2003 Civic VTEC engine cuts out above 3400 RPMs

Tiny
HBSB2CARPROS
  • MEMBER
  • 2003 HONDA CIVIC
  • 4 CYL
  • FWD
  • AUTOMATIC
  • 92,000 MILES
2003 civic vtec engine cuts out above 3400 rpms, all check engine, seat belt and key lights on dash stay lit.
Do you
have the same problem?
Yes
No
Saturday, January 1st, 2011 AT 7:00 AM

11 Replies

Tiny
KHLOW2008
  • MECHANIC
  • 41,815 POSTS
Get a scan done to retrieve trouble codes and let me know what they are.
Was this
answer
helpful?
Yes
No
Saturday, January 1st, 2011 AT 2:12 PM
Tiny
HBSB2CARPROS
  • MEMBER
  • 19 POSTS
Honda dealer says their $12,000 computer can't even sense or read the ecm to obtain codes.
Was this
answer
helpful?
Yes
No
Saturday, January 1st, 2011 AT 6:17 PM
Tiny
KHLOW2008
  • MECHANIC
  • 41,815 POSTS
Have you tried to use generic scan tools or manually try to retrieve for trouble codes?

At the DLC, use a jumper wire between terminal #4 and #9. Turn ignition on and note if the MIL blinks any codes.

When scan tools are not able to communicate with the PCM, perform the following test to see if you can come up with anything.
Was this
answer
helpful?
Yes
No
Sunday, January 2nd, 2011 AT 3:03 AM
Tiny
HBSB2CARPROS
  • MEMBER
  • 19 POSTS
I purchased one of those Honda civic engines from Japan from a distributer in the states, and had a mechanic install it. After his installation, the check engine light stayed on and engine cut out over 3800 rpms. The mechanic said he couldn’t do it but Honda Dealership had to re flash the ECM. Honda's diagnosis disagreed and found that my mechanic needed to swap Japanese engine's wire harness w/ removed engine’s harness. According to the engine distributer and Honda”Japan engine wire harnesses are not compatible with US civic engines." Honda also found a bad Vtec solenoid and a missing pressure sensor (which is normally located beneath Vtec solenoid in US engines). This engine also was missing a pressure sensor (not required for engines in Japan) and their harness didn’t have wires for pressure sensor required for US civic engines. After these corrections were made, engine ran fine for about 2 months. Until the latest symptoms appeared (check engine, seat belt and key lights stayed on. After Honda diagnosed this second complaint, the TAC stopped working. Can’t imagine why or what they did or maybe it’s just coincidental? My mechanic uses Mitchell’s online software and procedures and couldn’t find the problem. Honda’s dealer techs said they also checked for ground and voltage problems. Yet, neither shop has been able to find the problem. But, you think if I try your recommended tests that I should be able to find the problem?
Thanks. At my wits end.
Was this
answer
helpful?
Yes
No
Sunday, January 2nd, 2011 AT 6:33 AM
Tiny
KHLOW2008
  • MECHANIC
  • 41,815 POSTS
I need to get this question out to prevent it from hogging the forum.

Please do not reply to this question.
Was this
answer
helpful?
Yes
No
Sunday, January 2nd, 2011 AT 12:46 PM
Tiny
HBSB2CARPROS
  • MEMBER
  • 19 POSTS
I performed the measurement steps above. Steps 1-5 were ok, and says to then jump to step 10. The last step says "o volts" should be indicated by measuring between 5 & 7. I got 1.7 volts. Does that mean anything?
I was able to borrow a scanner, but it still was not able to read anything in the ecm. It says to check connections. I don't understand why it can't read (see) the ecm, when all the other tests check ok, accept for the 1.7v. How are we to know if the ecm needs updating? I guess the shop has to update if required? If I get wrecking yard ecm, won't I still have to have the dealership do something to it, like program the keys and anything else? Will the car still run in safe mode after initial install of newer ecm, before dealer reprograms?
Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 4:05 AM
Tiny
KHLOW2008
  • MECHANIC
  • 41,815 POSTS
If the 1.7 volt is achieved at Step 13, it indicates there is a short to power in the wire between DLC terminal # 7 and ECM/PCM (E23).

The test results for step 13 indicates if there is 0 voltage, wiring circuits are good and it could either be the ECM requiring updating or replacement. If the scan tool cannot communicate with the ECM, updating would not be possible.

If you replace the ECM which does not have the VIN input and immobilixer code rewritten to it, you would not be able to start.

When getting the ECM from the wreckers, the key for the same ECM has to come with it to be able to start the engine.
Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 5:51 AM
Tiny
HBSB2CARPROS
  • MEMBER
  • 19 POSTS
I did get a good check for continuity between pin 7 and e23(while unplugged from ecm) Should I check and see if there is a short between pin 7 & any other pin in e23 plug if there are no shorts between pin 7 and plug e23, where else should I be checking? I was able to subscribe to alldata as well. So with this and your help and the borrowed scanner hopefully i'll make headway to a solution. Thanks
Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 5:09 PM
Tiny
KHLOW2008
  • MECHANIC
  • 41,815 POSTS
Try checking for continuity between pin # 7 and ground and other terminals of the the ECM. Try unplugging sensors at the engine if there is any indication of continuity.

Btw did you try the manual bridging of the DLC terminal # 4 and # 9 to test if you are able to get the MIL to blink or stay on with ignition switch turned on?
Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 9:31 PM
Tiny
HBSB2CARPROS
  • MEMBER
  • 19 POSTS
Been reading blogs online, lot of info suggesting jumped timing belt. Can't understand why, if this is the problem, why scanner can't read the ecm. What is the best way to check if timing belt has slipped a tooth. Are these civics what they call an interference design, which means a timing belt issue could destroy the valves?
Was this
answer
helpful?
Yes
No
-1
Monday, January 10th, 2011 AT 12:34 AM
Tiny
KHLOW2008
  • MECHANIC
  • 41,815 POSTS
You need to remove the upper timing belt cover to check the markings. Valve cover would have to be removed.

Yes, it is an interference engine and the rev limiter cutting the engine at 3400 rpm can be due to the timing belt skipping, a faulty ECM or a bad TPS. Without communication between the ECM and scanner ot MIL indicating, speculating is not going to get you anywhere.

You would have to be able to communicate with the ECM to now if there are any sensor or control faults. Based on what has transpired, I believe the ECM is the most likely cause of it.
Was this
answer
helpful?
Yes
No
Monday, January 10th, 2011 AT 3:04 PM

Please login or register to post a reply.

Sponsored links