Solving RF Interference Issues in Sniper EFI Installations

[cmyer]

Perhaps you have encountered problems in your Sniper EFI installation and had someone tell you that you have radio frequency (RF) interference issues. And possibly your thought was, "Thanks a lot, but what does that mean?" Because I have a few of our customers who have faced this challenge, I wrote an article on our EFI System Pro Hangout blog that goes over the symptoms of RF interference, common targets & sources, and then tips for solving it. I thought I'd mention it here in case someone else might find it useful: Solving RF Interference Issues in Sniper Installations

As with much of what I write, I receive lots of insight & inspiration from this forum and the key contributors (I should at least mention Danny Cabral, S2H & REL3RD). I enjoy the fine-tuning of my understanding that is required to write a complete article. Further, I believe it's helpful for someone who is particularly beleaguered by a given subject to get the information in consolidated form. As always, if anyone sees anything that should be revised, or if you find value in the article, please post something in either location. :-)

[440roadrunner]

Well all RIGHTY then!! In addition, try to run signal wiring close to a ground plane, I.E. the firewall.
In stubborn cases, maybe a plastic car, you can buy shield braid and run it around the signal wires. Ground this and ONE END ONLY.
Don't disregard "wrenches in the works" like an alternator with a bad diode, which causes hash & spikes on the 12V bus.
It's not just EFI. A guy on another board was suffering interference between an MSD and the voltage regulator. The MSD caused the regulation to go nuts.
You mentioned rotor phasing. There has been cases where even factory distributors have phasing issues.
"Back in the day" before Al invented the internet, this showed up as cap carbon tracking. Some of us did not know much about that, then.

[cmyer]

Tracking Down RFI/EMI Issues

I recently collected some data from a customer's installation that graphically shows how dramatically EMI from the alternator can impact your EFI system. Most folks hardly give a passing thought to the alternator when trying to find EMI/RFI issues, but as you will see it can be pretty strong.

In this first picture, you see a great example of EMI riding on the RPM signal from the coil. Seeing this, you might immediately assume that because of the proximity to the ignition system and the obvious strength of the interference that it was coming from the there. And while that is a good guess, in this case it turns out to be incorrect. The light-green trace shows the throttle being depressed, and the bright red trace is the RPM increasing then decreasing with huge spikes. At the same time you see how this is affecting the ECU. The brown trace shows the ECU providing extra fuel for these fake RPM increases.

The customer didn't find anything suspicious in his ignition system, so he decided to do something that would quickly tell him the problem, he disconnected the alternator. As you see, that immediately eliminated all hint of EMI:

If RFI/EMI problems have you frustrated, consider reading my longer post on: Solving RF Interference Issues in Sniper Installations over on our website.

[Zmann]

Thanks Chris. This makes me wonder if Holley did anything to shield the ECU?
I have seen the Holley Sniper EFI assembly video (LINK), but can't see the ECU or cover.
I'm curious if some shielding like copper foil many electronics use inside the cover would solve some issues?

[HOTRODHARLEY66]

Yes, Holley absolutely needs to do something about shielding the ECU. With the size of the throttle body being so small, there is no reason why they couldn't make the front ECU housing bigger to accommodate some shielding method. I just went through a week of hell chasing every known path on a RFI issue on a Small Block Ford. The only thing that seems to have resolved it, is wrapping the plug wires on the distributor with ugly heat shielding. Looks stupid, but seems to work. The only way the car would run is with all the timing settings exactly the same. Either 15°, 15°, 15° at idle, cruise & WOT. Or even 20°, 20°, 20°. Or 25°, 25°, 25°. As soon as I changed the cruise & WOT timing settings to around 32° or 34°, the Sniper picked up serious RFI noise just from the timing moving as little as 10°. It's crazy how sensitive this unit is on a Small Block Ford with the distributor right out front. I'm going to say that if you intend to convert a small block or big block Ford, just save yourself the headache and buy the Holley Terminator unit.

[Danny Cabral]

I realize this shouldn't be you responsibility, but I'm sure I could fabricate a nice looking
shield that attaches to the ECU bolts with spacers. I'd first try the easier shielding tape.
https://www.redline-motorsports.com/products/rf-shield (Redline Motorsports RFI Shield - Sniper EFI)
Personally, I think something else is wrong: https://forums.holley.com/showthread...5559#post95559 (His original Sniper EFI forum thread.)

Just for reference: https://forums.holley.com/showthread...8868#post88868 (Sniper EFI Datalogger Interference/Noise Malfunction)
Sometimes it's in the alternator: https://forums.holley.com/showthread...050#post106050 (Related Forum Thread)
Sometimes it's in the distributor: https://forums.holley.com/showthread...9634#post89634 (Related Forum Thread)
It can also be caused by EFI wiring too close to high voltage components (spark plug wires, ignition coil, distributor, fuel injectors, etc.)
Read posts # 2 & # 4: https://forums.holley.com/showthread...-Datalog-Files (Related Forum Thread)

FYI: It's not safe to run an alternator with the wiring disconnected; remove the alternator belt (safe alternator test).
http://www.pvv.org/~syljua/merc/TooSeptST07.pdf (Alternator AC Ripple Voltage Test)
http://assets.fluke.com/appnotes/aut...e/beatbook.pdf (Fluke Multimeter Testing - Page 3 & 4)
https://forums.holley.com/showthread...0251#post50251 (Related Forum Thread)


View from 6:52 minutes onward for more information about the datalogger "Mark Data Points" feature: https://www.youtube.com/watch?v=BQdQ...ature=youtu.be (Holley EFI Proper Grounding & Wiring Practices - Devin Vanderhoof)
https://forums.holley.com/showthread...ta-Points-quot (Datalogger "Mark Data Points" Feature)
Additional information for others viewing this: http://forums.holley.com/showthread....2656#post32656 (Electronic Noise Interference, AM Radio Detection - Post #53)

[HOTRODHARLEY66]

I tried DEI gold shielding tape. It did nothing. I also fabricated a shield. It also did nothing. The only thing that worked was shielding on the plug wire ends on the distributor.

[cmyer]

Not to belabor this, but I received another customer's datalog today and it is another great example of interference. Here's what should grab your eye: Any time the throttle position is fixed, a change in RPM should roughly be mirrored by the manifold pressure. In the example below, you see that the throttle position is zero throughout. At the same time, the RPM is jumping around erratically between something around 1500 RPM & 600 RPM, yet the manifold pressure is generally following the top of the RPM trace (at or about 1500 RPM) and is ignoring the trips down to 600 RPM. This should tell you that, in reality, the engine is idling roughly along the top of the RPM trace and those bottoms are RFI/EMI ghosts.



It would be a bit easier to determine this in person, because you would look at the trace and immediate recognize that the engine wasn't surging like the RPM trace implies. The spikes on the duty cycle, however, are not make-believe. The ECU sees the changes in RPM and is actually making these huge swings in fueling to try and keep pace with what it thinks the engine is doing. Next step: Find & fix that RFI problem!

[cmyer]

OK, now I'm hoping someone might have some insights on one that has me stumped. You might have read post #3 and thought (as I did), okay, replace your alternator and you're all set. And the customer has. Three times. Twice with brand new 1-wire alternators and once with a Toyota alternator with external regulator. Each time, the same results: Noise on the RPM signal that is creating havoc. Each time, if you remove the alternator belt (or if you simply disconnect the alternator wire) the problem disappears. Reconnect, problem is back. So clearly it is related to the alternator.

Here is what we've tried and confirmed:
-Power/Ground for Sniper EFI System sourced directly at battery.
-Ignition wired correctly. No MSD CDI box, Summit Ready-To-Run distributor, yellow wire to coil (-).
-Standard 12 volt negative ground charging system. (Hey, the guys in Thailand so you check these things!)
-Voltage Drop @1500 RPM under load checked between Alternator output and Battery + Post (.38 volts, within spec).
-Voltage Drop @1500 RPM under load checked between Alternator ground and Battery − Post (.02 volts, within spec).
-All external electric circuits not required to run engine disconnected, problem persists.

I've called Holley Tech Support and Jason said he wasn't aware of anyone else with a problem like this. Any ideas?

[BrianSanDiego]

I don't know too much, but could it be that the distributor is putting out a bogus tach signal? Can he hook up an auxiliary tach (or already has one)? Might be his tach isn't playing nice with the alternator for some reason. If everything checks out and there's still noise on the power line, maybe try a +12V power filter on the feed to the Sniper EFI? The type you'd use on car audio equipment. I'd try to isolate wiring & components by possibly hooking up a second battery to run the Sniper EFI from, and jumping alternator power over to that, bypassing any of the existing wiring & grounding. Just ideas. Good luck.

[bingo2]

You might also try a dedicated ground for the alternator. Typically, as you know, the case is grounded through the block by the mounting hardware. However, if the block isn't grounded properly or the alternator mounting hardware doesn't provide a good ground path, there can be issues.

[BrianSanDiego]

Great idea bingo. It sure sounds like a ground loop issue.

[cmyer]

Thanks for the great suggestions. In addition, Jason at Holley Tech Support suggest that he move his alternator hot wire from the battery, back to the starter positive just to put some distance between it and the Sniper EFI power (sort of like what Brian suggested.) He already has the alternator power wire shielded, but I suggested also shielding the yellow wire to the coil (-). Jason also mentioned (as did Brian) trying a filter, so I suggested putting a capacitor in parallel with the Sniper EFI power/ground. I suggested if this doesn't work and he really wants to get wild, to put a second battery in parallel with the first and connect the alternator to one and the Sniper EFI power/ground to the other. Kind of outside the box thinking, but we just need something to change the pattern so we can get a clue where we're looking.

As I mentioned, the car & customer are in Thailand, but fortunately the customer is an English speaker or else this would be impossible. Thanks again for the suggestions, I'll post what we find.

[Firebob]

Was there ever any resolve to the RFI talked about in this thread? I'm experiencing almost the exact same situation. I did even try running a separate battery for the power feed for the EFI. The only thing that has seemed to make a difference is dropping off the alternator belt, as Danny mentioned in post #6. Doing that has allowed the engine to start & run.

[tdford5]

Same issues here I tried this out of desperation and it worked removed aluminum foil and truck stop running. So now to block this and look nice, any ideas/pictures?