Wire Shielding EMI or RF Noise?

[novassdude]

I would try a filter on the lines going to the tach and the data logger. Otherwise triple check all ground wires.

[Jackie McCracken]

Originally Posted by FED 387

try grounding the head to the block and the block to the frame or body with the old style braided ground straps--NEVER use solid core wire for ground wire --if you do use braided wire make sure it is of sufficient gauge and not only crimped but put a bit of solder on the wire and connecter crimp area to make sure it is really grounded ---

nothing wrong with solid wire and electrically it's actually better if done right. however, having said that stranded wire will not usually break because of vibration which solid wire will do, like a paper clip bent back and forth it will break at some point.


Some friends that run a sprint car had a similar goofy problem, make sure the distributor body is also properly grounded as the gasket will act as an insulator. You cannot have too many grounds, the more the better, including block, heads, intake, AND distributor.. and use the biggest wire you can physically get away with as modern ignitions pull a **** load of current. 12ga or bigger!


Jackie

[west coast]

So I have had the same problem as Bobby, my data logs for the rpm is very squiggly. I tried almost every thing listed so far and gave up a couple of years ago on this. One thing that I did do was go back to my tach which is a 2 channel with drive line sensor, down loaded it and my logs looked clean on RPM. one thing that was mention by Tom I am going to try. if anybody gets there's fix please post how you did it.

[Bobby Fazio]

Originally Posted by west coast

So I have had the same problem as Bobby, my data logs for the rpm is very squiggly. I tried almost every thing listed so far and gave up a couple of years ago on this. One thing that I did do was go back to my tach which is a 2 channel with drive line sensor, down loaded it and my logs looked clean on RPM. one thing that was mention by Tom I am going to try. if anybody gets there's fix please post how you did it.

Jim, I take it you are using the dual sync also?

I have the Autometer playback and the rpm looks better on there only because the frames per sec are much less than the FAST datalog. The tach gets its signal from ignition box, which gets its signal from the FAST ECU, which gets its signal from the distributor that I believe is the cause of the issue. I can see my tach needle start jumping around at higher rpm and when I use the playback feature on the actual tach I can see it as well. Putting a resistor in-line with the tach could fix the needle jump but it will not prevent the ECU from seeing a messy RPM signal which causes erratic timing and fuel decisions. I firmly believe a crank trigger is the only way to go with these aftermarket EFI setups.

[Signman]

Bobby
Have had the same problem with using the stock distributor with bank to bank and the MSD dual sync with sequential. Talked with David no longer with FAST and Jeff my trusted chassis dyno and tuner it is the distributor being driven by the cam seeing harmonics from valve springs etc. At least with a Ford it is driven on the front as opposed to SBC which is in the back and sees it all.


Crank Trigger cleans up most of the noise and provides more accurate timing would be the fix.

But racing stock it's a no go so do your best to shield the noise and deal with it. We can chase our tails with electronics to be more accurate therefore efficient therefore fast but still we need airflow to add fuel to tune.

Stock is not Stock
Stock is way cool
Stock Eliminator Serious Fun

[west coast]

Bobby yes i am using FAST dual syn dist. As far as my tach goes I dont use the replay on the tach just down load it to my computer via Tach facts software from autometer. I will check on the sample rate of both for a comparison will let you guys know what i find there.

Novadude Fast system datalogger is part the fuel injection ECU. So there is not a way filter it. will double check the schematic. let me know if I am wrong, may learn something new.

Signman are you saying David Page is no longer with fast, if so what happened and where did he go?

[Bryan Broaddus]

I've seen a few things I can't agree with on this thread. Just for a reference I have run an Electromagnetic Interference Test lab for the last 35 years or so and am a Certified EMI/EMC Engineer.

Solid wire is NOT as good a ground as stranded wire, Actually you should use 1/2 inch or larger braid for your ground straps. The stranded wire has more surface area which in turn leads to less inductance. Inductance is then enemy of a good ground. It opposes current flow and will develop a voltage across its impedance. Not good. You want the absolute lowest impedance you can get on your grounds.

Keep your grounds as short as possible. again this will limit the impedance of the ground and increase its effectiveness.

Too many units on a ground does NOT lead to R/F noise. If the ground is solid, like the negative battery post, then terminating all the returns there is a good idea. Now, if the ground you pick is a crappy ground and you stack up a bunch of devices on a bad ground then you can couple energy from one unit into the other via the ground.

The plastic clamp on Ferrite beads are a good idea on wires that you are afraid of coupling signals into. You can clamp then on the ends and even put a few on the wire spaced out evenly. This will not effect the desired signal on the wires but will soak up the unwanted RF energy that is coupling on the wires. Ferrites are much more effective on high impedance wires, like your tach lead but do not work much on low impedance leads like power wires. They can still have some effect but not nearly as much.

Shielded cable are only shielded if the ground is solid on both ends. In other words if the two units do not share a common ground then you may develop a voltage across the shield which can couple noise into the core wire under the shield.

You cannot ground something through an Anodized or painted case. So make sure to clean any anodizing or paint off of the area you are attaching the ground to.Clean bare metal is the best place to ground.

Adding extra grounds will not hurt you. So grounding the heads to the block and the block to the chassis is a good idea. Also for any ground you want a large CLEAN contact area. Threads, like through a bolt are notoriously bad grounds. You must rely on a large flat contact area.

[Dissident]

Mr. Broaddus,
I enjoyed your post. I wonder why more folks do not look at fabricating Faraday cages around either emitters or receivers to help defend against RFI/EMI problems. Your comments on grounds were spot on and unfortunately the average folks in the pits do not discern any difference in grounds accomplished by solid wires or braided cables. They are much more prone to apply whatever "bubble pack" solution might be at hand and most of those are boondoggles of one sort or another. Thanks for bringing some scientific methodology into the discussion.
Regards to All that like this kind of stuff,
HB2
Dissident

[Signman]

Originally Posted by west coast

Signman are you saying David Page is no longer with fast, if so what happened and where did he go?

There was a major announcement last week.
Edelbrock bought Compcams a while back.
David moved on to a leadership position at Edelbrock believe heading a team that looks at trends and future projects and products.

[Bobby Fazio]

I’m using copper tape as a shield on the cam signal, crank signal, and their shared signal ground. All 3 wires come out of distributor together along with 12v power and I have the shielding grounded at one end only. I separated the 12v wire right out of distributor as close as I could and put a ferrite bead on it. I also put ferrite beads on water pump 12v, alternator charging wire, and twisted the coil pos/neg wire from ignition box to coil and put one bead on the twist. Am I using them in the right spots? Should I put ferrite beads on the already shielded signal wires also? Thanks.