JP86SS

iTrader: (1)

Well all this talk about it finally broke me down to actually doing it!
Should have done this long ago but all good things...
Anyway, in thinking on how to do the VSS and DRPs I want to check some math with others who have done it. After going over all the info I could find, none of the drawn diagrams seemed right for the calculation as found for the chip. I made up a spreadsheet and diligently plucked numbers until I came up with these.

The RPM & Speed chart is figured on using a flip-flop for 50% DC but that also is at the cost of 1/2 the frequency. I have also added the ability to switch in a larger capacitor to reduce the ranges to get slow speed and cranking speeds.

Parts should be in in a week or so and I want to dive right into building.
just going to breadboard it and use jumper wires everywhere. Will post more info on what I'm doing, just don't want to make this post a book.

Does my math look right?
I keep thinking I got the freq wrong on the DRP. (factor of 4, div by 4 ??)

Edit: "Rb" has a bias resistor to set min control range.

jwscab

I'll have to look back at my notes and see what I did. I'll reply sometime this evening.

in the meantime, I'll PM you to try and send you an .xls file I have that has some rpm/frequency calcs in it.

JP86SS

iTrader: (1)

Thanks Joe,
You did send me a schematic a while ago that I referenced but found some things that didn't look like they were actually built/finished/tested so I figured to start from the begining and research from there.
The FF has no output as I discovered. Q2 nada. I added that and then started recalculating the frequencies but my head is swirling from the "how many per RPM" thing.
I made a sheet with the formula shown. Hz cut in half by flip/flop and should be right for 4 pulse per rev.

junkcltr

iTrader: (1)

JP,
Take a look at the Megasquirt Stimulator schematic. It sounds like the RPM simulator is very close to what you are building. It doesn't use a digital flip-flop though.

EDIT: It sounds like you are building a bench or something. One thing I tend to do sometimes is not use external hardware and do it with the code. For example, when I was testing the WBO2 to NBO2 code I simply would load the NBO2 RAM with a toggling value after the ECM did the actual A/D read. The rest of the code would think the NBO2 was toggling from the actual A/D value that was read. I could tune the NBO2 to varying AFRs and see how the code was responding. I would adjust the mV value and the rate at which it toggled by changing a counter value for the toggle rate.

JP86SS

iTrader: (1)

Yep, finally building one.
Excellent idea, thanks for that.

JP86SS

iTrader: (1)

Since there was another request for info I figured I'd update this post and get some info out there on what I'm doing.
The schematics are PDFs.
The RPM is confirmed to work but is not fully debugged yet.
Could just be my kluge board at the moment but when getting to 3K RPM my scanner stops logging. not sure what to make of that yet but the signals do work.
The VSS has not been connected to the optical input yet. hope to do that this weekend. Might need to add the flip-flop to that circuit as well.
I believe I did on my board but it is not shown on the schematic. Just do it the same as on the RPM one. The signals are cut in half by the FF, the values that are listed on the chart for the VSS are correct when using the FF.

The MAX232 circuit was built using a "MAX232ACPE" chip but did not work properly.
I think because the ALDL schematics show 10uf caps but the ACPE datasheet says I should be using 0.1 uf, still working that out and drawing pin lables need to be updated. Have been using the cable from my car to do my logging.

I have a nice little box with all my switches and pots mounted in and just beginning to wire it up. Will post pics as I go.
Main power supply is mounted in a 19" rack computer case that I put a wood plank in to mount termonal strip onto. ECM is all prewired to the terminals so making changes and connections will be easy in the future.
Terminals give nice test points for every ECM pin too.
Here's what I have so far. The pot values may change as I go along once they are all powered up. I also need to add resistors to them to set min/max ranges so codes are set unless you want to go there.
Thanks to Jwscab for some info to get me moving on this.
Jp

JP86SS

iTrader: (1)

Here is a pic of the board arrangement I was planning for individual RPM or VSS controls. Am figuring on making the boards modular so the can be stacked into a box or something. Rather than all on one board for now.
This way they can be put in a main box and have a small remote box with indicator and adjustment (on a cable) sit on the table next to the laptop.

JP86SS

iTrader: (1)

BTW, This is taking allot more effort to build that I first thought but I know it will be worth the effot. I'm learning a great deal about chips and crap just doing it. A nice learning experience and a hobby at the same time.

junkcltr

iTrader: (1)

Great work JP. That is the best part of this hobby....learning new stuff and getting your hands dirty all in the DIY way(non-commercial). I like the fact that you are building it all from scratch. I copped out on the bench and most of it is run from a Megastimulator. If I ever get done putting turbos on everything I am going to look into doing a software driven testbech with DACs & ADCs. At the rate I fab turbo setups, I should have the programmable bench done in about 3 years.

You'll love the bench once you have it running. It sheds light on so many unknowns. All for infinite MPG.

89 Iroc Z

I am confused, I am coming up with different numbers for the Hz, and this is how I worked through it.

- The distributor rotates at 1/2 the engine speed so it outputs 4 pulses per engine revolution (V8)
- Since Hz is cycles per second if the engine where rotating at 1 RPS or 60RPM, the distributor would output 4Hz
- So that’s a ratio of 4Hz per 60 RPM
- working out the ratio with 9000RPM as you listed in your calculation -> (4 x 9000) / 60 = 600Hz
- At 9000RPM as listed in your table it should be 600Hz not 1200Hz???

Another way:
- 9000RPM / 60 = 150RPS
- 150RPS x 4 pulses per revolution = 600Hz

Edit: I totally forgot how the flip flop divides the frequency in half, it just clicked, ignore this post.. opps

MonteCarSlow

Phew... that gives me some time to beat you to it then...

JP86SS

iTrader: (1)

I had dreams of that too but have to get through about 3 iterations before that will happen.

JP86SS

iTrader: (1)

Thanks, mostly because of available info is sketchy and I'm trying to do it right. Obviously that is costing more. I already have about $150 invested but should come out nice.
I'm at the point that I don't care about being secretive about the design.
If somebody is going to steal it for $, I think it will be obvious where the info came from. B & G have put all the stuff out there for the Stimulator and I have also referenced how they did things.
Hopefully this will provide feedback on what works and what doesn't. By the time several people have built some circuits the stuff will be clean enough for any DIY project.
If a commercial version comes out I doubt the cost will be as good as a DIY effort using the info here. Kinda kills the profit margin before it even happens.

JP86SS

iTrader: (1)

Thanks,
You did confirm my math for me.

89 Iroc Z

I am in the process of building my bench right now and have come up with a better way to generate the square waves for the speed and RPM input. I did some searching for some VCO (voltage controlled oscillators) and found the CMOS CD4046. The IC is a phase-locked loop with VCO. You can ignore the PLL stuff and tie it's inputs to ground and just use the VCO part of the IC.

I built a circuit with this IC that generates 0-1200Mhz (running through a FF) without the need for a Hi and Lo switch. I used 120k ohm on R1 input, R2 left open, 0.01uF between C1a & C1b, and a 10k pot inputting the voltage into VCOin. The IC generates a 50% duty cycle but I used a FF because it's CMOS and can only drive very small currents on it's outputs. You can also run the output through a TTL buffer IC or inverter IC and lower the output frequency range of the VCO by half but since I have tons of FF laying around I used that for now. I will post a schematic once I finalize the design for the RPM and VSS.

Also I wanted to confirm that the optical VSS is 2000PPM and the 4000RRM for the magnetic VSS.

JP86SS

iTrader: (1)

Sounds cool!
I'd be interested in seeing that range of adjustment. I'm having difficulty making the VSS work on mine. I think the level is not high enough
(or I burned up the chip soldering it in)
I belive you are correct on the VSS pulses, mag for sure is right at 4K and am 99% sure the optical is 2K.
I did end up using an ignition module (7 pin) from an old HEI I had. Never got to try the little circuit to provide the "test" pulse at startup.

89 Iroc Z

One more question when wiring up the RPM signal generator circuit do you connect it to the P & N terminals of the ignition module or the reference input to the ECM?

JP86SS

iTrader: (1)

Can be done either way IIRC but if you don't use the module there is a problem with the ECM "bumping" it on powerup to see a reaction from it.
There was a PNP transidtor circuit that one of theguys came up with that would give the reaction so the fault would not be set.
Using the module, I have it connected at the pickup coil connecions.
Works fine.
Don't forget to ground the module if using the 7 pin
That had me going for a while till I looked at what I was doing.

JP86SS

iTrader: (1)

Did you get this to work properly?
I'm still trying to make mine work with the magnetic input.
Works great on the optical, just has something "electrical" wrong with the signal strength I think for the mag setting.
In any case, I'd like to get the full range adjustment out of a single device.

jwscab

the magnetic pickup is a reluctor right? The optical one is most likely a square wave opto transistor in the sending unit, so it requires 5 or 12 volts on one side, and gnd on the other. so the signal goes high to low, or more or less 0 to 5 or 12 volts(depending).

the reluctor is looking for a zero crossing, that is, the signal might vary from -1 to +1 volt over each passing over the iron finger on the sensor, so the ecm looks for the '0'.

you have a couple of options. you might get away with AC coupling the signal, try something like....47uF or so...this should be non-polarized, so you can take 2 polarized caps back to back, they should go + - - +. that will drop average voltage of the waveform to allow the buffer to trigger in the ecm.

the other option is to feed the signal through an op-amp with a dc offset. the nice this about this is you can adjust the dc offset, for another level of flexibility.

JP86SS

iTrader: (1)

Absolutly right!
Don't know why that slipped my mind. Wants AC sine wave.
Thanks for the mental jog