Page 2 of 3

Re: Project Frankenstein - full ECU shield

Posted: Fri Mar 14, 2014 5:09 pm
by skeeters_keeper
Have you considered selling these as a "kit"?

Presumably if you order the parts in quantity it will be cheaper than everyone buying them individually. It would make it a heckuva lot easier for someone who wanted to try it too... Perhaps recommend that they buy an assortment of 0805 resistors/caps cheaply on e-bay, and you could provide the other pieces and the board for them to solder up themselves.

Just a thought!

Re: Project Frankenstein - full ECU shield

Posted: Fri Mar 14, 2014 5:11 pm
by AndreyB
This would also be a lot of manual packaging-labeling on my side, that would be CRAZY inefficient.

Re: Project Frankenstein - full ECU shield

Posted: Fri Mar 14, 2014 10:19 pm
by kb1gtt
Often with digikey or who ever, you can get them to do the "kitting" then you simply drop in the PCB with a package from them, and your done. However that's typically kind of expensive.

Re: Project Frankenstein - full ECU shield

Posted: Sun Mar 16, 2014 6:49 pm
by AndreyB
Ops, another minor issue - the diode part number is wrong, The BAR43FILM diode is a single diode with hi-side pin not connected. I believe the correct diode is BAR43ASFILM.

Re: Project Frankenstein - full ECU shield

Posted: Sun Mar 16, 2014 10:37 pm
by kb1gtt
The good news is that the incorrect diode does offer protection above 5V, which is what most people need and want. The bad news is that the desired diode which provides some below 0V protection is made from un-obtainium. A different diode will need to be selected to get both the over voltage and under voltage protection.

Re: Project Frankenstein - full ECU shield

Posted: Mon Mar 17, 2014 12:01 am
by AndreyB
BOM updated - new part # is 497-2516-1-ND

Re: Project Frankenstein - full ECU shield

Posted: Mon Mar 31, 2014 7:37 pm
by AndreyB
Q: what are these red, yellow and green jumper wires? do I need them?

A: red jumper wire is +5 going to 'VP' pin which is shared by all hi-side drivers. By running a +5 red wire you make these channels output 5 volts. I believe you can also run it to +12 if you need to, and you can ground it and convert these drivers into low-side.

Yellow is primary trigger channel and green is secondary trigger channel, these are here to let you choose between using the VR chip or op-amps for your input. If you have the VR cheap, use it. If you are too cheap and you know that op-amp would work for you, use the op-amp channels 11 and 12.

Re: Project Frankenstein - full ECU shield

Posted: Sun Apr 27, 2014 10:35 pm
by 20div0
First of all, congratulation that the kick starter is funded. I bought the Frankenstein PCB through it. Just started trying to order parts from Digikey.
There is a few parts that are back ordered but this part TC4427AVOA-ND (1.5A dual driver) seems to have a high minimum order (>600) to even put in an order.
A quick search shows that Mouser is the same deal. Any replacement for that for small quantity order?

Re: Project Frankenstein - full ECU shield

Posted: Sun Apr 27, 2014 10:44 pm
by AndreyB
20div0 wrote:There is a few parts that are back ordered but this part TC4427AVOA-ND (1.5A dual driver) seems to have a high minimum order (>600) to even put in an order. A quick search shows that Mouser is the same deal.
True, I have never seen this part available for small orders @ Digikey. I did have better success with it on Mouser, looks like you can order it now but it's backordered so there would be a wait. Another option would be trying to go with TC4426 - that's an inverting version of the same chip, we have support for inverted mode in the software so as long as you do not forget to apply the inversion while configuring that should work.

Another components which you have to source separately is the micro-sd card - somewhere here there is a link to the eBay. The BOM is know to be good enough, it is maybe lacking with the connectors but that's because everyone would have his own preference for the connectors - or total lack of them (soldering wires right to Frankenstein)

Thank you a lot for supporting the KS!

Re: Project Frankenstein - full ECU shield

Posted: Sun Apr 27, 2014 11:03 pm
by 20div0
So, a little more research... the closest part to it is the TC4427AEOA-ND which has a lower temperature range on the high side (85C vs 125C). Is 85 C good enough? Can I put a heat sink on it? The inverted version also has minimum order of 600 in digikey.

Re: Project Frankenstein - full ECU shield

Posted: Sun Apr 27, 2014 11:40 pm
by AndreyB
I guess we do not have much choice but to go with +85. Looks like the ones I have are the same temp rating.

Until someone tries that in Arizona in summer we simply do not know. Also depends on actual ignition module current, which might be lower than 1.5A

Re: Project Frankenstein - full ECU shield

Posted: Mon Apr 28, 2014 12:41 am
by kb1gtt
The lower temperature rating should be fine within some specific parameters. The internal resistance of this chip will generate a certain number of watts per P=(I^2)R. Then the data sheet's specified thermal resistance will tell you the amount of current that can be passed via this chip before the chip hits that max temperature spec. If we believe the datasheet thermal resistance, the 125C chip was limited to .125A. The lower temperature chip would have a lower max current limit. However if you add a heat sink, you can decrease the thermal resistance by an unknown amount, which then increases the current you can pass before you hit the temperature limits of the chip. Do you plan for more than .125 amps of drive? If so what do you plan for a max ambient temperature?

I tried to document the general environment design parameters here http://rusefi.com/wiki/index.php?title=Manual:Hardware#General_suggested_environment where I put the design max ambient at 50C. So I originally planned for a 75C, and this lower temperature chip would drop that to 35C, so I would expect you will either need a heat sink, or you'll need to draw significantly less than .125A.

Do you have a general system schematic you can share? If we review it we might be able to better help guide your choices in components. If you need some kind of external component, might as well get it all in one order if possible. I suspect you are using the hi/lo for ignition. If so, you are probably drawing less than .125A any how so you are probably just fine with the lower temperature spec and no heat sink. But I would need to know more before I can really predict if that will be OK or not.

Re: Project Frankenstein - full ECU shield

Posted: Mon Apr 28, 2014 1:56 am
by 20div0
I don't have any design yet. I wanted to support the project on kick starter and just getting started. I want to order the STM32F DISCOVERY board to play with, so might as well order all the parts. I will probably order the lower temp part for now since they are just a few bucks. Thanks.

Re: Project Frankenstein - full ECU shield

Posted: Mon Apr 28, 2014 11:23 pm
by kb1gtt
Sounds good. There isn't any real issue with the lower temp part, you might want to get a small heat sink if you can, such that when and if you use it, you have the option of pushing more current through it. Let me know if you have any questions. I'll see what I can do. Just FYI, I work the 40 and have many typically life constraints. So I might take a day or two to reply, and I might need an occasional bump if it gets lost in the weeds.

Re: Project Frankenstein - full ECU shield

Posted: Tue Apr 29, 2014 12:08 pm
by 20div0
Thanks a lot. I probably won't be controlling injectors for a while. My plan is to first read the RPM and control the solenoids in my SR20VE and also try to read the AC pressure sensor and turn the compressor off when it's too high. I will probably start a new thread when I get to it so I don't pollute this thread two much.

Re: Project Frankenstein - full ECU shield

Posted: Tue Sep 09, 2014 9:19 am
by Poseidon
kb1gtt wrote:So I originally planned for a 75C, and this lower temperature chip would drop that to 35C, so I would expect you will either need a heat sink, or you'll need to draw significantly less than .125A.
Just lerking here and am curious if you think this will affect me KB? Summer can easy hit 44C and then put that ecu in the car and skies the limit eh. Cheers M8.....
Jay

Re: Project Frankenstein - full ECU shield

Posted: Tue Sep 09, 2014 10:18 am
by kb1gtt
At an ambient of 44C and high RPM, this would be in the realm of unknown. I believe you will be ok but you are getting close to the edge of my design specs so I'm not quite sure. I would suggest aiming for significant heat sinkingand try to place in a cool area with air flow. I have noticed it's common for OEMs to put the ECU heatsink in the intake air filter. I'm not a fan but though I would mention as it was not something I would have though about before seeing it. If worst comes to worst we can add a fan to the heat sink.

Re: Project Frankenstein - full ECU shield

Posted: Tue Sep 09, 2014 12:58 pm
by Poseidon
kb1gtt wrote:At an ambient of 44C and high RPM, this would be in the realm of unknown. If worst comes to worst we can add a fan to the heat sink.
I was planing that to preempt and add reliability in Australian conditions these things will probably be needed anyway...

Re: Project Frankenstein - full ECU shield

Posted: Tue Sep 09, 2014 4:51 pm
by kb1gtt
I posted the design specs here http://rusefi.com/wiki/index.php?title=Manual:Hardware#General_suggested_environment My goal is for 50C ambient, so you should be good, but at 44C, you are getting closer to my design goals and we know if my predictions are correct tor not right now. Perhaps it's worth while to check the other goals and see if anything else jumps out at you. Frankenso has several temperature sensors, which plan to use for verification of the ECU's thermal design, and long term safeties.

Re: Project Frankenstein - full ECU shield

Posted: Wed Sep 17, 2014 7:23 pm
by LS1MonteSS
Hello All,

I just finished reading through this thread and all I can say is Holy Smokes!!

I really loved seeing this project develope and move along. I was wondering how this would work for a V8 engine. I'm assuming the engine it was tested on was a V6.
And are there any final cost projections. I intend to do the DIY thing with this project. And was trying to get a ball park estimate on cost.

I'll be following this clostly..

Congrats on the progress!

Re: Project Frankenstein - full ECU shield

Posted: Wed Sep 17, 2014 7:28 pm
by AndreyB
LS1MonteSS wrote:I was wondering how this would work for a V8 engine. I'm assuming the engine it was tested on was a V6.
And are there any final cost projections. I intend to do the DIY thing with this project. And was trying to get a ball park estimate on cost.
thank you for the warm words :)
So far all the engines to run rusEfi were inline four but v8 should not be any different. We have 12+6 channels on current boards and we are going to add a couple on the next revision, just in case.

As for the costs, please be more specific on what are you trying to estimate :) http://rusefi.com/wiki/index.php?title=Hardware:For_Sale should give you the idea in terms of hardware costs.

Re: Project Frankenstein - full ECU shield

Posted: Fri Nov 28, 2014 1:45 pm
by puff
mmkaay
Frankenstein_v011_pinout.jpg
Frankenstein_v011_pinout.jpg (354.24 KiB) Viewed 28239 times
which of these holes can be used as mounting holes and can be shortened to ground (to the metal case)?

Re: Project Frankenstein - full ECU shield

Posted: Fri Nov 28, 2014 1:47 pm
by AndreyB
I want to say all of them, all these holes were created equal.

Want for bolt head clearance with the soldered components though.

Re: Project Frankenstein - full ECU shield

Posted: Fri Nov 28, 2014 10:19 pm
by kb1gtt
Yes all can be used as long as the head does not touch an exposed pad

Re: Project Frankenstein - full ECU shield

Posted: Fri Nov 28, 2014 11:01 pm
by puff
the reason I'm asking is some of them are metallized, some are not - a bit confusing..

Re: Project Frankenstein - full ECU shield

Posted: Sat Nov 29, 2014 12:46 am
by kb1gtt
I was thinking they would all be metallized if they are not I should check the source it shouldn't matter if they have the metal centers or not. They are not grounded so if you have the metal centers it should not conduct to anything

Re: Frankenstein - cheaper ECU shield

Posted: Wed Dec 21, 2016 2:26 pm
by puff
few questions. soldering the wires from the board to the connectors would be rather tough - the wires would go right through connector holes in the enclosure. eventually, these wires (even from injector drivers) won't be thick.
1. on the injector driver part - there are four ignd pads. should there be four wires? should I connect all of them to the enclosure? or to a gnd pin of the connector? or to four pins of the connector and then right to the battery? but then, they would form loops, won't they? which is the proper way? and once again, should I run 8 negative wires to ECU and eight separate positive wires to main relay, or could that be one thick wire with taps to each injector? does it matter?
2. there's also some LED next to INJ01 output (D414, in series with D413?) is it just an indicator that INJ01 operates properly? not neccesary to populate?
3. there is a placeholder for some idc10 connector (probably) - cs4, cs3, cs2, cs1, mosi, miso, clk - what are these for? should I populate it?
3. which is the most suitable pin to connect ambient air thermal sensor on this board? should I connect it to an opamp? or directly?
4. which is the pads to connect the idle driver (step-dir)
5. if I have a separate VR board, which would be located in the same enclosure, is there a special pad on frankenstein to connect its output?
6. if I build an inductive or capacitance ignition sensor (to high-voltage wires, to see the current setup with dizzy and vacuum advance), I would also need a similar pad (russian mentioned, from the firmware standpoint it would be similar to the vvt angle sensor).

Re: Frankenstein - cheaper ECU shield

Posted: Wed Dec 21, 2016 5:14 pm
by kb1gtt
1. all iGND's would be best with their own wires going directly back to the battery negative terminal. This helps keep voltage drop and such RF kinds of issues off the MCU ground. You can combine them into one fat wire at the ECU but keep in mind that you may cause a GND issue.
2. Yes LED's are to verify that driver is working, as opposed to your wires. AKA failing injector with blinking light means wire or injector problem. However if no blinking light, then it means ECU problem. These are not strictly required, they are just helpful diagnostics tools.
3. I'll need to review this this evening. If I forget, please ping me and remind me.
4. see 3.
5. see 3.
6. I do not understand the question.

Re: Frankenstein - cheaper ECU shield

Posted: Wed Dec 21, 2016 5:45 pm
by puff
6. There's a risk of going offtopic... but the plan is to install a sensor on a high voltage wire of the first cylinder that would allow to get the current ignition advance map (angle vs rpm vs engine load) just to assess the efficiency of vacuum advance and to get the baseline table. And only then make an ignition swap.
So I would need to connect that sensor to some pad on the Frankenstein board.

Re: Frankenstein - cheaper ECU shield

Posted: Thu Dec 22, 2016 8:07 pm
by puff
how many pins from the discovery board do I need to run pololu stepper driver? step-dir-enable? pe13 and pe10 have seem to have almost no soldarable pads on the frankenstein?
to drive my ignition modules I don't need u601-603 drivers, but probably need four pads either directly to the modules, or may be to my level-shifter board (i haven't ruled out those issues when disco was shutting down during ignition tests, it might be the pins were not 'strong' enough for that module in noisy environment?)