EMI shielding material?

[shelburn61]

I'm having trouble with the electromagnetic interference from my MH ballasts disrupting my Aquacontroller. The ballast wires and AC comm wire all have to be run through a small whole in the floor down to the basement sump. The wires are not physically touching, but I am still getting interference. Can pvc drains conduct EMI?

What is the best material to use as a shield/sheath for the Aquacontroller communication wire?

[kcress]

Certainly PVC can do nothing for EMI.

Your comm wire should be cable that is sheilded twisted pairs.

You could also try a snap-on balun around the MH wires near the ballast to reduce the EMI leaving the ballast. Note: An example of a baluns are those lumps you see in cables like the cable into your LCD computer screen, etc. They make snap-on ones.

[stugray]

When you say AC "COMM" wire, wo you mean the serial cable to a computer or do you have the PX 1000 expansion box?

If you mean the serial cable, you need a "Ground Isolated Serial Connector".

Here's what kcress is talking about ( ferrite snap-on filters ):

http://www.mpja.com/prodinfo.asp?number=12805+FL

Stu

[shelburn61]

Stug,
I'm just talking about the standard telephone cable connecting the ACjr to the DC8. The Neptune guy said my only options were to physically separate the wires (won't work for me) or shield the phone wire...

So one of those filters on each ballast wire would block/absorb EMIs along the whole length of the wire?

[kcress]

Thanks stugray for that link.

[stugray]

Well, you could get some "shield braid" and pull the wire through it.

The stuff comes in rolls and looks like braided flat strap. When you push it together it opens up like chinese "finger-cuffs".

Its not easy to come by but I am looking at a roll of it, How much do you need.

Stu

[kgross]

The other thing do is if it uses normal phone cable replace with with a good twisted pair cable. Have a local electronics place use some CAT 5 or CAT 3 cable to make a phone cord out of, you will just have to have them install the RJ 11 connectors.

Kim

[stugray]

"you will just have to have them install the RJ 11 connectors."

I bought RJ-11 connectors, the tools & cable all at HD.

I'm not sure if HD carries shielded Cat 5/6 wire.....

Stu

[shelburn61]

<a href=showthread.php?s=&postid=13764497#post13764497 target=_blank>Originally posted</a> by kgross
The other thing do is if it uses normal phone cable replace with with a good twisted pair cable. Have a local electronics place use some CAT 5 or CAT 3 cable to make a phone cord out of, you will just have to have them install the RJ 11 connectors.

Kim

electronics place meaning radio shack?

[shelburn61]

I did some testing tonight. I'm running 2 identical Icecap 400w ballasts, but only one is causing problems. The wiring from the ballast causing problems is just 3 bare wires in fabric sheathing, but the other has heavy duty rubber-coated cord running from the ballast. Is it as simple as using the rubber-coated wire to kill the interference?

Stu,
I could use 6+ feet of the sheathing or should I just go with the Cat 5 cable?

[kcress]

I'd bet a buck just adding shield jacket won't work as the noise in this case is current based, and hence inductive, which cares not in the slightest about shielding.

[shelburn61]

<a href=showthread.php?s=&postid=13766623#post13766623 target=_blank>Originally posted</a> by kcress
I'd bet a buck just adding shield jacket won't work as the noise in this case is current based, and hence inductive, which cares not in the slightest about shielding.

You'll have to dumb things down for me a bit :) Are you addresing my last post or someone else?

[kcress]

Sorry shelburn61! I was referring to stugray's kind offer of shield braid. I doubt all the effort would be worth it.

'Current noise' is predominantly magnetic in nature. It couples into neighboring wires thru transformer action. The other form of noise is coupled in via mutual capacitance. Shielding works well for that.

[shelburn61]

kcress,

I'm running 2 identical Icecap 400w ballasts, but only one is causing problems. The wiring from the ballast causing problems is just 3 bare wires in woven nylon sheathing, but the other has heavy duty rubber-coated cord running from the ballast. Is it as simple as replacing the bare wires with heavy duty cord to kill the interference or is one ballast giving off more EMI?

The ferrite snap-on filters didn't help. Sheilding apparently won't work. Is my only option the Cat 3/5 wire??

[shelburn61]

DP

[hoosierpat]

You just need better phone wire. Get something that is twisted pair like any Cat wire. That is why they twist the wires inside around each other, it reduces crosstalk in the wires which is a form of interference. That may not be enough if the interference is high enough. I suggest a lead shield j/j.

[shelburn61]

<a href=showthread.php?s=&postid=13769820#post13769820 target=_blank>Originally posted</a> by hoosierpat
You just need better phone wire. Get something that is twisted pair like any Cat wire. That is why they twist the wires inside around each other, it reduces crosstalk in the wires which is a form of interference. That may not be enough if the interference is high enough. I suggest a lead shield j/j.

OK. does it need to be shielded or will regular Cat5 be enough?

[hoosierpat]

Hard to say. Finding well shielded wire probably won't be that easy. The wire is cheap though, just go to a big box home improvement store and you will be able to find regular twisted pair.

[shelburn61]

<a href=showthread.php?s=&postid=13770745#post13770745 target=_blank>Originally posted</a> by hoosierpat
Hard to say. Finding well shielded wire probably won't be that easy. The wire is cheap though, just go to a big box home improvement store and you will be able to find regular twisted pair.

yeah, just got to figure out how to get RJ11 plugs on there.
This is best I've found: http://www.amazon.com/exec/obidos/tg/detail/-/B000AA785A/ref=ord_cart_shr?%5Fencoding=UTF8&m=A3NMXBC4WIK5PI&v=glance

[kcress]

Just change your remaining snarl wired IC wire for the same jacketed cord the other one has. Probably cure your problem.

[DaveMorris]

A couple notes here:

Twisted pair wire has a minimal effect on outside interference.

Shielding in a cable does absolutely no good at all unless that shield wire is grounded properly. I doubt that the AC unit is wired for that or their own RJ11 cable would be shielded as well.

Rubber jacket has nothing to do with it either. The noise is probably coming from the ballast itself, not the wire. Can you swap wire on the ballasts? If you can, then you could see if the noise is associated with the cable or if it follows which ballast you are plugged into.

[shelburn61]

<a href=showthread.php?s=&postid=13771685#post13771685 target=_blank>Originally posted</a> by kcress
Just change your remaining snarl wired IC wire for the same jacketed cord the other one has. Probably cure your problem.

I think this will be my next move.

Dave,
I'm getting pretty conflicting advice here? Keep in mind we a talking small scale here..

The noise is certainly coming from the ballast, but it traveling through the ballast wire. Is it impossible that the rubber jacket could dampen the EMI enough to remedy my problem?

I've moved the ballasts 6+ feet away from the controller and any associated wiring. However, the outgoing wires from the ballasts have to pass within a few inches of the AC phone cable at one point.
Twisted pair wire won't resist EMI enough to help?

[kcress]

What you need to realize is that the threshold between your comm having a problem, and not having a problem, is tiny.

No the rubber has no shielding ability at all, BUT, the tiny added distance guaranteed by its jacket thickness may be all that is required to drop below the problem threshold. But wait! There's more! It also reduces the radiation ability of the ballast noise because the wires in the cable will be twisted and not making large radiative loops.

So as I say - try it.

[jdieck]

Are the two ballasts the same model? I found some Icecap may cause interference while other models will not.
Sorry to add one more variable but I would suspect electrical noise rather than magnetic.
Will you get the same issue running the AC on a battery?
If you do not there are lectrical filters you can conect the ballast into.

[shelburn61]

<a href=showthread.php?s=&postid=13775279#post13775279 target=_blank>Originally posted</a> by jdieck
Are the two ballasts the same model? I found some Icecap may cause interference while other models will not.
Sorry to add one more variable but I would suspect electrical noise rather than magnetic.
Will you get the same issue running the AC on a battery?
If you do not there are lectrical filters you can conect the ballast into.

Same model, at least both were manufactured within the last year. The only apparent difference is the outgoing wires on one have the rubber jacket. Not sure of the difference between EMI and electrical noise. Would the rubber jacket kill electrical noise?

Kcress,
It's not a distance issue. According to the AC monitor I'm getting the same level of interfernce with the wires 3" away or nearly touching. You might have a point with the twisted ballast wires though...

[jdieck]

<a href=showthread.php?s=&postid=13776408#post13776408 target=_blank>Originally posted</a> by shelburn61
Not sure of the difference between EMI and electrical noise. Would the rubber jacket kill electrical noise?

Electrical noise travels trough the wires from one plug to the other via your home wires, EMI travels trough the air.
Electronic ballasts internal circuitry transforms the 120 volts in the input to the voltage regulated for the bulb requirements but diferently from magnetic ballasts they also change the frequency of operation from 60 Hz to a range from 200 to 1000 Hz depending on the design. This higher frequency helps maintain a more stable arc and avoid flickering but can create high frequency noise on your power lines that are transferred to all other devices in your home. While most devices are not afected a controller works with very small voltage signals which sometimes are as low as the noise signals thus the interference.

[shelburn61]

<a href=showthread.php?s=&postid=13776497#post13776497 target=_blank>Originally posted</a> by jdieck
Electrical noise travels trough the wires from one plug to the other via your home wires, EMI travels trough the air.
Electronic ballasts internal circuitry transforms the 120 volts in the input to the voltage regulated for the bulb requirements but diferently from magnetic ballasts they also change the frequency of operation from 60 Hz to a range from 200 to 1000 Hz depending on the design. This higher frequency helps maintain a more stable arc and avoid flickering but can create high frequency noise on your power lines that are transferred to all other devices in your home. While most devices are not afected a controller works with very small voltage signals which sometimes are as low as the noise signals thus the interference.

Gotcha. The Neptune guy thought it was EMI and there is no difference in interference whether the ballast is plugged into the DC8 or a separate outlet so I guess he must be right?

The wiring passes through a hole in the basement ceiling within a foot of a large steel support beam. Could I use shielding and ground it on the beam?

[jdieck]

<a href=showthread.php?s=&postid=13776746#post13776746 target=_blank>Originally posted</a> by shelburn61
Gotcha. The Neptune guy thought it was EMI and there is no difference in interference whether the ballast is plugged into the DC8 or a separate outlet so I guess he must be right?

The wiring passes through a hole in the basement ceiling within a foot of a large steel support beam. Could I use shielding and ground it on the beam?

Once the noise is in your house wiring it does not really matter which outlet it is connected to unless they are conected to diferent phase then the noise has to travel all the way to the outside transforemer and back into the house.
On the shielding, although it can help I am afraid it might not be totally effective.
Usually beams are just supported to the concrete foundation without very good grownding, your best bet for grounding will be to conect it to the ground round pole of one of the outlets.
Do not forget to twist the wires onto each other and ground your reflector(s) if you are using metalic ones.
Also insure that the ballast body is also properly grounded and that the ground plug in the outlet is also really conected to ground on the electrical main box.

[kcress]

In all probability the noise is being projected from the random loops of your single stranded ballast wiring. A loop, electrically, being any space between the single strands. That is why a jacketed cable will help minimize it. It's also probably why your 'corded' one isn't a problem, if everything else is the same.

[shelburn61]

<a href=showthread.php?s=&postid=13778244#post13778244 target=_blank>Originally posted</a> by kcress
In all probability the noise is being projected from the random loops of your single stranded ballast wiring. A loop, electrically, being any space between the single strands. That is why a jacketed cable will help minimize it. It's also probably why your 'corded' one isn't a problem, if everything else is the same.

I'll let you know what happens

[shelburn61]

Last night I switched out the ballast wire for a jacketed cable. I noticed a very slight decline in interference numbers, but still not where I want them. Then I started moving the cords and ballasts around and noticed the interference numbers fluctuated a lot based on their orientation regardless of proximity to the phone cord. It was tough to find a discernable pattern to fluctuation of the numbers based on how I moved the cords so I was blindly using trial and error. Finally the arrangement that seems to work best is to just loop the cords from both ballasts together into a tangled mess. It seems straight runs of ballst cord produce the most interference?

[jdieck]

<a href=showthread.php?s=&postid=13782758#post13782758 target=_blank>Originally posted</a> by shelburn61
Finally the arrangement that seems to work best is to just loop the cords from both ballasts together into a tangled mess. It seems straight runs of ballst cord produce the most interference?
Yes, that is what twisting is all about. By twisting the cables around each other EMF is reduced significantly as one twist cancels the other thus reducing EMI interference.