Designing a Better Tach Amp
 

Well, thanks to some excellent reverse engineering work by OldPokey, I finally decided to get off my dead end and start the project I've been thinking about for two years, designing a replacement for the Tach Amp we all love to hate.

I've decided to use an 8-bit RISC microcontroller as the heart of the project. I have the breadboard complete, and have written the code to get the microcontroller talking to the tach gauge. I can sweep the needle to any value, and dwell at any point for any length of time.

Photos are below

1) Rough breadboard complete
2) Sensor interface built on a separate board.

Next step is to optimize the magnetic sensor interface circuit, then write the code to bridge between the sensor and the tach.

As an adjunct to this project, I'm thinking of designing a tachometer diagnostic tool. The tool would plug into the tach amp socket, and perform the following diagnostic tests:

1) Sweep the needle around the gauge to test the gauge and wiring
2) Indicate supply voltage within range with a green LED
3) Check magnetic sensor and wiring for open circuit - indicate with a yellow LED
4) Visually indicate pulses received from the pickup with a red LED that flashes at an engine speed dependent rate.

Using this tool, it would be possible to diagnose about 90% of the possible problems with the tach circuit. Would there be any interest in this as part of the tool rental program?

More updates as they develop.

Great idea!!

I'd be interested in this and, whatever you come up with tach amp-wise. And, I'm in Austin with an '83 if you need another test article.

I'm east of Dallas with an '82. No tach problems, yet anyway.

Wow, that's amazing. You could probably make some damn good money selling those on the Internet after you're through with the prototyping, too.

contact funola
he did a bunch of stuff with the tach amp earlier this year.

Thanks. I'm going to try to build up 2-3 more of these for testing, and I'd like to try them in a variety of cars prior to moving to a PCB.

Assuming of course I can get the first one working.

How about a 555?
 

I'm really impressed with the engineering you've done so far but it seems that you may be making this more complicated than needed. How about using a 555 timer in a monostable circuit as a Schmidt trigger? Something like this -

http://www.kpsec.freeuk.com/555timer.htm#monostable

talking about electronics... how about cruise control amps... :D

A several reasons....

1) The magnetic pickup produces a TINY signal. About 500mV p-p at idle, dropping down to something under 250mV at higher revs. I don't have enough space to build the amplifier I need to boost the signal to trigger something like a 555.

2) I hate RC timing components. I never have the right values around, and they drift all over the place with age and temperature. I'd much rather just change some software code to adjust timing.

3) I also get the flexibility to do some interesting things. I believe I'll have enough code space to build in a self-test. When you switch the key to the glow position, the amp will execute a self test routine that sweeps the tach needle to 5k RPM and holds it there for 1 second. It then drops to 2.5K RPM and holds for 1 second, finally sweeping to 0, at which point normal operation commences. I like the visual indication that everything is working. Makes me sleep better.

Plus, my microcontroller is cheaper than an industrial temperature range 555.....

Help me control the VNT please :)

Testers Needed - Improved Tach Amp
 

Hi everyone....

As a follow-up to my last discussion, my tach amp redesign project
has progressed nicely, and I am now at the point where I would like to solicit
some volunteers to help me test my Protype Rev2 (Photo 1 below).

This is a completely modern tach amp, based on an 8-bit RISC microcontroller.

Rev2 has been redesigned to now fit inside the tach amp canister. I have
the electrical components all placed on a PCB (Photo 2 below), but I'm holding
off on finishing the tracking until I get some "real world" testing completed.

Code is written and is working well in my car (300D, Model Year 1982, build
date 5/82). I would love to get a few volunteers to ride with this tach amp for
a week or so and feedback any observations. I'm looking for a variety of
cars (Ds and SDs) and I'm particularly interested if someone in a COLD climate
could lend some help.

If you're interested in helping out, please PM me with your year and model.
I have two of the Rev2 prototypes which should be ready to ship out by Thursday
or Friday of this week.

Seems like overkill to me. The tachometer is a glorified voltage gauge. Most of the tach amp problems are with poor contacts (Thats why most place a cigarette butt under the cap) or bad pickup sensor sending it a poor signal rather than a failure of the amp itself.

The cigarette butt trick worked once for me. For about 15 minutes. I've driven for two years with no tach. Until this week.

On my search for a tach I came across another good idea: A guy in Germany is selling brand new tachs which he modified to take the signal off the alternator. No amplifier needed at all. For those who speak German, here's the link:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=180198007241&ru=http%3A%2F%2Fsearch.ebay.com%3A80%2Fsearch%2Fsearch.dll% 3Ffrom%3DR40%26_trksid%3Dm37%26satitle%3D180198007241%2B%2B%26category0%3D%26fvi%3D1

Slightly overpriced, though, but this is only for the idea.

techguy512: The 'cigarette butt fix' is most likely the cause of my destroyed tach.

Read this thread about the alternator source: http://www.pointedthree.com/disc/forums/showthread.php?tid=5047&posts=25&start=1

Airbus - Can you elaborate more on this? What got destroyed? Would you like to try this new one?

Well, I bought the 300D with tach not working. Of course, these little things bother me, so I read about the cigarette butt fix and it worked for a week or so. Then one cigarette butt wasn't enough anymore and I stuffed another one in: worked for a day, etc., etc. At some point it didn't work anymore at all, so I bought a new amplifier. Now guess what: It doesn't work. I checked all wires, and I get the signal to the dash, 12V and ground are good, too, but no reading. I guess, I can't offer you my help, because the tach itself is shot...

Progress Update
 

Good progress to report on the tach amp replacement front. The unit is working well, and thanks to everyone that has helped with the testing.

The PC board has been released to the board house. To keep the cost down, I've used 4-week (!) turn-around, which puts boards in my hands around March 1.

All the test vehicles have been W123s. I'd really like to get a W126 or even a W116 represented. All you need is the OM617A engine, and the 2 inch tall 'trash can' on the inner left fender. Please note the 1 inch tall 'trash can' on some of the later model W126 and W123 indicates your car doesn't use a tach amp.

Any W126 or W116 takers?

are you still planning the diag tool? The tach doesnt work on the cd and I would rather not start swapping parts.

2nd priority. It suddenly got complicated when I realized that to do this right I need 6 I/O pins on the micro and I chose one that only has 5. My bad. There is a work-around but I'm not ready to tackle it right now.

I'd be happy to send you one of the protos for you to try. It has the self test that was to be part of the diag tool. If it works, you know exactly where you stand. If not, you've still learned something.

ok that sounds good, send me a PM with the details.

No activity in over a year on this thread... is he still planning to make these?

I've got a 126 with a non-working tach and just stumbled on this. I'd be interested in either making one of these myself or just about any option besides a new tach amp. I may try and rebuild it to secure all the contacts if I can get all the gunk out of the cap. Anyone talk to him, know if he is still working on this and/or looking for prototype testers? Mentioned above he wanted a 126 to test on. :D

-TC

I did follow through with this. As of today, there are about 36 units out in the field and working.....

I have one... Works great. Thanks techguy512!

Are there still more available? When my dad pulls his summer car out of storage in a month or so....he will possibly need one. (Tach is the only thing that doesn't work...)

I also have one. Worth every penny.

I assume you've had success in the 126 then? I don't forsee any problems/differences from a distance. Congrats on the success and great thinking!

There was a stupid minimum order quantity for PC boards. I probably have half of them left. I build up the units as-needed.

Someone asked if these work in the W126. Success has been achieved in the W116, W123, and W126.

My W123 tach does not work. Is there an easy check for this? Should I purchase one of your new gizmos?

Can you describe symptoms?

1) Is it dead all the time?
2) Dead most of the time but springs to life every couple of months for a few seconds?

If #2, there's a 90% chance the amp is bad.

If #1, it could be more than just the amp. Crank sensor at the harmonic balancer, or bad wiring, or bad gauge, or even fuse.

It doesn't hurt to try. If it doesn't work, just send it back.....

I've no tach operation in my SD. I've never tried to diagnose this yet, however I have two spare tach amps I've pull at yards.

However, I applaud your effort to redesign this.

I would like one please
 

If you can make one, I would buy one from you.
What steps do I need to take to make this happen?

My tach is in the #2 catagory - jumps to life now and again, but is all over the place. I would rather get a known working item than a used tach amp that might work for a while and quit.

I remember seeing these for sale on Ebay last year, but was skeptical. Then I found this thread, but was too late.

Please let me know if a purchase is possible.

Thanks

If you ever want to diagnose the system, you can borrow the diagnostic tool. It gets loaned out periodically, but when it comes back I'd be happy to send it to you.....

Can't you probe around with a voltmeter? Also I may be interested. If I can diagnose the problem (my tach has been completely non-functional) to the amp I'll PM you. Again great work and effort, much appreciated!

You can test the crank pickup by using a multimeter to read the millivolts produced while the engine is running. Set it to Ohms and test for wiring harness continuity from the amp plug to the tach plug.

If you still need someone to test a unit out on a W126 SD, send me a PM.

I have one of these little jewels.
 

If your tach is dead; get one. Took all of 15 seconds to plug it in and it has functioned perfectly for about 6 months. The start up sweep is pretty cool, too.

Probing around with a voltmeter will give you only the crudest indication of functionality ("yep, it's there" is about all it can tell you).

The crank pickup is a switched reluctance sensor. The amp puts a constant current through the coil, and measures changes in the current as the pin swings through the magnetic field. Peak voltage of the signal is around 500 millivolts. Picking signal out of noise is the trick.

Thanks. SunnyinTX helped me git it working last spring. It's all good.

If none of the units I have works, I can probe the pins with a scope to see if the signal is getting through. Does the amp load the signal or pull it up?

My SD when bought new,the owner had 3 amps put in.The Techs could not fix.Years later I did.What was happening is everytime you screw the amp on,it pushes the metal sockets and wires down.Mercedes should have allowed more flex in the harness.Solution is to wedge up the wires and metal sockets,so as you screw down the pins,the sockets don't move.

Using a constant current source, put 30 to 50mA through the pickup coil (pins 7 and 9 in the amp socket). With the scope on AC coupling, you should get a zero volt signal when measuring from pin 7 (gnd) to pin 9. When the engine is started, you should see a small 'blip' once per crank rev.

Tach Amp
 

I bought one about a month ago after not having a tach for almost a decade.
Works perfectly; utterly simple installation. Great piece of work!

price????????

I can't find the tach amp on the SD. Where'd they hide it?

Late in model year 1983 they removed it completely from the SD's, and replaced it with a system that uses the EGR computer to drive the tach. All they left was an empty plug with a cap over in on the left fender. If you have an empty plug with a screw on cap that is less than about 2" tall, you have the new system.

This system was also used in the 300d's but not implemented until later.

Interesting! I wonder if the SD's EGR computer uses the same chip as the 83 240D's EGR computer? I know the 83 240D EGR computer uses the same chip as the tach amp in an 83 300D turbo, that's how I was able to install a tach in my 240D (in place of the clock) by taking the tach signal from the EGR computer to drive it.

If a scope is not available, you can use a DVM in the AC volts setting to read the voltage. If you rev the engine and you see a proportional increase in voltage to RPM, it is likely the tach pickup coil is good. If you get no voltage then it is toast.

Yup, I saw that. I thought it was a diagnostic connector.