Getting knock at 6300 RPM
12-02-22, 12:22 AM
#1
Getting knock at 6300 RPM
I've got a street ported 13bt running a GT35. Its running a Megasquirt MS3X with the built-in knock module, but this isn't megasquirt specific, so I thought I would get a lot more traction here.
I've broken in the motor enough to finally take it to redline, and I've noticed that I am getting knock at ~6300 rpm reliably - every time, with both part or full throttle. I don't see any knock readings above about 30 anywhere else, but it will spike to 60-80 at that exact spot every time.
I don't notice anything abnormal in how it runs, its smooth all the way to 7k. I've also noticed that it seems to read more on the front rotor. Do you think this could be some sort of natural resonance?
Its running 10PSI, with an AFR of ~10.9, and timing at 11. I am using knock sensors from an RX8 attached to the stock location on each rotor housing. 10 heat range plugs, IGN-1A coils. IATs are under 100F, no EGT readings yet.
I've broken in the motor enough to finally take it to redline, and I've noticed that I am getting knock at ~6300 rpm reliably - every time, with both part or full throttle. I don't see any knock readings above about 30 anywhere else, but it will spike to 60-80 at that exact spot every time.
I don't notice anything abnormal in how it runs, its smooth all the way to 7k. I've also noticed that it seems to read more on the front rotor. Do you think this could be some sort of natural resonance?
Its running 10PSI, with an AFR of ~10.9, and timing at 11. I am using knock sensors from an RX8 attached to the stock location on each rotor housing. 10 heat range plugs, IGN-1A coils. IATs are under 100F, no EGT readings yet.
12-02-22, 04:51 AM
#2
Senior Member
Join Date: Jun 2010
Location: U.K - Instagram - copyninja_fd
Posts: 329
Received 153 Likes
on
89 Posts
My understanding of how well setup knock sensors work is that you have to actually induce a knock situation and when you can hear the engine ping/det then you pull back timing a few degrees. I've been successfully able setup knock sensors on piston engined cars. However, i've not bothered with knock control on a rotary because if the engine knocks/dets even once...its basically game over and you may as well restart the rebuild.
If you want to set up a true knock control strategy on a 13b you will more than likely need to have a spare 13b because you will 100% end up damaging the engine when setting this up correctly for your application.
You said 10psi 10.9 afr and 11 deg timing (whats the split?) is not going to cause any knock in my experience. For e.g. I run 12 psi at 11.0-11.2 afr with 15-18 deg timing and a -8 split which runs a 7.8s-8.1s 100-200kph consistently and reliably for almost 2.5 years. I highly doubt you are getting knock...because if your engine truly knocked....you'd know about almost immediately.
If you want to set up a true knock control strategy on a 13b you will more than likely need to have a spare 13b because you will 100% end up damaging the engine when setting this up correctly for your application.
You said 10psi 10.9 afr and 11 deg timing (whats the split?) is not going to cause any knock in my experience. For e.g. I run 12 psi at 11.0-11.2 afr with 15-18 deg timing and a -8 split which runs a 7.8s-8.1s 100-200kph consistently and reliably for almost 2.5 years. I highly doubt you are getting knock...because if your engine truly knocked....you'd know about almost immediately.
Last edited by rx7srbad; 12-02-22 at 05:14 AM.
12-02-22, 10:04 AM
#3
Moderator
iTrader: (3)
Join Date: Mar 2001
Location: https://www2.mazda.com/en/100th/
Posts: 30,884
Received 2,635 Likes
on
1,865 Posts
the knock sensors are microphones, they pick up noise, so you need to figure out if the noise is knock, or something else
you can try a few things, to make sure it is or isn't knock.
you might add fuel, add some timing split and then maybe retard a couple degrees, see if it changes/goes away.
mine has high knock at ~3000rpm, and its something mechanical.
you can try a few things, to make sure it is or isn't knock.
you might add fuel, add some timing split and then maybe retard a couple degrees, see if it changes/goes away.
mine has high knock at ~3000rpm, and its something mechanical.
12-06-22, 07:59 AM
#5
Racing Rotary Since 1983
iTrader: (6)
" if the engine knocks/dets even once...its basically game over and you may as well restart the rebuild"
this is too strong a statement.
all engines knock, all the time. it is a question of degree.
any single turbo rotary requires a 21st century knock system so as to manage knock. "manage" as in cutting timing on the next rotor face. the response needs to be timing, not fuel or boost as that takes secondS meanwhile 30+ rotor faces have sped by.
the normal knock setup procedure is to start out very conservative as to tune and observe knock readings. peak knock normally is in the peak torque area as that is peak combustion chamber pressure. since you indicate you get knock at 6300 part or full throttle it could be mechanical. a quick way to find out is to reduce timing in the area from 11 to 5 or 6 and do another run. if there is no change the problem is mechanical. if there is a change simply mod your timing map. the turbo rotary is way less sensitive than a piston engine w re to timing.
gasoline quality is VERY important as to knock. my engine developed knock around 11 advance (around 500 hp) out of the blue. i lowered the timing and the knock disappeared. i also picked up power a bit even though i was at 6 IGL since the knock vanished. it was the gas. i swapped in better gas, switched back to more timing and all was well.
"Howard is using them and disagrees" ?
similar to Bosch vastly upgrading their fuel injectors (from the EV1 to the EV14), Bosch also uprated their knock sensors which are the OE standard. when i discarded the dreadful OE turbo system in 2003 i took advantage of the newer style Bosch knock sensor. not knowing the resonance scale i bought them from Mazda since they were being used on the Renesis block. i figured it would resonate in the same neighborhood as my motor. they have worked perfectly since 2003. i now have 4 of them on my motor as the other two provide signal to my Link Knock Ears.
this is too strong a statement.
all engines knock, all the time. it is a question of degree.
any single turbo rotary requires a 21st century knock system so as to manage knock. "manage" as in cutting timing on the next rotor face. the response needs to be timing, not fuel or boost as that takes secondS meanwhile 30+ rotor faces have sped by.
the normal knock setup procedure is to start out very conservative as to tune and observe knock readings. peak knock normally is in the peak torque area as that is peak combustion chamber pressure. since you indicate you get knock at 6300 part or full throttle it could be mechanical. a quick way to find out is to reduce timing in the area from 11 to 5 or 6 and do another run. if there is no change the problem is mechanical. if there is a change simply mod your timing map. the turbo rotary is way less sensitive than a piston engine w re to timing.
gasoline quality is VERY important as to knock. my engine developed knock around 11 advance (around 500 hp) out of the blue. i lowered the timing and the knock disappeared. i also picked up power a bit even though i was at 6 IGL since the knock vanished. it was the gas. i swapped in better gas, switched back to more timing and all was well.
"Howard is using them and disagrees" ?
similar to Bosch vastly upgrading their fuel injectors (from the EV1 to the EV14), Bosch also uprated their knock sensors which are the OE standard. when i discarded the dreadful OE turbo system in 2003 i took advantage of the newer style Bosch knock sensor. not knowing the resonance scale i bought them from Mazda since they were being used on the Renesis block. i figured it would resonate in the same neighborhood as my motor. they have worked perfectly since 2003. i now have 4 of them on my motor as the other two provide signal to my Link Knock Ears.
The following users liked this post:
rx7srbad (12-06-22)
12-07-22, 07:44 PM
#6
If you're unlucky, software settings in the ECU could result in more ignition timing advance than you're seeing reported in the software. If you haven't already used a timing light to measure the timing, I would do that before anything else. Once you've verified that 10 degrees in the ECU is actually 10 degrees on the damper at high RPM, then I would do Howard's idea to remove a bunch of timing in case it's real knock.
If the ECU + knock sensor combination shows the same reading for a light-load sweep (no boost) and a full throttle sweep and you've verified the timing at the damper matches what the ECU software is claiming, I would simply stop trusting the ECU's knock readings. If you want to prove the readings are wrong, try unthreading the sensor from the engine, secure it somewhere it won't rattle around (maybe wrapped in a rag and tied somewhere safe) and record some part throttle sweeps to see if the spikes remain or go away. For what it's worth, I've never seen anything that resembles knock spikes on my stock-port engine with the stock knock sensor (AEM Infinity ECU), but the logged data shows the background noise get louder at full throttle and quieter when cruising or decelerating so I trust it's not completely useless.
If the ECU + knock sensor combination shows the same reading for a light-load sweep (no boost) and a full throttle sweep and you've verified the timing at the damper matches what the ECU software is claiming, I would simply stop trusting the ECU's knock readings. If you want to prove the readings are wrong, try unthreading the sensor from the engine, secure it somewhere it won't rattle around (maybe wrapped in a rag and tied somewhere safe) and record some part throttle sweeps to see if the spikes remain or go away. For what it's worth, I've never seen anything that resembles knock spikes on my stock-port engine with the stock knock sensor (AEM Infinity ECU), but the logged data shows the background noise get louder at full throttle and quieter when cruising or decelerating so I trust it's not completely useless.
12-09-22, 03:22 AM
#7
thats too retarded, advance your timing closer to 15-20 deg, 10 split. bet you knock goes away. do not retard any more, running less leading advance than split causes preignition of next rotor face, you probably on the ragged edge, depending on AFR
Last edited by gxl90rx7; 12-09-22 at 03:26 AM.
Trending Topics
12-10-22, 04:49 AM
#8
Arrogant Wankeler
Otherwise correct plug heat should cover residual electrode temp.
12-10-22, 05:06 PM
#9
Edit.. actually nm i reread op, he said it happens on free rev too, timing should have lots of advance there, unless timing map is weird. I would check trigger signals, does timing light jump around, does megasquirt show trigger errors
Last edited by gxl90rx7; 12-11-22 at 01:05 AM.
Thread
Thread Starter
Forum
Replies
Last Post
Harlan
General Rotary Tech Support
50
05-21-15 06:11 AM