The only thing I have done thus far is to try an inverter based upon a 7406 chip. Unfortunately it seems that a spike riding on the square wave put out by the ECM was too much for it and blew the chip.

I note that some of the newer versions of these speedos seem to be capable of taking positive or negative signals. At least that's what one or two ebay listings say. Which makes me wonder if there is a way to switch them. I was not able to get any info regarding that from the seller I bought mine from. He just said go to a motorcycle shop and have them install an adapter, in other words, make them figure it out. When I opened mine up I did spot an open pair of solder pads labeled JREV. I sure do wonder what happens when you jumper those pads together, and I think that is the next thing to explore.

 

Yeah, I was never a Dora fan so I'm not down with exploring on this thing. I'm still weighting the decision if I should just try my luck or leave it on my shelf to collect dust. Your F1 wiring diagram appear to be very identical to that of the F3. After I set up my wiring, I looked over your suggestions and they were almost a perfect match; I even ended up changing one of my wiring to mirror yours a little more. This unit had such high potential, to give so much grief for tachometer is very disappointing. BTW, what did you attach your speed sensor to? The sensor bolt is enormous, I'm thinking of fabricating a bracket to attach to the brake bolts, how did you affix yours?

 

That's what I did. I can't find a good photo in my archives of the bracket, but it just attaches to the lowest part of the left brake caliper mount. It was pretty simple, just a short length of 1" x 1/8" aluminum with a couple holes in it. I think I attached it to a hole where a clamp for the stock speedo cable mounts. Don't know if the F3 has the same hole or not.

I tend not to know when to give up on these kinds of things and if I do end up getting the tach to work right I will post info here. If not, I'll let you know where I found a good deal on a Koso RX-2N, because I really like the way all the other parts of the thing look and work.

 

Take a look at this, it looks like he placed a capacitor in the line for the pulse generator. I'm not familiar with the gs500 wiring but that looks like that can be the pulse generator line.

 

Yup, I've seen it. It doesn't really solve the problem with connecting to the pulse generator, which is that it reads double the actual RPM ( with or without that cap in there). Tonight I tried improving the grounding of the tach by running a very heavy wire (maybe 12ga?) directly to the battery negative terminal. That made no improvement over the stock chassis ground. The tach cuts out at just about exactly an actual 4000rpm (reading 8000 on the tach) and drops to about 1000 rpm when it does. I'm guessing this is because it gets confused by the missing pulses in the pulse generator signal. Regardless, it's useless to use the pulse generator signal because it reads twice actual when it does read.

I have tried pull up resistors, a transistor signal inverter, an inverter using a 7406 chip, caps, voltage dividers, diodes. None have helped. I have tried to input square waves from my iPad. No luck there either. The only signal I have been able to move the tach needle with is the pulse generator signal and a weird ground loop interaction with my scope that was somehow putting a 60Hz AC waveform into the tach input.

I keep reading things in the various ebay listings like this is for RPM positive pulse (fuel injection), this is for RPM negative pulse (carb), this is for either one, tell us whch you have, etc. So somehow the darned thing can be adjusted. A pic of the guts is attached.

Note the open jumper pads JREV (bottom, slightly left), J12V (top, just left of cable socket) the missing resistors and transistor R6, R7 and Q7 just to the right of J12V, missing diode D19 at the far left, and missing diode D13 far right, and missing resistor R25 just below the black circular tach needle stepper motor housing.

All those pads are there for a reason. It would be great if someone from Cqsansan who knew the design could step up and offer some guidance. The most immediately noticeable thing is JREV. I am guessing I can jumper it without blowing anything up and see what happens. Q7 and the two Rs nearby could be for a signal inverter, but I don't have enough info to say. J12V might be for pulling up the signal. Again, I don't have enough info to say.

 

OK I did some more research yesterday, looking at getting a Koso speedo as a replacement. In perusing the various installation guides on the Koso NA site I found one possible nugget. The manual for the DB-02R (which looks way more like a current day GP tach than their so called GP tachs) mentions that the signal must be at least 15V and not over 24V.

I went back to my scope and looked at the pulse generator waveform, the one that the tach can see, that reads two times too high and then the needle falls off to almost 0 rpm at about an actual 4000 rpm. The sine waves that the pulse generator puts out are about 16V peak to peak at idle and are at about 25V peak to peak at 4000rpm. They voltage continues on up as the revs go beyond 4000, I saw up to about 29V peak to peak. By the way this waveforn is +/- around ground, so at idle it is about +8V to -8V.

The square wave pulse from the tach output is a 12V square wave, but it has a big leading edge spike - persumably an inductive kick from the coil - and measures about 40-50V peak to peak including that spike. The wave is all above ground, that is it starts at 12V, spikes up to 50V, then drops to ground or maybe just slghtly above ground, then goes back up to 12V.

So a new possibilty is that the signal voltage from the tach output is simply outside the range that the tach can read. I will attempt to study this a bit more and report back.

 

Couldn't you put resistor's in there to bring it into the voltage range? I'm assuming that the pulse is driving the tach and the voltage is what's shutting it down. If that's not right then it might not work.

 

Still collecting data. It might be that a clamping zener diode would work. To further add to the confusion I just found that if I use a 0.1uF capacitor connected between the spiky square wave at the tach output and the digital tach input and then set the tach for 1 cylinder, that the tach works up to about 1000rpm, then abruptly cuts off the second you rev up any higher. So now that I am able to get some sort of response from the stock tach output, I need to figure out why it cuts out above a slow idle. With the cap in place the general shape of the spiky square wave stays the same, but the bottom of the square wave dips slightly below O volts. Don't know if that is significant.

Curiouser and curiouser.

 

YES! Got it! I had to go all Thomas Edison on this and try about 100 different things before I got the right combination.

One thing that helped from where I left off at my last post was changing the capacitor between the tach output and the new tach input to 0.01uF. Going from .1 to .01 I was able to see the meter respond up to a bit over 3000 Rs. I had my scope probe connected at the junction between that cap and the tach, and I realized that it was creating a ground reference for the input. When I disconnected it I lost the signal. Which meant that the tach lost it's ground reference through the probe. So, I connected a 100K ohm resistor from the juncton of the .01 cap and the tach input to ground. Bingo! Solid revs all the way up the scale. I had to fiddle with the cylinder setting, it seems that 2cyl is the correct one for this configuration.

So to 'recap', connect a 0.01uF capacitor between the stock tach output wire and the tach input wire of the SS182. I used a polypropylene cap rated for 630V because it was handy but anything 100V or higher is probably fine. Then connect one end of a 100K ohm resistor ( I used a 1/4 watt metal film resistor, carbon film is probably fine) to the junction of the cap and the tach input and the other end of the resistor to ground. You probably need to set the cylinder number setting in the tach to two cylinders.

Let me know if it works.

 

So what exactly did that do? Change the voltage to the proper range that the tach see's? Sort of confusing Doc.