Building a simple power strip with emi/rfi filter for treadmill

[EzRidr62]

Hi all,

After some research about treadmills causing DSL dropouts, and having tried and failed with many of the recommendations regarding induced line noise from the motor, I purchased a couple of emi/rfi filters that are small "line in/load out with ground" chassis mount boxes to be installed at the treadmill to cancel any noise being introduced into the house wiring. the price for these were right at less than $20 a piece including shipping off Ebay. They list for 52.14 each at DigiKey.

My son also has a treadmill and DSL in his house as well, (but has not hooked his treadmill up yet) so I bought a second filter to be pre-emptive.

The filters I bought I can not find the attenuation dB specs for. They are Corcom 20VSK6 filters, considered by their spec sheet to be able to provide "premium" attenuation for appliances, exercise equipment,etc. I have uploaded a sheet on this filter for reference.

My question is this:

Having purchased both filters for a decent price, and being the type of person who tends to overbuild something rather than skimp, can 2 of these filters be wired together either in series, parallel, or piggy-backed to each other to provide a better filtration method, or am I wasting my time and just wire up one and be done with it?

Thanks in advance for all who weigh in on this. I built a snowplow limit switch control system using help from you guys 2 years ago and it still works like a charm. The circuitry was simple,effective, and saved my $400 linear actuator from the same death it's predecessor met with from poor engineering from the manufacturer.

Cheers!

 

[JimB]

The filters I bought I can not find the attenuation dB specs for. They are Corcom 20VSK6 filters, considered by their spec sheet to be able to provide "premium" attenuation for appliances, exercise equipment,etc. I have uploaded a sheet on this filter for reference

.

The attenuation is on the fourth page of the datasheet which you have uploaded, Corcom call the attenuation "Insertion Loss".

can 2 of these filters be wired together either in series, parallel, or piggy-backed to each other to provide a better filtration method, or am I wasting my time and just wire up one and be done with it?

You could wire two filters in series and it would give greater attenuation.
BUT, to gain any real advantage, you would have to carefully consider other aspects like the screening of potential noise sources within the treadmill.

Just install one filter and call the job a good 'un.

Just as a point of interest, how come a treadmill affects DSL ?
You are not using some kind of "ethernet over mains" are you? If you are you get no sympathy from me for using such flawed technology.

JimB

 

[EzRidr62]

Just as a point of interest, how come a treadmill affects DSL ?
You are not using some kind of "ethernet over mains" are you? If you are you get no sympathy from me for using such flawed technology.

Oh my..... No. Although I am more of a ham when it comes to electronics, but I would never consider broadcasting another frequency through house current.

My quest began when, after a long hiatus from keeping the pounds off, (and needing to convert the lead in my a$$ back into iron in my blood) I began working out again. Needing some form of entertainment to drone out the monotony on the mill, I built a stand that allows the laptop to reside right over the mill. Oddly enough, shortly after the drop outs started.... but years before, it never gave me issues. I began to suspect a cranky router (we had other wi-fi issues when streaming at night) , so I replaced the b/g/n router with a 10/100/1000 unit.... higher speed must mean more sensitivity to noise because now the drop outs were quicker and more severe...I couldn't even load a static web page when the mill was on.

From what I've read when searching the Internet for "treadmill causing DSL drop outs" and such, I learned that when the treadmill is running, and you are trying to use the Internet, such as streaming music, video, or even requesting a web page to load, many people's DSL is virtually "down for the count". Nothing works. It seems that treadmills that were not considered "premium brand" like Pre-cor, etc. may not be supplied with line voltage filters inside to keep the variable-frequency drive motors from inducing lots of noise into the house wiring. At least, this is what I understand to be the reason. Treadmills have become these sophisticated units that vary the speed minutely from the point where one foot leaves off and the next makes contact with the tread, giving a more natural feel to the stride as if walking on hard ground. You can hear it in the motor as your foot contacts the belt, the motor increases speed slightly to compensate the drag of weight against the platen, and then slows slightly as the ball of your foot is behind you.

One thing is for sure. As soon as the treadmill motor starts, DSL quits working. I've tried different circuits that were far from each other, ran extension cords through the house to get the router and modem onto another circuit that was no-where near the treadmill circuit, made sure I was on opposite phases from the phase the treadmill was on, hard-wired cat5 cable from the router to the laptop, tried using a smart phone rather than laptop, all kinds of combinations. Only when I switch the smartphone to 4G instead of wireless network does it start working again. I can run 4G right on the phone, resting on the control panel of the treadmill and its fine, so I arrived at the conclusion that it was not RFI interference, but had to be induced through the AC and messing up the DSL. I even ran a new phone wire away from any AC voltage from D-mark to modem and still drops out. After I googled the problem I began to see that I was not alone..... however, my nature of curiosity refuses to just go out and buy a Tripplite Isobar power strip and call it a day. One forum I read a fellow called the treadmill company and they sent him a filter.... so I took that and started researching filters and found parts supply houses listing them specifically for certain treadmills...so I knew I was close.

I love solving issues using more "raw materials" than pre-built units that I have no idea how much filtering they do, plus I can easily build a small power strip myself, but I wasn't sure how well I was doing plan-wise without some help from people like you who are more well-versed on this subject.

I didn't want to double up a filter and cause other issues, like a drop in voltage or current flow to the mill which would fix one problem and cause a bigger one.

Thanks for the advice. I see the filter can attenuate quite a bit of noise now. Thanks for pointing that out.

 

[tunedwolf]

I suppose it's possible that your local electrical earth is poor. If your DSL drop is bonded, it should be, then anything with an RFI filter or leakage to ground will be wanting to use the earth wiring as an antenna rather than any leakage or noise being run off to ground as it should. The motor and control in your treadmill will radiate all sorts of harmonics right through your wiring in that case, which may find its way onto your incomming phone line if the drop is bonded. It takes surprisingly little disturbance to upset a DSL link

 

[EzRidr62]

To clarify, are you saying that I (may) have a poor connection to a ground source, such as the rod in the ground below the Dmarc? And thus, the noise is finding its way from the (safety) earth ground leg of AC wiring to the bond connection of the phone (DSL) line?

Or is it the other way around and the earth ground is not properly making connection to the mains house wiring thus allowing the safety leg of the house wiring to cross over to the phone line?

 

[tunedwolf]

Technically, it could be either, although I was meaning the connection from the protective earth wiring to the local electrical ground. About the only way to know for sure just how good the actual earth is there, a visual inspection aside, would be to conduct loop impedance tests. It may not be your issue at all of course, this is just an educated guess on my part given your description. But to clarify, all of your protective earth wiring runs from each outlet or node, back to the distribution board where there is an earth block. To this earth block all of your incomming services, such as water, gas, telephone etc should be electrically bonded. From there you will have a local electrical earth connection. Now the last part there could be furnished by the electricity supplier from their incoming cable, or it could be a physical earth rod driven into the ground outside with a copper running from that back to the earth block in your distribution board. Either way, if the connection from the earthing block to the local earth is poor then all of the protective wiring may act as an antenna which could induce noise into your telephone wiring if it is also bonded like it is supposed to be, or in fact it could find its way in by some other means of ingression. Likewise, if the actual earth is poor where the rod is driven could cause a similar situation. The only way to know that for sure will be loop impedance tests.

Like I stated though, it's just an educated guess on my part

 

[EzRidr62]

To this earth block all of your incoming services, such as water, gas, telephone etc should be electrically bonded. From there you will have a local electrical earth connection.

Gotcha.

It's worth a look to ensure all those connections are still where they were put oh so long ago.... I know where they all are but haven't really inspected anything since I installed separate ground bus bars in the main panels to meet code many years ago. Considering that SOMETHING had to have changed between the time I stopped using the mill (several X a week) a couple years ago until now, I only resorted to the filter idea due to a lack of thoughts on what else to look for (that could have occurred in the last couple of years). I used to stream Pandora back then and never had issues like I do now. Even with the router temporarily relocated into the same room as the mill, DSL still dropped like a stone. Years ago the router was on a different floor and never had a problem.

Thanks for the tip;

I really appreciate both of your feedback

 

[cct314]

If you have not already done so, I would "home run" your DSL connection to the NI (Network Interface) as this will go a long way toward making your DSL "immune" to what might be happening inside your house in terms of RFI/EMI.

In short, there are two ways to connect DSL to modem:

- Have it come into the dwelling, install a filter on every telephone and put the DSL modem at one of the telephone jacks.
- At the network interface, install a filter so that none of the wiring in the house "sees" any of the DSL signal (e.g. no filters required) and run a separate line from that filter to the DSL modem with nothing else on it.

In the first case, you have the problem in which you have stubs and taps that can cause all sorts of problems with standing waves/notches and potentials for longitudinal imbalances which means that EMI/RFI inside the house/wiring can be conducted onto the phone line and degrade the SNR of the DSL signals which can cause CRC errors, lost packets and force the unit to retrain.

In the second case, the filter effectively isolates the DSL modem from the electrical goings-on in the dwelling: I have seen the changing of the modem from the typical "filter at every phone" to "one filter at the NI" reduce the outbound carrier level by 6-12dB which indicates that not only are the losses significantly reduced in doing so, but there is the implication that the inbound signals would be less likely to be immersed in possible interference sources as well since they are also less likely to be attenuated - be it due to simple attenuation of effective loss due to longitudinal imbalance.

In my case, my original run to the DSLAM was 26000 feet and the technician just gave me the filter, which plugged right in at my NI: I then ran a 20 foot length of CAT5 over to where I'd placed my DSL modem, and the improvement in SNR and reduction in required signal was quite remarkable - as was the reduction in the number of retrains that I experienced in a typical week. Since then they have installed a DSLAM in my neighborhood (only about 1500 feet) but I still have the original configuration. If you have a standard-issue NI, the DSL filter simply replaces the original block containing the phone test/interface jacks - although I don't know off-hand where you'd get one if you can't get one out of the phone company. If your NI is like mine, there should be plenty of room to fit a standard DSL filter inside, though.

 

[KeepItSimpleStupid]

Time to ask a few questions?

1. How many phone jacks exist in the house?
2. How many phone things exist in the house.
3. Wiring? Quad, CAT3 etc?
4. Are you using multiple filters or a splitter?
5. Describe the NID bonding?
6. Describe the telephone wiring in relation to the treadmill.
7. Is it easy to run a telco wire from the "Distribution area" to the DSL modem?

I'll clarify once I get answers. I have a royal mess of non-home run telco, yet I have "pristine" DSL.

Comment: An ISOBAR gives you a protected equipment warratny IF you retain proof of purchase

 

[KeepItSimpleStupid]

cct has it nailed. It's likely you can get the phone company to install it for free.

 

[tunedwolf]

I use one similar to this on my line right at the drop: **broken link removed**
From there I have one line going directly to my gateway for internet, IPTV and the like, and the other into a punchdown block for all the internal phone points. The service provider could care less, although they might insist that "their" version of it be fitted

 

[KeepItSimpleStupid]

I had an instance of a VSD in a furnace not making X-10 work. A filter on the furnace fixed it. So, power line interference from a VSD is a "possibility".

If your using more than 3 filters. That's a probem. A splitter will ALWAYS improve DSL.

 

[EzRidr62]

I like the home run/filter at NI idea. I currently am running three filters for 2 cordless phones and 1 fax line at the HP print/fax/scanner.
Wiring is the typical mess of solid telco 4-wire the house was originally wired with plus additions from me. on the modem to NI side, no twisted pairs exist, so lots of noise might crossover.
The jack where the modem is now is approx. 15 feet from the treadmill.
No Cat5 has been used other than for the Ethernet cabling I ran from the router to different locations in the house.
NI is bonded to the same copper rod that the main service is bonded to just outside the foundation wall. All utilities enter the house in the same location.
The Dmarc box has both terminal lugs and rj11 jacks inside so perhaps I can use one of the supplied filters and place it into the box? It's weather tight.
And finally, yes with a little work I can run a home run to my modem.

As it stands right now however, I might relocate the modem and router to the basement near my 1 service panel that is fed by my backup generator; I have a split 200 amp service with 1 100 amp panel at the main entrance and another 100 amp disconnect, which then runs to the opposite side of the basement (some 75 cable-feet away) where my generator-fed panel is. All my awake time living areas (L/R D/R Kitchen/Bath) are wired into that panel. When power cuts out I hit the disconnect and backfeed that panel through an old 40 amp dryer circuit that has been rewired to the weathertight 220V generator interface plug. ( the house is a large L-shaped structure)

 

[KeepItSimpleStupid]

Try temporarily putting the DSL modem/router at the NID with the phones disconnected. You also won;t need a filter. This would be a quick test. One of your filters could be bad. So, you just have to plug the DSL modem into the test jack.

A reminder that you can remove the filters and use DSL as long as you don't use the phone. You can turn off the modem to use the phone.

The DSL modem does not care about polarity. I manged to find some CAT4/5 RJ11 mounded cables that did not have the twist to connect the modem. A standard telco cable would have the twist, but usually that's CAT3 (thin the grey cables).

Maintain the twists on anything that sees the DSL signal. Twists reduce EMI. Shielding reduces RFI.

==

I'd ask you telco provider if they would install a DSL filter at the NID for you. Mine provided a small length of outdoor rated wire to get to my demark point. At the very least, you could have a CAT6 splice box just inside. I brought the pairs to a Leviton punch down, but the telco still gets distributed form the carbon surge protector. That point feeds about 4 lines and a run to a second block with 3 or 4 more feeds.

Just make sure you get some outdoor rated cable from them to get inside the house. Make them possibly replace the cable coming in with CAT5.
You have to terminate. That's why I terminated to a 8-way leviton punch down. RJ31x alarm jacks (allows line seizure) would add a complication.

It's hard to troubleshoot. I managed to find appropriate push on terminal blocks that will fit a CAT6 splice box. Now, I can transition 4 pair quad to the punch down block and can disconnect each leg. I'm not quite there yet. One block is in place for proof of concept.

In the second junction block, I will ultimately transition to a patch panel for both telco and Ethernet For now, running wires is tough between blocks. Initially, all patches will be RJ45's and inserts will convert an RJ45 jack to a single line telco jack for new stuff.

I only have 3 wired locations far from my "network" stuff. That was done with free terminated cable. The network stuff is on a UPS and the DSL modem is powered by an EXTERNAL POE generated from the UPS.

This https://www.westsidewholesale.com/leviton-47609-tsv.html is similar to mine but it has the RJ31x security jack. Hindsight is 20/20.
They may have implemented it differently.

So, I have a fair number of CATx to QUAD transitioning to do and I could not find a commercial product to do it. Easy disconnection is a necessity, although I do need a test connector to have access to the pulled line or remove the spring clip locked wires.

The new blocks that Verizon insalled were super. It's so easy to troubleshoot and the quality is better.
Although it wasn;t this one, but you can see how potentially easy it is to trobleshoot.

Usually, there is a 1/2 ringer left in the NID and with the two blocks that I have it's so easy to:
1. just look at the naked line into the premise
2. Look at the line just with the 1/2 ringer
3. Disconnect the inside
4. disconnect the filter
5. Use the test jack.

Yep, it does require the "special" nose like aligator telco clips typically found on a but set.

OK, what's 1/2 ringer? It present a load of 1/2 REN to the ringing circuit so external tests can be preformed with the inside line disconnected.