HAM radio V Solar DC motor

[warnberg]

I recently had a Solar system setup for my pool, (Lorentz PS600) it runs the pool pump and is saving me a ton not having to run the pool pump all the time eating all that electricity.  Anyway what I have now encountered is a large amount of noise across the HF bands from it, I've done just about everything I can think of to eliminate or reduce the noise... additional grounding, relocating antennas, etc...  Anyone have a secrete to eliminate this noise?

Thanks for any and all suggestions..

Dave

[cdev]

Can you turn it off while you are operating? Pool cleaning doesn't have to be done continuously, does it?

Also, I bet the standard bypassing approaches would work to some extent. Another thing (the first thing I thought of) was sheathing the wires externally in something and grounding it.

Considering the risk to any kind of device posed by lightning, as well as the possibility somebody digging might hit it, if its buried - putting cable in conduit seems like its likely a good idea, and probably required by electrical codes.

One thing that might be interesting, is seeing what capacitance does to the moise.. If the cable is currently a wire lying on the ground (sounds like it may be) try taking a yard or two of aluminum foil, in essence making a big long "feed through capacitor" along the line, and then grounding it to see if it reduces the noise (I'm not suggesting that asa permanent solution, just as a test!) You could use an RTLSDR and the "rtl_power" binary the osmocom driver comes with  to visualize and that way, get a quantitative idea of what is happening spectrum wise. It can sample huge swaths of spectrum (if you had two RTLSDRs you could run it to sample from below the AM broadcast band to >1700 MHz, and dump its findings to a CSV file, then you can use the data viz app of your choice to examine your data. There are scripts that can turn it into charts, frequency and amplitude (mapped to color) over time.

[warnberg]

Yes, I can turn off the pump and the noise does go away, the pump is used to circulate the pool water as well as generate chlorine from the chlorination system (salt system) so I could turn it off and do from time to time but too frequently and my pool will turn green.

I'm have both a SDRplay as well as a FLEX-6500, I can see the spikes(interference) in frequency... interesting to watch it change as the sun goes behind a cloud or come from behind one... that was my first clue as to what was causing the issue.  So as to your idea as to what is happening the big kicker here is it's variable with sun intensity.  I can easily throw a notch filter on the signal at a certain frequency and remove the noise with  the FLEX, but it's temporary as the interference drifts.. that is the challenge.  So I go back to attempting to eliminate the noise as much as possible.  I honestly think it's coming from the pump it's self (DC variable speed).. was thinking of a wooden box, line it with something to shield it maybe??  Not sure that would work hence the post for idea's..

thanks

Dave

[cdev]

You could put the pump in a grounded, well ventilated metal box and use high current carrying capacity feed through capacitors to ground for where the DC leaves the box.

[ConKbot]

I'm guessing you dont have conduit running between the motor and the box? Since its a BLDC motor, youre going to have a ton of switching noise and harmonics, if running on normal 4-core (3 phase + gnd) household wire. Replace the lead running to the motor with a shielded (with an actual braid) something like 7777T21 on mcmaster-carr comes to mind (4 wire, foil and braid). Though it doesnt mention outdoor or wet environment use, so that specific cable may not be the best. Clamp-on ferrites are easy to install, so may be worth a shot before you replace the cable. 

[warnberg]

Actually the system wiring is all in conduit, from panels to breaker box (underground for some time), from breaker box to the power pack (PP600H off 99% of the time), from the power pack to the controller (PS600) from the controller to the pump..

This is the lorenz system http://www.floridasolarpump.com/solar-pool-pump-ps600/

[ConKbot]

I should have specified, metal conduit?  I'm guessing probably plastic so it can be water-tight for cheaper. Assuming the pump controller has the drive electronics, I'd definitely be looking at the long unshielded cable with high current square waves being sent down it as the source of your emissions. How long is the run from the controller to the pump?  And I suppose you should also take a look at the DC inputs to the pump controller as well. If the decoupling is insufficient, you'll get noise emissions from the current varying on the input, but that will be easier to suppress with adding caps than shielding.

[warnberg]

From controller to pump is about 6 feet and yes the run is in a plastic conduit, inside is 4 wire runs, one ground and a L1, L2 and L3 for each "phase" as they reference it in the installation manual.

Thanks

Dave

[Ian.M]

That's your problem right there.  Its an inverter drive to a three phase (BLDC) motor that acts as a 6' antennae for the inverter's PWM frequency.  Lovely sharp edged pulses, lots of inductive kickback, and you probably couldn't do much worse to wipe out the bands below VHF if you were running a rotary arc transmitter.

Adding suppression caps to the motor wiring is highly likely to kill the drive electronics, so your only hope is shield the s--t out of it.  Use screened cable through metal conduit, and you'll need a metal screening box round the pump and the controller.   It might be easier to relocate the controller next to the pump and put one screening box round the whole lot.  Filter the s--t out of the DC in where it enters the screening box.

[warnberg]

Thank you sir...  I've been doing some research as well... 
good info about the issue here: http://www.solar-electric.com/reducing-electromagnetic-interference-pv-systems.html/

So I may be trying a combination of things:
1) shield the pump in metal louvered box grounded
2) change the plastic to metal conduit from controller to pump AND try twisting the wires inside the conduit going to try 1 twist per inch.
3) adding Ferrite chokes, cores, or beads inside the conduit on the wires between the controller and the pump

[madalf71]

It's a problem with inverter drives, particularly single phase input to 3 phase.
Typically there'll be filtering caps between phase to earth....however the initial charge time/current of the caps is often enough to trip RCDs Residual Current Devices.
But yes there are other ways around it. Try giving ABB or Bosch/Rexroth tech a call and ask what they do.