Didn't make it there yet, but will soon.

 

Popped open the old MSD. Even more generic then I thought. Circa 1976 design. Uses an lm2917n for a freq to voltage conversion. Then uses a voltage comparator. The RPM pill does not nock any voltage off the the tach signal. There is a duel RC filter circuit off of the tach line with fixed values on it. Completely non adjustable. The RPM pill helps match what voltage is needed againts the one comming from the lm2917n to trigger the VNP14n's to flop there outputs.

My bad was counting on freq to freq for triggering. The RPM module(resistor) being part of an RC circuit for the oscillator (this is where rons 00 misunderstood my post) to make the comparing signal which would have been more accurate. MSD did it cheaper and did freq to volt converting then volt/volt comaparing.

Here's a scan/print/wrote on/rescan of the MSD and a pinout/layout of the old lm2917 chip and how they use to use it for driving tachometers.

 

Interesting, I think I see whats going on, maybe. I'll have to get into it in the morning, though.

 

What did he say? Now if it was computer language I would be all set.

 

Oh, sorry NOSJohn. Lemme say it dis way. The MSD uP(lol, not realy) takes the bit rate of the tach line in binary. It then converts the binary over to Hex. It then does a 2's compliment addition of a set value you want to look for a non positive and non negative result, meaning zero, or a match. I was guessing a Bin to Bin addition, the MSD ended up doing a bin to hex then a hex to hex addition.

LOL, all faux but does it help any? j/k

 

First, I am not an Electrical Engineer, just a hobby guy trying to understand, so some of next input may be off.

Zener: keeps voltage from car (12v in)constant? Was wondering how this was done. Thus a linear voltage that allows Pills (resistors) to act correctly. Seems 12v is the constant if my math is correct.

Heterodyne meter: each resistor gives it's own signature and/or voltage/frequency at heter meter.

What does the small Micro Farad capacitators do?

Tach Input: input frequency is changed to a voltage signal.

Ok, BMN correct me where wrong.

 

Zeiner diodes act pretty much like a pass through regulator. Much cheaper and simpler then a true voltage regulator. They are used in the tach line filter to keep flyback and kickback voltage, which can get quite high, from damaging parts. Not needed on cars that have a tach signal generated by the PCM. Only needed to clean up noisy/dirty signals from negative side of a coil.

Hmeter compares two frequencys togethor and it's output is usually the difference between the two. Not used in MSD or BMN

Tach signal is a frequency that is eithor read/computed or converted to a voltage then read/computed to see if it is the disired level. It can be any voltage you could imagine, all that is needed is the Hz or period of the signal.

Ask away, I'll try and help you figure it out.

 

I have a msd window switch on the way. so i should cut the wires to put it in 4 cylinder mode? (Still a newbe)

 

Yes, cut away.

 

Unless you are doing this just for science, it is pretty easy to get a fairly accurate resistivity reading for a desired RPM range. It is not a completley linear relationship, but we can do some guestimating and get pretty damn close. When graphed as RPM vs Ohms, it is actually two straight lines with a small change in slope.
4570 (6k pill) - 4780 (6.2k pill) = 210
4780 (6.2k pill) - 4990 (6.4k pill) = 210
These three resistance values ARE linear. So a 6.1k pill would be 4570 + 105 = 4675, and a 6.3k pill would be 4885

5210 (6.6k pill) - 4990 (6.4k pill) = 220
5430 (6.8k pill) - 5210 ( 6.6kpill)= 220

Again, another straight line. 5210 + 110 = 5320 for a 6.5k pill, and 5210 + 110 = 5310 for a 6.7k pill.

From the small sample of pills that we have, it is difficult to say what would be necesary for a 6.9k pill. Of course if we had a 7k pill it wouldn't be difficult to determine. This is something that is going to be rather difficult to come up with a simple equation. The best you could do would be to plug the numbers in to a computer and have a best fit line calculated. However is not going to be a very friendly equation to work with. My guess is that the pills are only accurate to with in 15 or so rpms anyway. If you follow my reasoning so far, the relationship is more than likely a smooth non-linear one, but due to keep things simple MSD has decided that between rpm ranges (say 6000-6400), 210 ohms = 200 rpms and from 6400 - 6600 rpms, 220 = 200 rpm. It is also possibe that due to the small sampling we have it only appears to be 2 straight lines. These resistors are probably only accurate to within 1% if we're lucky anyway. If we called and asked MSD how they figure resistivity values for pills they more than likely stuck a pot on the thing and wrote down where things happened and made a graph. Making an equation for something like this sounds cool, but it's not really practical IMO. This is "close enough for government work" Although given enough data points a pretty simple if/then program in excel could come up with a value using the same logic as above.

Edit, I just checked some of mine.
6000 = 4.58
6200 = 4.79
6400 = 5.00
6800 = 5.45
Slightly different readings. But all within 1% (.36 % @ 6800 rpm) of each other.

Hope this helps, knowing the values of more pills you could interpolate and get pretty damn close to any rpm value you wanted.

BTW, I did get a degree in electrical engineering but I haven't used any of it since I graduated, so I am way out of practice. And it's 4:30 am. I'm out.

Craig

 

Hey, thanks a lot. I've got thre 3, 4, 5, 6000rpm pill kits and think I will whip out the excel. Some good ideas.

 

Well did you figure it out yet? If you did I will send you 5$ to make me a 5900 pill.. let me know.

 

still working on it. not forgotten though. lots of stuff going on, but will get back to it soon.
Robert

 

When you know PM me I will be glad to give you my money insted of a company making 98% profit.

 

http://members.aol.com/blownf1/pill_rework.html

 

violator3434, thanks alot, this makes my task dissapear. I checked my numbers against these and they match up, so I was on right track. Once again thanks, Robert.

 

no prob