COP Ignition Development

-
Coming right along! Can't wait for the on car tests! I have an early Hemi on a stand that would be a nice platform..... ;)
 
One of my part orders arrived, so I started to place parts on a prototype printed circuit board. The board that came with the box has plated holes on each side at 0.1" spacing. Most of my parts other than the transistors and Arduino Nano circuit are surface mount. The surface mount parts can be placed on the square plated holes.

Here is a possible layout that will be used to construct the circuit. I use CadSoft Eagle to generate the board layout.
 

Attachments

  • COPbrd.jpg
    28.4 KB · Views: 371
Great work. Do you have the tools to install the SMT components? I have never done so, but people who do it say for small volume or rework, they use a tiny solder paste applicator to extrude a dab on each pad, viewed under a magnifier. They place the chip as best they can, heat all the pads until they melt, then surface tension centers the chip in the pads. Much harder than thru-hole components, but when you reach high volumes it is all automated and surface mount becomes cheaper. Many chips today only come in SMT versions. At least you don't have any ball-grid array chips. Those must be extremely hard to install by hand.
 
I use a temperature controlled iron and non-magnetic stainless steel tweezers. Using steel is a problem, the parts stick due to residual magnetism. I take care to keep the part aligned, and flush to the board for the first solder joint. For small parts like resistors and capacitors the iron can be quickly moved from end to end, to get the float for de-stress. Parts are soldered quickly, a second or two at most. It would take pages to explain, I am not an expert either.
View attachment 20160202_125943.jpg

A fine solder wick is handy to remove excess solder from QFP, and MSOP fine pitch leaded parts. I do not use QFN or BGA type parts.

I hired an automotive contract manufacturer to build my ECU 12 years ago. Here is the company: http://www.pgftech.com/ The reliability has been 100%. If the COP ignition goes into production, it should be contract manufactured.

My prototypes serve only for initial test evaluation, and learn a few things, prior to the PCB development. I do not like to waste money, so I achieve to get things correct the first time.
 
You are certainly above and beyond abilities I have. This is interesting, please keep posting progress. Just got the last of my teeth pulled, so not doin' much this week
 
BillGrissom said:
There was an almost-product, termed "Rabid Gator" that was a "spark control" box for use w/ a GM 8-pin HEI module, but the guy got a better day job and no time to see it thru. I recall TrailBeast has been pinging you for a similar box, but this approach moves it up a big notch.
Click to expand...

I remember this too! too bad, looked kinda cool
https://www.yumpu.com/en/document/view/9763808/gm-hei-timingcomputerinstructionmanual-rabid-gator-


I was able to SMT solder chips with the flood method: Apply thin flux and lay very thin solder wire on top of pins, use iron to melt solder in a fluid motion across the pins. with luck, the flux keeps the solder on the pins with no bridges, if you get a bridge, just use solder wick to absorb the bridge. Its suprisingly efficient.
 
I learn by doing, made a few minor changes in the build. I discovered the MAP sensor purchased a few years ago was in a different package. It is a 2.5 bar so will work for turbos too. I was able to fit it in, but using nylon screws to avoid high voltage, next to transistor.

Next week I hope to be bench testing with 8 plugs. I have pick up my toys by Saturday, expecting visitors with numerous children.
 
Kit, it is always a pleasure to stumble upon one of your threads. Always interesting, useful and excellently engineered. I look forward to following the evolution of this system. Thanks.
 
If Kit can supply the SMT boards with pre-tinned soldered pads (usually an option when you order the bare boards) you can literally flux and lay the SMT chips on the pads and pop the entire thing into a toaster oven and watch the chips solder themselves to the board, aligning themselves and everythimg! Just watch the time, as soon as you see the solder start to grab, cut the power then crack the door and let it cool down over about 30 minutes. That or the old heat gun.......works too but is harder over a large board. Kit Ill PM you
 
The wires were added to the board and test fit in housing. If this becomes a reality, a sealed panel mount PCB connector will eliminate all the internal wiring. Building the circuit has provided good information on how to improve the geometry. If the width is increased slightly, the coil drivers could be placed four on each side. Depending on how much room the connector takes, the enclosure length could be reduced.
View attachment 20160218_101529.jpg

At this time the enclosure is just for the prototype. Building it was not easy, so not planning a second at this time. The initial test of applying 14V, for test of communications, and sensor inputs went well. I have tested over an input range from 7V to 18V. The unit switches current to ground to energize external devices such as coils and relay that are powered by an external feed. The current supply to the unit measured less than 0.05A. The coil and relay supply voltage is external to the unit.
View attachment 20160218_150448.jpg

If any metal working folks have a simple chemical way to turn polished aluminum anodize, dull black, please advise. I have thought of shooting with flat black paint then wiping of the tips of fins, but not sure how that might look involving the corners. Also not sure how well paint will stick.

I have installed a 2x4 Molex MX150 sealed connector on the coil drive wires, but my crimps are not perfect. I have place an order for a correct MX150 style crimp tool. I was hoping to use one of my many crimp tools, but that did not work well enough to continue.
I ordered a Molex 63811-1000 service grade crimper, I wanted the 63811-6000, but it is $300.

Next is to build a coil cable test jig with signal taps, for bench testing using spark plugs. If that goes well the next step is making a harness for on car tests.
 
You can buy a product called 'alumiblack' or 'aluminum black' made by Birchwood Casey. It's typically only used to touch up small areas and will likely not turn out very even on a larger part. Careful application is necessary to keep from simply making a mess. The resulting finish is OK for durability. Probably as good as spray paint, but doesn't look like paint. Fun fact: also works on most stainless to turn it black (ish).
 
Phreakish, thanks for info.

mguner, while a hemi would be nice, my 38 year relationship with my 273 Barracuda, might win the first test. The system should work with 4, 6 or 8 cylinder engines regardless of distributor rotation. For dual plug engines, two units could be used.

The crimp tool arrived. When I opened the package I was somewhat disappointed, but using the tool changed my mind. It has 13 different dies, one for the conductor, the other for the insulation grip. I found it is possible to start with a larger die, then move down to get the perfect crimp. Once I gain experience, I will know what die to use for specific terminal and wire sizes. It is a light duty crimper, for small terminal connectors.
View attachment 20160218_183313.jpg
 

Attachments

  • 20160220_101929.jpg
    69.4 KB · Views: 237
It rained yesterday so time was available to finish the system wiring connections for bench testing with power supply, plugs, drill turning distributor and mity-vac on MAP port.

The setup works, except for communication with PC is disturbed by the open plug sparks related EMI. There are a few things that can be done to reduce the EMI, they include using resistor plugs, making short plug wires (using full length cables), and making plug covers from copper screen. It is the PC side with the problem, so a USB cable with EMI filters may help too.

View attachment 20160304_135732.jpg
 
I changed the plugs from very old N14Y, pulled out of my Barracuda, to NGK BPR5ES, a similar plug with internal resistor. That did the trick without other changes, now the communications works without interruptions. I will try to capture a video later, of the plugs sparking, and user interface reporting RPM, MAP and ignition timing.
 
Awesome! I guess the resistor plugs are cheaper than a Faraday cage! :)
 
Kit, I think you contacted me about some heat sinks a while back, are you still interested?
 
Resistor plugs go with modern electronic ignitions. The internal resistance is about 5.6K Ohms. I plan on using them with short spiral core plug plug wires. The resistance not only helps reduce EMI, it also reduces the current, prolonging spark duration. The spark discharge is less singular, and appears as finer multiple discharges between electrodes.

I think the old used plugs were contaminated. Even though they were properly gapped, the spark quality varied by plug. The sparks were taking weird paths. I have four new resistor plugs bought by mistake, and used those to replace the problem plugs. The communications problem went away, so I went out and purchased new plugs. The sparks look good, and are uniform for all plugs.
 
You must log in or register to reply here.
-

Back
Top