Tach resistors

[Eleven]

I had a fellow test my tach and he said that my thermo resistor (at least that is what I think he said) was burned out. Since there was no writing on the part, he could not replace. Does anyone know the numbers of this piece or have a schematic that would show it? Many thanks!

 

[alpine_64]

I had a fellow test my tach and he said that my thermo resistor (at least that is what I think he said) was burned out. Since there was no writing on the part, he could not replace. Does anyone know the numbers of this piece or have a schematic that would show it? Many thanks!

I think Tom Hall might do new tach internals.

 

[Eleven]

Thank you Michael, do you know how I reach Tom?

 

[Tom H]

Tracy,

Here's a link to a schematic.

http://www.classictiger.com/techtips/motach.html

The thermistor is clearly noted here as the device next to R2. It is an unlabelled device. It would surprise me that is could be "burned out", which would imply it was destroyed by some overload. I would expect that any overload would also destroy the meter movement. The purose of this thermistor is to provide some temperature compensation for the tach. I have not figured out if it is designed to compensate for temp effects of the meter movement itself or on the electronics, such as the transistors. Could be both. But the tach should still operate- but not so accurately - if the thermistor was removed.

I have a spare tach of unknown quality. If needed I could measure the thermistor and see what are its characteristics. I could measure its resistance at two or three temps. But if someone else already knows these it would be nice to have that info.

Tom

 

[RootesRacer]

Ive seen the thermisters crack, this would make it a full or intermittent open.

Tom, It would be very nice to know the temp coefficient of the resistor.

Otherwise I am still up for the task of designing a retro board to eliminate all that funky circuitry.

 

[bernd_st]

Forget about the Thermistor and put in a 56 Ohm resistor instead. It will work perfectly.. The Thermistor the biggest over engineered part of the Smiths tach...

 

[Tom H]

Can you explain further? I assume you mean put a 56 ohm resistor in place of the thermistor, i.e. in parallel with the 55 ohm resistor already there? If so, why not just replace the 55 ohm with a 27 ohm.

But more importantly can you explain why you feel the thermistor is not needed. Do you know what value res it is at say, 0 deg, 25 deg and 45 deg C? And do you know if it is there to compensate for the meter movement (copper coil) or deoes it compensate for the temp dependence of the Germanium transistors, which would change to V level at which the transistor turns on and off.

Since most of us only drive our Alpines in decent weather,with temp between 60 and 100 Deg F, I suppose the temp stability is not that big a deal.

Thanks, Tom

 

[Eleven]

Boy, am I in over my head. I'll try to link this discussion for the fellow who is working on my tach. Maybe he can tell me more so that I can accurately pass it on. I truely appreciate you fellows interest in this!

 

[Tom H]

Tracy, at least be sure he gets the schematic. And know that a couple of us have a curious interest in the subject, besides wanting to help you.

And you should understand that the device your guy says is bad is simply a temperature sensitive resistor. It is a device that is deliberately designed to change resistance as the temperature changes. Some thermistors are used to measure temperature. But this one is in the circuit to compensate for and correct for undesired temperature effects in some other part of the circuit.

I'm still hoping to hear from someone who knows the specific characteristics of the one in our Tachs. And also to hear more from bernd.

Tom

 

[Eleven]

Thank you. I will pass on everything I learn on this. I really appreciate your help!

 

[Tom H]

OK, so I did some measurements on the tach and the thermistor.

The Tach meter movement itself is about 72 ohms (at 68 Deg F, room temp)
The thermistor itself, then when paralleled with the fixed 56 ohms is:
450 Ohm @ 40F (5C) = 50 Ohm total
210 Ohm @ 68F (20C) = 44 total
170 Ohm @ 81F (27C) = 42 total
112 Ohm @ 95F (35C) = 37 total
66 Ohm @ 113F (45C) = 30 total

Note that a moving coil meter responds constantly to a constant current, but the circuit applies a Voltage (pulsed, but a volatge pulse). In order to maintain calibration, the resistance of the coil plus series resistance added must remain constant. But the meter coil is copper which has a positive temp coefficient of 0.4% per deg C. So the neg tempco thermistor is an attempt to compensate for the Copper coil.

Over the above noted temp range, the coil resistance varies from 68 ohms (@40F) to 79 ohms (@113 F). If the series resistor were fixed at 44 Ohms (instead of the 56 Ohms plus thermistor combo) the total resistance (Series Res plus coil res) would vary from 112 to 123 ohms over those temps. Assuming the Tach was calibrated at around 68F, that would cause the meter to read about 3.5% High at 40F. And it would read about 5.7% low at 113 F.

With the compensation factor of the thermistor this reduces the variation (and in fact reverses it - a bit of "overcompensation") to a range of about 118 (@40F) and 109 (@113 F). Again assuming the tach is calibrated at 68F, this causes the tach to read about 1.7% Low at 40F and about 6.4% High at 113 F.

Soooo. It seems that the compensation from the thermistor may help in the range from 40 to 80 deg. Probably about what the English designers had in mind. But for some of us who drive in warmer climates, or those of us who only drive their Alpines on warm sunny days, and who have a hard time keeping the driver's compartment cool, this tach compensation does not seem to help much. In fact it may hurt. So I think Bernd has a valid point. You can leave it out. Maybe replace the 56 ohm resisitor with a 47 ohm. And maybe calibrate your tach at some elevated temp, close to what your behind-the-dash temp gets to when driving.

I wish I had done this earler. Next time I have my tach apart I think I will leave the thermistor in place, but I'll put a 33 ohm resistor in series with it, to reduce the compensation by about half, which should get it to just about even.

So Tracy, tell your guy to replace the 56 Ohm with a 47 ohm, and remove the thermistor. Then have him calibrate the tach at some temperature that is close to what you have inside the driver's compartment most of the time.

And Rich (Rootes), sorry but I'm still an analog guy!

Tom

 

[RootesRacer]

>>And Rich (Rootes), sorry but I'm still an analog guy!

You thinking perhaps I am Rich?

 

[Tom H]

Nah, I meant you, Jarrrid. I keep losing track. Senior moment.

Tom

 

[Eleven]

Amazing! Tom, thank you for all of this!!

 

[bernd_st]

Coming back to the original topic :

1) The Thermistor -as correctly stated earlier - was intended to compensate mainly for colder temperatures which are not really relevant for our normal Beam operation. Would say they are cracked or broken in around 70-80% of all Smiths tachs

2) Just unsolder the remains and put in a 56ohm resistor instead. Of course it has the same effect to put a 27Ohm instead of the 2 x 55Ohms however because of the specific PCB layout I find it easier and less time consuming to just exchange the broken thermistor.

3) 47 Ohm would work as well. The delta is only minimal and the instrument has to be calibrated with the poti anyway.

4) For calibration you need a dedicated pulse generator with a reference rpm meter (make sure it´s referring to the correct number of cylinders).

P.S. I´m rebuilding these Smiths tachs as a Hobby so everything above is experience based...

 

[Tom H]

Bernd,
Good and valid points all around, except note that there can be significant error at high temp, not just cold, unless you do the cal near the temperature it will operate at. I would expect most people would do the cal at a normal lab temp of about 74 F. But the tach can easily get over 100F in Summer or Arizona, and the uncompensated meter would read about 4 % low.

I have rebuilt 4 of these tachs so my knowledge is from experience also. I have opened up about 7 tachs and have not seen any thermistors that were broken. All of them have quite a bit of flaking of the epoxy coating, but all were working. If you have a precision Pulse generator you can skip the reference RPM meter. I use a crystal controlled digital synthesizer (borrowed from work) as my pulse gen. I made a special version of the outsde part of the coupling Transformer to better match the 50 Ohm impedance of the Pulse gen to the Tach. I wound about 20 turns of small wire where the normal 1 turn of the ignition wire goes. Not sure if more or fewer turns would be better, but this works OK.

I set the Pulses at 200 Pulses per second (4 cyl) and adjust the pot for 6000 RPM. Then I check it at 4000 RPM (133 PPS) and 1500 (50 PPS) where the cal dots are on the SV dial (just outside the regular dial marks). I think the original factory adjustment is done at 1500 because that's where the cal dots are closest together, and then the other 2 points are checked. I guess this is better to have your most accurate readng near idle where you often adjust the carb and ignition. But I always figured I am more worried about the accuracy near red line. On the SIV tach the cal points (dots on dial) show adjust at 1300 and check at 3500 and 5200.

I'm glad you posted about how useless the Thermistor is. Until then I always assumed it was important. Now I see it helps a little, but actually hurts a bit at High temps.

Have you ever done any temp testing on a complete Tach? I might try that someday.

Tom

 

[puff4]

OK, so I've been reading this discussion with great interest, particularly as (1) I have *no* electronics knowledge or experience and (2) I have a flaky tach. When it gets very hot outside (80+ degrees) the tach begins to read *much* higher than it should, following exponentially along the scale. That is to say, it reads reasonably close to the actual idle speed, but then as I speed up it starts to take off and at the point at which it should read around 3000 it's reading probably around 4500, etc. Is this a busted thermistor?

 

[Tom H]

Kevin,

Could be the thermistor. But the error would be a constant percentage. So if it reads 4500 at actual 3000, then that's reading 50% high. And if it were a thermistor problem it would read about 50% high at idle also. But that means reading 1500 when it's actually at 1000. But some might say 1500 is "reasonably close" to 1000.

I'm sending you an e-mail.

Tom

 

[bernd_st]

Bernd,
Good and valid points all around, except note that there can be significant error at high temp, not just cold, unless you do the cal near the temperature it will operate at. I would expect most people would do the cal at a normal lab temp of about 74 F. But the tach can easily get over 100F in Summer or Arizona, and the uncompensated meter would read about 4 % low.

I have rebuilt 4 of these tachs so my knowledge is from experience also. I have opened up about 7 tachs and have not seen any thermistors that were broken. All of them have quite a bit of flaking of the epoxy coating, but all were working. If you have a precision Pulse generator you can skip the reference RPM meter. I use a crystal controlled digital synthesizer (borrowed from work) as my pulse gen. I made a special version of the outsde part of the coupling Transformer to better match the 50 Ohm impedance of the Pulse gen to the Tach. I wound about 20 turns of small wire where the normal 1 turn of the ignition wire goes. Not sure if more or fewer turns would be better, but this works OK.

I set the Pulses at 200 Pulses per second (4 cyl) and adjust the pot for 6000 RPM. Then I check it at 4000 RPM (133 PPS) and 1500 (50 PPS) where the cal dots are on the SV dial (just outside the regular dial marks). I think the original factory adjustment is done at 1500 because that's where the cal dots are closest together, and then the other 2 points are checked. I guess this is better to have your most accurate readng near idle where you often adjust the carb and ignition. But I always figured I am more worried about the accuracy near red line. On the SIV tach the cal points (dots on dial) show adjust at 1300 and check at 3500 and 5200.

I'm glad you posted about how useless the Thermistor is. Until then I always assumed it was important. Now I see it helps a little, but actually hurts a bit at High temps.

Have you ever done any temp testing on a complete Tach? I might try that someday.

Tom

Tom,

nice to appearantly meet a real expert for tachs on the forum. Unfortunately I don´t have a professional pulse Generator therefore I built a simple device myself. It is based on one of the earlier rebuild articles and uses a Timing IC555. Accuracy is generally good but it´s certainly not a 100% reference. Therefore I like to keep the reference RPM Meter. But you are right there is a direct proportion between munber of pulses vs. needle deflection.
Usally I do calibration with 4 measurements at 1500, 3000, 4500 & 6000 rpm. Of course impossible to adjust to 100% matching needle indication, but with some exoerience you can at least come close (best compromize). My simple generator is normally connected to the hot line of the input transformer which goes to the electronic circuitry. Coupling done by a 100nf/100kOhm serial RC buffer.

Unfortunately have not made any Temp curve measurements since over here in Germany we are only talking about normal driving temperatures between 15-25°C . At least in that range the Tach works accurate enough without the Thermistor.

Have rebuild between 10-15 units so far and another 20 or so are waiting - as time allows ...

 

[Tom H]

This past weekend I did some temperature testing on a SV tach. Simple setup: Large cardboard box, with a hole for a hair dryer, a thermocouple thermometer, a stable signal generator, and a SV tach. I found that with all the original circuits in place the tach reads about 3.5% Higher at 110 F than at 70F. 6220 vs 6000 RPM. Not as bad as I expected based on my "back of an envelope" calculations. This tells me that the thermistor does not "Overcompensate' as much as I thought, and also suggests that there are other thermal factors besides the tempco of the moving copper coil in the meter movement.

So I repeated the test with the thermistor removed from the crcuit. WOW! Without the thermistor providing any thermal compensation, the tach reads about 12% low at 110F vs 70F. 5330 vs 6000. I will redo these tests more carefully and with a more precise signal generator, but I want to warn all as soon as possible NOT to delete the thermistor.

I have seen many of these thermistors look like they are shot, but I think they are still OK, usually. There is often a thin coating on them that flakes off, and the device then looks shot or burned out, but I'll bet they are still useable. On one that I have the lead has broken off on one side of the chip. But I think it can be simply resoldered on.

From my research so far, I have not found a replacement thermistor, so we had best save the ones we have.

I'll report more when I have finished my tests. I think I can even come up with a way to "fine tune" the compensation.

And I will also do some tests to see how much warmer the tach gets when the gauge light is on. I'll bet it makes 110F gauge temp more common than we thought.

Tom