Page 1 of 1

noise filtering on PIP inverter

Posted: Sat, 11 Apr 2020, 21:30
by vincent.p
Hi all,
i use a set-up made of PIP5048GK fed by 12 x270W panels and stored by a bank of 8 Trojan batteries T145 (260AH)
This makes typically a good job for the price but...
I noticed a really high level of EMC disturbances with a lot of electric noises conducted on all the lines. because my main hobby is Ham radio, i like clean frequency spectrum !
I'm electronic enginneer and used to deal with that kind of stuff in my daily job, so i decided to install a set of filters.
So i put a Schaffner filter FN2200 25 33 in series with the panels just before the inverter connection. Then the troubles started.
The inverter started to make switching noise and the main 30mA breaker in front of the inverter, cuts the supplies because of current leakage.
Since the Schaffner filter has some 1Mohm resitors in it connected to earth i understood that becasue of non insulated PV string, something went wrong.
I opened (hardly) the filter and cut the internal connection to earth to remove the problem. Then the inverter worked again, and the differential breaker stays one.

My main problem is despite the fact i also put AC mains filters before and after the inverter, plus the modified one for the PV input, i still have a lot of noise especially
between 7 and 21MHZ. Frequency Spectrum is desesperant...

Do you experienced the same and is some of you reduced succesfully the RFI conducted disturbance from PIP inverters ?
Thank for your feedback
Kind regards,

Re: noise filtering on PIP inverter

Posted: Sat, 11 Apr 2020, 21:49
by jonescg
Why would you need to filter the panels? They are a very smooth DC supply...

Re: noise filtering on PIP inverter

Posted: Sat, 11 Apr 2020, 22:30
by vincent.p
In fact, any voltage/current converters acts as a switching powersupply.
During switching high currents/voltage are switched at a high frequency, generating disturbances that are emitted and conducted.
From the PV point of view, yes it is smooth, but believe me, the inverters inject a lot of RF and unwanted signals back to the PV cables.
If you want to check, just put parallel and close together your USB/RS232 data cable and the PV cable on the inverter side. If sometimes you loose the data, you will know why.
On my side i can tell you that my radio piloted clock can't synchronize as long as the PIP is working. Switching off the PIP, it comes back. Andd the clock is 15 meters far from the inverter including a concrete wall.
For EMC test, it seems that MPP solar is never doing conducted measurement on the string imput and the battery input too.
Every one can check here of the reports.
Conducted Test are only done on the mains access. input for string and battery are never tested and even not connected from the set-up schematic...

Their standard test configuration is powered by mains and used to power a laptop...that witdraws probaly less than 50 watts...
All the curves got are already borderline... i can only imagine what it should be if everything is connected.
So, this is why we pay a low price.

Re: noise filtering on PIP inverter

Posted: Sun, 12 Apr 2020, 07:40
by coulomb
I note that the PIP-5048GK models (Axpert VM III) have the 500V max solar charge controllers, so as well as the switching noise, they have the 230VAC (chopped up a bit) superimposed. In battery mode, I guess the inverter proper noise is superimposed as well. Is it quieter in line/bypass mode?

I wonder if a model with a 145V max solar charge controller might be quieter.

It might be worth considering an external MPPT, somehow choosing a quieter model. It would connect directly to the battery instead of to the noisy 400V bus, so that would be two factors leading to a quieter radio environment.

Re: noise filtering on PIP inverter

Posted: Sun, 12 Apr 2020, 08:59
by paulvk
I wonder if a model with a 145V max solar charge controller might be quieter.
They appear to be as my radio is still usable and running off the PIPs , but they still have a lot of RF getting out.
RF interference is something that has been getting worse over time.
I might have to get my spectrum analyzer out to see what is getting out.
One thing that may help is an old "E" core transformer and running the incoming PV cables through the core
old microwave ovens are a good source of these with easy to remove winding.

Re: noise filtering on PIP inverter

Posted: Sun, 12 Apr 2020, 11:42
by T1 Terry
We had serious issues with the early PiP 24v PWM controller models and sorted it by looping each cable through a ferrite choke. I note that the Projecta mains chargers do the same thing inside the unit so they must have had issues as well.
I wonder what voltage the solar input is, both open circuit and Vmp, could it be the MPPT clipping the input to avoid over current on the output? Although 3kW of solar should be even close to its capacity .... but it still might have a high voltage problem. Once the battery charging side goes into PWM mode, fast switching can cause a lot of noise on a radio AM band

T1 Terry

Re: noise filtering on PIP inverter

Posted: Sun, 12 Apr 2020, 18:31
by vincent.p
Thanks for your comments,
to me the lower the current switched , the lower the RF noise. High Ldi/dt (high current change for small time) generates high voltage across inductors made by any length of PCB tracks or wiring cable coupled to each other like a transformer. With a Lower voltage MPPT, the switched current is higher and should generates more noise in theory. Of course it depends, also on the design itself.
If the behahiour is practically seen better with low voltage MPPT, it's a clue also about probably better design.
Finally, the worse case should be then when running on battery : lower volatge higher current and finally higher noise.

Regarding my set-up, the 3KW are never reached. Open voltage is max 460V, and Vmp is 375V. I agree that the 460 are too high regaridng the working range of the MPPT (up to 450 and 500V max). Then it is only a matter off not switching off/on the PV under full light.