[gator68428]

I recently upgraded to the QH from the Tweecer because of intermittent relay flipping and errant MAF calculation reading (MAF kg/hr goes exorbitantly high occasionally randomly and car sputters way rich, while MAFV logging normally). With QH it's doing the same thing. After messing with the connection to the J3 port it seems its just a contact issue that can be improved somewhat after wiggling chip on J3 port. Mind you i have cleaned the J3 port several times; it was clean for the QH and Tweecer. But problem is still there, just in different degrees of intensity depending on chip position. The history of my setup though is that i was using the F3 Moates chip before the Tweecer or QH and had plugged and unplugged the chip numerous times each re burn, over the years. My guess is that this has worn down the electrodes and is why i'm getting intermittent connections between the QH and Tweecer.

So my thought is that if i can make better connection at the J3 port these issues will go away and my ECM is likely fine? Anyone solve anomalous issues with better contact?

To be sure I just purchased a reman ECM with a fresh set of electrodes (hopefully fresh or at least better than my existing one...) and now won't be plugging and unplugging nearly as much since i have a QH and it should set me straight for a while.

Anyone second this logic? Seen contact issues at J3 port cause anomalous signals and relay flipping?

[gator68428]

BTW, the F3 STILL works flawlessly when i put it in. It has the snuggest fit on the J3 port. But the QH is a close second (and brand new; F3 is 5+yrs old), with the Tweecer having the least tight of a fit. So I'm a bit surprised the QH has issues similar to Tweecer...

The existing ECM I got from Mass-FLO EFI (now Pro-M Racing) back in 2006. I've heard leaky capacitors etc. could be an issue leading to intermittent issues, however, b/c of the flawless operation with the F3 (no relay filpping or hiccups of any kind), it seems the ECM is likely fine.

So beware of worn electrodes on the J3 port (my best guess as to the culprit so far). An explanation is that these circuit boards weren't necessarily designed for repeated plug-in and -out cycles, so try to minimize it if you can.

[cgrey8]

I'm not saying your problem isn't the J3 connection, but what you are describing sounds more like misconfiguration in the tune.

For instance, functions SHOULD ALWAYS have the top and bottom value in the 1st column represent the largest and smallest possible value representable and the values in between must always flow from largest to smallest (1st column only). There was a post recently talking about an exception for the MAF curve on older EECs that have a 16v entry for the top value. However I, personally, will be keeping my tune with a 16v entry in the MAF curve.

Go through every function in your tune (including scaling functions) and make sure they all adhere to this requirement.

Then go through the scaling functions you've modified and make sure you aren't entering values larger than the strategy supports for the 2nd column.

[gator68428]

Your suggestion is easy enough to check by looking at the calibration file. But we still have to reconcile the fact that the F3 running the same tune runs flawlessly.

MAF function largest x value is 16V.

Cgrey, would you mind looking at my calibration file?

If it is the J3 port electrodes wearing out this is valuable information that should be disbursed and my help many people.

[gator68428]

I went through and found that two functions had the max x value not at max allowable value:

fuel accel enrichment vs TP, max x was 5.08; changed to max 5.1
fuel wot multiplier vs rpm, max x was 8000, changed to max of 16384

Other than that they all looked good. All had values flowing in increasing direction from min to max x.

Will update tune with this and see what happens. But at minimum i will be adhering to best know practices to reduce chance for issues. Thanks.

[cgrey8]

Make sure you also check the minimum possible values. They can be tricky since you might think an RPM column would be minimum of 0 however minimum allowable is something like -8128 or something like that. How could you possibly have negative RPMs? Obviously, you can't under normal conditions. But if a corrupt value proliferates through the system, not representing those negative values in that function could send the processor "off-into-the-weeds" well beyond the area in memory where that function resides. So to limit the damage of corrupt values as they are processed, you need to have all possible values represented in your tunes from top-to-bottom.

I'm at work and don't have access to BE right now. But I'll try to remember to look at it when I get home.

As for the J3 port wearing out, it may get dirty, but unless you are plugging and unplugging devices multiple times a day, it's not likely it would wear out. If you remember, Atari and Nintendo cartridges used edge connections just like the EEC has and when they misbehaved, it was almost always due to dirt and corrosion. Back then, we used 70% Isopropyl (rubbing) alcohol solution and a Q-tip but today I would use a RESIDUE-FREE solvent.

I guess it is possible a solder joint could be weak and intermittently making contact and thus causing issues. But that's most likely to be on the tuning device. You won't fix that by cleaning the J3 connectors.

I've heard of people on the forum even taking sand paper to the J3 port. That seems awfully excessive to me. I've NEVER had to do that with an edge connection before. But if there is some kind of heavy buildup muck on the connectors that solvent just isn't getting off, I guess it is possible that would help. But I would make that a last resort kind of thing. And the fact that the chip works fine leads me to believe it's likely not the J3 port. Buildup that solvent couldn't get off would likely affect any J3 device.

[gator68428]

Let me clarify what I mean by "worn": Most contact electrodes have a thin coating of another metal to enhance contact reliability; gold is a very typical one. Once this layer wears off contact can become more difficult due to the less ideal properties of the underlying (usually much less costly) metal; typically Al, Cu or Mo. For example, oxidation can happen where an insulating layer is formed. This is something a solvent will not help with. But don't take my word for it, for further reading see excerpt below*. I was not implying the entire electrode was worn through to the board. This explains why sand paper can help, as it would abrasively remove oxide or corrosion buildup. But in the end it is merely a band-aid to a more fundamental issue.

If we're being logical here: because the tune worked flawlessly with the F3 means its NOT the tune. So check that off possible causes.

Cgrey8, do you happen to know that the J3 port and Atari and Nintendo games utilize electrodes with the identical composition and material thicknesses? Or were you merely pointing out your observation? FYI, manufacturers will not utilize any more costly materials than they deem necessary for a given use. I.e. I can guess that a game system meant for repeated abuse and plug-in and -out cycles would be designed with thicker surface coatings for longevity than a sealed J3 port that's intended for infrequent use.

*
“The intrinsic properties of the materials chosen for the contact are important for increasing the lifecycle of the contact. For instance, due to its low electrical resistivity and low sensitivity to oxidation, gold is widely employed as a contact material… In general, contacts are desired to have excellent electrical conductivity for low loss, … appropriate hardness to avoid material transfer and chemical inertness to avoid oxidation.” [1]
[1] P. G. Slade, “Electrical Contacts: Principles and Applications,” Second Edition (2014)

[cgrey8]

...Cgrey8, do you happen to know that the J3 port and Atari and Nintendo games utilize electrodes with the identical composition and material thicknesses? Or were you merely pointing out your observation? FYI, manufacturers will not utilize any more costly materials than they deem necessary for a given use. I.e. I can guess that a game system meant for repeated abuse and plug-in and -out cycles would be designed with thicker surface coatings for longevity than a sealed J3 port that's intended for infrequent use...

No I was making reference to the problems we always had with those consoles that forced us to always have to blow into them to get them to work. Or if they got bumped, that was enough to ruin a great game if the cartridge or console had some age on it. It got worse in a smoker's house. I wasn't trying to imply that the materials or the specs were the same. But the concept is the same.

[gator68428]

Well the "new" ECM came in and the results are in.

Visual inspection of the "new" one compared to the old one--the J3 port electrodes specifically: the new electrodes were uniform in color (appears to be Al contact material) and the old electrodes were definitely worn through the surface contact coating and were mixed copper and silver colored, likely due to the scrubbing over the years from the F3 chip, but were otherwise clean.

So i optimistically loaded the QH in to the new ECM and took it for a test drive. To my disbelief the thing was flipping the relay and hiccuping all the way down the street! Had to turn around 3 houses down and head back. And had to restart the engine couple times in the process.

And sure as the sun, the F3 chip worked like a champ after popping it back in.

Both the QH and F3 are running the same tune. This tune had the two updates I described as a result of Chris's recommendation. I also checked the tune for the min x values of all functions and they looked good (all were at the lower limit).

I'm happy to say I've divorced myself from the "worn" electrodes marriage I was in for the past week.

Now I'm all ears if anyone has any ideas. I may call Moates just to see what they recommend (why F3 and not QH???). BTW thanks Chris for your help so far. But I am officially baffled at this point and am desperate. Luckily i can use the F3 chip for the time being.

[jsa]

which mode & tune slot are you using on QH ?

I had a similar problem with my car not running on QH but running fine on f3.

Turned out the EEC side of QH was faulty in my case.

You say you bought QH because you had relay flipping with tweecer as well ???

Time to look for an intermitent wiring connection in the car, I think.

Do you have a Jaybird ?

[gator68428]

Mode 1 tune 1 and 4 for the QH. Have toggle switch between two tune options.

No Jaybird.

The Tweecer had issues in the past. I wasted hours of my life from the get-go with the Tweecer b/c of lifted pins on its memory IC. So it was easy to blame when it stopped working and F3 always did.

Maybe there is some hardware difference on the ECM side and more wires are being used in the harness when a dataloggable device is being used? If there are any differences that's where i should look. I don't know if there are...

What specifically was the issue with your QH?

[jsa]

Try Mode 2 Tune 5, all other Tunes cleared and no toggle switch.

It is possible to test the J3 side of QH with a Jaybird.

The QH was perfect on the USB side, but the J3 side output was a garbled mess, thing ran on half a cylinder half the time. Some sort of fault on the QH output hardware. Raised an RMA and it is sorted.

I'm not thinking about extra harness wires.

I'm thinking about four things.

First, when QH is in use, is the EEC sitting in exactly the same spot, have you moved the EEC to fit QH ??? If the EEC was moved the harness connections have moved.

Second, have you got a USB cable in QH while QH is fitted. Could be EMI picked up on the USB cable.

Third, same for toggle wiring - EMI

Fourth, connections in the wiring loom or to the relay, could be a bit loose, randomly make or break as they see fit.

[gator68428]

Just heard back from Moates. They said:

"You replace the capacitors? Both the QH and TwEECer draw a lot more power than a simple chip. Screwed up power supply caps will play into that."

Sounds reasonable. Right now that makes the most sense. Looking for a difference and power consumption is one of them apparently. Not sure why reman'd ECU would need new caps tho... But worth a look i suppose.

[cgrey8]

The replacement EEC, if freshly reman'd should have new caps in it already. Or was it just a different EEC, not a remanufactured EEC?

[cgrey8]

Looking over the tune, here are the things that caught my eye as things I found odd or just plain wrong.

• You've got the special idle scalars set to -79. I don't know that there's a problem with this, but I'm curious why.
• You also have the tune setup for forced OL mode. I'm assuming this is because you don't have HEGOs? Why not? If you are wanting this to be a daily driver, you really should have HEGOs to allow the EEC to be adaptive. Otherwise you just have a very expensive glorified carburetor.
• The MAF curve looks odd. It maxes out at 3.9v??? What kind of MAF is that? Most any MAF is going to be usable well above 4.5v...most go well into the 4.8-5v range. Even stock MAF curves generally go to 4.7v. So something seems just wrong here.
• FN331B is misconfigured. Top cell is 5.08, not 5.1. I don't know that this is causing a problem, but it is technically misconfigured.
• FN367, the Injector Offset curve, is stock A9L. My guess is that's not right. The stock A9L offset curve is pretty high for most aftermarket injectors. It was probably high for the injectors that came on the engine stock, but EEC tuning wasn't what it is today.
• FN320A doesn't look right either, but that hardly matters with a tune setup for Forced OL.
• FN1329 is different than stock. However being this is a MassFlo conversion, the firing order of your engine might actually be different. In fact, it probably is different. But I don't know your engine's firing order right off hand. But it is worth you at least explicitly indicating it so we can compare whether what's in the tune is correct. People that run Ford EECs on Chevys sometimes forget that Chevy numbers their cylinders different than Ford...and I think the #1 is even on the opposite side of the engine. But what goes in their tune is 1-8-4-3-6-5-7-2 (std V8 Chevy firing order) and that's not right.
• FN1327, you have all injectors acting as a single bank. While this is OK, it's not ideal. Ideal would be running 2 HEGOs so you can run DS and PS separately and thus get better fueling control. My stock Explorer 302 had DS and PS fueling almost identical. However my 331 stroker has about a .025 lambda difference between DS and PS. If I ran them as one bank, one side would always be richer than the other.

What's odd is that this tune works perfect on your chip, but like crap on a tuning device. I don't have a good explanation for you on this one. Since I've never run my truck with just a chip, it's always run with either a TwEECer or QH, I couldn't say with 1st hand experience whether there is a difference between the two. But I do have to say I'd be surprised if there were given the number of people on this forum that HAVE done that with no difference in performance/behavior.

[LeadHead]

We had extremely similar issues. Turned out to be a ton of interference from the ignition system. New cap, rotor, high quality ignition wires and installing some radioshack clip-on EMI ferrite cores onto the USB cable fixed the issue.

[gator68428]

Thanks for the input jsa:
-Shifted ECU slightly to account for QH protruding more... New mounting pos is same for QH and chip (now as in: when the mentioned issues were happening).
-It is a newly purchased Reman'd ECU. I agree, seems there should be fresh caps.
-I suppose i should unplug the USB cable take off toggle wires to see it is EMI noise

Thanks for looking over the tune, Chris, but I don't think it has anything do with this issue... But here are my responses:
-I don't think I've messed with special idle. May have been that way from Mass-FLO?
-Running open loop because i don't have the transients dialed in yet and lean tip in issue is annoying the crap out of me. It's not a daily driver anyways. Just to and from Track days and AutoX. Works very well for that. And i have a WB and can tweak the MAF whenever it needs it to keep cruise in the region of 14.5-15. A little rich on tip in makes for nice driveability. I disagree about the glorified carb statement. With the MAF, IAT, BAP and other sensors, can still factor in changing atmospheric conditions, its better than a carb. After reading over the literature and EEC programming document, Adaptive is not what most ppl think it is. The main benefit of adaptive is accounting for sensor drift over the life of the car and as a fudge factor for fundamental failing of the computer model... B/c I have WB and ability to tune the MAF, i can keep that (sensor drift) in check (while running in OL b/c lean tip in issues magnified by CL). And adaptive will not help in any way with transients.
-I cut the MAF off there (at 2000 kg/hr) because i will never flow more than 1500 kg/hr. If stray voltages arise and my fuel flow is limited to that (req'd amount of fuel for that amount of air), then so be it.
-FN331B i corrected to 5.1. Don't know why or how it was 5.08
-Thanks for mentioning about the injector offset curve. I uploaded that from spec file sent to me from Pro-M. Trusting it's right but no way to know...
-Firing order is correct otherwise i would know... Firing order for Pontiac is 18436572. Numbering convention for Pont. is driver side front to back (1,3,5,7), pass side front to back (2,4,6,8) --much more intuitive if you ask me ;)
-I have long tube headers and moderately radical camshaft and cannot use the narrowbands accurately with their recommended proximity to head. I proved this by running in CL and monitoring LAMBSE while logging WBO2 also. The narrowbands mistakenly think the engine is leaner at low loads than it is and add way too much fuel. Only under higher loads and higher rpms do they agree with eachother (this is when leaked intake charge into exh becomes less of an issue due to cam lobe overlap). So i scrapped the narrowbands. Luckily my WB has simulated narrowband output, and only have the one, so i sent that to bank 1 and set it up to run only off that one. CL and adaptive worked great under steady state cruise conditions, but made the transients worse. I was hoping to be able to switch between CL enabled and force OL tunes with toggle switch and QH but... That way i can flip to open loop if transients are annoying me then go to closed loop while cruising and back to OL on the track.

Thanks for the feedback Leadhead:
-It seems like EMI noise is where i should look next. I will try to unplug usb from QH and see what happens

[cgrey8]

...Running open loop because i don't have the transients dialed in yet and lean tip in issue is annoying the crap out of me. It's not a daily driver anyways. Just to and from Track days and AutoX. Works very well for that. And i have a WB and can tweak the MAF whenever it needs it to keep cruise in the region of 14.5-15...

That's reasonable.

...I disagree about the glorified carb statement. With the MAF, IAT, BAP and other sensors, can still factor in changing atmospheric conditions, its better than a carb...

Yep, better than a carb. Glorified carb.

...After reading over the literature and EEC programming document, Adaptive is not what most ppl think it is. The main benefit of adaptive is accounting for sensor drift over the life of the car and as a fudge factor for fundamental failing of the computer model...

That may be true as how the factory intended adaptives to work. But for DIY tuning, it has the added benefit of helping to dial-in error that the DIY tuner didn't smooth out in their tune. That may not be Ford's original intent to assist DIY tuners with "learned" feedback, but it works for that. To deny yourself the benefits of it purely on principle because you would be using it in a manner not exactly as intended is....well, I don't have a polite word for that. But given this is not a daily driver, I can also get that learning adaptives just aren't necessary.

...B/c I have WB and ability to tune the MAF, i can keep that (sensor drift) in check (while running in OL b/c lean tip in issues magnified by CL)...

This does mean that you have to monitor the engine on a regular basis. Again for a driver to the track-only vehicle, that's hardly an issue. But for a daily driver, that might get old and just not sustainable for some people. And BTW, I'm not brining any of this up to sway you from your strategy. I'm stating it more for the benefit of others that will read this post.

...And adaptive will not help in any way with transients...

True.

...I have long tube headers and moderately radical camshaft and cannot use the narrowbands accurately with their recommended proximity to head. I proved this by running in CL and monitoring LAMBSE while logging WBO2 also. The narrowbands mistakenly think the engine is leaner at low loads than it is and add way too much fuel. Only under higher loads and higher rpms do they agree with eachother (this is when leaked intake charge into exh becomes less of an issue due to cam lobe overlap). So i scrapped the narrowbands...

You are correct that with a radical cam, you cannot run CL at idle and low RPM/Load conditions for the reasons mentioned. However I wouldn't have scrapped them for the RPM/Loads that they can be used at. But that's just me. Again for a to-track only driver, what you are doing is probably just fine.

...Luckily my WB has simulated narrowband output, and only have the one, so i sent that to bank 1 and set it up to run only off that one...

Ahh, that's what you were going for.

...CL and adaptive worked great under steady state cruise conditions, but made the transients worse. I was hoping to be able to switch between CL enabled and force OL tunes with toggle switch and QH but... That way i can flip to open loop if transients are annoying me then go to closed loop while cruising and back to OL on the track...

What happened with that? That sounds like a perfectly reasonable solution. That and spending some more time on the transients to understand why they aren't behaving correctly as-is and then fixing them to work better.

...It seems like EMI noise is where i should look next. I will try to unplug usb from QH and see what happens

I never had this issue until I did my CNG conversion and soldered in tune-switching wires to a relay that would flip me between CNG and gasoline tunes. The presence of the wire that did this switching was enough to cause the CPLD on the QH to freak out and often would just plain lockup while driving down the road requiring I cycle the ignition key to get the engine to come back. Shortening the leads and wrapping that wire around a torrid ferrite bead solved that problem. It hasn't been an issue since. You might also try aluminum foil tape to help shield the exposed portion of the QH to see if that makes any difference.

[gator68428]

What happened with that?
[gator68428 wrote: ...CL and adaptive worked great under steady state cruise conditions, but made the transients worse. I was hoping to be able to switch between CL enabled and force OL tunes with toggle switch and QH but... That way i can flip to open loop if transients are annoying me then go to closed loop while cruising and back to OL on the track...]
That sounds like a perfectly reasonable solution. That and spending some more time on the transients to understand why they aren't behaving correctly as-is and then fixing them to work better.

That's my plan! BUT.... We're in this whole debacle b/c I can't get QH to run reliably.

Once i get over that hurdle (getting QH to work), I will keep working with the transients b/c I do believe it can be done for my app, just gotta keep tweaking!

I do appreciate a smooth driving vehicle. It is not purely a track car, but one I enjoy driving where ever it may be. I'm going for the Pro-Touring thing and would eventually love to compete in an Ultimate Street Car comp. Just had my first track day in the car this past weekend after working the bugs out of all the mechanical upgrades at AutoX's over the past few years. I'm hooked! It was awesome and the car is fast and forgiving and outran some trailered vehicles.

Also: Moates mentioned about checking my grounds. Will do that (again) b/c it makes sense that something needing more power (QH vs F3 chip) would be more sensitive to that.
But will def try everything relating to shielding/remove any EMI.

[gator68428]

I finally got back around to tinkering to try to get QH working without hiccups. I had the F3 chip in there over the past few months during two track days and two autocrosses and some cruising. Had plenty of non-computer relating tuning (susp, brakes, etc.) to worry about that took priority over messing with the QH. Good ole trusty F3 working like a champ as always. And the car held up to all the track abuse, which was nice. Now however, I have unlimited dyno access and need the QH for tuning.

Of all the feedback I've received so far, electromagnetic interference seems to make the most sense. Because the main diff between the F3 and QH is the bulkier chip and external wires that could act as noise antennas. So I took the below advice and started shielding everything.

We had extremely similar issues. Turned out to be a ton of interference from the ignition system. New cap, rotor, high quality ignition wires and installing some radioshack clip-on EMI ferrite cores onto the USB cable fixed the issue.

I re-installed the QH and added a ground wire to the case metal of the ECU, shielded the protruding edge of QH with Al foil (insulating tape first, so don't short pins together) and wrapped the USB wire and my toggle switch wires with foil as well; and made sure all foil was in electrical contact with the ECU case ground. Ferrite beads were already installed on the USB wire.

After a full weekend of dyno testing there were ZERO hiccups. After months of headache, only a half hours work of shielding completely fixed it! Thanks for chiming in LeadHead, you nailed it.

[decipha]

excellent, there was a fellow on the efi forum too that moved his usb cable to a permanently mounted location and began having issues with the communications dropping out during logs, he moved it back to just the temporary across the seat type install and it went back to functioning perfectly

shielding is very important with this devices and especially using a high quality usb cable such as those that are supplied with the qh