3rd Generation Specific (1993-2002) 1993-2002 Discussion including performance modifications and Technical Support Sections.
Sponsored by:

Testing and Calibrating the Speedo with a Signal Generator

Thread Tools
 
Search this Thread
 
Old Dec 20, 2024 | 08:58 PM
  #1  
RGF's Avatar
RGF
Thread Starter
Full Member
Liked
Loved
 
Joined: Mar 2022
Posts: 51
Likes: 21
From: Queensland
Testing and Calibrating the Speedo with a Signal Generator

Following my post about Testing and Calibrating the Tacho with a signal generator, I was asked about the speedo.

Some people might get perverse pleasure constantly removing and replacing the instrument panel but I am sure that most people only want to go there when there is an issue, get everything working on a bench and replace it just the once.

So, while I had everything out of the car I gave it a go. It is easy to set up, but then you enter a rabbit hole. I think that I have mapped it out but, full disclosure, I have only done this to my FD, won’t get to confirm things on the road for quite a few months, and some of the information here is just an educated guess derived from snippets posted on this forum and elsewhere confirmed to some extent by observations on my own 1999 RS.

Connecting up – the easy bit.

You first need to get access to the speedo circuit board. There are plenty of helpful posts as to how to get to this point. Then hook up a 12V source and signal generator. Fortunately, there are quite a few through hole mounted components on a board of this vintage so there are exposed leads you can hook onto.

AlienR2 has mapped out the input stage of the board: https://www.rx7club.com/interior-ext.../#post12212569. R18 is a through hole resistor, and you can squeeze between the circuit board and gauge face to get an appropriate test hook onto it, a standard CRO probe in my case.

I used these connection points (marked-up borrowed photo):




Set-up:







Too easy! A +/-1V sine or square wave will now drive the speedo.

Caution: The Odometer did not illuminate with 12V power and I thought no more of it. I am assuming the globe is separately powered. I am hoping that the odometer itself was also powered down, otherwise I have clocked many, many hours at 180kph.

Restoring factory calibration.

This had me stumped for some time. Re-read the disclaimer in the 2nd paragraph.

Mazda, and of lot of other Japanese speedos from this era, have the words “60mph=1024rpm” for imperial speedos and “60kph=637 rpm” for metric speedos on the gauge face. (1024 is a significant binary digital number but forget that thought as there is nothing digital about it in this application). That number refers to the shaft speed of the speed sensor. The speed sensor outputs 8 pulses per rev, so 1024/60*8 = 136.53Hz = 60 mph calibration point and 637/60*8 = 84.93Hz = 60kph calibration point.

I fed 84.93Hz into my metric speedo and it was spot on 60kph.

Use VR1 to adjust as necessary and you are factory calibrated – assuming that is your zero point has not been shifted due disassembly or some other mishap.

Custom calibration

All speedos are calibrated identically at the factory regardless of final drive ratio. Adjustment for diff ratio is performed by changing gearing to the speedo sender. This post https://www.rx7club.com/2nd-generati.../#post12617361, shows a 7 start worm driving a 23 tooth pinion which will reduce drive shaft speed to the sensor shaft by the ratio 7/23. I am assuming that this post shows a US manual sender for a 4.1 diff which clocks up with my subsequent calculations but if this is not the case the subsequent calculations here are wrong.

Later Japanese cars with 17” wheels, i.e. 4.3 diff ratio use a different sender but the same worm on the output shaft. I’m guessing it runs 24 teeth. I am confident enough with this guess that I am not going to take my speedo sender out to double check.

The automatic gearbox sender for the 3.909 diff is a different part number again. Beyond that I am ignorant.

Changing gear teeth adjusts the speedo reading in about 4.5% increments which you may or may not be satisfied with. We can make finer adjustments with VR1.

Calculations.

For the inhabitants of first world countries living in the 21st century, I’ll start with metric using my near full tread, road going Michelin Pilot 4 255/40R17 rear tyres and 4.3 final drive. Substitute your own parameters

Tyre rolling circumference. There are 3 different ways you can choose from:

Calculation: Pi*(0.255*40/100*2 + 17*25.4/1000) = 1.997m

Michelin Data book: 831 Revs/mile = 1609.34/831 = 1.937m

Measured with a fibreglass tape = 2.006

The US Michelin data book lists 831 against every 255/40R17 tyre so is the least believable against my Spanish made rear tyres. I am going with 2.006

Lets calculate for 100kph and scale up and down from there:

100*1000/2.006/60/60 = 13.847 wheel revs per second.

Then through the gear ratios and multiply by 8 pulses per rev:

13.847*4.3*7/24*8 = 138.93 Hz

For the third world inhabitants living in the 19 century, lets stick with the metric rolling circumference but go for a 4.1 diff, 60 mph with a 23 tooth drive gear on the speedo sender:

60*1609.34/2.006/60/60 = 13.371 wheel revs per second.

13.371*4.1*7/23*8 = 133.48 Hz - Quite close to factory calibration.

From there it is just a matter of scaling the frequencies up and down and adjusting zero and VR for your best fit.

Zero is adjusted by pulling the pointer off. It is a very tight press fit so takes a lot of force (see other threads), replacing it so it is firm, then rotating it against the stop past full scale. If you go too far, repeat. When satisfied push it home being careful not to leave it fouling on the gauge face. I use a couple of business cards as a spacer.

I was happy with my initial calibration, so I left VR well alone. It is likely to be quite sensitive. These are also a low reliability component. You may wish to go for a precision multi-turn trimmer and bend the leads to suit. From experience, some trimmers out there are rubbish. I'll name one of them again as payback for wasting my time: Suntan brand = Piece of ****. Go for a known brand like Bourns 3296X-1-204-ND

Last edited by RGF; Dec 21, 2024 at 02:15 AM.
Reply
Related Topics
Thread
Thread Starter
Forum
Replies
Last Post
anderml1955
1st Generation Specific (1979-1985)
3
Apr 26, 2022 02:55 PM
dcarroll95
3rd Gen General Discussion
3
Nov 13, 2018 09:14 AM
Acesanugal
2nd Generation Specific (1986-1992)
6
May 21, 2009 09:27 PM
black1991
2nd Generation Specific (1986-1992)
3
Apr 24, 2009 02:13 PM
Devy55
1st Generation Specific (1979-1985)
11
Dec 22, 2008 10:39 PM




All times are GMT -5. The time now is 03:09 AM.