Dew heters destroys PowerLine network signal

[Magnus_e]

Hi.

I have just finished building my obsy, and yesterday it was time to install the scope, cameras, etc, but I found an issue.

I was struggeling with network loss, and when I got a connection the speedtest was showing me that I had a full 20 -> 30 Kb/s connection....

As I was going to wrap things up for the night, I tried powering off all equipment and re test network speed, and just like that I had 10Mb/s! So I powered one ting on at a time, re tested speed, and when dewheaters was powered on ~30Kb/s. I did some more testing and I'm now 100% sure it's the dew heaters killing my network speed.

So what is going on? The dew heaters work well, and does heat up. I have had no fuses triggered, and cannot find anything wrong. the dew heaters is a sort of resistor, so could it be that the powerline signal gets 'shorted' out in a way it cannot handle, or should I look for something else?

 

All help greatly appreciated

Magnus.

[steppenwolf]

Does the dew heater have a controller? If so, this will almost certainly be a pulse width modulation (PWM) type and this could be inducing a small current on you network cable is the two are in close proximity to one another.

[Magnus_e]

It has a Hitec Astro (manual) controller?

It's one temperature controll with four outputs. No inputs or similar.

I toght the temperature was just controlled by adjusting voltage?

[D4N]

18 minutes ago, Magnus_e said:

It has a Hitec Astro (manual) controller?

It's one temperature controll with four outputs. No inputs or similar.

I toght the temperature was just controlled by adjusting voltage?

Adjusting voltage is very inefficient and wastes a lot of power whereas a PWM is a much more efficient way of reducing the heat output.

Always good to ensure power cables are kept separate from data cables, ferrite cores can also help.  Use screened cables for data, Cat5e or better greatly reduces interference.

[Macavity]

+1 for the above. I found that anything that "chops" DC (Heater control,
focus control even!) can put noise onto shared supply lines. Also via lead
proximity etc. I'm pretty sure the (my) Hi-Tech Controllers are PWM!  

Aside: My obsy runs on Mains or 12V 90Ah rechagable Caravan Battery, 
but still gave poor network rates / random disconnections. Eventually,
I bought myself a couple of 12V Laptop adapters, which give ~15V DC
I put this through 12V stabiliser BOXES (inc. filtering, fuses, switches): 

BUT having a *clean* supply for "sensitive" equipment helped a lot.

[SkyBound]

No, those controllers use a PWM controller, which alters the voltage by switching the power on and off very fast, so for example if it turns the power off for 1 ms and on for 1 ms then the voltage would be 6v if using a 12v input, if it turns on for 2 ms  and off for 1ms the voltage at the output would be 8v, and so on......depending how long it is switched on  and off for determines the voltage.... it's actually switched much faster than my examples above in reality....

because of this it does cause noise on the lines so that is probably what is causing your issue..

ferrite cores are a very very good idea, I use them on all my Astro cables, just to make sure there is no noise issues.

hope that makes sense

[Magnus_e]

Thanks

All of that does make sense!

As I'm using Powerline network, it's no chance of getting clean power. (the network is on the mains in the entire house)

I have my dewheaters on a 12v 20 amp supply that is also used for opening and closing roof.

Would @Macavity solution work to isolate the 12v supply from the rest? I guess that can be a solution?

[Gina]

I've heard that these Powerline devices are very prone to interference and also give very low speeds compared with what you would expect.  Best solution is to run a CAT6 cable and use an Ethernet link.  You can get outdoor grade CAT6 cable quite cheaply from reliable suppliers like Amazon - that is what I've had to do.  WiFi was intermittent and about on it's limit I guess.

As for dew heaters, when needed (which isn't often I've found) I connect directly to a DC supply.  I make my own low power dew heaters.  I now have a redundant HitechAstro dew heater control box!  PWM has a habit of causing interference on image capture cameras too.

[Skipper Billy]

I dont know the science behind it but since I swapped from one leisure battery to two - one for clean supply (camera etc) and one for dirty (anything that switches on and off - mount, dew heaters, filter wheel, focuser etc) - my subs have definitely been cleaner.

[tekkydave]

Probably interference as suggested above. Powerline devices can work very well - I get 165Mb/s TX speed and 132Mb/s RX speed between my office and the router downstairs using Netgear Powerline AV 200 devices (according to the manufacturer's utility).

I can't, however get the network signal anywhere which is off the up/downstairs power rings as they are isolated in the consumer unit by an RCD. This filters out the network signal as 'interference'.

[Magnus_e]

I will have to figure out what to do, an upgrade is needed...

The reason I do not run cat6 cable is that I would have to drill a hole in the house to get it out. Not sure how to go about that.

It would be nice with a faster network, but at the moment it's usable. (without the dewheaters)

[steppenwolf]

Quote

The reason I do not run cat6 cable is that I would have to drill a hole in the house to get it out. Not sure how to go about that.

Much easier than you might imagine (unless you live in a castle or 300 year old flint cottage!). Start by drilling a standard hole with a masonry drill bit from the outside through a mortar join then use a long masonry drill bit to go through both outer and inner walls (assuming a standard cavity wall).

The comments above seem to confirm my original suggestion so I would tackle this as a first priority by perhaps powering the dew heater with a 12v battery as a trial.

[Magnus_e]

I'll start planning for a cat6 cable

I could try to install my 12v batterybank so I can use the dewheaters if I need to.

However it's not much room in the obsy for me to install the battery (service door for the electronics is a future upgrade)

[Peter Drew]

I would drill through a vertical pointing so that you could drill "uphill" somewhat. Helps to stop rain getting in.  

[Macavity]

What do you really need "Fast Networking" for tho'? Fortunately
an (ahem) "DIY enthusiast" had left quite a lot of RANDOM holes
in the side of my house, some which have proved useful! lol. 

GIVEN a hole or two, e.g. Observatory control (Even Telescope
Control!) can be achieved with lowish speed (Cat[n] cable) links.
Even Coax. These MAY be sufficiently immune to random noise.

Whereas I am something of a fan of *hardware* (cable) control
I now conceded that WiFi plus *TeamViewer* can also provide a
view (+ mouse clicking) with a PC screen from *either* End? 

Clearly if you have GigaBytes of Data, it might be nice to transfer
this via FAST (fibre optic?) link. But not *strictly* necessary?  

P.S. I would say that "remote control" has taken much more 
of a toll on my budget + *exasperation* than anything else!
In retrospect, better spent on actually DOING Astronomy?

[Magnus_e]

Well, my obsy is running linux servers with INDI.

Everything gets sent over the network. Telescope, focus controll + images from my dslr and qhy guidecamera. It would be useful with fast network.

I'm also going to image remotely over internet (I live an hour away from the observatory).

At the moment the powerline network is the limiting factor. I have 75/15Mb/s ADSL at home, and my parents have an insane 300/300Mb/s fiber network

The broadband supplyer here was taken over by a big fiber company that supply 40/40 -> 1000/1000Mb/s. All existing customers of the old company got 300/300 for the price of 40/40

[Gina]

It's quite easy to drill a hole through a window frame for Ethernet cable   Or go up through the ceiling and out through the eaves.  Ceiling probably plasterboard and easy enough to drill though mucho dust!

[Magnus_e]

Thanks for the tips.

Ethernet cable is first on the list of upgrades.

I did manage to do a full imaging sesdion of Andromeda 

but it would be nice with some more speed. The server used almost a minute and a half to upload a 10 min sub from the canon eos 550d to my laptop....

[Magnus_e]

Well, sometimes one gets lucky!

I was about to see where I should drill a hole in the house, and I found a ready with flex tube

I have never seen it but the previous owner probably wanted a power socket there. The flex tube ended at a power 'box'.

So then it was just to shorten it a little, add a cat6 ethernet socket and pull the cable trough!

Did a speedtest on my Odroid, and I'm getting 300Mb/s Very pleased whit that!

--2016-12-03 16:33:31--  http://speedtest.tele2.net/500MB.zip
Resolving speedtest.tele2.net (speedtest.tele2.net)... 90.130.70.73, 2a00:800:1010::1
Connecting to speedtest.tele2.net (speedtest.tele2.net)|90.130.70.73|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 524288000 (500M) [application/zip]
Saving to: '/dev/null'

     0K .......... .......... .......... .......... ..........  0% 18.9M 3m42s
    50K .......... .......... .......... .......... ..........  0% 18.4M 3m45s
   100K .......... .......... .......... .......... ..........  0% 34.3M 3m11s
   150K .......... .......... .......... .......... ..........  0% 43.5M 2m47s

511800K .......... .......... .......... .......... .......... 99%  333M 0s
511850K .......... .......... .......... .......... .......... 99%  319M 0s
511900K .......... .......... .......... .......... .......... 99%  261M 0s
511950K .......... .......... .......... .......... ..........100%  306M 0s
512000K                                                       100% 0.00 =14s

2016-12-03 16:33:45 (301 Mb/s) - '/dev/null' saved [524288000/524288000]

500 MiB in 14 seconds

[celkins]

You really shouldn't run network cables in the same trunking/conduit as mains - there's a possibility of interference, but it is definitely a risk from an electrical safety viewpoint; each cable is probably only rated around 300v to a common, so it only takes a single fault in the insulation to leave you close to danger

[Magnus_e]

2 minutes ago, celkins said:

You really shouldn't run network cables in the same trunking/conduit as mains - there's a possibility of interference, but it is definitely a risk from an electrical safety viewpoint; each cable is probably only rated around 300v to a common, so it only takes a single fault in the insulation to leave you close to danger

Hi.

I guess previous owner had planned for a outdoor socket, but the cable was newer installed.

All there was is a empty flex tube going to a power coupling box. I shortened the flex tube, so the cat6 cable is no where near mains lead.

 

Thanks for the heads up though.

[IanL]

It is the frequency of the PWM cycle that will be the issue. At a practical level it does not matter for a dew heater whether the frequency is high or very low. In terms of interference with wired and wireless networks, a low frequency is much better.

See this in respect of the cheap LED dimmers many of us use:

http://smokedprojects.blogspot.co.uk/2013/09/led-dimmer-pwm-hack.html

Personally I have never had any interference issues with networks or imaging gear despite using high frequency PWM on a single supply. The only thing I do is run the USB and 12V cables in separate bundles, but I guess I have been lucky.

 

[Daz69]

Just an idea, not even sure that it's made, but can you get shielded cat6 wire? OR, try and shield the dew heater wires with copper foil, or swap the wire running from the PWM to the heater with a shielded RCA phono lead, which is fairly simple and cheap. Even a guitar/amp fly lead, I know they are shielded as I use them. 

 

If drilling through a wall, use a thin long drill bit and go outside in. Once through, step up to the larger bit. That way you won't get a large "break-out" on your plaster inside. Avoid drilling through eaves as described previously, as a lot of older houses used asbestos for the soffits.

[SilverAstro]

Thinking out loud here :-

The PWM chopped voltage doesnt have to travel all the way along the dew heater feeder wire/cable,  up to the resistive load of the dew heater itself.

Consider that bit of cable as an integral part of the DC load, so put a big capacitor for smoothing and a small capacitor for high freq and  transient supression across the output of the controller box ?? ( you might even get away with just one small cap. depending how sensitive your data cctry is ! )

I dunno the details of your particular PWM cct, so you may want to put a small series resistor between the controller output and  the filter caps. Doesnt have to be big in relation to the resistance of the heater so it will not waste much power. In an obstinate case put the controller and the caps in an aly box as well and earth it !

What you are trying to achieve is smooth  DC on the cabling to the dew heater band.

Alternative : put the controller in an aly box and feed the two wires from the controller up a shielded twin cable and secure the shield to the box, earth it and put filtering caps on the power wire supplying the controller to stop HF noisy stuff getting out into the system backwards.

Good luck.