When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Yeah I did think it may have been done for ease of assembly. Although is having two speed sensors easier than having two clusters? Hard to say . That wouldn't explain how the aftermarket 300kmh gauge faces work accurately, though. If the odo did change with calibration and you adjusted the calibration pot right down to scale the gauge from 180kmh to 300kmh to utilise the new face scale, how would the odo count right? It wouldn't, it'd read low as it would be unaware of the face change. Unless the odo isn't affected by this calibration and the calibration is just to get the analogue face to read right.
I'll find out when my car is back together so then we will know for sure . Or I may rig up my drill and some rpm sensors to test the theory, that's likely an easier route than trying to map out the speedo board.
You're running the rx-8 crown and pinion right? So if the odo isn't corrected your odo would be out by 8.29%, so only 0.83 miles over a 10 mile journey, not overly noticeable unless you do a longer trip. But if compared to GPS would be quite noticeable over a 100 mile trip or so.
I am running the RX-8 4.44 ring and pinion. I have tested over long drives with my GPS and the mileage is accurate. But, it is quite possible that the numbers lining up on the speedo don't equate to them being correct in the odometer, especially with a good deal of change, like changing R&P and going to a 300 KPH speedo, etc.
Apologize for the bump on an old thread but for those with JDM cars did this work across a range of speeds? When im thinking about it, it seems like the scale between MPH speeds and KPH speeds wouldnt be the same. So you could calibrate a single speed of the KPH gauge to read MPH but wouldnt it be inacurate the further you get from that speed in either direction?
Apologize for the bump on an old thread but for those with JDM cars did this work across a range of speeds? When im thinking about it, it seems like the scale between MPH speeds and KPH speeds wouldnt be the same. So you could calibrate a single speed of the KPH gauge to read MPH but wouldnt it be inacurate the further you get from that speed in either direction?
Yes I had a broken 300 speedo and changed over the face to my working 180 speedo . Calibrated it with the trimpot on the side , works like a charm . Did 2 of my friends JDM speedo too with 300 faces . Everything works as described
Yes I had a broken 300 speedo and changed over the face to my working 180 speedo . Calibrated it with the trimpot on the side , works like a charm . Did 2 of my friends JDM speedo too with 300 faces . Everything works as described
Bumping this up with a special thanks to Dale for this excellent write up. I’ve had the 4.3 rear for about 15 years and this was always on my to do list when first seeing this thread many years ago. I finally got the courage to pull the cluster and attempt the work. (Another thanks to RuffRotor’s YouTube video for the excellent step by step instructions on cluster removal). I had the benefit of one of those mobile speed radar things on a backroad near where I live. With 245/40/17 tires, my speedometer was showing 60 mph when the radar indicated 54. Did the adjustment with the screw and no kidding, it is very, very sensitive as Dale pointed out. I gave it more than a slight push clockwise and it was now showing about 43 mph when the radar indicated 60. Took about 4 or 5 more adjustments to finally nail it at 60 on the speedometer and 60 on the radar.
Bumping this thread to thank Dale for posting this up, and to provide some additional info for those of you who prefer doing their speedometer calibrations on a bench instead of the on the road. In full disclosure I haven't gotten around to doing this procedure on my FD yet, but I will as soon as I can get my hands on a suitable signal generator and an oscilloscope, which are both needed to do this on a bench.
Basically I did a little bit of reverse engineering of the FD Service Highlights and Body Electrical Manual (BEM) to figure this out...
This is a block diagram of the speedometer circuit from the FD Service Highlights document
What we can glean from this diagram and from elsewhere in the applicable BEM and FSM troubleshooting/test procedures is that the OEM speedo sensor is essentially an AC generator that produces a sine wave signal with 8 positive peaks per full rotation of its shaft. We can also glean that its peak output voltage is around ~5 VAC. I derived that 5VAC figure from this speed sensor test procedure in the BEM:
To bench test and calibrate the speedometer, we need to do the following:
1. Apply +12VDC power to pin 2D, and a ground return to pin 1E.
2. Use an AC signal generator to generate a sine wave at various frequencies (I'll get to that soon) and connect that to pin 3C (output terminal on signal gen) and pin 3A (output return terminal on signal gen).
3. To verify the signal generator output, we'll connect an oscilloscope to monitor the signal to ensure it's outputting the correct frequency, amplitude and sine wave shape.
So now that we know the amplitude (~5VAC) we need to know how to set the frequency. And that info can be easily derived from this table in the Service Highlights, which I annotated to show some math...
So if you go to the yellow highlighted 60MPH data, you just need to convert the "pulses per minute" data into "pulses per second" or Hertz (Hz) to get the frequency needed to result in a 60MPH speedo reading. Keep in mind the figures in this table are based on OEM gearing and tire sizes.
To calibrate your speedo, you'll just set up the signal generator to produce frequencies at a few different speeds on the table, adjusting the calibration pot to get it as close as you can.
Very nice info!!! This would allow for someone to do a whole build and calibrate for every change all at once. Ie wheel and tire combos and rear end ratios.
Last edited by gracer7-rx7; Jan 23, 2024 at 01:17 PM.
Reason: removed spammy links
I've been running that setup for a while now, VERY happy with it. The big thing I haven't been happy with, though, is the speedometer error. It's close to 10% or so, which doesn't sound like much, but when the speedo says you're doing 90, you're actually doing 80. I guess it can keep you out of a speeding ticket, but it's REALLY annoying.
So, I've been looking into how to fix this problem. There are devices that can modify a digital speedometer signal to compensate for R&P changes, tire size changes, etc. Mighty Car Mods just put out a video this week on a digital speedometer corrector that got me thinking about this -
(BTW, if you guys haven't started watching MCM, do so! Really neat web series).
While a digital corrector is one good way to do it, I really wondered if there was some other adjustment on the speedo itself. 240SX's have a number of traces that can be cut or wired to correct the speedo, for example, so I decided to dig in.
Pulled my instrument cluster out and removed the outer covers. First off, I'm running my original, stock, US speedometer - it's the original one to the car. Along with that is a '99 6 o'clock tach and a JDM oil pressure gauge.
You can (obviously) see the speedo on the right. If we get over to the right side of the cluster and take a careful look, we can see a small blue poteniometer (pot) -
OK, so we've got an adjustment pot. Fun! Now what to do...
First off, I drilled a hole in the white plastic body of the instrument cluster for a straight shot at the pot with a screwdriver. If you look in the second pic, you can see the hole. I drilled a few holes next to each other to make a larger hole, that made it possible to not only have a screwdriver in there to adjust, but to also look in at the pot and watch it as I adjust it.
The pot has a WIDE range of adjustment. Here's the rules of adjusting the pot -
- It is very sensitive in that a small turn will easily change 3-5 MPH.
- Turning CLOCKWISE will REDUCE the speedo reading.
- Turning COUNTERCLOCKWISE will INCREASE the speedo reading.
OK, so we now know how to adjust it. How do we know how MUCH to adjust it?
You need a speed reference. The good thing is this is easy. Most car GPS's will give a very accurate speed on their display. I found even better than that to get a speed app for my phone. I have an Iphone 4, I got an app called "SpeedBox Lite". It was free, it displayed the realtime speed in MPH, KPH, or even knots, was easy to read, and kept the phone from going to sleep/locking while it was active. This worked PERFECTLY for me. I'm sure there's an Android equivalent app, just search around.
OK, armed with that and a freshly-drilled hole in the side of the cluster, I put the cluster "glass" cover back on to protect the gauges (I didn't put it all back in the full instrument cluster shroud) and plugged in the 4 connectors to the back. I just set the cluster in place and brought a #2 Phillips screwdriver to adjust the pot.
To start off with, I cranked it all the way counterclockwise and went for a drive. Going down a residential road I saw that my signal was WAY too high - the speedo was showing 60mph and I was MAYBE doing 20-25! So, that's the wrong way. Pulled over, shut off the car, cranked it the other way. Now it was showing barely any speed when driving - like 5 mph or something. Good.
So, I kept doing iterations of that - see what my error was like compared to the GPS, pull over, shut off car, make adjustment, start car, and test again. You may be able to adjust it while the car is running, but I didn't want to try that.
I eventually got it to within 3 mph at 30mph. Getting close. Now, tiny adjustments are key - like BARELY bumping the screwdriver into a turn. Once it seemed pretty close at 30, I went to 40, got that close, then to 50 and got it dead-on. At that point you should be within a good margin, if it's accurate at 50, you're going to be accurate up higher, but you could keep tweaking if you so desire.
Now, here's the biggie, I think this might be useable to re-calibrate a KPH speedo to read in MPH. So, it would say "50 KPH" but you'd be travelling at 50 MPH. That is HUGE. If that works, you could buy a JDM speedometer (of which there are MANY more that are in good condition!), calibrate it, and have an accurate speedo!
It makes sense that Mazda would just make one speedo mechanism, put the appropriate faceplate on, then have a factory worker calibrate it to that face. That makes MUCH more sense than entirely different speedo "guts" for each market. This could also explain how there are many JDM aftermarket speedos that read up to 300 KPH - they just bought the same mechanism, got a faceplate made, and calibrated it.
Most likely next weekend I'll have a chance to test that theory - a buddy of mine has a bad speedo and has been running a JDM speedo (reading in KPH). So we'll see!
Anyhow, this is VERY exciting and a VERY simple and easy fix!
Dale
my one is broken… tried to adjust it a bit and it just came right out. The hole was drilled before so, I guess previous owner really did a job on it..
tried to fix it by soldering it out, but not a chance.
If anyone know what it’s called so I can replace it, that would be awesome !
I have a 2002 Bathurst with the white plates so hard to find a new one…
HIGHLY recommend sending your cluster to Michael Gagne, he is a regular on the Facebook RX-7 group. He can 100% fix and also replace failing components and calibrate so you have a good, reliable cluster.
I've had mine done by him - top notch work and he has a whole setup to in-depth test and calibrate our clusters.
. That wouldn't explain how the aftermarket 300kmh gauge faces work accurately, though. If the odo did change with calibration and you adjusted the calibration pot right down to scale the gauge from 180kmh to 300kmh to utilise the new face scale, how would the odo count right? It wouldn't, it'd read low as it would be unaware of the face change. Unless the odo isn't affected by this calibration and the calibration is just to get the analogue face to read right.
