Project JAGuar for FreeEMS

[DeuceEFI]

I know the schematic doesn't reflect Fred's suggestions, I will update the files tonight along with other suggestions...

But it is a start...

On a side note my first PCB was a bad photo-etch, so it was karma preventing me from messing something else up... LOL

Andy

[Fred]

Awesome!!! OK, firstly, I meant no offense to Sean either, his code IS going to be the mainstay of the future of FreeEMS, unless he dies or pulls out or something. The thing is, that there is a LOT of other work that needs doing and if I spend time on that, the really important stuff suffers. Hence I'm being hard nosed about it.

Right, you'll be happy with semi sequential fueling, I'm sure. I ran proper single shot batch on my 4 cyl truck with big injectors and it idled pretty smoothly.

If you're only running one position/rpm sensor, you could hack the code (very easy to do) and run port T to another fuel driver or ign driver. This is only useful if your sensor is on the cam, though, as otherwise you can't make any more use of it anyway.

Mimicking puma/dfh is dangerous as both have lots of issues. A HUGE amount of time (from at least 7 people) has gone into making Ravage schems near perfect. If you want a reference of what is right, go there. It's a better source than the top of my head right now, pretty much.

A7 is the correct port for the fuel pump, however it's easy to change it to any other port if that was wrong in your spin 1.

Re the caps, go 35V tant on the inputs to the reg(s) and 25v (or even less) tant on the outputs of the reg. Electros never last, esp not with any heat. The huge caps on the DFH design are my fault for asking for them. Puma copied that. My mistake, my cost correcting it now :-)

The main thing the 144 pin chip gives you is 8 more ADC pins, so it's not a huge loss to not be using it. Otherwise, the software won't even know which it's on.

T0, T1 input connections depend on the decoder in use. Both should have the same circuit, and I'm hoping you went with A2 mode MAX992X chips?

O2 and TPS are currently just for logging, same for MAT, IAT is used for speed density. Later I'll move it to a RAT var (Real Air Temp) and provide some transfer functions to move from IAT, MAT, Ambient, time after start, time since last running, temp at last run, etc to the RAT value. Use a GM IAT somewhere with good clean air flow away from exhaust manifolds, close to the front of the car, preferably near the intake pickup location/filter. If you have a MAT too, hook it up and let it get logged out of interest's sake. It will be interesting to me, at the least.

AAP also only logged for the time being.

O2 2 may not even have code attached, but is ULTRA simple to add, so if you really have two (why?) then that can be logged also.

The ADC pin outs are correct, I just checked.

Re following FreeEMS standards, they are under development :-) RavAGE is going to be at the very least the closest possible thing to it, however that doesn't really guide you on what you "have" to do and what you "should" do and what you "can" do (or not). I need to write that stuff up in a nice formal latex doc at some point, "soonish".

For your injectors, I'd recommend a FET each and driving two FETs with two resistors from each of the three port T pins. That should get you a decent result with some luck. You can equally drive 3 FETs and load two injectors onto each. Whatever is easier, both should work well. Roll with 1k value or maybe even lower, for your FET to CPU connections, lower if you're going to put two on one CPU pin and leave the other three T pins unloaded. The limit is for the set of 8, so this is a legitimate thing to do.

I would suggest finding out what your crank/cam patterns look like so we can make sure it's supported in time for you to run when you want to.

I'll try to find time to check out the schematics soon. I'm sure there will be things to improve, but I'm somewhat loaded up getting RavAGE 100% right at the moment. I'll suggest that Preston and Dan and others take a look and see what they can spot.

I hope this all helps :-)

Fred.

[Fred]

I know the schematic doesn't reflect Fred's suggestions, I will update the files tonight along with other suggestions... But it is a start...

I just took a quick look at the PDF and I'm not sure what you mean by that, care to explain? :-)

Fred.

[Dan]

Reviewing your schematic pdf now mate.....

[Dan]

OK so things I have found on a sheet-sheet basis when I reviewed your schematic PDF are:

Sheet 1 (Jaguar.sch)
Nothing I can visibly see, it is just a linker sheet to me anyway so I have no comment on it.

Sheet 2 (Outputs.sch)
It appears your LED resistor value is a tad low (I calculate that you are allowing 15mA through your LED?). Is your 1N4004 diode for back-emf snubbing going to be suitable enough that you wont get noise/crap/etc on your 5V rail through your LED circuit? (most LED's have low peak reverse voltage ratings). Also, out of curiosity, which FET's are you using?

Sheet 3 (Injectors.sch)
Same comments as Sheet 2 above with regards to LED peak reverse voltage levels and the value of the LED current limiting resistor.

Sheet 4 (PowerSupplies.sch)
Too much bulk electrolytic capacitance in my opinion. The LM2937 reg's have internal reverse polarity protection so no need for your 1N4004 diode's either. How are you planning on switching the '5vdc-switched' regulator on and off? is it just with a 12V switched input?

Sheet 5 (inputs.sch)
I think you analog temp input circuits are incorrect as the bias resistor may be in the wrong place along with a few other things (capacitors, etc). Can I suggest that you copy RavAGE analog input conditioning circuits? This goes for all your analogue inputs. I havent looked at the MPX4100 datasheet, but if it is similar to the 6250 and 6115 that I am using in RavAGE (which I think it is) then you should have those circuits the same as RavAGE too I think. I wont comment on your ignition tach input/timing advance output circuit at this stage, if at all.

Sheet 6 (cpu.sch)
You will probably find that you cant get a nice 16MHz crystal that requires 22pF load caps. 18pF/12pF is probably a better choice for crystal availability. I wont look at this sheet in detail yet, but I assume you based the PLL, load/run, clock and other circuits off RavAGE? If you did so, then they should be correct. I havent checked the pinout of your cpu IC though.

Sheet 7 (ft232.sch)
I think you should opto-couple your USB comms as per RavAGE also. Remove the transistor approach. Also fully isolate the USB circuit from the Jaguar circuitry (ie. grounds, etc). Check out the current Ravage.pdf in my git repo if you like.

I look forward to seeing you develop this mate, nice work!

Dan

[Fred]

OK, a few things:

2.2 k resistors for some adcs are too high, 1k max, lower is better, 470 is in ravage.

Not sure what Dan means about bias being wrong.

2.49k for bias is fine, but any value around that is good, so if there are cheaper ones with similar precision at more common value points, all the better. I think we should go to higher precision in ravage for those and the battery ref ones too. Speaking of which:

39k and 10k for battery ref is wrong, go to 3.9k and 1k for that, and make them high precision, 0.1% or something, or tune the software values after you set it up. 1% minimum, though. If you don't have a resistor to the ADC pin from the divider centre, you should add one too.

I think the map sensor circuit looks OK.

This is your design choice, but you might want to add schottkys to the rails for the ADC inputs.

PLL values don't look like a tuned set? I have some posts and/or a thread on this somewhere.

CPU caps should be 0.22uF not 0.1uF and it's wise to parallel some 10uF tants with some of those depending on pin loading too. Though that is optional and we don't have it yet, and nor do various other boards.

Copy the load run stuff from ravage as future SM versions are going to remove pull up from their config and require hw pull up. Future CEL is on that pin too, so optionally put an LED on it.

I didn't check the CPU config stuff, but check that it matches ravage yourself.

FTDI looks bus powered, which is good, optos are a winner idea, but not necessary unless you're going to use an inverter with your laptop.

power supply input diodes, if you choose to use some, should be schottky for minimum drop, otherwise cold cranking could be a problem.

re the power stuff, just go key switched and power both inputs from a solid relay feed. The wake up stuff is pretty complex and only needed for A) fancy hot start stuff and B) sd card logging. There is no software for it yet, so best to hold off on that. Dual regs is still nice to keep the CPU safe.

agree with dan re LEDs and resistors and reverse V

I'm also curious about your FET choice. Obviously (?) they need to be logic level at the least, and protection is nice to have. VND/VNB/VNP omnifet parts are sweet like that. Other stuff will work too, though. Normal fets require a driver to function properly.

OK, enough for now, that ought to keep you busy :-)

Fred.

[Dan]

yeah I might of had a mind 'fart' RE: 'bias resistors being incorrect'

please disregard that comment....

[DeuceEFI]

I know the schematic doesn't reflect Fred's suggestions, I will update the files tonight along with other suggestions... But it is a start...

I just took a quick look at the PDF and I'm not sure what you mean by that, care to explain?

Fred.

I was just referring to your suggestions after my initial posting of the pics with the HUGE electro's... I will remove them and use your suggestions as well as Dan's... Thank you both for getting me on the right path

Also, it appears that my basing my schematics off of PUMA and DFH was incorrect... I will take a look at the latest Ravage pdf and update my drawings accordingly.

I will digest all your suggestions and work on updating my drawings and post again once I have updated all the schematics or if I have further questions. I really appreciate all the suggestions and welcome any others that come to mind.

[Dan]

No problems at all! I hope what we have said helps!

I am keen to see your revised schematics!

[DeuceEFI]

I'm also curious about your FET choice. Obviously (?) they need to be logic level at the least, and protection is nice to have. VND/VNB/VNP omnifet parts are sweet like that. Other stuff will work too, though. Normal fets require a driver to function properly.

OK, enough for now, that ought to keep you busy

Fred.

The FETs I chose for the injectors were originally what I pulled out of the GM ECU, manufacturer unknown and I haven't been able to cross reference the markings "435M9395" on DPAK footprint. They are logic level gates from my testing and drawing out their configuration from the original PCB that I pulled with the engine. In my testing with a 1k resistor from the CPU to the Gate with a 100k resistor to ground and the Source connected to ground it only pulls 50uA

I found FQD12N20LTM 9Amp QFETs or VND5N07 5Amp OmniFETs that have the same DPAK footprint so they can be switched easily.

For the Fuel Pump relay out, I found a VNN1NV0413TR OmniFET rated at 1.7A in a SOT-223 footprint.

You and Dan have given me a lot of great information that I will incorporate in my design

I have also attached a pic of the car Jaguar is being designed for...

Andy