1/4 scale V8, first project.

[Lakc]

With coils like that you dont need to hook the scope leads up to anything, just lay them nearby and set the scope voltage gain. The early GM HEI ignition used a tin can formed to the top of the coil as an inductive pickup.
Resistor plugs are not going to help, its the coil primary kick at around 400v mixxing with the secondary kick thats putting voltage across every wire within 3 feet. You have a classic RFI problem here, time to put on your electrical engineer thinking cap.
The insulation on the plug wires looks good, grab some metal (brass preferably) screen and wrap the coils up with the plug wires sticking out the top. Ground the screen. Same with the coil drivers (CDI coils?). What your making is called a Faraday cage.
Fit all the powers and grounds to the ignition with several different sized small capacitors, 1nf, .01uf, etc. Dont bother ordering a bunch, scrap them from that old VCR your never going to use again.
Make sure your grounds all go to a good mains ground. Use a 12v lead acid battery instead of any mains transformer or voltage supply. Batteries can suck up RFI really well if given a chance.
Next, get some ferrite beads from anywhere (like aforementioned VCR) and wrap the cable to the laptop in them. You can also wrap some grounded screen around the cable as well, it might help.
If it still doesnt work, dont fret, they can be tricky but solvable, with more ingenuity then coin. Know any ham radio operators in your area? They can be a great source of help if they have ever fought the "RF in the shack" battle.

 

[keith5700]

Thanks Jeff.
I'll see if I can get the scope to pick anything up just from the lead.
The stuff I've read elsewhere had a trace of the 12volts supplying the ecu. When the spark fired there was a spike of 80volts going up it.
I don't know enough about it at the moment to work out if the spikes are being generated from RFI in the airspace, or being generated in the actual engine and then being fed back to the ecu via the earth wires.

I'm going to try caging the coil packs tonite. Cheers.

 

[doubletop]

Keith

Still tagging along in the background watching this in awe. I saw your post from yesterday and for some reason woke up this morning thinking about it and Faraday cage and good earths came to mind. Looks like Jeff has beaten me to it and some more.

As you won't be able to afford the gold sheet baking foil will do, (in fact I was told of a case of baking foil making into space....) Not on this though..

Anyway back on earth I did also think all that wiring could do with a bit of a tidy up. It could be adding to your EMC woes.

Pete

 

[propclock]

Just my 1.414 cents use and old cheap radio for your troubleshooting.
You will get instant feedback with your hands free to try new stuff.
Just use your ears.
Much easier than looking at scope traces.
Resistor plugs were only made to stop am radio interference, I think.
Just have the radio going during any testing.

Ferrites are my favorite solution for the tough ones.
But the radio has helped me many times with rf interference
even servo motor tuning. All pwm systems are broadcasters . Of course a coil
is a single pulse

For what it's worth.

Protect your electronics until you have down to a reasonable noise level.

 

[nfk]

Hi Keith,
I've 'megasquirted' an old (1980!) Honda Accord a few years ago (and still running!) and found the same spark resistor issue you'r having right now.
After months of lecture and head scratching, I end up installing an EDIS 4 ignition controled by the mega ECU.
The RFI noise went away only after using the modules. (and i've tried all)
http://www.megamanual.com/ms2/EDIS.htm
For your application you would need an EDIS 8.

One of the benefits of using it is you won't need a distributor.

I can share i few notes about the megasquirt if you need any help, i've spend quite a lot of time troubleshooting and setting it for my car.
Fuel maps could be a pain!

Cheers,
Norberto

 

[keith5700]

Ok, one last go to get a post up without it getting lost!

I've tried the foil shielding, and ferrite rings, and capacitors on the ecu +ve, and 8k resistors on the plugs so far, but nothing will make the ecu communicate with the laptop when the engine starts turning.
Those of you who have had any dealings with stand alone ecus will know how difficult it would be to program fuel and ignition tables when you can't see where on the table the engine is running!

From my research so far, and talking to the seller, it seems the interference is ground based, ie feedback into the ecu, from the engine to ecu earth wires. ?? I've no idea.

Anyway, I got fed up with that and so put some best guess maps in for ignition and fuelling, and after an hour it started.
At the moment it runs at around 2000 rpm, regardless of where the throttle is, so it's obviously still a mile out.

On the plus side I now have a working electronic fuel injected V8. On the minus side, I wish I'd never bought the Megasquirt ecu.

Norberto, thanks for your post, I'll look into the Edis module. It would be great if this resolved the issues, although I'm still a bit skeptical that anything is ever going to cure it.
The ecu will control the engine, but it won't show any live info on the screen, which, really, is the whole point of it.
Then, to change any settings I have to reboot the ecu and restart the tuner program on the laptop.
This starts to wear a bit thin after a while.

A few pics, notice the ptfe spark plug doesn't look like it's going to last very long.

 

[Lakc]

Well, progress is progress, you have at least proved your concept of getting an electronic fuel injected motor running.
There are nearly an infinite number of ways to tackle RFI problems, and most of the solutions are very dependent on the actual situation. I once used a ground rod in a well watered potted plant, so its difficult to tell much beyond the basics without "being there".
Take a break and celebrate your success!

 

[propclock]

Those sure look like inductors and not resistors to me. I am probably wrong though.

 

[doubletop]

Ok, one last go to get a post up without it getting lost!

......................., and talking to the seller, it seems the interference is ground based, ie feedback into the ecu, from the engine to ecu earth wires. ??

Keith

I think we are all suggesting a similar thing here. Get everything that should be at ground 0V well strapped down to one common point, including the foil. Don't daisy chain things together and use decent sized cables and make sure of good connections. When you've got say 25Kv in the system even a few milli ohms of resistance in a connection can have a good few volts across it when most of your PC components are running on just 3v.

You'll get there

Pete

 

[doubletop]

Well there's me being a smart arse and then I looked at your post again

Pete

 

[Lakc]

Those sure look like inductors and not resistors to me. I am probably wrong though.

On the sparkplugs? They seem to be the old 1/4 watt resistors, although they may be wire wound type and still have some inductance. Not something you want to put on a tuned circuit in a sensitive radio receiver but for this purpose probably not an issue.
The two round whiteish components taped together appear to be coils or capacitors for a CDI type ignition?

 

[keith5700]

The spark plug resistors are ordinary carbon 8k type. The 2x capacitors are connected between the live feed to the coil packs and earth. The Megasquirt manual recommends putting them on, as per full sized practice, although I'm wondering if they should be electrolytic type.
They are non polarised, 25uF 400 volt type off a motor.
The resistors made no difference so they are now off.
The caps make no difference whether they're on or not.
I've seen an edis 8 for £60 so may try that, but haven't got money at the moment, so I'm just making a bench for the Rivett lathe.
Cheers.

 

[keith5700]

Well, I finally got it to work, by using a different laptop. I've just put a video up on youtube so hopefully this link will work -

[ame="http://www.youtube.com/watch?v=7EGXo5HwIHk&feature=youtube_gdata_player"]http://www.youtube.com/watch?v=7EGXo5HwIHk&feature=youtube_gdata_player[/ame]

 

[John Rus]

Beautiful engine! Glad you were finally able to make it run properly.

So what is your next project, the throttle I guess for now?

Cheers,
John.

 

[dsage]

Very nice. Sounds really good. (although I'm partial to the un-muffled sound myself)
SO get us up to speed here Keith.
What about the smoke, carburetor, plugs, burned head gasket (or whatever), ignition coil, points vs electronic, fuel injection etc. etc. etc. that you have been writing about.
Seems you jumped a few chapters somewhere. Maybe I missed something.
Inquiring minds need to know how all that was solved.

Good work.

Sage

 

[LongRat]

How stable is that idle?! That machine is butter smooth, it runs amazingly Keith! Plus I got the first comment on your Youtube vid!
Now I'd also love to hear the technical parts about how you solved your problems.

 

[keith5700]

Hi all, there's not much missed out, I gave up on it for a while after the Megasquirt problems, and did something else for a while.
Last week I got my old laptop out and tried that on the Megasquirt, and amazingly it worked, ie. no lost communication between Msquirt and laptop.

So then I just carried on from where I left off, put some approx fuel and timing maps in and fired it up.

It still puts a lot of smoke out, but that's only going to get fixed with different pistons and rings.

Oh one thing I changed was to control it by throttle position sensor instead of manifold pressure sensor. There's nowhere near enough manifold vacuum to use the map sensor option. There is a small trim pot on the throttle shaft acting as the throttle pos sensor.

I've learnt a lot along the way, and when I've got a bit more time I'll put all my thoughts down on here, like what worked ok and what I would change next time around.

I'm starting the V10 at the moment. Just roughing out the pistons first. I'll start another thread for that when there's something to show.

I need to do some fine tuning on the v8 next, but I need a fuel cooler first. The engine dies after a few mins as the fuel heats up from the pump, and starts to aerate and thus drops the pressure.
Cheers.

 

[FannBlade]

WOW! i can't stop watching the video. Sharing it with everyone I know.

Keep up the great work!!!!!

 

[Art K]

Keith,
Longrat wasn't kidding when he said it ran butter smooth. This is an amazing accomplishment. Stock cars around here used to have the fuel line run through a can filled with ice, good luck with the fuel cooler.
Art

 

[doubletop]

Keith

What else can I say? - Outstanding!!!

Pete