, and now twice on the street in the last week.

All power cuts out during high(ish)-rpm pulls. Feels as if someone turns the key off for half a second or so. Each time it's happened, I've immediately let off the throttle. By the time I even think to look at gauges or anything, it's back to normal. I can then repeat the exact same acceleration without any problem. The car itself still has power, just the ECU cuts out.

In the .msl attached, it happens about 2/5ths of the way in. You can see the red line in the scroll bar in megalogviewer, same as if the car stalls or is shut off while logging. Looks like just over 6k rpm and 9.6 psi of boost when it happened.

Any suggestions on where to start looking?
Attached Files
File Type: msl
2015-08-18_22.46.35.msl (2.69 MB, 188 views)
Old 08-19-15, 11:02 PM
  #2  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
Voltage seems to wander a bit on throttle, is that normal? Battery is new, cables are new.

Here's another log from the drag strip a few weeks ago. Happens halfway through the log, at 7k rpm this time and 8.1 psi

Could it be noise on the CAS line? Grounding problem from the MS panel?
Attached Files
File Type: msl
2015-08-03_15.38.15run6.msl (282.8 KB, 355 views)
Old 08-20-15, 08:41 AM
  #3  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
I should probably include some info on the car:
-87 Rx-7
-Megasquirt 1, board V3
-Engine is mostly S5 turbo, with s4 housings and front cover. Just rebuilt less than 2k kms ago
-Stock s4 turbocharger, manual boost controller set at ~10.5 psi
-Tune is fairly solid, car runs great, compression seems good, fires up quickly every start
Old 08-23-15, 10:47 AM
  #4  
Engine, Not Motor

iTrader: (1)
 
Aaron Cake's Avatar
 
Join Date: Feb 2001
Location: London, Ontario, Canada
Posts: 29,789
Likes: 0
Received 108 Likes on 91 Posts
Is it possible you can point out where in the log the problems are located? Can't find them.

But I'd look towards an electrical issue:

-ground at Megasquirt panel
-main relay
-broken/bent DB37 pins

Along those lines.
Old 08-23-15, 06:33 PM
  #5  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
here's a screenshot of the 2015_08_18 file right at the moment it happened. Do you not see the red line in the scroll bar (circled) when you view it?

I also just noticed the 'MARK 002 RESET' text there (also circled)

I've had the car out twice this week and it hasn't happened again


Old 08-24-15, 12:49 PM
  #6  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
Here's my megasquirt panel ground. Is this a bad location since it's on that bracket on top of the actual metal of the car?

Old 08-25-15, 07:11 AM
  #7  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Originally Posted by distr0
Here's my megasquirt panel ground. Is this a bad location since it's on that bracket on top of the actual metal of the car?
The MS should be grounded to the engine not the chassis.
Old 08-25-15, 10:54 PM
  #8  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
so after not happening at all for the last week, it happened TWICE tonight. Babied the car home afterwards since i'm worried that one of these times something catastrophic is going to happen.... I'm leaning towards a noisy power source (alternator) since there's so much fluctuation of voltage on throttle in the logs. I'm planning to get the alt tested asap.

muythaibxr interesting, what's the thought behind that, is it just a 'cleaner' ground? would ALL things ground there (the whole MS fuse panel) or just the ground wires from the 'squirt itself
Old 08-26-15, 06:59 AM
  #9  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Yes, the factory ecu grounds to the engine for a reason. I grounded a Megasquirt to the chassis once just for the hell of it, and it caused all kinds of problems... every time I hit the brakes the CLT and MAT signals would change readings and the TPS signal would jitter and cause unwanted AE, which would stall the engine.

Proper grounding should be:

- Battery grounded to chassis and then to starter
- Engine grounded to chassis.
- MS and any associated grounds go to engine, preferably as several smaller gauge wires all the way to the ring terminal where they connect to the engine ground bolt under the intake manifold. Don't just join them all to 1 big wire then run that out to the engine.
Old 08-26-15, 08:44 AM
  #10  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
wouldn't that mean that sensor grounds (CLT, IAT, etc) would run from the sensors all the way in to the car to the megasquirt panel, then all the way back out into the engine bay to the block? wouldn't it be better to just run short wires direct from sensors to block in that case? not trying to disagree at all here, just to make sense of it all haha
Old 08-26-15, 12:04 PM
  #11  
In Full Autist Cosplay

iTrader: (1)
 
Black13B's Avatar
 
Join Date: Nov 2002
Location: Alberta
Posts: 2,330
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by distr0
wouldn't that mean that sensor grounds (CLT, IAT, etc) would run from the sensors all the way in to the car to the megasquirt panel, then all the way back out into the engine bay to the block? wouldn't it be better to just run short wires direct from sensors to block in that case? not trying to disagree at all here, just to make sense of it all haha

Your battery has been relocated hasn't it?

What does your circuit look like.

Is the circuit breaker on the positive side or the negative, how big are the wires, where do they go, etc..

If your chassis is grounded properly, then grounding to the chassis for the MS shouldn't matter - As long as your chassis is a massive single chunk of iron and not a unibody that has multiple layered panels that is spot welded and riveted together, that may or may not have been exposed to the elements for the last 30 years..

Where the ground is located isn't as key as how sufficient the ground is.

Easiest way to test if the ground is your problem is to run a cable directly from battery negative terminal to the ground point for the ECM temporarily, and see if the issue goes away.

If you are in doubt about the grounds at ALL, redo them.

Battery to block, block to chassis, chassis to battery (which is best achieved at the starter area). Those 3 are more than sufficient, and all 3 should be a heavy enough gauge to support the starter load on their own. Then you won't have any ground issues.

The next question will be is the chassis ground contacts sufficient from the ECM point to the main chassis ground which gets back to the battery. It's safe to assume it is, but something as critical as an ECM ground should be as close to the ground loop as possible... No assumptions neccessary..

Last edited by Black13B; 08-26-15 at 12:06 PM.
Old 08-26-15, 06:49 PM
  #12  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Originally Posted by distr0
wouldn't that mean that sensor grounds (CLT, IAT, etc) would run from the sensors all the way in to the car to the megasquirt panel, then all the way back out into the engine bay to the block? wouldn't it be better to just run short wires direct from sensors to block in that case? not trying to disagree at all here, just to make sense of it all haha
Sensor grounds need to reference the MS as 0v. Those should go to the MS sensor ground. It is mainly about making sure 0v is consistent during all parts of engine operation. The chassis is not a good place to do that. Keeping the factory ground locations and cleaning them up is the best way to do things on this car.

See this link: http://www.msextra.com/doc/pdf/html/...re-1.4-13.html

Last edited by muythaibxr; 08-26-15 at 07:06 PM.
Old 08-29-15, 10:19 AM
  #13  
Engine, Not Motor

iTrader: (1)
 
Aaron Cake's Avatar
 
Join Date: Feb 2001
Location: London, Ontario, Canada
Posts: 29,789
Likes: 0
Received 108 Likes on 91 Posts
In complete agreement with the above. I used to tie all my grounds to the chassis near the ECU, then run that point back to battery ground. I was reliable, except when it was not, and it meant that everything had to be PERFECT not to see jitter or voltage differentials.

Now I tie to a common point at the 'Squirt mount, insulated from chassis, then run several 12 gauge wires from that point to the engine block. Of course the engine must be properly ground. Two 4 gauge wires or thicker, one from the starter area to chassis. One from another convenient point (I like the front iron where the accessories mount) to chassis.
Old 08-29-15, 11:08 AM
  #14  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
i had just one large (4ga i think) ground from starter mount to chassis. last night i added a second one from the rear iron to chassis, into the car at the same point as the megasquirt ground. that's about all i could do with the supplies on-hand. i'm going to try drive it now, and if it still acts up i'll have to dismantle things some more and run independent grounds.

wish i had known all this before i snipped everything short and grounded all to the ms panel

battery-wise, it's relocated to the rear bin, negative terminal runs to where the rear seatbelt would bolt to. circuit breaker mounted at the battery, cable run from there to a bulkhead fitting on the transmission tunnel near the starter, then from under the car there's a short run to the starter and another cable up to the fuse box under the hood

here's an in-progress pic of the battery setup

Old 08-30-15, 09:44 AM
  #15  
Engine, Not Motor

iTrader: (1)
 
Aaron Cake's Avatar
 
Join Date: Feb 2001
Location: London, Ontario, Canada
Posts: 29,789
Likes: 0
Received 108 Likes on 91 Posts
You can ground to the MS panel, but just make sure the terminal is isolated from the chassis and then you have a few 12 gauge grounds from that terminal out to the engine.

Or run a big ground wire from the engine, to that terminal, back to the battery negative.
Old 08-30-15, 04:10 PM
  #16  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Originally Posted by Aaron Cake
You can ground to the MS panel, but just make sure the terminal is isolated from the chassis and then you have a few 12 gauge grounds from that terminal out to the engine. Or run a big ground wire from the engine, to that terminal, back to the battery negative.
Multiple smaller wires from the MS to the engine with no intermediate attachments works best.

Ken
Old 09-05-15, 09:40 AM
  #17  
Engine, Not Motor

iTrader: (1)
 
Aaron Cake's Avatar
 
Join Date: Feb 2001
Location: London, Ontario, Canada
Posts: 29,789
Likes: 0
Received 108 Likes on 91 Posts
Yeah, I know that's the preferred method.
Old 09-07-15, 07:57 PM
  #18  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
argh, happened again last night. so my plan is to run a cable from the battery negative terminal to the megasquirt panel,and another from ms panel to engine, then remove the current cable that went from the ms to chassis. i think that jives with what you guys were suggesting?
Old 09-07-15, 09:11 PM
  #19  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Originally Posted by distr0
argh, happened again last night. so my plan is to run a cable from the battery negative terminal to the megasquirt panel,and another from ms panel to engine, then remove the current cable that went from the ms to chassis. i think that jives with what you guys were suggesting?
No not really.

Ground from battery should go to strut tower then to starter. Engine should have another ground from the rear iron to the firewall. The MS should ground to the engine and only the engine.
Old 09-07-15, 09:37 PM
  #20  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
Originally Posted by muythaibxr
No not really.

Ground from battery should go to strut tower then to starter. Engine should have another ground from the rear iron to the firewall. The MS should ground to the engine and only the engine.
ahh ok... i was going by what Aaron suggested here, unless i misread:

Originally Posted by Aaron Cake
Or run a big ground wire from the engine, to that terminal, back to the battery negative.
right now by battery negative goes only to the rear seatbelt bolt (battery is in the rear bin) so i wondered if that could be part of the problem... do i need to add a cable all the way from the rear bin to the front strut tower?
Old 09-07-15, 09:40 PM
  #21  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
any opinions on whether the voltage fluctuation in the logs (bouncing between 12.5-13.5 V rapidly under load) is normal?
Old 09-08-15, 07:16 AM
  #22  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Originally Posted by distr0
ahh ok... i was going by what Aaron suggested here, unless i misread: right now by battery negative goes only to the rear seatbelt bolt (battery is in the rear bin) so i wondered if that could be part of the problem... do i need to add a cable all the way from the rear bin to the front strut tower?
I would attach all cables the way the factory did. The strut tower is closer to the engine than the ground point you chose. That makes it better.
Old 09-08-15, 07:16 AM
  #23  
MegaSquirt Mod

 
muythaibxr's Avatar
 
Join Date: Sep 2004
Location: Maryland
Posts: 4,721
Likes: 0
Received 1 Like on 1 Post
Originally Posted by distr0
any opinions on whether the voltage fluctuation in the logs (bouncing between 12.5-13.5 V rapidly under load) is normal?
Not normal.
Old 09-18-15, 12:09 AM
  #24  
Senior Member

Thread Starter
iTrader: (1)
 
distr0's Avatar
 
Join Date: Dec 2010
Location: St. George, ON
Posts: 536
Received 12 Likes on 9 Posts
finally had some time to run new grounds. removed the existing single wire that went to chassis and ran multiple 12g from ms panel to engine. took the car for a quick drive and had no issues so far, but the logs still show some voltage fluctuation and drop at high rpm. im leaning more and more towards faulty alternator (although i just had it tested and it passed)
Old 09-27-15, 09:22 AM
  #25  
Engine, Not Motor

iTrader: (1)
 
Aaron Cake's Avatar
 
Join Date: Feb 2001
Location: London, Ontario, Canada
Posts: 29,789
Likes: 0
Received 108 Likes on 91 Posts
Do yourself a favour and get a new FD alternator with warranty.


Quick Reply: Megasquirt Random loss of all power. Log attached



All times are GMT -5. The time now is 07:19 AM.