FSAEracer03

This thread is great. Seriously, this has to be the single best thread on Rennlist in recent history. Transaxle, nice hard data! John, I'm looking forward to seeing the Knock Scope demonstration and such.

...I'm actually curious as to what circumstances allow for 87 octane gas in a safe environment as far as knock and temps. So far we've heard: never, stock boost, and a couple other answers.

Who wants to datalog this one??

Evan70

Ahh, now I see why I get different voltage readings. My SBC aux input is really just a voltage meter, and I'm not sure how fast it samples. So, it would look like I'm just catching a portion of the signal and not the whole thing. Fortunately, it does have a graph display function so I can get a general idea of how much knock I'm seeing. I know it takes multiple samples/second, but I'm sure it's nowhere close to 1000. Probably somewhere around 10 samples/sec at a guess. Anyway, for basic knock monitoring I think this may be a good solution. It won't give full data, but it'll at least let me see when I am showing knock.

So the next question, is there any kind of inexpensive standalone digital counter available that we could try hooking up to pin 15?

Transaxle

Take a simple one chip computer from AVR, PIC, etc. - bring #15 to TTL level, connect it to the external interrupt or timer pin of your computer and write 5 lines of code to count the knocks and to bring the results to your laptop or to a local display.

There are some cheap development systems for these type of computers available. Costs should be around 50$.

Zero10

Interesting developments in this thread.
What is the output voltage of pin 15? It should be able to handle the current of a LED, no?

Evan70

It peaks at 11 volts at start up, but after that I've never seen over 5 volts. Typical output that I see is from 1 to 3 volts, probably due to the fact that I'm only catching a brief sample of the actual spike since they are so brief. If you were seeing significant sustained knock, then you should see the 5 volts.

Zero10

Cool, I might just wire a LED to it with a small resistor. The brief 11V spike won't do appreciable damage to the LED. As long as it's for less than maybe 0.25 seconds.
I made a pretty bold statement that 87 octane is just fine for a 951, and I need to back it up with some kind of data.

pstickne

Manual says 96 RON (92 CLC/AKI) so, call me conservative but ...

Zero10

Everybody called me crazy, I know.
I ran 87 octane in my N/A as well, I got flamed for a little bit, and now it's the norm for most people on here.
Nobody will believe me that it's safe unless I can prove it is (or teach myself a lesson that it is not).

Ahmet

I ran 87 octane in my 84 NA for quite some time without any adverse effects years ago.

I've ran 87 and 89 in stock and chipped 951s, stock seemed to be fine. I blew a headgasket at I think it was 17 psi with no-name chips on an 86 turbo. I was hoping for a broken piston, but didn't get that lucky

PS: Knock is pretty hard to hear on a noisy 951, especially with an aftermarket exhaust.
Ahmet

pstickne

That is right -- NA is 91 RON (87 CLC/AKI) but Turbo is 96 RON (92 CLC/AKI).
At least that is what page 104 of the 944 Manual from '86 says...

mark944turbo

I dont think an LED will work, the signal seems too fast. You would need some kind of flip flop circuit to hold power to the LED for long enough, if I recall an old thread on this subject correctly.

Perhaps an EE person with a little time on their hands will come up with what we need, if not, I will be able to when I go back to school in mid January.

Transaxle

Take one of these systems:

http://www.active-robots.com/product...tmel-avr.shtml

There is no easier and cheaper way to get a knock monitor.

Zero10

If you feel like programming one of those. I just learned how to program the PIC family of microcontrollers, no way I'm learning another, lol.
I am not certan if the signal would be too fast. The only way to know is to try. Other than that, you could use a 30-40Hz oscillator on the reset of the flip flop, the LED on the output, and of course the knock signal on the input. That way the light would stay on for a 30th or a 40th of a second every time. Or rather it would tell you if there was a knock in every 30th or 40th of a second. That shouldn't be too fast to see.
I just don't happen to have any flip flops on my shelf, but I do have a ton of LED's and resistors.

Where can I locate this pin 15? I might give this a shot soon.

Jfrahm

Inlet air temp has a lot to do with knock sensitivity. In cold weather, you might be able to run 18psi on 85-87 (R+M/2) octane gas. In hot weather you might not get away with 18psi on premium. Every car is different and gas isn't always what it claims to be, so if you are pushing the limits you should have some knock detection. Otherwise one bad tank of gas might cost you a motor.

Also you probably will not get knock until your head has warmed up. Do not think you are OK if you do not hear knock in the first few minutes of driving.
Likewise on the dyno.

-Joel.

Zero10

I suppose that air temp is another thing to consider. The air here is cooler than for 90% of rennlisters (although not today, it's a nice 14*C), but the pressure is lower.

I don't ever get on boost until my oil starts to warm up (warm idle oil pressure at ~ 1.5 bar), so the head should be good and warm by then too.