DFH - Defacto FreeEMS Hardware in KICAD

[jharvey]

About the grounded FET, damn. I had changed it be it didn't save. Sorry about that and good catch. We've had a bunch of storms come through lately, from time to time KICAD has had some trouble saving, and I don't understand why. I think it might be related to the power glitch's.

About the switch, I put it in the schematic as a switch, but I was planning on three pads / holes that can be solder jumped or wire jumped. So not a real switch, just three holes / pads and a small amount of PCB real estate.

[Fred]

So not a real switch, just three holes / pads and a small amount of PCB real estate.

Good idea, if you make them pads with holes, but make them square and close together you are golden for whatever.

Fred.

[davebmw]

In work i have designed many PCB's for Microwave applications (3G cell phones) at these frequencies the signal needs piping to its destination and the clever placement of caps simply does not work.

Nice, I'd bet you get into microstripping and such on a regular basis. I've got a little experience up in that end of the world, I've used GUNN diodes for 10GHz comm, and I've helped a guy up the road with the backhaul on a simulcast repeater network. We tuned the microwave backhaul with little flakes. Lots of fun. Any how it's lower frequency here, but you might find this project interesting.

http://www.scottyspectrumanalyzer.com/msaslim.html

Layouts were a bit easier in the early days of the one military product i oversee because we were using alumina substrate. When they decided to do a cost down version, all hell broke loose as we had to go back and rethink the whole layout, every bend put mitres here and caps there all because they wanted to change to bog standard FR4! the reflections you get are astounding! bloody skinflints!

10GHz is proper point to point line of sight stuff, My work projects range from 1Mhz to 60 MHz, 900MHz to 2.4GHz, and 2.4GHz to 6GHz. Thats as much as I am allowed to say.

That looks like an awesome project, delving right into the black art there!

Anyway back to the task at hand eh?

[Fred]

Indeed :-)

Back to those 70 Ohm resistors for a second :

3 watts each * 6 = 18 watts (call it 20)

Add staged injectors on another board in the same case and you have 40 watts and enough draw to drain the battery fairly quickly if you leave the key on and fry the dash board while you are at it. We simply can't do this realistically and nor do we need to. In fact, Flacid could be onto something with the consistency thing. If you just switch on the injector the opening rate is predictable and can be tuned for easily (defaults will probably do well enough) but if you have it priming like that, I wonder what the switching time curve looks like? Either way, we just can't have it like that, it simply won't work.

Another thing I just thought of :

With the current sensing, what are you hoping to diagnose?

If an injector physically dies but is electrically still sound it will appear identically to the circuit.
If you just want to check that an injector is hooked up and functioning an LED will tell you that it is being switched and the fact the engine isn't running on N-1 cylinders will tell you that the injector is both firing and in good shape. You can also unplug them one at a time to see if it changes the operation of the engine. Lastly, you can feel them switch physically and hear them too. It's really easy to diagnose injectors blind IMO. I don't see what the feedback will actually gain you. In the event that one dies while you are driving it you WILL know about it anyway and there is nothing the EMS can do save cutting power to all of them (which you almost certainly don't want, think on motorway, or pulling out in front of oncoming traffic etc).

I don't see any reason for having that at all really. Can you illustrate the situation/problem you hope to solve with this feature?

Fred.

[davebmw]

Some observations:

Power dissipation constantly in the injector and the resistor = waste heat, waste battery power = waste fuel.

warmer injector = increased resistance in the winding + Reduced intake charge cooling.

Switch on times may increase by a fraction of a Pico-second but you will loose that gain with warmer fuel.

Closing times will increase due to the slower decay of magnetic flux in the injector.

the current feedback is a great idea for COP ignition for detecting which cylinder you have a misfire in, but pretty pointless on injectors. this can be more elegantly achieved by spark sensing on the HT ground return, this can be used to detect differing levels of ionisation in the cylinders. hence consistent problem with injectors consistent misfiring.
Buy which time i have no doubt, we would all be under the bonnet with a laptop and DMM anyway, given 5 minutes we would have the problem by the bollocks!

Extra components = bigger board = more money = more to go wrong = more to debug = more headaches.

Can we think about using OEM auto-protected FET's that the best OEM's have been using for the last 20 years.

1 part 3 pads, thick tracks, minimal effort maximum gain, reliable as the British weather is shit!

[Fred]

Can we think about using OEM auto-protected FET's that the best OEM's have been using for the last 20 years.

Did they exist 20 years ago? I'm not so sure. Still I agree that a single part is a good idea, however I see no harm in a higher value resister and Jared's simulations show improvements in noise output with one installed. Ditto the cap. Assuming they still function at higher resistance values.

1 part 3 pads, thick tracks, minimal effort maximum gain, reliable as the British weather is [shizzle]!

Yep, sounds very good doesn't it.

1 part 3 pads, thick tracks, minimal effort maximum gain, reliable as the British weather is [shizzle]!

What are you talking about?? It's a BRILLIANT day down here in Northampton Shire! (tongue firmly in cheek)

Fred.

ps, so brilliant that the diyefi test mule got a bit sideways :-) I couldn't resist, especially given that I'm not insured for it and I've been here over 12 months so my license isn't even valid anymore LOL. Naughty lad.

[davebmw]

Did they exist 20 years ago?
MOSFET's no but Bosch used protected Darlingtons in the Motronic 1.3 i believe, principle is the same: its a switch that has lasted the test of time.

It's a BRILLIANT day down here in Northampton Shire! (tongue firmly in cheek)
Oh I see a certain Kiwi has been reading that book how to win friends and influence people


ps, He! He! He!

[Fred]

Don't worry, you POHMs don't even know what rain is. Trust me. When all that rain hit last summer and I was watching the news I was in stitches at the cops carrying people over about 6" of water and telling tales of how bad it was. The rain wasn't even heavy or sustained, the country is just so damn flat ;-) Anyway, point is, it rains way more at home, but never snows and doesn't get as cold.

Fred.

ps, fishtail on the handbrake in 2nd gear ;-)

[jharvey]

Here's freeEMS 1.0 A.09

http://sourceforge.net/project/showfile ... _id=618983

I've included the release notes with it, mostly fixed some things here and there. I've also added a TODO list. I know some folks have commented on some topics. I may have missed the comment, or forgot about. I'm using the TODO list as my reminder about what to focus on. If you had a thought and it's not on the TODO list let me know.

Here's a copy of the release notes.

A.09
- added the change log and TODO list
- fixed injector drives again, this time it saved correctly
- changed 70R to 2K
- ran BOM tool to generate a BOM called freeEMS_1.lst
- semi automated template convert process, now edit the tempates, then run convert-template.sh to get the real schematics
- added footprint to many resistors
- change VR circuit some

Heres a copy of the TODO list

--- ToDo list

A.09 - future steps include
- create PCB combined modules like Caps and DPAK and TO220 combos
- inversion of signals like RPM input
- automate PDF creation (script)
- automate zip file creation (script)
- QUCS resimulate with 2k instead of 70R (can't seem to get it to work stable over 100R. I'm sure it will work stable in real life.)
- LED's to ouput drivers, perhaps RPM input signal as well
- add footprints to symbols
- correct signal paths for CPU board, (done mostly at PCB layout)
- add vendor numbers to field1 or field?
- correct "Xk"'s and "C"'s from BOM, (figure out what those values should be, and replace the Xk's and C's with the right value)

This log is a reminder for me to remember what to work on. I seem to get on something then forget the other parts, so I'll try to remember to place items to work on here, then some day they will be removed from here, and found under the changes log.

Note to self or anyone that uses windows to remove the /n/r from a windows edited file, use dos2unix inputfile outputfile. You will need to do this if you edit the custom scripts by hand.

[Fred]

Note to self or anyone that uses windows to remove the /n/r from a windows edited file, use dos2unix inputfile outputfile. You will need to do this if you edit the custom scripts by hand.

Note to self or anyone that uses windows to remove the /n/r from a windows edited file, use dos2unix inputfile outputfile. You will need to do this if you edit the custom scripts by hand with an extremely poor editor like notepad. Those using a decent editor like VIM or textpad or whatever will have no problems with this. In fact, even wordpad does it right :-)

Fixed it for ya :-)

Fred.