randy merritt
Donating Member
Re: Holley HP install and tips
OK. I have been screwing around with the knock retard function of the Holley, and I may finally have it figured out (I think). I used the stock knock sensor in the stock location. The frequency is set to 5.4 kHZ, and the sensitivity is set to 10 (after several trial and error settings). With the sensitivity set much higher, the knock counts go up dramatically. I drove the truck around, and was getting knock counts of 30-50 at idle and low speeds, and counts of 10-30 at WOT and cruise. The Holley does not start pulling timing until the knock counts exceed 80. Here is an excerpt from the Holley EFI forum for a guy that seems to know what he is talking about:
"I usually just look at a lot of datalogs, and use my ears to listen as well.
Start with the value very low. When you look at a log, if you don't see any noise, turn it up just a little bit at a time until you do.
I try to keep timing safe and listen for audible knock.
Then you can watch in your datalogs for a sudden spike in the noise, it will be considerably higher then the rest of the noise.
It will be very apparent if you actually heard some audible knock, and then go look at the log and see a spike or multiple spikes.
If you have seen & heard actual knock, but it didn't trigger the knock timing decrease (above 85% knock signal),
then you know you can just turn up the gain until you have a pretty good amount of noise showing,
so that those knocks are a loud enough signal to trigger your knock timing table.
I find that typically, if you see 15% noise at part throttle, you probably have more than enough to see actual knock spikes if & when they happen.
I'm at 65% sensitivity, with my LS1 one wire knock sensors in the side of the block (iron block), and barely see any noise at part throttle (5% ish).
And for my car, it's enough for real knock to show up as a big enough spike to trigger my knock table."
The only real way to tell if the knock retard is working is the make the truck knock, which I am not going to do. I will set the max retard at 10 degrees, and just run it, as this is just a safety device anyway and my truck has a known good tune. The only thing I am concerned about is a meth failure, but I think my meth safety switch, which I rigged up to limit boost if meth is not flowing, should have me covered.
OK. I have been screwing around with the knock retard function of the Holley, and I may finally have it figured out (I think). I used the stock knock sensor in the stock location. The frequency is set to 5.4 kHZ, and the sensitivity is set to 10 (after several trial and error settings). With the sensitivity set much higher, the knock counts go up dramatically. I drove the truck around, and was getting knock counts of 30-50 at idle and low speeds, and counts of 10-30 at WOT and cruise. The Holley does not start pulling timing until the knock counts exceed 80. Here is an excerpt from the Holley EFI forum for a guy that seems to know what he is talking about:
"I usually just look at a lot of datalogs, and use my ears to listen as well.
Start with the value very low. When you look at a log, if you don't see any noise, turn it up just a little bit at a time until you do.
I try to keep timing safe and listen for audible knock.
Then you can watch in your datalogs for a sudden spike in the noise, it will be considerably higher then the rest of the noise.
It will be very apparent if you actually heard some audible knock, and then go look at the log and see a spike or multiple spikes.
If you have seen & heard actual knock, but it didn't trigger the knock timing decrease (above 85% knock signal),
then you know you can just turn up the gain until you have a pretty good amount of noise showing,
so that those knocks are a loud enough signal to trigger your knock timing table.
I find that typically, if you see 15% noise at part throttle, you probably have more than enough to see actual knock spikes if & when they happen.
I'm at 65% sensitivity, with my LS1 one wire knock sensors in the side of the block (iron block), and barely see any noise at part throttle (5% ish).
And for my car, it's enough for real knock to show up as a big enough spike to trigger my knock table."
The only real way to tell if the knock retard is working is the make the truck knock, which I am not going to do. I will set the max retard at 10 degrees, and just run it, as this is just a safety device anyway and my truck has a known good tune. The only thing I am concerned about is a meth failure, but I think my meth safety switch, which I rigged up to limit boost if meth is not flowing, should have me covered.
