Topic

I managed to find some time today to get some calibration data for the speedo and rev counter for the my TF. I plan on replacing the dials sometime and don't want to try and guess the position of the pointers when I have to remove them to fit the dials. Some people have been know to use their GPS data on their phone while driving their car to try and set the correct position for the pointers, I wanted to try and avoid this and get them bang on. This data may come in useful for anyone wanting to attempt this sometime.
First I had to find a signal generator capable of producing a pulsed square wave signal of sufficient amplitude to drive the speedo and tacho. Fortunately Ebay came up trumps again and this little baby cost all of £2.78 from China - it looked like it met the spec for the job!

Capable of 1 to 200khz and putting out a signal of 12 volts, it looked perfect.

So it was time to connect up the signal generator and get the oscilloscope linked into it so we could see what frequency gives what RPM.
The oscilloscope is a device to accurately measure waveforms, this Picoscope connects to a laptop PC and gives good results!



Now its time to adjust the signal generator to give a nice round RPM value, I chose 5000 rpm for easy maths. This is easily set by adjusting a potentiometer on the signal generator PCB.



Then with the oscilloscope channel 1 (in blue) connected to the signal generator I got the trace of the wave on the screen. As a check I used the Oscilloscope's on-board signal generator to match the frequency of the signal generator that is feeding the tacho, then I read the frequency off from the on-board generator to get an accurate value of the frequency, shown by the red trace underneath the blue one.



the frequency of the pulses to give 5000rpm is 167Hz or 167 pulses per second. This converts to a nice calculation of 10000 pulses per minute for 5000 rpm. So the TF tacho picks up 2 pulses per rev from the engine!



Next it was a double check at a different RPM value, this time I chose 3000 RPM.



And the Oscilloscope trace gave a nice linear value of 100Hz or 6000 pulses per minute!



So now I have accurate frequencies to calibrate the tacho, next its time for the speedo...

First I had to work out where to connect the signal from the signal generator, a check with the electrical schematics showed the input is on pin two on the green socket on the rear of the instrument cluster. The yellow wire is the signal coming from the signal generator. The red and blue wires are the 12 volts needed to power up the instrument unit.



Now the signal generator is turned on and a useful speed set on the speedo, in this case 100MPH.



Now the waveforms are matched with the Oscilloscope's on board generator and a frequency of 111 Hz is determined for 100MPH or 6660 pulses per minute.



Then its time to check the linearity by picking another speed, this time 50MPH



And the frequency noted as 56Hz or half that of 100MPH - as one would expect.





Then as I couldn't resist it, I took the speed up to max - not many of you will see a TF with this speed on the clock!



Note that the Odometer is gradually ticking away as I have done the speedo tests - This one is actually only 5000 miles more than on my car, so if I use this speedo, I may change it over when the miles match.

So there you have the frequencies for the TF instruments. I can now remove the pointers from the clocks and put them back on and check the calibration is bang on when it comes time to fit the new dials...I am hoping to fit some rear illuminated ones - if they are still available from a source I found last year!

Sundance

Clever stuff - not that understood much of it. Haven't used an oscilloscope since A-level Physics days.

We know where to come for electrical issues now though

cairnsys wrote: wrote:

Clever stuff - not that understood much of it. Haven't used an oscilloscope since A-level Physics days.

We know where to come for electrical issues now though

The signal generator is way of substituting the crank sensor or speedo sensor inputs and the scope tells me what the frequency of these signals are to give a set speed or rpm reading. Now when I remove the needles, I can make sure that they go back in their original place and that the speedo and Tacho will be 100 percent accurate.
I was hoping to fit a set of Metal Monkey custom dials with a backlight, but unfortunately they contacted me yesterday to inform me that they are now not making dials anymore, so I need to source another supplier - and Plasma dials are NOT an option, I just dont like the in your face look of them!
The project may be shelved until I can find a source - unless I can make them myself!

Sundance

Most people have seen the rev counter dance and the mph needle go off the rickter but wow that's amazing...it's a long time since I've seen a fuel gauge on the max.

Brilliant article you have posted here.

It gives me loads of the infomation I am looking for.

What I am planning to do is re-purpose an instument panel from a TF for installation in a Robin Hood Kit car.

I've got the rev counter sorted but hadn't got a clue which pin to use for the speedo input. Now I know.

Just a couple of questions, do you know if I can use a reed switch as the sensor, I was planning on attaching one near the crank pulley for revs and another near a wheel hub for speed and distance.

And, do you know if there is any way to calibrate the speedo to correct for different wheel diameter?

Finally, would you mind telling me where you managed to get hold of the schematic for the IP

Many thanks

John

Hi John,
Glad you found my article useful. I determined that the Tachometer looked like it picked up two pulses per RPM. If you use a reed switch and a magnet, you may need to double up on the magnets to give you the two pulses. Unless of course the TF rpm sensor pics up through a 2 to 1 gearbox.
Anything that will switch the 12 volt output should work, but the transducer must be capable of switching fast enough. I suspect that reed switches may not be quick enough as I am sure that they are mechanical switches encapsulated. If they are Hall effect transistors then you may be ok.
I would think an industrial proximity sensor may be a better option, but remember it needs to be a PNP one if its a three wire switch.
The electrical schematics have been attached to this reply for you to save. I have to admit that the MG-Rover schematics are the best wiring diagrams I have ever used!
Finally I have a pin out diagram I made for the fuel and temp gauges, I could post it up for you if I can find it. They do work differently to each other.

Good luck with your project..

Sundance

PF03FBC wrote: wrote:

….. in a Robin Hood Kit car......

Hi and welcome to the forum John

Which engine and gearbox are you using?

James

Wow ! Impressed doesn't even come close .

Excellent piece of work which I found really useful.
I have installed a cruise control onto my TF from Audiovox in which I used CJJ's guide to aid fitting. This is the third time I have fitted one of these cruise control units on different cars, but before I have used magnets to act as the input to the control servo. This time I tapped into the VSS circuit but using the CJJ guide of 2000 pulses per mile the cruise control would only work up to about 50 mph, not engaging at a higher speeds. This calibration data shows that the pulses per mile should be set to 4000. After resetting to 4000 ppm the cruise control now engages at all speeds and works perfectly.

There are a couple of us on the MG Experience forum fitting the MGF/TF EPAS into MGB's, we think we have most things sorted but would like a couple of things confirmed, the first is whether the EPAS will work with only one pulse per engine revolution, we think it will except the 3000 rpm cutout will now be a 6000 rpm cutout.
The second is do both engine and speed inputs have to be 5 volt square wave? or are they 12 volt square wave? We are using the Siemens 09.185.326 12 volt to 5 volt square as used by GM on the Corsa EPAS
Here is a link to what we are doing;;
https://www.mgexp.com/forum/mgb-and-gt-forum.1/mgf-power-steering-on-mgb.4178408/page-3

G'day MGB281,
I'm looking at using MGF power steering on a Triumph TR7 V8 conversion, I have the bits now I need the time, the poor old Tr is at the end of the project queue.
You can adjust the level of power assistance to road speed If you use one of these:
Jaycar KC5384 Digital Pulse Adjuster
in conjunction with a KC5386 LCD Hand Controller (unfortunately Jaycar don't have this listed on their site, it may no longer be available. You'll need to contact them to find out. Possibly there is an alternative.)

Jaycar Australia: www.jaycar.com.au 1800 022 888

Jaycar New Zealand: www.jaycar.co.nz 0800 452 922

Jaycar United Kingdom: www.jaycarelectronics.co.uk 0800 032 7241

Jaycar USA/Canada: www.jaycarelectronics.com 1800-784-0263

No matter how many digital pulse adjusters you and your mates have, you'll only need one hand controller as you only need it to set the digital pulse adjuster.

I can email you more information in a pdf file if you're interested, just PM me your email addy

Cobber
My personal message to you does not seem to work, here are the contents
Because Phil Harry in Melbourne is using a standard MGB engine and transmission he is wanting to use a sensor on the prop shaft to give his speedo pulse, a single pulse per prop shaft revolution will give about 3300 pulses per mile which is about 18% slow. Due to me changing the engine and gearbox this is not an option for me so I propose to fit a five tooth ring on a road wheel to get 4200 pulses per mile which is about 5% fast. We think that we will be ok with those pulse rates but we still do not know what voltage these pulses should be, is it 5 volts or 12 volts?
Likewise with the engine RPM sensor we were both going to use the coil to source a signal which on a four cylinder engine should give two pulses per engine revolution. Will this work? or will the voltages pulses vary to widely? Should we put a 2 tooth ring on the crankshaft and another sensor to read it?
We are both trying to avoid pulse generators with control knobs.
Philip