Ben's 914 TURBO Official World Premier, 'cuz it didn't blowd up. Options

[airsix]

Ok, so it isn't as cool as Chapman's 914-6 Turbo 3.0, but it's a turbo 914 all the same. Ladys and Gentlemen, at about 10:00 PST I got boost and there was much rejoicing. I took the car out, filled it with premium, tuned it with the laptop a little to make sure it got plenty of fuel under boost, and the mission was a sucess. Boost starts to come on at ~2,700rpm and is going full-steam by 3,500rpm.

I stand (ok, sit) here before you my brothers to declare that yes indeed, you can turbo a 914. (IMG:style_emoticons/default/clap56.gif)

Details:

Engine: '73 1.7, stock internals
EFI ECU: Perfect Power MIC3 (Cost ~$500 to get on the car and running 3 years ago)
Misc EFI stuff: Subaru 1.8T injectors, Holly throttle body off a 3.0 Chrysler, BMW 5xx fuel pump, foam motorcycle aircleaner (temporary)
Turbo: IHI off 1.8L Subaru (including wastegate)
Misc Turbo stuff: Bosch blow-off valve (for a Saab I think). Cheap. Plastic. Works great.
Total cost to add turbo: $175 (EFI was already on the car. Use Megasquirt+turbo and you can do this for <$400 easy)

I tried to modify my fuel pressure regulator with a boost-reference port like the in the dune-buggy.com article but I broke off a wire guage drill bit about 1.5" deep into the bolt. Dang. Only had about a quarter-inch to go too. So I just bolted it back on the car and set the fuel pressure at the 30psi I have the ECU turned for. I tried to crank up the injector pulse width under boost to compensate, but they hit 100% duty cycle at 5,000rpm because at that point they only have 20psi effective pressure (because there's 10lb of boost by that point). So I'm keeping it under 5k rpm until I get a rising-rate pressure regulator. The wideband O2 says mixtures are just fine below that point. I estimate I'm getting 110hp (That's not a WAG. I really did the math) which doesn't sound like much but is sure better than the 80hp I had before. I would still like to add an intercooler which could put it in the neighborhood of 130hp without getting too boost-happy. Even as it is now I would say it's 100% more fun to drive. I'm really happy about it. I'll get the pressure regulator issue squared away so I can run it up to 6k and I think it'll keep me happy for quite a while.

Rick said "This thread is worthless without pictures" so here's the best I could do. If you missed the earlier thread I'll throw in one of the earlier pics too. I tried to make a video but all you hear is wind noice and the blow-off valve between shifts.

[lapuwali]

I meant the FETs powering the injectors, which should be mounted directly to the MS board and may get fairly hot. The resistors limiting the current to the low-impedance injector should get hot, and should be mounted far away from the ECU.

Having the cover off means you're exposed to electromagnetic interference from the ignition system. The same thing that causes static on the car radio if you don't run resistor wires and/or plugs can cause havoc with any computer system. Every component on the board, esp. the metallic leads on things like resistors and capacitors, will act like a tiny antenna and pick up the EMF coming off the plug wires. A metallic case will help shield the ECU from this interference.

When you see the reset, you see the VE table is empty (or scrambled)? Hmm. Do you have to re-upload the VE table to get it to work? This is ringing a bell. What *should* happen in a simple processor reset is that it will simply boot again, and copy the config data from FLASH to RAM, and it runs out of RAM. When you edit the tables, it just edits the RAM copy, and when you "save", it's copying the data from RAM back to FLASH. A reboot will literally take milliseconds, so you'd normally have a hiccup while it's booting again, but then it will keep running after it's re-copied the FLASH data to RAM (which will have been erased during the reset). If you see the copy in FLASH going away, then the processor is somehow corrupting or clearing the FLASH. I seem to recall this being a known problem that's been fixed with newer code. I'll have to do some reading on the MS boards to see if I can find this again...