EZ-F knock retard!
04-08-2008, 08:29 PM
#17
Inventor
Rennlist Member
Rennlist Member
Thread Starter
Here's how it looks. I used a superbright green LED (10 candela) to indicate when the retard is active. The Knocklite hides in place of the ashtray, so it is still in reach for (programming).
I'd like to make a two timer circuit, one for the retard, and one to briefly power a red LED in the unused [SERVO] light (PS failure warning, next to [TAIL LAMP]), to see if the retard is actually stopping the knock.
It took some doin', but the whole thing plugs inline at the EZ-F. The air temp sensor is deleted, and it's wires reused for the knock sensor.
I'd like to make a two timer circuit, one for the retard, and one to briefly power a red LED in the unused [SERVO] light (PS failure warning, next to [TAIL LAMP]), to see if the retard is actually stopping the knock.
It took some doin', but the whole thing plugs inline at the EZ-F. The air temp sensor is deleted, and it's wires reused for the knock sensor.
Last edited by PorKen; 04-09-2008 at 01:43 AM.
04-08-2008, 11:44 PM
#18
Rennlist Member
It's funny, I read Porken's threads and I only half get it. He explains it well, but I'm an idiot when it comes to electronics. I just ask, okay Ken, just tell me what to do and I'll do it. Seriously though, if I read and re-read etc... I start to get it, kind of.
I think my '84 Euro has a knock sensor. But I don't know much about it.
Anyway you cut it, I still like reading your posts, Ken.
I think my '84 Euro has a knock sensor. But I don't know much about it.
Anyway you cut it, I still like reading your posts, Ken.
04-09-2008, 03:37 PM
#19
Inventor
Rennlist Member
Rennlist Member
Thread Starter
Jim,
Unless someone put it in, pre-87 engines don't have a knock sensor.
Prior to this, I've never made an integrated circuit based gadget... that worked! It hurt my head learning all the basic facts about transistors, diodes, the 555 timer chip, even which way to wire the LED. I printed out a stack of circuits, with notes, and double, quintiple checked the orientation and pinouts of all the components as I put it together. I was still amazed, however, when the LED came on the first time!
Just like changing the timing belt the first time, it seemed like an impossible job. After it's done, you say: "that was easy, why didn't I do that sooner?".
I'm already adding features that I want. 556 dual timer chip, maybe add a transmission switch (002 545 451 4) from the S4 autos, so it retards to save the clutches and stop from spinning the 265's when it shifts at full throttle. I'm making more power than an S4, so I should have one, right?
I did a quick search, and the prices aren't too bad. There's even free PCB CAD programs!
Unless someone put it in, pre-87 engines don't have a knock sensor.
Prior to this, I've never made an integrated circuit based gadget... that worked! It hurt my head learning all the basic facts about transistors, diodes, the 555 timer chip, even which way to wire the LED. I printed out a stack of circuits, with notes, and double, quintiple checked the orientation and pinouts of all the components as I put it together. I was still amazed, however, when the LED came on the first time!
Just like changing the timing belt the first time, it seemed like an impossible job. After it's done, you say: "that was easy, why didn't I do that sooner?".
I'm already adding features that I want. 556 dual timer chip, maybe add a transmission switch (002 545 451 4) from the S4 autos, so it retards to save the clutches and stop from spinning the 265's when it shifts at full throttle. I'm making more power than an S4, so I should have one, right?
I did a quick search, and the prices aren't too bad. There's even free PCB CAD programs!
Nice work Ken.. this is cool stuff.!
04-09-2008, 10:23 PM
#21
Burning Brakes
Ken,
How may these tweaks do with aftermarket chips?
My crude understanding is you:
1) are tricking the temp II to see an apparent air temp of 32 degrees F (i.e., avoid 3 degrees retard at operating temp).
2) Avoid 20 degrees retard as 2/3 throttle.
Does this mean, w/ chips, we get a net 7 degree advance? What we would normally get, w/ chips, on a cold day before an increase air temp?
I'm thinking if, w/ chips and tweaks, this combo is "too close for comfort," then your development of the shorted octane loop mitigates this problem.
How may these tweaks do with aftermarket chips?
My crude understanding is you:
1) are tricking the temp II to see an apparent air temp of 32 degrees F (i.e., avoid 3 degrees retard at operating temp).
2) Avoid 20 degrees retard as 2/3 throttle.
Does this mean, w/ chips, we get a net 7 degree advance? What we would normally get, w/ chips, on a cold day before an increase air temp?
I'm thinking if, w/ chips and tweaks, this combo is "too close for comfort," then your development of the shorted octane loop mitigates this problem.
04-10-2008, 05:07 PM
#22
Inventor
Rennlist Member
Rennlist Member
Thread Starter
John,
The aftermarket chips are already advanced (~4°), so there'll be too much advance if you add another 3 or 5°. IE. 9° or more! They are not advanced at idle, however, and a little advance smooths the idle.
Temp II is the coolant temp sensor. I'm making the EZ-F think it is -20°F, to which adds 5° to the idle, and cruise maps, only (not WOT map). If you make it think it's 50°F, then it adds only 3°.
Temp I is the air temp sensor. Stock, the EZ-F only looks to see if this input is faulty, IE. shorted (wires connected together = octane loop), or open (sensor disconnected). When the air temperature is over 122°F, the resistance is so low, the EZ-F considers this a shorted wire. Either way, it subtracts 3°, but only when there is low vacuum (high load), or the WOT switch is on.
At WOT there may be more or less advance or retard in the WOT map, versus the cruise map, at different rpms - there isn't a blanket 20° retard. The WOT map is RPM based only (1 dimensional), and doesn't take load (vacuum level) into account. (The idle map is also only rpm dependent.)
I disconnect the WOT input, so when it's not at idle, it's always in the cruise map which uses rpm and load (2D) to determine the amount of advance to use. This map is roughly 5° retarded from max power overall so it will likely never knock, even on regular. (Using the cruise map also means I can use the warmup map to fool the EZ-F into adding 3 or 5°.)
I fool the EZ-F into thinking the air is always cool with a fixed resistance in place of the Temp I sensor. The knock sensor circuit shorts the wires (octane loop) for a few seconds if there is knock detected.
Using the stock map, and adding 5° all the time, I could 'retard' by a total of 8° if needed. IE. take away the 5° addition, and use the octane loop to subract another 3°. So far, even using 87 octane, the 3° is enough to stop knock, however.
The aftermarket chips are already advanced (~4°), so there'll be too much advance if you add another 3 or 5°. IE. 9° or more! They are not advanced at idle, however, and a little advance smooths the idle.
Temp II is the coolant temp sensor. I'm making the EZ-F think it is -20°F, to which adds 5° to the idle, and cruise maps, only (not WOT map). If you make it think it's 50°F, then it adds only 3°.
Temp I is the air temp sensor. Stock, the EZ-F only looks to see if this input is faulty, IE. shorted (wires connected together = octane loop), or open (sensor disconnected). When the air temperature is over 122°F, the resistance is so low, the EZ-F considers this a shorted wire. Either way, it subtracts 3°, but only when there is low vacuum (high load), or the WOT switch is on.
At WOT there may be more or less advance or retard in the WOT map, versus the cruise map, at different rpms - there isn't a blanket 20° retard. The WOT map is RPM based only (1 dimensional), and doesn't take load (vacuum level) into account. (The idle map is also only rpm dependent.)
I disconnect the WOT input, so when it's not at idle, it's always in the cruise map which uses rpm and load (2D) to determine the amount of advance to use. This map is roughly 5° retarded from max power overall so it will likely never knock, even on regular. (Using the cruise map also means I can use the warmup map to fool the EZ-F into adding 3 or 5°.)
I fool the EZ-F into thinking the air is always cool with a fixed resistance in place of the Temp I sensor. The knock sensor circuit shorts the wires (octane loop) for a few seconds if there is knock detected.
Using the stock map, and adding 5° all the time, I could 'retard' by a total of 8° if needed. IE. take away the 5° addition, and use the octane loop to subract another 3°. So far, even using 87 octane, the 3° is enough to stop knock, however.
04-10-2008, 05:30 PM
#24
Inventor
Rennlist Member
Rennlist Member
Thread Starter
S4, no. GT, yes. 
I'll probably boost instead, though. 
BTW: instead of/along with an ignition retard, you could use the KnockLite + relay to control a blowoff valve if there's knock.
There are programmable systems for boost control, but this whole setup, KnockLite, S4 knock sensor, relay circuit, cost less than $200 - DIY.
I'll probably boost instead, though. BTW: instead of/along with an ignition retard, you could use the KnockLite + relay to control a blowoff valve if there's knock.
There are programmable systems for boost control, but this whole setup, KnockLite, S4 knock sensor, relay circuit, cost less than $200 - DIY.
04-10-2008, 06:37 PM
#26
Nordschleife Master
Lets pitch in and get Ken a GT and he can be making all these goodies for our cars!
Ken, your the best sales pitch for the Shark Tuner yet! You've proven there isn't anything you cant do with that thing. (Not sure if it relates to this thread but you know what i mean)
04-10-2008, 06:46 PM
#27
Race Car
Ken is working around having to use a sharktuner by making clever little modifications to the engine management system in order to fool it to provide more ignition advance. This stuff would not be that useful for an '87+ 928 since it has knock sensors and the EZK can afford to provide more advance by default.
Dan
'91 928GT S/C
Dan
'91 928GT S/C
04-10-2008, 06:49 PM
#28
Nordschleife Master
IT WAS A BLANKET STATEMENT!
I'm well aware that EZK provides 2 knock sensors, his innovating is generally what the comment was about. In this case his innovation is of no help to me, thats why i thought the suggestion to get an S$ would be great and beneficial to ME, at the expense of the OB guys of course.
I'm well aware that EZK provides 2 knock sensors, his innovating is generally what the comment was about. In this case his innovation is of no help to me, thats why i thought the suggestion to get an S$ would be great and beneficial to ME, at the expense of the OB guys of course.
04-10-2008, 06:54 PM
#29
Race Car
IT WAS A BLANKET STATEMENT!
I'm well aware that EZK provides 2 knock sensors, his innovating is generally what the comment was about. In this case his innovation is of no help to me, thats why i thought the suggestion to get an S$ would be great and beneficial to ME, at the expense of the OB guys of course.
I'm well aware that EZK provides 2 knock sensors, his innovating is generally what the comment was about. In this case his innovation is of no help to me, thats why i thought the suggestion to get an S$ would be great and beneficial to ME, at the expense of the OB guys of course.
Dan
'91 928GT S/C
04-10-2008, 07:08 PM
#30
Inventor
Rennlist Member
Rennlist Member
Thread Starter
If Roger pushes more PKsn'rs out the door, I can afford a LH2.3 car, someday... for R&D!
The SharkTuner would let me avoid having to play tricks on the EZ-F, and make an ignition cruise map that more closely matches the (resonant) airflow with the exhaust, cam timing, etc, I have now. It could also be programmed so that there would be more than one level of retard, as there are 5 cells for the air temp retard, but only two are used in the stock map.
The SharkTuner would let me avoid having to play tricks on the EZ-F, and make an ignition cruise map that more closely matches the (resonant) airflow with the exhaust, cam timing, etc, I have now. It could also be programmed so that there would be more than one level of retard, as there are 5 cells for the air temp retard, but only two are used in the stock map.