When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Hi there yotatechers! I've got a 1989 Toyota sr5 pickup with a 3.0.
I recently did a swap and I have been having issues getting it to run right, I finally got it to a mechanic friend because I could not figure out my problem. We have found that my timing is being messed up by the computer. When the truck is idling it is about 10 deg retarded (after top dead center) and when you rev it up it goes to about 15 deg btdc??? It seems like an awful big jump in timing.
I do have one engine code which is 52 (knock sensor) but we have tried to disconnect the wire going to the ecu, as well as grounding out the connection to the ecu, both with no effect.
Could it be that my ecu is bad from the truck sitting so long before I did the swap? Or could it be the knock sensor?
If someone has had the same issue or has a suggestion as to what could be wrong please let me know!
Thanks,
Alex
When the ECU "loses touch" with the knock sensor, it no longer knows if the engine is knocking. If the ignition had been set too advanced (so engine was knocking), it could easily blow a hole in a piston. So what does the ECU do? It dramatically retards the timing to be SURE it's not knocking (to save the engine), and then helpfully tells you it just did that by throwing the code.
If the ECU has lost touch with the knock sensor (broken or shorted wire, bad shield, bad sensor), would it help to disconnect the knock sensor? Ground it?
First, pay your mechanic friend (so he can spend the time to figure out how to correctly address this problem), or get a new friend. What you've done so far is only risking damage to the engine.
Second, search "code 52" or "knock sensor" here. You'll learn that 90% of the time you've got a bad "pigtail." The part is only $18 (including shipping), but you have to remove the lower intake manifold to get to it. Not trivial, but no way around it.
Good luck!
(I just now see that you got pretty much the same answer a few hours earlier from RJR in your other thread. It's worth it to pay attention to what he says.)
the knock sensor signal is represented in ac, so unplugging it will lead the dc part to sense it is disconnected. you would need to know the resistance of the knock sensor and use that value resistor(s) between the sensor wire and ground for it to operate normally without sensing any knock, but you don't want it to run like that. unless someone knows the resistance, tell me where you unplugged it and i'll take a reading on mine at the same spot.
... you would need to know the resistance of the knock sensor and use that value resistor(s) between the sensor wire and ground for it to operate normally without sensing any knock, ...
LOL! Go ahead, try it. Then you'll know that won't work.
LOL! Go ahead, try it. Then you'll know that won't work.
let's just review exactly how the knock sensor works:
first, the ecu outputs a bias voltage, probably 5, 10, or 12v. it goes to the sensor, which has an internal shunt in the form a of a normal 1/4-1/8watt resistor you would find at radioshack, etc connected to ground. also, in parallel is a peizo plate nearly identical to a peizo tweeter, just like a microphone and speaker are nearly the same, but microphones can use a bias in place of, or in conjunction with the permanent magnet. all audio signals are ac, and that is all the peizo "mic" emits in the bias current. internally, the ecu has something like a 220-2k resistor, either connected to a subcircuit, or directly to the one of the logic ic's. when the sensor is unplugged, it will reach 5v (and this initates the code) or it will be like .5-2v nominally, and the logic will do nothing, or it will reach zero, from a short in the wire, and initate an opposite fault condition. if there is a subcircuit, it may output to the logic ic the same info, therfefore storing the same code and any data exactly the same if it is a short, or an open circuit fault.
i fail to see how matching the knock sensor shunt resistor value, and eliminating the peizo feedback would not effectively bypass the sensor.
... i fail to see how matching the knock sensor shunt resistor value, and eliminating the peizo feedback would not effectively bypass the sensor.
Which would make you the first person (well, on this forum at least) to successfully bypass a knock sensor!
Since bypassing the knock sensor could destroy the engine in short order, do you know WHY any engineer would design it with a shunt resistor?
P.S. I was kidding about the "go ahead and try it." If your hypothesis could possibly work, I wouldn't want anyone running their truck with the risk of unlimited knocking.
i can't recall the sources for all my info, but i took the liberty of googling some things. first is a pic of what is inside this style...... actually, following the image, i found an article on them...... http://www.azosensors.com/Article.aspx?ArticleID=50
next, i remember seeing images where they have been bypassed......
directly out of a diagnostic bypass thread on another forum.
i certainly agree that you never want to run the engine down the road and it wouldn't work in newer intuitive engine modules, especially bolt-through ones. *(in fact, i recall those ones have diodes, which require more voltage than a true rms meter puts out to even register outside of a diode check, but i won't go into that) however, if you put in the correct size resistor and this old efi ecm sees a proper voltage drop, it should return the timing to normal. i prefer this test at times because you could have a deteriorated piezo mounting and good shunt resistor and it would freely be allowed to detonate.
in case you are wondering, i've been repairing (sometimes designing and prototype testing) electronics and vehicle circuitry since the mid-90's and have been a mechanic for 16 years, not counting schooling.... so every once in a while i'll have a unique tid-bit.
Last edited by AKHeathen; 09-02-2015 at 11:10 PM.
Reason: bbcode
Since bypassing the knock sensor could destroy the engine in short order, do you know WHY any engineer would design it with a shunt resistor?
proper bias voltage to eliminate electrical feedback, much like an old hi-fi microphone. plus, it gives the computer a better way to monitor the circuit.
You’ve suggested that owners with the “dreaded” code 52 can clear that code, and return the timing to “normal,” by disconnecting the knock sensor and replacing it with a suitable sized resistor (one that matches the DC resistance of a good knock sensor). You describe (without ever seeing) the internal circuitry of the ECU that detects that resistor. You concede, of course, that such hypothetical circuitry would only detect an open or short in the knock sensor circuit, and would not detect “a deteriorated piezo mounting” or a broken piezo crystal or anything else that would interrupt the sensor output.
First, does Toyota monitor any sensors that way? Yes, they do. Code 21: “Open or short in heater circuit of heated oxygen sensor for 0.5 sec or more.” Codes 22, 24, 31, 32 and 41 are all similar. Toyota knows how to detect an open/short in a sensor, and they clearly describe it in the manual. Each of those codes could be “bypassed” with a suitable resistor, but you’d be silly to do that.
Okay, how about code 52? “With engine speed between 1,600 rpm - 5,200 rpm, signal from knock sensor is not input to ECM for 6 revolutions.” Sure doesn’t sound like the ECU is looking for a resistance, does it?
When you said “however, if you put in the correct size resistor and this old efi ecm sees a proper voltage drop, it should return the timing to normal”, it made me think you’re not clear on what the knock sensor is there for. It’s not some sort of emergency brake that is only expected to be activated zero or one time in the life of the vehicle, and which can be “reset” with a resistor. Instead, the ECU is constantly advancing and retarding the timing to get it just right: advancing until it gets a signal from the knock sensor, and then retarding until the signal just goes away. You always have the maximum advance possible. A resistor won't provide that signal.
What happens if the knock sensor fails? (short, open, or just a broken piezo crystal) As the ECU advances the ignition, it knows about where it “should” get s knocking signal, and if it gets too far past that it realizes something is broken. It’ll give it 6 revolutions to be sure, then throw the code.
Well, if a resistor wouldn’t bypass the knock sensor, could some sort of AC voltage source fool the ECU into thinking the knock sensor is still there? No, no more than a 0.9v voltage source could fool the ECU into thinking a missing O2 sensor was there. The ECU isn’t looking for a static condition (resistance, voltage), it’s looking for a change in response to ECU inputs. It’s the failure to see the change that triggers code 52 (and 14, 25, 26, 42 and so on).
To be clear, I had nothing to do with the actual design of these trucks, so I have to rely on Toyota publications and common sense understanding of how circuits work.
Why do I care? While 90-95% of code 52s can be traced back to one component (the pigtail), there are still 5-10% that turn out to be much more difficult to fix (member AZSpeed claims he couldn’t cure that code until he replaced all the capacitors in the ECU!) Some of those members are close to desperate for a solution, and if they thought a simple resistor would be a panacea they could risk damaging their engine chasing after a non-solution.
I think we all SHOULD make suggestions for fixes, even some that are long shots. We have the combined knowledge of all the members to help filter the sure-fire fixes from the well-give-it-a-try fixes. Even better, we have some very smart members (like yourself) who are willing to give something a try and report back on what they found (the resistances, the part numbers, and so on). But when you start a post with “let's just review exactly how the knock sensor works:”, and follow it with something that is anything but exact, it makes me, well, sad. When you’re not sure something will work, consider starting your post with something like “Has anyone tried measuring the resistance of a knock sensor, and then bypassing ....”
So please keep on making suggestions from your own experience, but try to a) tie it back to something from Toyota (we have the FSM online, so that’s mostly easy), and b) please don’t say something is “fer sure” if you’ve only tried it 2-3 times. Or not at all.
08-31-2015, 08:59 AM
