Did you disconnect the battery while performing the service? If not, I would do that so the ECU can start "fresh" with the clean TB.

 

Yes, hopefully at a minimum you had the key out of the ignition when fiddling with the throttle body.

If you attempt to work the throttle by hand from under the hood with the ignition on, or engine running, you're in for big trouble with the electronic throttle control system.

However, you can work the throttle plate by hand all day long provided the power is off.

Joel

 

Any update on this issue? We're you able to solve the issue?
i just cleaned my TB and now engine is running rich with short FT about 14% at idle.
i am not sure if idle learn procedure would work in this case.

 

Sorry for bringing this thread back from the dead, but I am also having the same issue.
Cleaned the throttle body and now when am at park and i tap the accelerator the RPM's goes up then comes down then goes up again and comes down to 750-850 RPM's. and I also feel that little kick from the throttle when coming to a stop (feels a bit harder to stop).

Does anyone have a solution for this?

I have a 2006 EX

 

I believe in the throttle body there in a small almost pin sized hole that is the idle control sensor (bottom surface of TB). If plugged your idle may search. Either way, when you step on the gas, in park, the idle will go up and come back down. It will also "search" as it settles to the preprogrammed idle as it cannot just instantaneously return to 750-850 rpm. The surge could be an electronic issue as the throttle is electronically controlled, not mechanically controlled. Checking that is best left to one who knows how to run an ohms meter, knows the values required and what they stand for. This is worse case scenario...could also be that the idle is high and that can cause the surge at stop as well. Same fix...clean TB, make sure idle sensor clean & clear, also disconnect battery to allow computer to reset & relearn and lastly checking resistance of the circuits.

 

I'm going to bump this thread.

I cleaned my TB today and am experiencing the same thing. I cleaned it as per the factory service manual. I removed it and cleaned it with a shop towel soaked in throttle body cleaner. Reinstalled and am seeing about 100-150rpm over the target idle on my scan tool. Target should be 650 and I'm seeing 700-800rpm along with the surge when the engine decelerates. It drops to about 1100 rpm, then surges to 14-1500 rpm and then slowly drops to about 800rpm.

I've tried all the relearn steps that are posted all over the Internet. When I look in my scan tool it says that the idle relearn is "complete" even after the battery has been disconnected and before you even start the car. I don't think disconnecting the battery does the same thing as resetting the PCM with a bidirectional scan tool.

Another post on a Civic forum I found talks about throttle position sensor A & B and that they should be equal. If they're not then the car needs a TPS calibration. Mine are A=0.71V & B=1.57V. However, it's my understanding that the TPS B sensor is just for error detection so not sure why it would matter if they were both exactly the same. Even still, you need a full bidirectional scan tool to do a TPS calibration.

I'll keep an eye on it and see if it eventually fixes itself or not.

 

Interesteing, I've reset idle on many Odyssey/Accord/Infiniti vehicles all the same way without the scan tool and no issues at all, no idle hang, no idle surging.

Disconnect battery positive for 20 min, turn on headlights to fully drain.
Reconnect only after turning off lights
Start vehicle with absolutely no load at all and let idle for 10-15 minutes.
Turn off, wait a few seconds
Turn back on and start, go drive around and no idle hangs or odd behavior and no CEL

 

...

Disconnect battery positive for 20 min, turn on headlights to fully drain.
Reconnect only after turning off lights
Start vehicle with absolutely no load at all and let idle for 10-15 minutes.
Turn off, wait a few seconds
Turn back on and start, go drive around and no idle hangs or odd behavior and no CEL

Tried this exact procedure. I'm anywhere from 50-150 RPM above target idle and can't get rid of the surge on engine deceleration.

Frustrating that one can't perform simple routine maintenance without this kind of goofiness. I see lots of reports of this exact symptom so I know I'm not the only one. I hope it goes away on its own.

 

I've done that with exception of removing battery negative rather than positive and I didn't leave the headlights on the whole time the battery was disconnected. I did turn them on and off while disconnected.

I'll try it again using your exact steps. I'm at a loss as to what to do if it doesn't fix itself. I don't think I damaged the TB in any way by following the service manual to clean it. I manually opened and closed it while off the vehicle to clean it. I'm not real keen on cleaning it on the car with a brick on the gas pedal. If that plate closes with your finger in there it will darn near take it off.

Carguy above metions that there is a small hole for the idle but there is no such small hole. It looks like the electronic throttle plate controls the idle rather than an IACV.

 

What did you remove and/or what were the steps you took to clean the throttle body. ( Be as detailed as you can)

 

Pretty easy job really...

1. Unplugged MAF sensor
2. Unplugged MAP sensor
3. Removed air cleaner box top
4. Removed coolant lines from bottom of TB and plugged them
5. Unplugged 6p connector from TB
6. Removed 2 nuts and 2 bolts holding TB to intake.

Took TB over to the bench and used blue paper shop towels wet with CRC throttle body cleaner to wipe down inside of TB until clean. I manually opened and closed the throttle plate while cleaning inside the TB and around the edge of the plate. I was careful to open and close it slowly and not let it snap shut. I also wiped down the inside of the intake manifold the same way.

Then I simply reversed the steps to put it back on. I've got no MIL and fuel trims are at 0-1% at all rpm ranges so I don't think I have any leaks.

Could the VCMuzzler affect the idle relearn process? I can't imagine it would but is something I've thought about.

 

Dont mind me too much, i am attempting to keep lineaer with you

1. Normal but not needed
2. Normal but not needed
3. Needed so yes I see why on this
4. Normal but not needed (NOTE: You should bleed coolant whenever a coolant line is disconnected, not doing so can cause a bit of an idle flux in some cases)
5. Normal but not needed
6. Normal, however did you use a new gasket when reinstalling throttle body? (If you did not this may cause a slight air leak, enough to flux a little)

Careful with the blue shop towels as sometimes they tear and could "jam" the throttle body slightly. Good job on the complete cleaning, did you ensure to line up the little tabs on the intake pipe to air box and airbox to tb when reinstalling and ensure all connections are good and tight?

A slight leak will cause these conditions you have but won't immediately trigger a CEL either. It'll take a day or two in order to trigger. The only time you'll cel on these vans for a air leak is if the leak is severe. (Dont ask me how i know........stupid mistake)

 

I did not bleed coolant as I only lost a few drips. I pinched the lines with vice grips and then immediately plugged the lines when they came off. I had one of the same lines of when I adjusted my valves a few months back and didn't have any problem then. I left the throttle body attached when I removed the intake for valve adjustment.

I didn't use a new gasket. The gasket looked perfect so I don't think that's the problem but I'll probably pick one up and replace it just to rule it out.

I didn't leave any shreds of the blue towels in the plate. I checked and double checked before I put it back on.

The tabs on the flex pipe and the throttle body line up perfectly and the clamps are tight. I've had that stuff off on numerous occasions.

If it were just the slightly high idle I'd suspect an air leak. The surge on engine rpm deceleration is what's weird. Would an air leak cause that?

 

1. Gasket = IT could cause a bit of an issue along with coolant in some small cases, Manual calls to replace that gasket every time but due to experience, I do understand what you mean when you say it looked perfect too. I've had on a few occasions where the gasket looked perfect but when replaced the car ran better. The part is cheap too so its cheap insurance

2. Awesome haha, i had one time where a shred was left and now i triple check that every time before use

3. Hmm okay, yeah those are annoying and on a few occasions i've had the clamps get tight but then after a few hours they worked a bit loose. Resulted in a bit of a vibration at idle (A/C used masked it by raising idle)

4. High idle would show on a bigger leak but a small one unlikely, also you'd hear a air rushing noise. Surge on decel looks like would be cause from a leak or now that I think of it too, there may be some amount of cleaner still present that was lodged in crevices.
Think about it, you rev up and the MAF see's (x) amount of air coming in and thus throws fuel so you make power. You let off so the throttle plate shuts (But not all the way since it controls idle, so it opens up a bit) and there is still a in rush of air (albeit unmetered or metered slightly) and then the air flys into the intake and down into the heads. Thus the o2 sensors may see it and throw more fuel for a split second thus instead of coming down smoothly and resting, we now come down, and throw back up for a split second since air then settles.

I'd inspect the TB gasket, bleed coolant just to be absolutely sure, check any external vac sources such as the line on pass side (PCV to intake) and the air line from valve cover to air intake hose for any leaks. I'd even go so far as to remove the intake entirely (If we want to try and rule out the throttle body gasket) and then reinstall the intake. While we have the intake off i'd give the bolts/nuts on the throttle body another 1/4 turn just to ensure we have good seal at that point too. Then reset ECM again and fire up

 

Well, I've checked everything again this morning and can find no leaks. I haven't replaced the gasket yet but inspected it again and it's perfect. I sprayed water and TB cleaner around the gasket looking for any RPM change while running and can't affect any change in RPM or in O2 sensor data. I'll get a gasket later today or tomorrow, if possible, but I'm pretty sure that's not it.

Coolant level is perfect and there doesn't seem to be any air in the system.

The last thing I'd want to do is remove the intake an introduce a new variable. The intake was put back on with new gaskets a few months ago when I did the valve adjustment. There were no idle issues after I did that job.

Watching my scan data, if I rev the engine to 3000 RPM, I can see that the target idle throttle valve angle will show something like 4.8 degrees. When I let off the gas, the target angle goes up to maybe 5.4 degrees and then slowly drops down to it's idle setting of around 2.12. This tells me that the PCM is re-adjusting the target idle throttle valve setting during engine deceleration, causing the surge I'm seeing. The throttle valve is doing everything it's being told to do--I just don't know why it's being told to do that. I wish I had a way to do the TPS calibration with a bidirectional scan tool. I hate to pay to have that done only to find out it doesn't fix the problem, though if you read the thread I linked above, two separate people on the Civic forum claimed to see the same issue and the calibration fixed it.

 

Do electronic throttle have hefty spring to close the throttle?

 

Yes. It's internal to the electronic portion.

 

John Clark, hope you can fix it. I had the same problem after cleaning the TB last year and I ended up bringing the car to the dealer and have the TB replaced. I drove the car for a week with the RPM surge thinking it takes time to relearn and while I was searching the internet up and down for a solution, but after a week the issue was still there that's when I told myself I have had enough.

 

Well, that's not good to hear. I just have a hard time thinking the throttle body is damaged from a simple cleaning via the method the service manual says to use. I'm going to try the TPS calibration, I think. One of the guys on the Civic forum said he used Foxwell NT510 had the capability and it fixed his. The last thing I need is another scan tool but I'm not spending $1000 to have the dealer replace my throttle body. Checking Bernardiparts, though, a new OEM throttle body is only $100 but it would likely need calibrating after installation too. I just bought an MD802 scan tool but the Foxwell NT510 supposedly has bidirectional capabilities that you purchase on a "per car" basis so that might be the way to go. It's on Amazon so I could probably return it if it doesn't work.

 

Looks like good idea! I drilled down and it looks like extra manufacturer costs $90 each. If this scanner works, it is not a bad deal at all for people like us who only have to maintain few makes of our vehicles. Please keep us posted on your opinion of Foxwell NT510 I would love to hear how you rank it as compared to MD802.

 

Well, I bit the bullet and ordered the Foxwell NT510 from Amazon Prime with 1 day delivery. I bought it based on this post, as well as the Amazon item description, and Foxwell Website that says it does bidirectional function tests.

Let's hope this works. A new throttle body is less expensive than the scan tool so if it doesn't work I'll order a new throttle body. This is mind boggling. Some say after a month or so it fixes itself but I'm not convinced. We'll see I guess. It's only money, right?

The MD802 works well but is what got me into this mess. I did the throttle body test in the service manual, looking at the Rel TB Sensor angle, which showed my throttle body was above the desired angle at idle and needed cleaning. I should have left it alone, I guess, but at 150K it was bound to need it sooner or later. I've seen cleaning recommended at half that mileage.

Why does trying to save money cost so much?

 

And then there's this:

Attention: 8th Generation Civic, 3rd Generation City and Accord owners (DBW related) - 180157 - Page 2

I spent a significant amount of time researching BEFORE I did the cleaning and didn't find half the stuff I'm finding now.

 

Interesting. No one in that thread, however, has explained why the system doesn't come to equilibrium over time.

 

Well, I just got back from driving it about 40 miles of stop and go, highway, and freeway driving. No change. If the vehicle is rolling the idle will stay up at about 1500rpm. When you stop it will slowly come down to about 800rpm with the AC on. It should be 700 with AC on. The car will stay at 30-40mph without your foot on the gas. Very annoying.

I sure hope the TPS calibration is an option on the Foxwell scan tool and that it solves this problem tomorrow. $100 for a new TB isn't bad but I've seen reports of people replacing the TB's with new ones and having the same problem.

 

I hope it works for you.

The link you provided came from a forum in Pakistan. The person who wrote it up got it from his dealer. If you want find trustworthy source, can you ever do better than Pakistan or a car dealer?

So please don't blame yourself for missing that link in the past.

 

Yeah, I know it's not the most reliable information. However, it would have been enough for me to decide not to bother cleaning it. It's just strange. There's a guy on YT who cleaned his electronic TB on an Accord (looks identical to this one and on the same engine) the same way I did, bolts it back on and starts it up and it works fine. The results are just all over the place.

 

OK, well I think I've figured this whole thing out. Everything that everyone reads and says on the Internet is wrong except for the one 25 year tech on the Civic forum. The Pakistani guy is also partially correct. he's correct in that cleaning a very dirty TB will give you the problems he notes (and I experienced) afterward if you don't reset the TPS data in the computer BEFORE doing a relearn procedure. He's wrong in that ONLY the dealer can do it. It does take a bidirectional scan tool, however.

I received my Foxwell NT510 today (Thank you Amazon Prime one day delivery!) installed the software on my laptop, registered the device, and downloaded my one free car software--Honda software. That all went very smooth. If anyone is interested in this scan tool I suggest Amazon as noted above.

When I plugged into the car with it I was surprised (well, not very) how it was almost identical to the Autel MD802. However, I had a lot more options in the OEM data lists than on my MD802. There was an option to reset the PCM/ECU. This is what the factory service manual says to do. However, the service manual states that if you use a generic scan tool to reset the PCM/ECU then you have to relearn the crankshaft position sensor (CKP) data after you do the idle relearn since the generic scan tool resets everything. The HDS has a special reset just for the throttle body. So, while it's nice that it has that feature I didn't use it. CKP sensors do go out occasionally so that may come in handy down the road. It's a pretty easy procedure to relearn CKP but I wanted to see if it would do it without having to go that route.

There is another menu item called TP Sensor Calibration check. I selected that and it asked if I wanted to check the TPS stored data. I said yes and it then asked if I wanted to reset the Throttle position data. It even gave a warning that said if I didn't clean the throttle body before doing this that I could experience problems. I went ahead and told it to clear the data. It quickly said it was done and for me to now clean the throttle body. That was it!

I turned off the key and then turned it back on again and did the idle relearn procedure that is found in the manual, as well as all over the Internet. 3000 RPM for two fan cycles or 194F degrees, then let it idle for 5 minutes. If the fan comes on during the 5 minutes you're not supposed to count that time toward the 5 minutes.

After doing the 3000 RPM and then closing the throttle to let it idle, my idle dropped right down to 650RPM and was perfectly smooth. I pulled up the idle learn complete/incomplete data PID and the idle learn still said incomplete. I let it idle the requisite 5 minutes (not including fan time.) Once done, I turned off the key for a few moments, then fired it back up and checked the data again. To my surprise it still said the idle learn was not complete. My thought was that it's possible this is where the VCMuzzler can screw things up. I think the computer is looking for 194 degrees on the ECT 1 for this idle learn cycle to complete as I've read numerous statements in the manual about looking for temps on ECT 1. So, rather than pull the VCMuzzler (and setting a code) I just pulled the radiator fan fuse and ran it until I got 194 on ECT 1. It was about 215F on the ECT 2 but, heck, my radiator fan fuse was burned out for who knows how long anyway. I just let ECT 1 hit 194, then dropped the engine to idle and plugged back in the fan fuse and let it idle another 10 minutes. If I were to do this again I'd remove the VCMuzzler before starting the whole TB cleaning and relearn procedure. Then once everything is done and the car is cold again, reinstall the VCMuzzler.

During the 5 minute idle time I noted a glitch on the NT510 where it was showing the low speed fan control in the ON condition the whole time even though the low speed fans were cycling normally about every minute or so. I plugged in my MD802 and it showed them cycling and showed the idle learn as completed. I plugged in the NT510 again and it showed the low speed fan on continuously and the idle learn still incomplete. I'm guessing this is a glitch in the Nt510 software, however, the perfect idle results speak for themselves. I wish I had plugged in the MD802 right after I reset the TPS data to see if showed "not completed." Maybe next time.

So, in the end, everything works perfectly now. The idle is perfectly smooth at about 650 +/-15 RPM, well within spec which is +/-50.

It is my argument that there is NO way to perform an idle learn without clearing this TP data first, either with the TPS Sensor data clearing or a PCM reset via the scan tool. My theory is that most of the people reporting on the Internet are cleaning their TB's long before 150K when I did. Mine had a pretty good build up in there and there was a significant change in airflow when I was done. If you clean it more often there won't be as much correction in the computer for the carbon build up and you may not have the problem I encountered. That's my argument as to why some have the problem and others don't. It's pretty clear to me, though, that an idle relearn can't really happen without clearing that data.

I contend that if you clean the TB regularly, like every 30K miles you probably will get very little correction in the computer and you will have no issues after cleaning. If you wait 150K like I did then the correction has to be reset. A battery disconnect does NOT clear this TPS data as I tried it numerous times--for as much as two hours.

So, while it cost me $166 to learn this lesson I'm not unhappy (though it did stress me out for a couple days.) I extremely pleased that I was able to get past it myself without having to go to the dealer. I'm sure it would have cost me $100 or so just to have them reset and relearn and so for another $66 I have a tool that does LOTS more than just a scan tool, even the MD802. It has all kinds of tests for activating components like fans, evap solenoids, AC clutch, and much more. It even has options for replacing PCM/ECU's. Additional manufacturers are only $90 so if I get in a bind and need something more than the MD802 for my GMC or my VW I should be able to have it covered for $90. Even without the specific car software it comes with OBDII preloaded so you can scan any car with it. You just don't have the OEM data unless you purchase it for that manufacturer.

There is a lot of "noise" on the Internet about TB cleaning. What I've learned is that 99% of it is wrong. I've got to try and thank the 25 year tech on the Civic forum that posted about clearing this data, as well as the guy who said he did it with the NT510. Those two posts were the solution to the problem. I just added in the detail to explain it all.

 

OK, well I think I've figured this whole thing out. Everything that everyone reads and says on the Internet is wrong except for the one 25 year tech on the Civic forum. The Pakistani guy is also partially correct. he's correct in that cleaning a very dirty TB will give you the problems he notes (and I experienced) afterward if you don't reset the TPS data in the computer BEFORE doing a relearn procedure. He's wrong in that ONLY the dealer can do it. It does take a bidirectional scan tool, however.

I received my Foxwell NT510 today (Thank you Amazon Prime one day delivery!) installed the software on my laptop, registered the device, and downloaded my one free car software--Honda software. That all went very smooth. If anyone is interested in this scan tool I suggest Amazon as noted above.

When I plugged into the car with it I was surprised (well, not very) how it was almost identical to the Autel MD802. However, I had a lot more options in the OEM data lists than on my MD802. There was an option to reset the PCM/ECU. This is what the factory service manual says to do. However, the service manual states that if you use a generic scan tool to reset the PCM/ECU then you have to relearn the crankshaft position sensor (CKP) data after you do the idle relearn since the generic scan tool resets everything. The HDS has a special reset just for the throttle body. So, while it's nice that it has that feature I didn't use it. CKP sensors do go out occasionally so that may come in handy down the road. It's a pretty easy procedure to relearn CKP but I wanted to see if it would do it without having to go that route.

There is another menu item called TP Sensor Calibration check. I selected that and it asked if I wanted to check the TPS stored data. I said yes and it then asked if I wanted to reset the Throttle position data. It even gave a warning that said if I didn't clean the throttle body before doing this that I could experience problems. I went ahead and told it to clear the data. It quickly said it was done and for me to now clean the throttle body. That was it!

I turned off the key and then turned it back on again and did the idle relearn procedure that is found in the manual, as well as all over the Internet. 3000 RPM for two fan cycles or 194F degrees, then let it idle for 5 minutes. If the fan comes on during the 5 minutes you're not supposed to count that time toward the 5 minutes.

After doing the 3000 RPM and then closing the throttle to let it idle, my idle dropped right down to 650RPM and was perfectly smooth. I pulled up the idle learn complete/incomplete data PID and the idle learn still said incomplete. I let it idle the requisite 5 minutes (not including fan time.) Once done, I turned off the key for a few moments, then fired it back up and checked the data again. To my surprise it still said the idle learn was not complete. My thought was that it's possible this is where the VCMuzzler can screw things up. I think the computer is looking for 194 degrees on the ECT 1 for this idle learn cycle to complete as I've read numerous statements in the manual about looking for temps on ECT 1. So, rather than pull the VCMuzzler (and setting a code) I just pulled the radiator fan fuse and ran it until I got 194 on ECT 1. It was about 215F on the ECT 2 but, heck, my radiator fan fuse was burned out for who knows how long anyway. I just let ECT 1 hit 194, then dropped the engine to idle and plugged back in the fan fuse and let it idle another 10 minutes. If I were to do this again I'd remove the VCMuzzler before starting the whole TB cleaning and relearn procedure. Then once everything is done and the car is cold again, reinstall the VCMuzzler.

During the 5 minute idle time I noted a glitch on the NT510 where it was showing the low speed fan control in the ON condition the whole time even though the low speed fans were cycling normally about every minute or so. I plugged in my MD802 and it showed them cycling and showed the idle learn as completed. I plugged in the NT510 again and it showed the low speed fan on continuously and the idle learn still incomplete. I'm guessing this is a glitch in the Nt510 software, however, the perfect idle results speak for themselves. I wish I had plugged in the MD802 right after I reset the TPS data to see if showed "not completed." Maybe next time.

So, in the end, everything works perfectly now. The idle is perfectly smooth at about 650 +/-15 RPM, well within spec which is +/-50.

It is my argument that there is NO way to perform an idle learn without clearing this TP data first, either with the TPS Sensor data clearing or a PCM reset via the scan tool. My theory is that most of the people reporting on the Internet are cleaning their TB's long before 150K when I did. Mine had a pretty good build up in there and there was a significant change in airflow when I was done. If you clean it more often there won't be as much correction in the computer for the carbon build up and you may not have the problem I encountered. That's my argument as to why some have the problem and others don't. It's pretty clear to me, though, that an idle relearn can't really happen without clearing that data.

I contend that if you clean the TB regularly, like every 30K miles you probably will get very little correction in the computer and you will have no issues after cleaning. If you wait 150K like I did then the correction has to be reset. A battery disconnect does NOT clear this TPS data as I tried it numerous times--for as much as two hours.

So, while it cost me $166 to learn this lesson I'm not unhappy (though it did stress me out for a couple days.) I extremely pleased that I was able to get past it myself without having to go to the dealer. I'm sure it would have cost me $100 or so just to have them reset and relearn and so for another $66 I have a tool that does LOTS more than just a scan tool, even the MD802. It has all kinds of tests for activating components like fans, evap solenoids, AC clutch, and much more. It even has options for replacing PCM/ECU's. Additional manufacturers are only $90 so if I get in a bind and need something more than the MD802 for my GMC or my VW I should be able to have it covered for $90. Even without the specific car software it comes with OBDII preloaded so you can scan any car with it. You just don't have the OEM data unless you purchase it for that manufacturer.

There is a lot of "noise" on the Internet about TB cleaning. What I've learned is that 99% of it is wrong. I've got to try and thank the 25 year tech on the Civic forum that posted about clearing this data, as well as the guy who said he did it with the NT510. Those two posts were the solution to the problem. I just added in the detail to explain it all.

Ran i to the same issue after doing a Throttle Body cleaning on my Odyssey. The scanner was showing idle relearn pending as well. Googled the forums and found your post. You are bang on, can’t redo Idle relearn with resetting PCM unless you have HDS and also need to remove muzzler until idle is set. Thanks for the post. It served as a confirmation before I started to loose my marbles.

 

Now that it's working normally and I'm able to watch the scan data while working normally, and it's a bit more obvious why it was doing what it was doing. Basically the TB was working just fine. The TP Sensor A and TP Sensor B don't have to be equal, as they aren't equal now.

I notice that normally when you rev up the engine it drops back down to about 1000-1100, stops, and then smoothly drops down to it's target idle speed of 650 or 700 depending on engine load (AC, lights, etc.) When I was watching the scan data graphs on those two TP sensors yesterday I would push on the accelerator pedal and let go. They would jump up, drop down, then a small jump up would happen followed by a slow decline down to the predetermined voltage. After the TP data clear and re-learn I see the same pattern but the ending idle TP voltage is lower, meaning the throttle plate is closing more.

So, basically, the TB was doing exactly what it was doing before I cleaned it but the throttle plate was still open the same amount only now letting in more air because the carbon was gone. The surge I was seeing was happening at a higher RPM but should just be the short pause in the deceleration, followed by a smooth deceleration to idle. Now that it's calibrated properly it does just that. That means that opening and closing it manually off the car didn't hurt it in the least.

Hopefully, this is all helpful to someone and I'm not just rambling. I like to understand why things are the way they are so that next time I'm smart enough to outsmart it!

 

Even though I have a 1999 van, I appreciate your write up. I would be needing this fundamentals if I ever decided to touch my throttle in 2012 Acura TL.

I am seriously considering NT510 but a the last count I had about eight to ten of the various scanners :-(

 

...I am seriously considering NT510 but a the last count I had about eight to ten of the various scanners :-(

Yeah, I hear ya. I have an Innova 3130 that I've had for 6-7 years and it has worked well. It's easy to use and shows live data for generic OBDII. However, recently I decided I wanted a better scan tool and after much research I decided to purchase the Autel MD802. It's a nice scanner that does OEM data for pretty much all makes and models but has no bidirectional capability. Since usually that capability is only found in $800+ scan tools I decided the MD802 was the best for what I could afford.

Once I got that I found all kinds of things I could do with it, including checking the throttle body relative position which checks for carbon build up in your TB. I found I was almost double the spec TB angle given in the service manual so I decided I better clean it. That's how I got into this whole mess.

That's when I found the post about the NT510 having the capability and with a low cost. However, the MD802 cost me $235 and the NT510 cost $171 with shipping so I'm at close to $400 in scan tools just in the last few weeks. The Autel DS708 is a full bidirectional tablet scan tool that does most all the manufacturers and is what I originally wanted. Only problem is it's $800-900 and $500/year for updates if you need newer year model vehicles. If I add another one or two vehicles to the NT510 I'm getting close to that DS708 price range.

So, in hindsight I should have just bought the DS708 that I wanted and had all my bases covered. It was just hard to justify at the time.

However, for those who don't have one and are looking for a scan tool for their personal cars I highly recommend the NT510.

 

Thanks for posting the detailed info. Wish these things were easier to work on! For what its worth, I clean mine out once a year as part of my maintenance routine. No issues but I do mine a bit differently. I turn the ignition key to on, put a brick on the gas pedal to open the throttle body, spray it out and turn it off, then let it sit. Start it up and no fuss, no muss. I don't know if that's the 'proper' way to do it but it works for me.

 

That's one method I've seen. However, the factory service manual specifically says to remove it for cleaning and to NOT spray cleaner directly on the TB. There is molybdenum grease on the shaft bearings that they don't want removed.

As far as whether you do it using the brick method or the removal method it really doesn't matter for calibration. It's a safety issue. If that brick slips off or something happens and that plate closes it will take your finger off. It's a VERY strong motor and, from what I've read, has an ice break mode where it's strong enough to break ice if it thinks it has built up. I think what matters on the calibration is how much it has built up when you clean it. If you clean it yearly then you're never getting enough built up for the computer to correct for it. Subsequently, when you clean it you're not really removing enough carbon to change the calibrated airflow. Yearly might be the best method if you don't have the scan tool to reset the TP sensor data. If you do it that way, though, I'd recommend not spraying cleaner directly on the TB plate and shaft. Wet your rag and just use that to wipe down the TB. That's how the manual says to do it and it cleaned mine up quite easily using that method.