Project JAGuar for FreeEMS

[Fred]

Another suggestion, why don't you move the injectors south to T456 and add one on T7, then run T3 out on the XOR? That way your board could be used for wasted spark setups with sequential injection on a four cylinder as well as your intended purpose. No pressure. I just have an uncontrollable urge to generalise :-)

[DeuceEFI]

Another suggestion, why don't you move the injectors south to T456 and add one on T7, then run T3 out on the XOR? That way your board could be used for wasted spark setups with sequential injection on a four cylinder as well as your intended purpose. No pressure. I just have an uncontrollable urge to generalise

I like the suggestion and have added it to the schematics and to the BOM.

T0 is Crank RPM
T1 is Cam Position
T2 is for DIS Advance Signal (PWM 0-5vdc square wave)
T3 is now DIS Bypass voltage output
T4567 are now Injection Outputs

Looks like EDIS won't require any additional components...

I just committed/pushed my changes to github including the Jaguar-Schematic.pdf file for everyone's viewing pleasure

Andy.

[Fred]

What are "mopars" and do they really use 9.3k? Reference? These values aren't super critical, btw, standard 2.4k will work for near everything, only the tolerance matters, and the data in the device matching the combination. The exact value just shifts the region with the best accuracy up and down in the temperature range, no big deal unless you're under 30 foot of snow in northern Alaska :-)

I'm afraid that I don't understand your jumper setup for the injectors. In particular, I don't understand what P21 is for at all. The text says hook pin 1 of P21 to pin 1 of P21, which is also confusing (you're trying again, aren't you?). Can I suggest the following configuration:

Three 8 pin headers:

1. T4*2, T5*2, T6*2, T7*2
2. 8 injector FET inputs
3. PORTB

In that order, and right next to each other, then you can jumper very easily between the FETs and the CPU pins and not confuse me and the next guy :-)

Of course, it takes two more fets and some board space to do this in a clean way, but I think it's worth it. It's your project, though.

A few more things:

You don't need to pull down pins that you're not using, no problem there, so no need for the resistor array thing for port B if the FET circuits themselves have the 100k on them and/or you're using port T.

I gave you bad advice re input resistor values to FETs, change them back to 1k. 470 would be ok for 3 fets on 3 cpu pins on one port with the other 5 not being used or lightly used. If using all of them, which is now an option, it's an excessive load, and if paralleling up onto 1 pin for two, it's a 235 ohm load, which is way excessive.

Caps look good on the CPU but you left an old label about C5 on the right.

The 100k is on the wrong side of the 1k in the fuel pump drive circuit.

[second to] Final note, using port T for an on/off switch is a waste, but works fine. It doesn't make sense to add an extra digital IO pin for this when other apps will be happy about what you've done. No change required, just noting it :-)

[final] What about spare analogues for:

oil pressure, oil temp, ambient temp, radiator outlet temp, etc?
ground and power monitoring other than the standard brv?

And what about a couple of FETs on the 16 bit PWM channels for idle/boost/fan control?

[post final] I'm not going to check on this, but you should try to mimic this physical layout with your connector, ie, heavy on one side and light on the other (roughly rpm, analogue, low current ign, high current outs, power, ground, in that order), and the same on the PCB behind it: http://stuff.fredcooke.com/RavagePinOutAttempt1.png

Looking good! :-)

Fred.

[DeuceEFI]

I like the suggestions, I will work on the schematics after dinner.

[DeuceEFI]

New committed changes have been push to github.

What are "mopars" and do they really use 9.3k? Reference? These values aren't super critical, btw, standard 2.4k will work for near everything, only the tolerance matters, and the data in the device matching the combination. The exact value just shifts the region with the best accuracy up and down in the temperature range, no big deal unless you're under 30 foot of snow in northern Alaska

MOPAR is generic for Chrysler, Dodge and Plymouth automobiles here in the USA. I have noted on the schematic to use 2.4k 0.1% Metal Film resistors and use FreeTherm to adjust if necessary.

I'm afraid that I don't understand your jumper setup for the injectors. In particular, I don't understand what P21 is for at all. The text says hook pin 1 of P21 to pin 1 of P21, which is also confusing (you're trying again, aren't you?).

Not trying to confuse you, just trying to keep it on the 3" x 5" form factor so that it still fits my needs I think you might have mis-read the text, I have edited the text to make it a little clearer (hopefully)...

Can I suggest the following configuration:

Three 8 pin headers:

1. T4*2, T5*2, T6*2, T7*2
2. 8 injector FET inputs
3. PORTB

In that order, and right next to each other, then you can jumper very easily between the FETs and the CPU pins and not confuse me and the next guy

Here again not enough room on the PCB, but I did give them the option to use the 4 Port T outputs T4-7 (P20) and run wire jumpers to Injector FET inputs (P34) Pins 1-4 for the first 4 FETs... Actually now that I'm reading what I typed in the text box, I forgot to mention how to set it up for 4 cylinders... LOL...

They can also follow the new text to set it up for Semi Sequential for 6 and 8 cylinder engines by installing some jumper shorting bars between FET input pins (if they install a 0.100" pitch pin header).

Of course, it takes two more fets and some board space to do this in a clean way, but I think it's worth it. It's your project, though.

I have added to two additional FETs, so now it can support up to 8 cylinder sequential injection. I had a couple of buddies wanting to use this for their V8 Mustangs...

You don't need to pull down pins that you're not using, no problem there, so no need for the resistor array thing for port B if the FET circuits themselves have the 100k on them and/or you're using port T.

Cool, I have removed that from the BOM.

I gave you bad advice re input resistor values to FETs, change them back to 1k. 470 would be ok for 3 fets on 3 cpu pins on one port with the other 5 not being used or lightly used. If using all of them, which is now an option, it's an excessive load, and if paralleling up onto 1 pin for two, it's a 235 ohm load, which is way excessive.

No worries, I have changed them in the schematic. I'll update the BOM after sleeping...

Caps look good on the CPU but you left an old label about C5 on the right.

Whoops, took care of that this time around.

The 100k is on the wrong side of the 1k in the fuel pump drive circuit.

OK, that's fixed as well now.

[second to] Final note, using port T for an on/off switch is a waste, but works fine. It doesn't make sense to add an extra digital IO pin for this when other apps will be happy about what you've done. No change required, just noting it

OK, let me think on this one, I may have to re-arrange things to use another port on the other side of the CPU, it is kind of crowded on that side

[final] What about spare analogues for:

oil pressure, oil temp, ambient temp, radiator outlet temp, etc?
ground and power monitoring other than the standard brv?

Might have to hold off on those due to real estate, could put in pin headers to bring them off board to an expansion board as the case I have will allow for multiple PCBs, or we may need to do a full SMD board, as we are about out of room using thru-hole components...

And what about a couple of FETs on the 16 bit PWM channels for idle/boost/fan control?

Well, I didn't add it to a PWN channel, but I added a fan (or other relay driver FET) to PA5 and I added a 3 pin header for Ports PP0-2 to go to an expansion board if someone wanted to try it.

[post final] I'm not going to check on this, but you should try to mimic this physical layout with your connector, ie, heavy on one side and light on the other (roughly rpm, analogue, low current ign, high current outs, power, ground, in that order), and the same on the PCB behind it: http://stuff.fredcooke.com/RavagePinOutAttempt1.png

I agree, I'm just waiting for my free samples of the TE Connectivity Super Seal 1.0mm and the AMPSEAL connectors to arrive to see if they will even fit at the end of my aluminum case. Peter has suggested I look at another one from JAE Electronics as it is a 2 row instead of 3 or 4 row like I have been looking at, so I'll get back to you on the connector.

Looking good!

Thanks, hopefully once we run out of wish list items and are comfortable with our decisions, I'll print out a proof on paper to check component spacing before I etch a test board...

[Fred]

OK, looking at the latest, just pulled:

I understand your jumper scheme now, however you shouldn't say "for sequential" or not, you should say "to use PORT T pins with current code" and "to use PORT B pins with future XGATE code" and point out that 4 = max until then.

I just noticed that the SM stuff is slightly wrong in both this and Ravage, I'll have to think about that and provide a fix to both projects. Don't print a board until I do that :-)

If ford really uses 27k resistors in their ECUs and MOPAR really use 9.3k resistors in their ECUs then it's likely a good idea to use a closish value to these. I just find it a bit surprising as they would be fairly noise sensitive. Nippon is 2.7 AFAIK. Bosch, 2.2? GM 2.49, hence 2.4 suits all of those just fine and I recommend it. You'll need to configure the code with FreeTherm or the data with a tuning app in all cases, so you should reword that a bit.

I can't see anything else wrong right now!

Fred.

[DeuceEFI]

I understand your jumper scheme now, however you shouldn't say "for sequential" or not, you should say "to use PORT T pins with current code" and "to use PORT B pins with future XGATE code" and point out that 4 = max until then.

Ok, I have made changes to the wording to sound more like the above...

I just noticed that the SM stuff is slightly wrong in both this and Ravage, I'll have to think about that and provide a fix to both projects. Don't print a board until I do that

Ok, I'll hold off, just let Dan and I know what should be changed on the circuit

If ford really uses 27k resistors in their ECUs and MOPAR really use 9.3k resistors in their ECUs then it's likely a good idea to use a closish value to these. I just find it a bit surprising as they would be fairly noise sensitive. Nippon is 2.7 AFAIK. Bosch, 2.2? GM 2.49, hence 2.4 suits all of those just fine and I recommend it. You'll need to configure the code with FreeTherm or the data with a tuning app in all cases, so you should reword that a bit.

Ok, I also changed the wording around that area of the Input.sch file, check it out when you have time, no rush...

I can't see anything else wrong right now!

Well, that was then, but I have made a few changes, so you are free to retract that statement and it won't hurt my feelings

I have also updated my notes sections on most of the schematic sheets with things we have discussed in the forum.

[Fred]

The serial monitor change required is to add a 1k resistor from the PA6 pin to the ends of the existing two AND change the 2.4k direct pull up to 10k. This keeps the MCU safe if the switch gets moved while the firmware is running and outputting a high/low alternating signal which firmwares as of today do.

Note, TA is exempt from this as the switch has a 1k on the lower side of it and the switch is attached t the pin directly. so it needs a slightly different approach and forms a voltage divider, which is less than ideal, but works well enough.

Fred.

[Fred]

Other than the SM change, I think it looks pretty good. I'll get Preston and Dan to look it over as soon as they can. Hopefully that doesn't hold you up much, but it's pretty safe to move forward with a PCB design once the above change is done, I think. Fingers crossed that we didn't miss anything :-)

Fred.

[DeuceEFI]

Other than the SM change, I think it looks pretty good. I'll get Preston and Dan to look it over as soon as they can. Hopefully that doesn't hold you up much, but it's pretty safe to move forward with a PCB design once the above change is done, I think. Fingers crossed that we didn't miss anything

Cool, I have updated all the documents and even included some PNGs of the board from KiCAD.

BOM file is docs/Jaguar.ods

I have placed helpful hints throughout all the documents to help with the Assembly document which I will start on while awaiting others to review before going much further...

Well, I may print out the front and back of the PCB to paper to place components on to test placement