What's new
CK5
Register an account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members.

Alt tach sig for old six two...

DieselWarrior

DieselWarrior

1/2 ton status
Joined
Jun 3, 2010
Posts
678
Reaction score
0
Location
LowBuck-LowTech
Id like to develop a tach interface (Since the 'other' one is friggin $65!) using a Picaxe embedded microcontroller (Yeah, Im a bit of a techn0 n3rd too).

Ive been reading about the tach sig from an alt that is equipped with such an output, I think it was P or W or something.

I doubt my '82 has such, but I did read that you can tap into one of the diodes to fetch the same sig on an alt not so equipped with the output. I read it takes "about 30 minutes" to do. :whistle:

Question... Since Im an alt dummy, anyone care to give detail on how this is done? (Pix would be valuable at this stage :popcorn:)

If I can get the sig, I can build the interface for it and we can use standard (Read: $50) gasser tachs, ACCURATELY.
The other option I am thinking, but adds expense to the project, is to use an LCD display that interfaces with the
Picaxe (About $30??) to display in digits, the RPM. I should be able to use a 16mhz micro for fast results. (?will have to see?)
The display option would be really neat, and I think all in all, would cost less than the 'other interface' alone!

Any help would be greatly appreciated, and Id be willing to recruit contributors to field test the interface at no cost to them.
(At the end of the test, they keep the interface) This would be offered to someone that "goes out of the way" to help
with fetching the sigs, Id need only 2 or 3 in CONUS for this offer to work.

I have a stroboscope to optically get ACCURATE RPMs, and I can borrow an oscilloscope to visualize the pulses.
I 'think' the rest would be rather elementary. (Now watch, since I said that, it will end up being an undoable task! HA HA!)


Thanks much!
Andrew
 
OK ...If you do this, and pull off for $100 or so, I'll buy one. I would love a three or four digit LED digital tacometer. I don't do PICs or Stamps so I was thinking of doing the same thing using a dimple D/A converter and some bargraph LEDs with enough resolution to show a 50 rpm change. That would require 80 LEDs The thing about gas car tachs is that they redline at 8,000 or better. No big diesel should ever go over 4,000 so adapting the w terminal to a car tach is not IMHO a good plan. Your competetion is a company named Tiny Tac (tinytac.com) which used a peizosonicsensor on an injector line to get a count. I would greatly prefer seven segment or bargraph LED over the seeven segment LCD they have.
 
Check out the PicAxe, its a LOT cheaper and simpler to get into!

My idea was to count the pulses in a given time frame and calculate the RPMs from that. Using a PicAxe compatible LCD display, I can serial out to it the results (hence, display the RPM on a nice backlit LCD)

The possibilities are "endless", like setting a red LED to light/sound a buzzer/etc when you exceed "x" RPMs, etc..

This is of course, in the concept phase. My concern is the sampling rate and how fast I can get the display to refresh. Id like something without a delay and shows "almost real time" RPMs, at the minimum. If this cannot be accomplished it would not be worthwhile to invest a lot of time.

The only way I can find out, is to start prototyping one!

I cant see it "selling" for more than $100 for a finished and assembled product at this time.

In small quantities, I can build my on PC boards, program the chips, etc.. So this wont quite be something you see at Jeggs or Summit, more than likely 2 or 3 at a time available on ebay. :popcorn:
 
I am running a factory tach off the engine speed sensor installed on the oil pump on my 6.2L. I built a simple transistor circuit to convert the sawtooth waveform to a square wave and it works just fine. The ESS makes way too much voltage, but the frequency is correct for the tachomenter. Run the signal into the base of the transmitter and hook your tach up to the emitter with the collector connected to 12V, bingo, 12V square wave as required by the tach.
 
There is a speed sensor on the oil pump? Isn't the oil pump inside the pan? My 84 is 49 state version and I don't have a ECU. I found a digital tach (Autometer Sport Comp) but it needs an input pulse train.
 
My apologies, that should read vaccum pump! The vac pump with the ESS is from a cali spec 6.2L, or from a 6.5L with a DB4 injection system (trucks without a vac pump and DB4 have a sensor installed on the filler piece to run the oil pump off the cam)
 
So... I need to replace my old vac pump with a newer model, which has the ESS built in. Do you have a "year" that would have this feature?
 
my 91 6.2 doesnt have a vacume pump at the back of the engine,it just has 2 wires coming off the piece that is in place of the vacume pump.can this be used to run a tach?it appears to hook into the cam and oil pump.
 
sledheadak said:
my 91 6.2 doesnt have a vacume pump at the back of the engine,it just has 2 wires coming off the piece that is in place of the vacume pump.can this be used to run a tach?it appears to hook into the cam and oil pump.
Click to expand...

It sounds like you have the engine speed sensor version of the oil pump drive. You will have to build as stated earlier in the thread build a Simple Transistor circuit to convert the output into something the tachometer can understand. I would PM Russell to see if he can get you a parts list or diagram to what he did.
 
I have not given up on the project, I am working on learning more about the Picaxe microcontroler and the ability to output to LCDs... When I have something more to post, Ill do an update.

Andrew
 
OK...I am thinking I will buy the injector line clamp on sensor from Tiny Tac and put a scope on it to see what kinda of signal it makes. If it is a peizoelectric sensor, it might make a petty good pulse. After I do that I'll re-ponder my bar-graph LED design. The other day I put my hand on one of the injector lines and I could feel the pulses at idle pretty well.
 
maynardogle said:
OK...I am thinking I will buy the injector line clamp on sensor from Tiny Tac and put a scope on it to see what kinda of signal it makes. If it is a peizoelectric sensor, it might make a petty good pulse. After I do that I'll re-ponder my bar-graph LED design. The other day I put my hand on one of the injector lines and I could feel the pulses at idle pretty well.
Click to expand...

I can tell you that's what it is.
I just didn't want to go thru all the hassle to make my own circuit so I bought the whole package.
I even got more than one size clamp for my different engines in case i wanted to swap it from vehicle to another.
 
TTT recognized...

I am learning about the PULSIN command, you can specify a duration for the MC to count pulses... from that some simple math to calculate RPM... and I also have just gotten some 7 segment LCDs for dirt cheap, they will only display "numbers", but for $3 each, they were right on target.

My junker is at the shop having head gaskets (sigh) replaced... when I can get her back Ill be able to start experimenting with the project more... Maybe a small magnet epoxied to the crank pulley and a Hall Effect sensor to send pulses to the MC.

I hope this might turn to a project I can "sell" either as a kit or assembled maybe. Will have to see how it progresses.


Ill be back with updates as soon as I can!
Andrew
 
diesel timing pulse adapters are a peizo clamp. and you hook the timing light up to it. works great. totally easy to verify timing on mechanical engines. the one I've used was a Snap-on/ Blue point unit. with an El cheapo timing light.
 
I know, its been FOREVER ago I started this.

Ive graduated to full blown PIC microcontrollers and running LCDs no problems.

I am still trying to figure a way to sample the revolutions of the engine, with a very simple technique, not needing to replace this, or that.

I am looking at Hall Effect Sensors to see if one can be used on the crank with a small (but strong) earth magnet epoxied to the pulley, and a hall effect sensor counting the pulses in a time domain to calculate the revolutions.

I WILL post back with any further progress, but Ive got to finish working on some other stuff first...

Andrew
 
You must log in or register to reply here.

Latest Posts

Top Bottom