Ok,

I'm trying to fix my tach by soldering two 100k ohm resistors between pins 4 and 10 on the back of the tach, to try to get as close to 193k ohm as possible.

I'm new to soldering and something I don't understand is happening.

When I practised soldering the two resistors together, I got 198k ohm. Cool. So I practised soldering the two resistors to a junk PCB, and I got some weird readings. Individually, they measured as 60k ohm, but together they read as only 44k ohm... What??

I'd been playing with the same resistors a couple of times, so I thought I might have over heated them or something.

Pretty sure of my technique, I soldered two new resistors to pins 4 and 10 on the tach. Each measured as only 77k ohms, and together as only 113k ohms... Why?? I'm pretty sure I didn't subject them to excess heat this time.

Reading around online, it seems you can't measure resistance of a component when its part of a circuit. Sure enough, I cut my practice resistors off the junk board and remeasured - 198k ohm...

So, erm... have I got this right? I assumed the resistors would have to measure 193k ohms in situ as part of the circuit. I'll be sticking the tach in the car tomorrow to see what happens, but can anyone explain what's happening with these resistance values?

Many thanks for any help,

Neil

 

Hi Neil ,

What you are seeing is normal and there is a very easy explaination .

When the resistors are on the bench , yes of course they will measure at their rated resistance , But , when you solder them onto the board they are now in a circuit with several other components and those components have resistances of their own that will "load the circuit down" if you will , by their presence . Notice how the resistance is always lower when they are on the board ? Yep , that's cause your not only measuring the one you just soldered in , your also now measuring it and all the other things it's now connected with .

Never forget that when you parallel resistors , the number will get smaller , as the smallest number , zero , is a perfect conductor (straight wire) and the higher the number the "less of a connection" a resistor is . Thus , two 60K ohm in parallel will become 30K ohm (a "better" connection) , whereas those two 60K ohm resistors in series would be 120K ohms (a "lesser" connection) .

Hope that made some sense of it for ya ...

 

Thanks Mr. Bird!

Yes that does start to make some sense. I was just a bit thrown when I saw numbers I wasn't expecting. I'll chuck the tach in the car tomorrow and see what happens - the proof is in the pudding!

Cheers,

Neil

 

I have full confidence that your tach will work perfectly when you put it back .

Make sure to update here when you do get it put back in the car

 

When you try to measure something that's part of a circuit, you measure not only the thing, but also the circuit around it. Therefore any result you get by doing that, is meaningless.

 

Ah well,

I plugged the tach back into the car, confidently reassembled the dashboard, and fired up the car.

The tach jumped to 1 on the rev counter... and stayed there, stuck.

At least the needle used to go up and down before!

So, umm, any ideas where to go from here?

 

Damn ....

I got two ideas and I fear neither will be cheap . There is of course the buy one used option , and then there is a gent I've seen around these forums who runs a bit of a garage workshop operation where he repairs them . I've never used the guy's service so I don't know how much he charges but if you do a search in the for sale ads and find a guy selling rebuilt tachs I have read several good posts concerning him in the past .

And I was really hoping you had it solved . All does look proper with your solder , yes ? You don't have too much and have bridged a couple of traces that aren't supposed to be ?

 

Thanks for the advice,

I'm not an expert, but I'm pretty sure I didn't bridge any adjoining connections. I'm not completely surprised it didn't work, as my tach didn't quite behave like others. Although pins 4 and 10 were open, my tach didn't peg - it just overread, which is weird as all other tachs with 4 and 10 open should peg the tach immediately. The other strangeness, was that one of the other resistances across the chip was double what it should be...

Unfortunately, being in the UK, getting that guy to fix my tach will probably cost as much as buying another one...

Is it possible to buy a new chip and simply replace the existing?

 

You are now out of my area of knowledge as I have never attempted any more than the resistor repair to one of these tachs . The Chip may well be replaceable , just it may well not be what's wrong with your tach , and only someone with experience working on them would know .

 

Why don't you try cutting the resistors apart (de-solder the junction of the 2 resistors) thus returning the circuit back to it's original state. See if the tach functions as before or if you created a new problem. Go from there. GL!

 

Electronics as old as these often have issues with cold solder joints and dried out electrolytic capacitors.
I agree, undo what you just did, and then re-flow the components so all of the solder looks nice and shiny. Replace the capacitors and see what you have after that(keep in mind they have negative/positive leads). Good practice anyway.

IIRC, someone on here used a variable resistor for their repair, and adjusted it while in the car.

Another option is to get a wave generator and set it within range of what a tachometer would see. an oscilloscope helps too if you have access to one. If you're not that fancy, then take a lead off of the car (or any car) and test from the engine bay.
If you're sure yours is broken, then pick up a tach out of a different car to play with, then move onto this one and compare.

Some of the chips on these boards are proprietary, some are not. The latter can be replaced if you feel it necessary. There are numbers to cross reference and/or look up on places like digikey.com etc. Heck even the proprietary ones are probably common, just covered up or striped off of the top.
General tachometer circuitry can be found as well and you could reverse engineer the GM one you have with some effort. When you look at it subjectively, the circuit just converts a pulse to a metered output.

Lots of options depending on your curiosity and determination/resourcefulness.

 

Thanks for the advice and links guys,

As a first move, I'll pull the tach back out and desolder the resistors to see if I've actually made anything worse!

Very good idea to re-solder all the contacts - must be worth a shot. I assume I'd just warm the existing contact slightly and add a touch of solder? (I am new to soldering!)

As for the capacitors - would obtaining new ones be a simple matter of reading the numbers off them and ordering new ones?

As for the signal generator, would some cheapie tat like this do?
http://www.ebay.co.uk/itm/DDS-Functi...MAAOSwu4BVjdJd

I also found this diagram in another thread:


but the pin numbers don't seem to tally with this diagram (I soldered between 4 and 10 as on here:


Not sure I should be looking at these as I feel even more out of my depth!

 

Totally different chip. The chip you jumpered 4 to 10 on is just a resistor network (resistor circuits on a chip for sake of ease of install and acreage). The other is a function generator with support hardware on a chip. If you don't understand electronic circuitry and what the various components do, you are probably better off disregarding that circuit diagram. Do as advised above, go back to square 1 and determine that no damage has been done, then try the modification again and double check your work. HTH and GL!

 

The 1st drawing is National Semiconductor's "suggestion" of a "typical" circuit using their part. There's no particular reason that anyone creating an actual product would exactly copy that. In fact, in most cases they wouldn't; the values in their schematic are probably for a 4-cyl with a 0-6000 RPM range, that being their "typical" application further down in the text of the datasheet. Furthermore, there's no particular reason that Rx in an actual as-built product would match the R #s in NS's "typical" circuit, either.

The 2nd drawing is the thick film resistor network that GM decided to implement their design with. IMO a VERY poor choice, even given the limitations of technology at the time. (yes I'm an electronic engineer, so I deal with such things for a living) For some odd reason, probably lack of adequate power dissipation capability, it fails all the time. The resistor you're working on, called R4 in GM's design, is the one in the charge pump feedback circuit, which is R2 on the NS schematic.

I seriously doubt you hurt anything.

 

Oh... and yes, that function generator should work OK.

600 RPM = 40 Hz
3600 RPM = 240 Hz
6000 RPM = 400 Hz

 

Thanks guys - really awesome advice!

Hopefully I'll get a chance to pull the tach back out of the car this week, and then I'll see where things stand.

 

Edit: You can disregard what I posted below. It's been a while since I looked at a tach board and didn't realize they are one sided, and much more delicate than I remember. I changed some of the other so it's more appropriate for the electronics in question.
Most of the components on these boards are called through hole because the legs extend through the circuit board. The legs pass through what's called a barrel which is more or less a copper tube that allows the solder to flow in and cement the leg in place. The lines extending from the barrels are called traces and the board itself is typically fibreglass.
-What is important is that you need to heat up each connection enough to liquefy the solder all the way through the barrel, and then leave it with a shiny surface. Many times manufacturers don't heat up connections enough which results in poor connections from the start.
The other side of things is that you don't want to heat up the barrel or the components too much or they can become damaged from too much heat. The ideal situation is to have a soldering iron that can transfer heat quickly so you can get out before too much heat is applied over all. From the earlier info, too much heat can damage the traces and burn the fibreglass while the chips can be damaged internally. Technology this old is a lot more robust than current electronics but it's something to keep in mind.
That said, once you get the hang of it, it's pretty easy to do consistently.


You'll want to try to find capacitors as close to what you have already. That isn't always easy to do but places like Digikey should be able to help you with that.

Not sure if any of that helps but I tend to be more theoretical than practical. I've had the advantage of factory parts vs. sourcing and repairing whatever I've come across. On that end I do more out of curiosity since the time involved never makes it worthwhile financially.

 

These boards don't even have through-hole plating... they're absolute bare-minimum single-side.

I'd look at all the contacts with a good magnifying glass, like a jeweler's loupe. If they aren't cracked, I'd recommend leaving em alone... too easy to create solder bridges, overheat pads & traces and lift them off the substrate, etc. I'd also suggest, if you DO solder something, get some flux; rosin, NOT the stuff you use for plumbing. Then, when you get done, wash it all off with lacquer thinner. You'd be amazed at how much more professional it looks, even if you're not a pro, if you clean all that stuff up when you're done.

 

Yep, you're right Sofa. I'm at fault for quickly writing out some general info while going by memory. Probably a bad idea at my age. lol
Looks like the electrolytic cap/s were on the 90-92 boards only.
I also went and grabbed a couple of old tachometers and see now what you're talking about.

I do agree on the flux, it can even help remove solder bridges if the solder is kept to a minimum.

 

I did look quickly over the board when I was working on it, and the solders didn't look visibly bad - although I think one was slightly discoloured (sort of greenish?).

I didn't know to look for cracking though, I was looking for dullness. When I get the board back out, I'll check for any evidence of cracking and only make good those solder joints that seem to be affected.

I've been using a draper solder wire that already contains flux.

One thing that did puzzle me about my tach was that it only over-read, even though R4 was completely open - I understood the open ones should peg on startup. Does anyone know if that particular symptom might point to a particular type of failure?

 

Ok,

I finally got a chance to tinker with the car. I pulled the tach out and desoldered the resistors I'd connected across pins 4 and 10. Tach went back to roughly how it was before - high idle followed by pinging to the top and back again when blipping the throttle. Re-soldered the resistors and reinstalled the tach and back to where I was before - the needle would sort of settle to between 0 and 0.5 on key-on, then jump to '1' and stay there solidly while the engine ran, regardless of engine speed. Back to square one!

Should I bother trying to obtain an LM1819 chip and replace it?