Discussion of Number 10

[Peter]

Well it's 12:19AM here, and I managed to get the new engine, and larger injectors installed in a weekend. Got out on the interstate, and they've closed the road about six miles out of town. So I'll probably be missing some class tomorrow after working really hard to get done in a hurry.

The new injectors make it idle and run fairly rough. I couldn't get Venom to tell me much of anything useful about their injectors. I'm guessing the injector dead time is quite a bit longer. Venom/Python has lost any future business from me. They absolutely refuse to send me a datasheet on their fuel injectors. The only guy I talked to that would give me any numbers said .7 to .8mSec generally works pretty well for them, but it depends a lot on the circuit you use. They're rated at 720cc/min, but that has to be the 85% rating due to how rich it is there.

[Fred]

That sucks! I guess fuck around with it and see if you can make it decent and start building a flow bench to find out what the deal really is. It would also verify that they are in good working order and evenly matched.

BTW, the dead time is dependent on the driver topology used. Oh, which reminds me, did you upgrade to darlingtons yet? If not, you could be failing to open them properly leading to really inconsistent fueling. In any case, even if you did, the darlington setup will open a fraction slower than a FET + PWM setup, and know knows what their "0.8 to 0.7" tests were done with.

There is every chance that they don't have the datasheets available too. Especially if they're just buying them from someone else and reselling. Googles...

"VENOM Performance Fuel Injectors are 100% flow tested, balanced and matched to within +/- 1.5% of their nominal pound/flow." they should provide the sheet for your particular injectors! Bastards. And this is just lol: "Solid copper terminals are used on each connector to increase current flow to the coil."

I'm tempted to phone them and abuse them on your behalf. If I do, I'll record the call. TOLL FREE 800.959.2865

Fred.

[Peter]

It's 11:45 and the road is still closed. So here I am.

That sucks! I guess fuck around with it and see if you can make it decent and start building a flow bench to find out what the deal really is. It would also verify that they are in good working order and evenly matched.

Honestly, I don't really want to take the time to build a flow bench. I've considered sending them off to be tested, but I haven't talked to anybody about it yet. I'd want them to use my circuit to do the testing so that I have the most accurate results. I found these people http://www.witchhunter.com/injectorserv1.php4, but it's a little more than I'd like to spend. So for right now I'm back to the stock injectors, because they run really nice with the default setup.

Oh, which reminds me, did you upgrade to darlingtons yet?

No the darlingtons are suppose to get here Wednesday afternoon. I can't believe I screwed that up, but that's what I get for trying to use free shit, and not fully understanding what's important. But mostly because I didn't bother asking anybody that knows what's going on.

I'm tempted to phone them and abuse them on your behalf. If I do, I'll record the call. TOLL FREE 800.959.2865

You're welcome to if it'll make you feel better. Maybe you can convince them not to run off some of the aftermarket efi community. I need to start recording my calls. The tech lines for Venom and Python are the same number. The part number that's on the box is 21304-720-6.

I think you were right about not revving this engine to 7k. There was about 2-3 tablespoons of super fine nonmagnetic metal in the oil pan. Probably babbitt or aluminum, but at any rate way more metal than I've ever seen in an oil pan of a running engine. Possibly a blessing in disguise, because I'm going to be a bit more careful with it from now on.

This is about all the higher the engine lift would go in the cold weather. I ended up overfilling it with some boiling hot hydraulic jack oil to get the valves working.


My mom took some pictures with her camera that probably costs more than my car. So I figured I should share.


My dad helping me. Probably because he doesn't want the car in his garage for a month while I screw around getting it done.

[Fred]

I bet you it will work a lot better with darlingtons in there! Likely they weren't opening properly at all. That middle picture deserves to be higher res, it looks really crisp :-)

I don't recall the comment about not revving to 7k, but it sounds like you were doing that! I thought redline was 6k? :-/ What is the stroke on it? You're likely lucky it didn't throw a rod out the side if you were doing that to it.

I'll post about a flow bench in your area and get back to you.

Fred.

[Peter]

Well I decided I wanted to see how much boost this turbo setup would make. First I put a tank of 92 octane in it. Then I jacked up the boost, and tuned it a few times. After I was happy with how it ran I made a few hard pulls onto the interstate. I was really happy with how much power it was making, but the pistons weren't.


So I lightly honed the three cylinders with cracked pistons, and stuck the pistons out of the engine with the cracked block in it. I'm waiting on some head gaskets to put it back together, but it should be running by Thursday if I get off my ass.

[Fred]

Did you forget that you don't have control of the ignition timing? :-/ That sucks man! Because you're running OEM ignition you have to run it rich as fuck, and yes, it'll make less power/torque rich, but it will be less likely to do that. How lean was it? Did you heed the advice that Dan and I gave you a week or two back? Or did you roll with your plan?

[Peter]

Did you forget that you don't have control of the ignition timing?

Forget? ... No .. ignore? Um, maybe a little bit. I suppose it could have been pre-detonation, but I didn't hear it knocking. 19.2psi of boost and drinking a little didn't help the situation at all.

:-/ That sucks man!

Sad day, but 1. I was asking for it, and 2. it's not that big of a deal. Some spare parts, a set of head gaskets, and two to three days of work depending on how hard I work.

How lean was it?

I'm embarrassed to say. You knew I was eventually going to accomplish this.

Did you heed the advice that Dan and I gave you a week or two back? Or did you roll with your plan?

Kinda, it didn't melt the pistons, but it didn't live.

I use to have a snowmobile that would break the ring lands off, I jetted it off of a pyro with about 12psi of boost. It had a failure about every week to every other week, but the only time it would break ring lands off is when there was a ton of fresh powder. So I put a knock sensor in it, because some people were telling me that it was pre-detonation. I could make it pre-detonate, but it would still break the ring lands off without throwing a pre-detonation fault. Taking the boost down to about 3-4psi would make it live, but it wasn't as much fun to ride.

Overall, not that bad of a day. I got to have some fun with my car. Just waiting on some parts to put the engine back in. Had several drinks with a nice lady I really like. Now I'm going to bed.

[Fred]

I totally understand the "blew up engine, had great day" thing :-) I like your attitude.

Will you tell me in private how lean you tuned it?

The thing with melt vs crack is this: If you have minor detonation (pre-ignition is a different thing, btw) the shock waves knock the skin of cool gas off of the piston (and chamber) and remove a layer of insulation thus over a few cycles allowing it to get really hot and then melt. If it's not so minor you just break things outright with no chance of such overheating before other major damage is done.

If you want to play on 20psi on high compression and stock parts there are a few things you need to be aware of:

With OEM timing you're going to have to be rich rich 10:1-11:1 in boost
With pump gas (92aki ain't that high) you're going to have to be rich and retarded
With high octane you can get away with murder and make massive power BUT you will eventually do one of two things: thermally stress the pistons causing them to come apart regardless of knock (your snow mobile situation probably) and/or physically stress the rods causing them to come apart (the situation with my 2.0 engine at 400hp for around 3 months abuse).

Did you get a datalog out of any of this tuning??? :-) <tries to salvage something>

Nice lady sounds good, perhaps one day I can meet her and you can tell me other stories about your youthful carelessness :-)

Fred.

[Peter]

I totally understand the "blew up engine, had great day" thing I like your attitude.

lol The first time I pulled that engine it took me three days. Now it takes me about 4 to 6 hours. Of coarse I've also lost track of how many times I've pulled it out, but I think it's around 12. Two of the pistons are actually tighter than the ones I pulled out even after I honed the scoring out of the cylinders.

Will you tell me in private how lean you tuned it?

I was aiming for around 12:1, but I have a feeling it was leaner than that.

If you want to play on 20psi on high compression and stock parts

I did find out that Acura makes a forged pistons for a newer version of this engine. The only problem is that it has an 11:1 compression ratio, and the cast pistons are 10:1. So I might buy one, and see if I think I can get away with milling off the top of the piston.

Did you get a datalog out of any of this tuning?

I tried, but apparently I didn't get the device opened or the start log box checked in CuteCom, because I have an empty file.

Nice lady sounds good, perhaps one day I can meet her and you can tell me other stories about your youthful carelessness

I don't know if she'll still be around by then, but I hope so. Do you have a timeline for your next trip around the globe? Like 3 to 5 years out something?

I'm thinking it might be time to fix all my little problems this weekend. Since it'll be easier to work on it with the engine out.

I was also looking at parts to setup the ignition. I kind of want to test the BIP373 against an IGBT. Do either of those require a driver chip, or can I just throw a square wave at them? I'm pretty sure the Ravage board has a few MOSFET drivers, but I don't think those qualify as MOSFETs. This is the IGBT I'm looking at. http://search.digikey.com/us/en/product ... -ND/812232

[Fred]

I was aiming for around 12:1, but I have a feeling it was leaner than that.

Yeah, that would be a good place to aim for IF you had appropriate timing, but you don't, so it's not.

The only problem is that it has an 11:1 compression ratio, and the cast pistons are 10:1.

OK, 10:1 is NA material, and with boost, 8psi max on 92AKI fuel. More is not going to be a healthy mix and require overly retarded timing and overly rich mixtures. Going to 11:1 is marginal NA on 92AKI.

So I might buy one, and see if I think I can get away with milling off the top of the piston.

Machining the domes off is crazy, they're only barely thick enough as is, and to remove enough to drop compression to the 9:1 or 8:1 that you need for your pump gas boost aspirations (without h2o injection) you couldn't machine enough off without finding the rings, losing all quench/squish, making a hole in the middle :-p Look for something with a matching bore, pin size and pin to crown height from something 16V and use those. Or stay at 10:1 and keep your boost low, or your octane extreme, or both, as situation calls for it (diff tune and fuel for track for example).

I don't know if she'll still be around by then, but I hope so. Do you have a timeline for your next trip around the globe? Like 3 to 5 years out something?

Yeah, maybe three years from now. Hard to say. I imagine I'll be in the US briefly before then, but don't know.

I'm thinking it might be time to fix all my little problems this weekend. Since it'll be easier to work on it with the engine out.

Great :-)

I was also looking at parts to setup the ignition.

Planning on going back to semi-sequential and waste spark and full control? If so, give me some warning so we can knock together a decoder for you.

I kind of want to test the BIP373 against an IGBT. Do either of those require a driver chip, or can I just throw a square wave at them?

Neither will work adequately from the CPU directly in a safe way. Buffer with some standard 20 or 25mA max continuous logic IC and use a 200 ohm resistor. However, even that won't really be enough because the IGBT has lots of capacitance in the gate and the BIP is a current amplifier and will like more than ~10mA you'll get from 5v with 200 ohm.

I'm pretty sure the Ravage board has a few MOSFET drivers, but I don't think those qualify as MOSFETs.

BIP is a darlington, but 3 stage, still requires current, though, so it's just about voltage and resistance for the BIP. The IGBTs are the same as mosfets from a driver point of view. We use the fet drivers in ravage to get suitable voltage and current and polarity with an off the shelf part.

This is the IGBT I'm looking at. http://search.digikey.com/us/en/product ... -ND/812232

Way cheaper on future:
http://octopart.com/parts/search?q=IRGB14C40LPBF

OK, that looks to have a logic level input, so you could drive it directly from the micro, BUT, there are complications. You need to split the ground from these and the CPU apart and once you do that, there will be a significant ground potential difference between the two circuits. At that point 5v ain't 5v anymore, so you can run into switching issues because of that. So a fet driver with 200 - 1000 series resistance and 12v to the input with isolated grounds would be a good choice. V clamp on input is 10 - 12V so you won't get much current from 14.4 to that with a such a series limiter. Make sense?

Gotta ask, though, why do you want to drive coils with such parts? Why not use some existing ignitors? You're not planning to put them in the cabin and run cables to the coils from the ECU are you? If so, don't expect any support from me, that's a fucking terrible idea. If you're going to use these things, put them in their own well heatsinked box in the engine bay, directly between ground point and coils and run signals to them. Doing anything else is screaming "hey, noise, come and fuck with me" :-)

Fred.