code 52 **video inside** problem

[ccguy]

i did a 1987 7mgte swap into my 80 toyota 4x4. runs great minus the hesitation at 2k. just like the video shows.
code 52. i replaced both knock sensors, rewired both knock sensors with shielded wire, tested continuity of wire and it is good from the sensor plug to the ecu.
i can rev it over by feathering it but it will not rev over 2k under load. throws the code and acts the same if they are unplugged as well. so ecu????
i am going nuts.
what ideas do you have???
http://www.youtube.com/watch?v=GjEoxA2D_80&feature=youtu.be

 

[IwantMKIII]

If you have easy access to a buddy's extra ECU, see what happens. A long shot but easy to check, I had a very similar problem, was a poor grounding issue; the car acts possessed and nonresponsive when the grounds are just on the cusp of not working at all and poor but usable.

 

[Victor Charlie]

All grounds should be retaped and Reinstalled with a new bolt and dielectric grease. I would also want to verify that the knock sensors you put in actually work on a different supra. I recall something about a code 52 logic diagram. Ill hunt a bit.


Sent from my DROID BIONIC using Tapatalk

 

[ccguy]

i will double check my grounds from the ecu to the chassis. a friend has one but it is 3hrs away. if the ecu is not solidly grounded do you think this could be it?
the knock sensors were on a good motor. also the ones i pulled i hooked up to an ohm meter and made contact with the threaded metal end to a metal table and hit the table with a hammer and the ohm meter moved with each hit.

 

[Victor Charlie]

I found the info and updated myself. The ecu gets asignal from the ks that is a 7khz band pass filter of engine noise. It shows small amplitude ac signal for normal ignition, high amplutude for knock. If the signal is not recognizable you get a code 52. A short or ground will do it. But so will a weakened signal from a cheap ks - use toyota part or one scavenged from a working engine part out. So will over or under torquing the KS. 52 inch lbs is correct. If overtorqed once the ks can be damaged. Or incorrectly grounded shielding. Ground shielding at one end. I would think to the ecu sensor ground, but I havent checked. You need a quality signal showing engine activity to avoid the code, not just any signal. Only with an oscilloscope can you test this.

Theoretically, vibration near 7 khz could also create jamming interference. Something lose on the block that vibrates two cycles for each engine rotation, a plausible scenario, could cause a code at 3500 rpm, for ex. Make sure everything bolted to the block is solid. Check valve covers, for ex. This is maybe a far fetched theory, but compatible with tech training info from toyota. Ive seen the ecu program code dissassembly, and this is compatible with what I remember of that.

Sent from my DROID BIONIC using Tapatalk

 

[Victor Charlie]

I also checked the ECU program code today, briefly. It seems that the knock sensor feeds into a separate control unit that receives info also from the ECU on when a cylinder is at TDC and returns info on Engine vibration in the form of good, bad or indifferent, but not as specific levels of knock vibration. The ECU can only set a code if the knock sensor control unit, not the ECU, identifies a problem with the quality of signal from the knock sensor. The knock sensor control unit is a black box to us, so I can't tell what type of signal triggers a code 52, but I think what I said above is reasonably correct.

 

[jetjock]

^ Some is right, some not. Doesn't matter though since all one can do is replace the sensors, ECU, or fix the wiring. And don't forget code 53...

 

[3p141592654]

.... returns info on Engine vibration in the form of good, bad or indifferent, but not as specific levels of knock vibration...

Actually the knock sensor MCU returns none, mild, medium or heavy knock. The ECU responds accordingly. That said, with a code 52 present the ECU will retard timing a maximum of 15 degrees from nominal, and in reality the maximum value is pulled form a table which is for most conditions less than the maximum 15 degrees. So, I would think your hesitation is from something else. Did you by chance remove the fuel dampener from the fuel rail?

 

[AJ'S 88NA]

Actually the knock sensor MCU returns none, mild, medium or heavy knock. The ECU responds accordingly. That said, with a code 52 present the ECU will retard timing a maximum of 15 degrees from nominal, and in reality the maximum value is pulled form a table which is for most conditions less than the maximum 15 degrees. So, I would think your hesitation is from something else. Did you by chance remove the fuel dampener from the fuel rail?

I've been wanting to know that for years "code 52 present will retard timing" I thought so but was never confirmed that that is what the ECU would do , pull timing.

Thank you very much 3p

 

[Victor Charlie]

I've been wanting to know that for years "code 52 present will retard timing" I thought so but was never confirmed that that is what the ECU would do , pull timing.

Thank you very much 3p

Thanks Jon. Actually that's a more precise statement of what i meant. presumably 2 bit input, 4 levels. It is also clear that the ECU signals back to the MCU piston TDC timing, at least. Wading through the code was slow going after two years not looking at it, so I'm not clear on one thing: does the MCU flag the bad signal condition that generates a code 52, or does the ECU determine this? E.g. process level of vibration or knock in relation to piston position and determine if the signal is modulating in phase with the engine, as expected when the engine is running, to determine if there is a valid signal. I apologize for asking a question that is answered in the code. but time is short this week and it's of general interest to understand this correctly. Thanks.

 

[Victor Charlie]

thanks for the food for thought, JJ

 

[ccguy]

Actually the knock sensor MCU returns none, mild, medium or heavy knock. The ECU responds accordingly. That said, with a code 52 present the ECU will retard timing a maximum of 15 degrees from nominal, and in reality the maximum value is pulled form a table which is for most conditions less than the maximum 15 degrees. So, I would think your hesitation is from something else. Did you by chance remove the fuel dampener from the fuel rail?

if i am correct the fuel dampener is on the banjo bolt on the inlet side of the fuel rail (looks like it has a plastic hard hat on)???
no i kept that setup. on the fuel return side of the rail i went from the fpr straight to a hose back to the tank return. i did not use the j tube that is factory in between the drivers motor mount with he small orifice in it.

 

[3p141592654]

Yes, that is it. It was a long shot, but you can get weird resonances without the dampener in place.

You can get issues if the AFM intake is not bolted to a properly sized air filter adapter providing laminar air flow into the mouth of the AFM, and make sure the intake is away from any external air turbulence from the engine fan. Your video doesn't show under the hood, so just guessing right now. I assume you checked fuel pressure and other basic things like the injector wiring and coil/ignition wiring.

 

[ccguy]

i picked up another ecu. but it is for a manual.
ARE THE PLUGS THE SAME (BASICALLY) ON THE MANUAL AND AUTO ECU??
just want to make sure nothing is big time different before i plug up the new manual tranny ecu.

 

[ccguy]

so i put in the manual tranny ecu and it still stumbles like crazy. shit i dont know maybe re wire the knock sensors.

if it was detecting knock would it retard the timming this bad? it basically shuts off the engine (fuel or spark??)

 

[3p141592654]

Did you read what I wrote? ::stickpoke

 

[ccguy]

yup i did read it......i have a baffle protecting the air filter form any hot air from the rad. and the adapter for the filter is the same size square as the factory one that goes into the afm. i will have time tomorrow to tinker some more with it.

 

[ccguy]

ok so messed around with it for a little bit.
i unplugged both of the knock sensors at the ecu to make sure that it was not actually picking up a bad knock. it acted the same as when they are plugged in. so i dont think it is actually detecting a knock. it still threw the code 52 at the same rpm range. also i used a mechanics stethoscope on multiple places on the block and head and revved it up and could not detect a knock or ping.

SO.... time to rewire it again and see what happens.
fuel psi is good. afm is good and unobstructed, ecu is good, knock sensors are good so, crap i sense wiring in my near future.

 

[supraguy@aol]

For clarification(for me), does code 52 represent that a knock is detected, or that there is a problem with the knock sensor system?
If it is the latter, then unplugging the knock sensors won't verify anything, because even if you had a perfectly working system and unplugged the sensors, I would think that you'd trigger a code 52 anyway.

 

[3p141592654]

52 is a failure of the knock system, resulting in the maximum ignition retard of up to 15 degrees being applied all the time, which will hurt performance but it will still run just fine.

52 does not mean knock is detected. In normal operation the system is always listening for knock and advancing the timing to be on the edge.