Dip Switch Settings on a VDO Tachometer

[aeromech]

[exvwmd]

I have had ALL the same problems with my VDO tach. It NEVER was accurate,not even close from the very begining.{off by as much as 1000rpm's too low} Called the main office where I bought it and they told me to set it to the 6 cyld setting. Can you imagine that? I did it and now it is ONLY off by about 400 rpm's higher at 65 mph. Great huh? I am done with ALL vdo gauges {pressure and temp gauges are off too.} Any one know a better brand for VW gauges?
_________________
Mary Jane Muggles Lives!

[telford dorr]

[aeromech]

I don't dispute anything you said Telford. Obviously you have a lot of knowledge here. The only thing I'm saying is that I've install two of the VDO's into buses and never had any issues with accuracy or noise and I never did anything in the electrical system to counter that. In one bus I actually did have electronic ignition running too.

[Air_Cooled_Nut]

[telford dorr]

[telford dorr]

[Traveling Writer]

The VDO instructions/PDF is WRONG.

The correct settings are: 123; UP DOWN DOWN

I had it set according to the instructions, the needle got to 500 rpms at good revving. Reset the switches (vdo in place) with a tiny mirror and screwdriver. Works like a charm. Needle is steady as a rock with the diode in place. I used a diode that ends in 07, not 05.

And how do you edit the title of your post?? I'd like to add that the problem is solved for future searches...Everett?
_________________
Cheers,
Davi

1977 2.0 FI Westy
2011 1200cc Yamaha Super Tenere
1976 Feet (they work surprisingly well)

[Air_Cooled_Nut]

[ScottK]

[aeromech]

Here's some new information and I may need to eat some crow.

I have an engine test stand and a control box with a VDO tach as well as oil pressure and voltage gauges. It's got a low oil pressure idiot light, a green light to show that ignition is active and a starter switch. I haven't used this very much... it's pretty new.

Today I test ran a type 3 engine and the tach was erratic. I was thinking in the past that this was probably caused by the control box being too close to the coil and that I needed shielded cable but after today's discussion I'm wondering if I need the diode previously mentioned. What do you guys think?

[SteveStevieStevens]

Don't some of the newer alternators fix the radio whine issue with a diode on the back of the alternator? I know the one I recently bought did.

That might be the difference between those who had the needle jump and those who didn't.

I'm far from an expert though.

[Air_Cooled_Nut]

[aeromech]

[Air_Cooled_Nut]

No, it's mounted in the stock clock location, up in the dash, as any good gauge should be
_________________
Toby http://www.aircoolednut.com/
Did I mention that I'm an original Darksider?
'72 VW Squareback, 2007cc, GB 5-speed, rag top; '76 VW Riviera Penthouse Sundowner 2.0L; 2015 Audi S5 Cabby w/Stage II APR; '06 Ducati Sport Classic 1000; '14 Ducati Diavel Strada
The First Invasion

[aeromech]

[telford dorr]

VDO Tachometer Fix

OK, got a VDO tach (333051D) for my 71 bus. Wanting to avoid post installation hassles with this thing, I did a bench test on the unit. Set the switches, hooked it up to a 12 volt power supply, and applied a 12 volt square wave from a Tektronix 114 pulse generator. When the frequency of the generator was 133.3 hz, the tach indicated 4000 rev / min. Wohoo! It works, and the calibration is right! Also checked at 2000 and 6000 rev / min. Both were close enough.

"OK, self, this should be a piece of cake install," I say. Little did I know...

So install the unit in the dash, wire up to ground (at the dash junction) and +12 volts (at the idiot light supply junction.) Run a length of #16 wire back the the engine bay, dressing it in with the air cond harness, tie-wrapping it up securely.

Of course, there's always side projects: welded up the sheet metal screw holes for the front cover pan, re-drilled, painted bare metal and re-installed the screws so the pan was nice and tight again.

So, time for a test drive. Start it up. Stupid tach is jumping all over hell. Spin the engine up a little, and the tach pegs at 7K. Damn!

So check everything: power and ground measure out at 14.2 volts, with the engine running, so that's OK. Put a 'scope on the points signal line: looks OK.

Fig 1 - Raw points Waveform

[Note: the labeling for the vertical axis in the waveform above is incorrect - it's actually 50 volts / division.]


As you can see, the points voltage spikes to 230 volts positive to 80 volts negative, with quite a bit of oscillation. This is a normal waveform.

Installing a 1N4005 diode in the signal line, per the VDO tech note, didn't fix it.

So, what's with the tach? POS! My only conclusion is that their circuit design is marginal to poor, and can't handle the real world. On the other hand, since it looks good in the dash, and works well on the bench, maybe we can fix this. So I set about to create my own "front end" buffer circuit.

Fig 2 - Prototype board

Fig 3 - Prototype close-up

Fig 4 - Schematic

As you can see, the circuit is a simple circuit, based on the ubiquitous, good for almost anything LM555 timer chip. In this case, we're using it strictly as a Schmitt-trigger inverting buffer. The real work is done in the preceding input signal conditioning circuitry. The prototype was built on a chunk of perf board and cobbed up to the engine so measurements could be made. I'll post pics of the final packaging when it gets done.


The following Circuit Description is for those interested in exactly how the circuit works, so that you can modify or adapt it to your needs.

The raw points signal enters through resistor R1. R1 has a high enough value such that there is virtually no loading on the points signal. R1 and capacitor C1 form a low-pass filter which removes most of the ringing from the waveform. Diodes D1 and D2 clamp the signal to between 0 and 12 volts.

Fig 5 - After one level of filtering

As you can see, most of the nastiness has been removed from the waveform, but not all. We can do better.

R2 and C2 form a second low-pass filter, removing all remaining ringing from the waveform.

Fig 6 - After two levels of filtering

So now we have a nice clean waveform, but the signal impedance is way too high to connect to the tach, so the LM555 chip buffers the signal. This chip can supply about 200 ma of drive, so you should be able to easily drive the tach. Also, this chip can take up to 18 volts of supply voltage, so it should easily survive in an automotive electrical environment. Just in case, resistor R4 was added to act as a fuse, and to smooth out the input power a little. C3, C4, and C5 remove any input power noise. The chip inverts the signal polarity, but the tachometer doesn't care.

Fig 7 - Final output to the tach

Just for fun, R3 and the LED form a built-in static timing light, in case you have to statically time the engine in the field.

Parts for this circuit used to be available at Radio Shack, but they seem to want not to be in this business any more. It should be easy to find parts on-line (in the USA, Mouser Electronics, if nowhere else.)

This pic shows the prototype board cobbed into the engine bay for development.
Fig 8 - Prototype in engine bay

Fig 9 - Closeup of prototype in engine bay

This pic shows why the board was cobbed in where it was - too hard to get probes to the coil...
Fig 10 - Hidden ignition coil

Fig 11 - Circuit under development

Since adding this buffer circuit, the tach display is completely stable and smooth, without a hint of jumpiness.

Side note: the lamps in the tach were way too bright at night, compared to the other dash instruments. This was solved by adding around 18 ohms in series with the feed wire to the lights. You can use an 18 ohm 2 watt resistor. I used four 4.3 ohm, 1/2 watt resistors in series (because I had them in stock), covered with a length of heat shrink tubing.

So, you may ask, this seems a little complex for what it does. For me, it solves several problems: (a) makes the tach work flawlessly, (b) avoids running a multi-hundred volt points signal all over the bus, incurring electrical noise in the audio equipment, and (c) isolates the signal line from the points so that any failure or loading in wiring downstream won't affect engine operation. In the future, I plan on adding a circuit to drop power to the AC compressor above a preset speed, so that you can go fast when needed without worrying about the compressor over-speeding and blowing up.

Note: I did not open the tach to reverse engineer it's innards and find out why it works so poorly out of the box, as it's a new tach and that would have involved mangling / removing the outer trim ring. If anybody has an old tach that they would be willing to donate to the cause (non-working is fine), I'll open it, document it's innards, and post the results here. I'm sure it's modifiable internally, so it will work right on a Bay without an external circuit. If the tach isn't working, I'll try to fix it, free, assuming parts are available, and return it to you. Email me if interested.

[aeromech]

Do they give Nobel prizes for VW bus engineering?

[kjspgd]

Telford-

This is a cool and timely bit of information.

I've got an old harley 2stroke tach mounted (don't ask. PPO... but it's grown on me) and I've been prototyping an inline flip-flop get the tach to only receive every other pulse. I saw some weird results (not quite divide by 2) and thought it might have been related to sloppy signal. I'll try the low-pass filter and move up to the full on buffer if needed.

[telford dorr]

Should work fine. Instead of the 555, you can go to 4000 series cmos logic - a Schmitt hex inverter and a flip-flop; buffer the ff output with paralleled leftover inverters. Let me know how it works out.
_________________
'71 panel, now with FI
'Experience' is the ability to recognize a mistake when you're making it again - Franklin P. Jones
In theory, theory works in practice; in practice, it doesn't - William T. Harbaugh
When you're dead, you don't know you're dead. The pain is only felt by others.
Same thing happens when you're stupid. - Philippe Geluck
More VW electrical at http://telforddorr.com/ (available 9am to 9pm PST)