Jump to content

5th gen - Low voltage on MAP sensor - replace gray ECM connector?


vfr800_red

Recommended Posts

  • Member Contributer

Got some great fuel injection diagnostic advice here recently, so I figured it would make sense to check back here. 98, 5th gen VFR. My FI light is on, solid, not blinking. I reset the system, it went off, but after riding some more, it has re – appeared. Following the procedure in the factory service manual, I have one blink, which is a MAP sensor. Per the manual, page 5-14, the voltage between pink and ground is within range, as is the voltage between pink and green/orange, but the voltage between light green/yellow and green/orange is low, 4.15 V, when it should be 4.75 to 5.25 V. Per the manual, this is either and open or short circuit in the green/yellow wire, or loose or poor contact on the ECM connectors. Because I’m not getting 0 V, I don’t think it is an open or short circuit. The contacts are clean, and I sprayed them with a little tuner wash, but that did not help. It would seem I have loose contact at the ECM connectors, and the way to fix this is to purchase a wiring harness from a salvage yard and use it to replace my existing gray ECM connector by cutting its wires and soldering. But maybe I’m completely wrong. Any advice is appreciated.

Link to comment
Share on other sites

  • Member Contributer

don't kill messenger! MAP sensor rarely fails.

 

Double-check and verify its output by using rubber-T and tap into MAP sensors's vacuum-line at intake manifold. Measure actual vacuum at idle with warmed up engine.

 

Measure 5v reference-voltage at ECM itself. If not within spec there, may be problem with ECM. If 5v reference-voltage is not same at MAP-sensor connector, there's wiring issue in between.

 

Don't cut harness to repair, will introduce too many additional variables. Most of time, issue is with resistance at terminal ends. So replacing wiring in between doesn't solve issue. Inspect connector terminals on both end closely with magnifying glass. Clean all connectors with OxGard 1st and re-test voltages at both ends. Also measure resistance of wires end-to-end.

Link to comment
Share on other sites

  • Member Contributer

As Danno rightly states. FIRST make sure all the vacuum hosing to the MAP sensor is good. No cracks, breaks, kinks, blockages or leaks.

 

Your MAP, BARO, IAT and TP Sensors rely on a stable 5v VCC from the ECM this is on your Pink Wire and confirmed as good.

You have also confirmed the Signal Ground being the Green/Orange wire is good.

 

Confirm you have good continuity ie zero ohms for the Light Green/Yellow wire between the MAP connector and the ECM connector.

 

After thoroughly checking the vacuum hoses and the Light Green/Yellow wire for continuity. Suggest you apply some CRC 2-26 or WD-40 or a small amount of Ox-Gard to the pins and sockets of the MAP sensor. And maybe a light spray of WD-40 or the CRC 2-26 at the ECM plug for the MAP wires. Don't recommend using "Tuner Wash" for sensitive electrical connections!!

 

Failing that, fit a new MAP sensor. Or swap it with the Baro sensor.

 

Good Luck.

Link to comment
Share on other sites

  • Member Contributer
11 hours ago, vfr800_red said:

but the voltage between light green/yellow and green/orange is low, 4.15 V, when it should be 4.75 to 5.25 V. Per the manual, this is either and open or short circuit in the green/yellow wire, or loose or poor contact on the ECM connectors. Because I’m not getting 0 V, I don’t think it is an open or short circuit. 

When doing this test, disconnect ALL sensors: MAP, BARO, IAT, ECT, TPS. These all share green/orange signal-ground wire and if there's issue with these sensors or their wiring, it would affect readings at MAP-sensor connector. 

 

If MAP-sensor connector test is OK, re-connect those other sensors one-at-a-time to see which one pulls down voltage. Then inspect that sensor, its connector and wiring closely.

Link to comment
Share on other sites

  • Member Contributer

A little embarrassing but the problem was the two tubes which connect the bottom of the airbox to the starter valve were disconnected. Re – connected those the trouble code has gone away. Actually, it is surprising how well it ran when those were not attached. Haven’t ridden much lately because of health issues but a couple times did have a sudden loss of power when these were disconnected.

Thanks guys!

Link to comment
Share on other sites

  • Member Contributer

Good news thanks for the feedback. Glad you have the VFR's health back to full strength, and it was a simple fix.

Hope you can get your own health back to 100% asap, keep covid safe.:fing02:

Link to comment
Share on other sites

On 7/25/2021 at 1:37 PM, vfr800_red said:

Actually, it is surprising how well it ran when those were not attached.

While you're learning stuff... the explanation for why your bike was still able to run pretty well with a MAP sensor problem is contained in Chapter 21 of the Service Manual.  In this magical chapter all is revealed, including the fact that the ECM (Honda PGM-FI) only uses the MAP sensor during low throttle openings (i.e. small throttle position sensor opening angles).  As you wind on more throttle the ECM switches modes to a pre-programed fueling method (usually referred to as Alpha-N).  This mode does not reference the reported MAP sensor value because at higher RPMs small engines (like motorcycle engines) deliver wildly chaotic MAP sensor values.

Link to comment
Share on other sites

  • Member Contributer

  

On 7/26/2021 at 4:42 PM, GreginDenver said:

While you're learning stuff... the explanation for why your bike was still able to run pretty well with a MAP sensor problem is contained in Chapter 21 of the Service Manual.  In this magical chapter all is revealed, including the fact that the ECM (Honda PGM-FI) only uses the MAP sensor during low throttle openings (i.e. small throttle position sensor opening angles).  As you wind on more throttle the ECM switches modes to a pre-programed fueling method (usually referred to as Alpha-N).  This mode does not reference the reported MAP sensor value because at higher RPMs small engines (like motorcycle engines) deliver wildly chaotic MAP sensor values.

 

It's actually other way around. If you look at datalogs of MAP-sensor, it varies tremendously at idle, with swings of over 100% between cylinder pulses.

 

uc?export=download&id=17249G51j1LHzum_nz

 

As load & airflow increase, manifold-pressure stabilizes.

 

uc?export=download&id=1QBa6PRcOv6XWNuSBU

 

Problem here is that it's TOO stable under high-loads, and MAP-sensor output doesn't match actual air-flow or load. For example, at 60% throttle and 5000rpm, MAP-sensor output may be 1.27v. Then at 100% throttle and 10000rpm, MAP-sensor output will be similar 1.19v. Even though actual air-flow volume and load on engine is dramatically different.

Link to comment
Share on other sites

16 hours ago, DannoXYZ said:

It's actually other way around.

Actually...

Chapter 21 of the VFR800 Service Manual says...

 

 

Screen Shot 2021-07-27 at 12.54.40 PM.png

 

And then Chapter 21 goes on to say...

 

 

Screen Shot 2021-07-27 at 12.55.07 PM.png

Link to comment
Share on other sites

  • Member Contributer

Manual writers. They've never been wrong before with translating technical info into Engrish...

 

Sounds a lot like finding info about planet Earth and universe by going to some old book written long time ago. When you actually physically measure the planet, it shows a completely different picture...

 

One only has to physically touch the bike and use instruments to measure things to get real empirical data. Major issue with ITB is not having large enough manifold after throttle-butterflies to combine vacuum from all cylinders to average them out. Have you connected oscilloscope (or even basic voltmeter) to MAP-sensor and measured its output at various loads & RPMs yourself to see what it's actually doing?

 

There's also many ways to do alpha-N blending, You can combine them with addition of MAP vs. TPS tables. Or you can multiply them. Also have a mixing-table in between with at sliding-scale 0-100% weighting from each. There's also separate blending of fuel vs ignition-maps. Lack of technical detail really doesn't help. They don't say what these tables represents, is it sensor-output response? Is it injector pulse-widths? Ignition-advance?

 

uc?export=download&id=1byxZ9sqcIo7dRKzVi

 

Link to comment
Share on other sites

  • Member Contributer

Here's some further references.

 

http://support.moates.net/theory-alpha-n/

Quote

4. In conjuction with Speed-Density and some kind of blending algorithm.  This approach is often used with very large camshafts that pull little vacuum at idle.  Basically, TPS and MAP are allowed to contribute varying amounts to the overall load calculation.   Net result: more stable and meaningful load index close to idle when MAP sensor readings are unstable.  Found on the Electromotive TEC3 among others.

 

 

http://electromotive.com/wp-content/uploads/2017/04/TEC3r-Manual-1.11.pdf

Quote

- TPS/MAP Blend  


  Engines with high-overlap camshafts or throttle-per-cylinder setups (Hilborn-style) often experience erratic MAP sensor operation at low RPM’s; particularly around idle.  The main problem is that these engines will either produce a non-consistent or weak MAP signal at low RPM’s.  To compensate for this, the TPS/MAP Blend feature should be used.  This feature allows the low, constant voltage of the throttle position sensor to be blended with the erratic, weak MAP signal at low RPM’s.  In effect, it “tricks” the TEC  into thinking that the engine is operating at a lower manifold pressure (higher vacuum) with a consistent MAP sensor voltage.

uc?export=download&id=1nzu9s8KXeZiAMac05

At low-RPMs where MAP is unstable, TPS contributes 65% to load-calculations. By 4000-rpms, TPS weighting is 0% and MAP is used 100% for load-calculations.

 

 

http://www.msextra.com/doc/pdf/Megasquirt3_TunerStudio_MS_Lite_Reference-1.5.pdf

Quote

uc?export=download&id=12wLrxnLfbl1Jvne9y

At low-load/low-RPMs when MAP is inconsistent, TPS is primarily used. As load & RPMs increase, TPS is used less and less. VE-table has lower low-load zones set for TPS and upper high-load ones uses MAP.

Link to comment
Share on other sites

I have worked with MegaSquirt/MicroSquirt fuel injection conversions since 2005.  Yeah, I know there are many ways to combine parameters to achieve good FI.  Yes, I think Honda PGM-FI is a bit old-school compared to modern day FI ECUs but it got the job done for a whole bunch of Honda products for a lot of years.

Link to comment
Share on other sites

But here's a question: Have you done the "Flapper Mod" on your 5th Gen VFR800?  If the answer to that question is "yes" then we should not discuss anything vehicle-related (it just won't be productive).

Link to comment
Share on other sites

  • Member Contributer

Nope, prefer to leave everything stock. Honda really did good job on EFI systems for these bikes.

 

Were you using an Arduino to decode MAP pulses to simulate cam-sync signal to do sequential fuel/ignition without needing cam-sensor on Megasquirt?

Link to comment
Share on other sites

2 hours ago, DannoXYZ said:

Nope, prefer to leave everything stock. Honda really did good job on EFI systems for these bikes.

 

Were you using an Arduino to decode MAP pulses to simulate cam-sync signal to do sequential fuel/ignition without needing cam-sensor on Megasquirt?

No, both of my FI project bikes are "batch" injected.  I am using an Arduino to sample a MAP sensor attached to each throttle body, the Arduino then sends the lowest value to the MicroSquirt via a DAC.  This effectively simulates the vacuum stability of having a large central intake manifold.  This setup isn't my creation, a guy named Dennis who lives in Germany thought it up and he was nice enough to help me copy it.  If I remember correctly Dennis tried sequential on his 750 race bike but it didn't deliver any improvement in performance (for a lot of extra effort) which I think is a common conclusion people reach when they're dealing with a motorcycle engine.  The RPMs are so high it makes a mockery of the attempt to impose the higher order that is sequential.

Link to comment
Share on other sites

  • Member Contributer

Ah yes, that's what it was! Found some of your post from years ago about that.  

 

I'm working on turbocharging 2-stroke twin now and gonna try getting around inaccurate low-speed MAP signals by using large automotive-style intake-manifold connecting both cylinders and using single throttle-body. If still not good enough, I'll switch to draw-through MAF. We'll see... :)

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...

Important Information

By using this site, you agree to our Privacy Policy.