Radio Noise from PV System

[PNjunction]

Wow, my Icom R75, and Alinco DXR8T seem to have no problem.

What kind of radio are you using, and how long is the dc run from the battery? Are you hearing some sort of pulsing? If not, perhaps your panel wiring is acting like an antenna itself, and the noise isn't from the Sunsaver per-se. Is the noise consistent across a broad spectrum, or only on specific frequencies? Are we talking HF, or maybe the am broadcast band? I'm assuming you have definitely nailed it by disconnecting everything, including the Sunsaver, and tested by only running from the battery and that single run of dc back to your receiver....

I'll borrow a friend's Yaesu FT-897D portable, and take it right on top of my Sunsaver with a pair of cans and see if I can force the issue to hear it myself.

[niel]

actually some noise appearing at some frequencies is normal. it depends on many factors and the noise just may be below your radios noise floor and is why you don't hear it, which is a good thing. it does not mean that somebodies stuff, be it ham or solar, is defective if it appears even though it is undesired.

[PNjunction]

Borrowed the 897D transceiver and ran some quick lunch-time testing..

With the transceiver running on it's own power, I heard no noise coming from the solar setup with the sunguard or sunsaver PWM cc's.

With the transceiver sharing the same battery as the CC's, I was extremely pleased not to hear any broadband hash. However, I was able to track down a small number of noises coming from the CC, although I think it is from the supporting circuitry, and not PWM itself, which for the morningstars are a duty cycle on/off switch up to about 300hz which is basically audio. I discharged the batteries a bit prior to the test to make sure that the duty-cycle would be active. Being able to quickly connect/disconnect the CC and switching amongst internal and external power on the 897 made it easier to identify.

During the short test time available, I was able to identify an unstable CC buzz with the same characteristics on 2.034 Mhz, 3.775, 7.083, 14.160 mhz. There were quite a few weaker ones up in the 10 meter band which I didn't log. Still, with only a small handful of freqs to deal with, there is NO way I'm going back to the 1980's non-temperature compensated on/off shunt controller. I normally don't run from batteries that are under charge anyway, so it's kind of a moot point, but I wanted to know anyway.

I might end up doing some rf-choking and capacitor bypassing, but I can easily live with just a small handful of problem freqs that I don't use anyway. My biggest fear was broadband hash, and that is really not the situation here. Whew!

[niel]

that is good news of it not be a problem for you, but the tests weren't necessary as i knew there was some to be found if one looked for it. i simply said that no mppt controller is noiseless. i never implied for you to go pwm or intended to badmouth all mppt controllers. my point is that it is simply a fact that there is some noise and does not mean that there's something wrong. even some layouts could exacerbate the noise that is there. in some cases it does fall onto desirable frequencies with enough strength to warrant further efforts.

[daniel]

Hello:

I have come to the conclusion that it is a really bad idea to put ferrites on the wires going into or out of the charge controller. These ferrites increase the inductance of the wire and if they are large enough they can stop the FET transistors from turning on and off correctly. It would take fairly large inductors to work with a high current system but even small ferrites are going to make it harder for the FET transistors to switch.

Daniel

[Vic]

It is good news that you are happy with the noise profile of your PWM CCs.

Would guess that the more power that the CC is delivering to the batteries has an effect on the PWM noise. PWMs should be fairly quiet, but fast edges lead to improved efficency, but fast edges can also cause noise issues.

Often the polarization on the receive antenna, its capture area -- relative size, proximity to PV array and so forth are large factors. This is a large factor in Line-Conducted noise.

PN, you may be measuring mostly Radiated noise.

Often, True SineWave Inverters create more noise on HF than PWM CCs.

As an aside, many 12 volt HF transcievers do not make full power when operating from 12 volt batteries which are not under charge at that time ... some are much better than others. Some have very poor transmit IMD specs when Vbat is below 13.5-ish volts. YMMV, Vic

[BB.]

Welcome to the forum Daniel,

I tend to think that the whole ferrite on controller leads is a bit more complex...

First, there is the ferrite materials. Ideally, you want to look for "lossy" ferrite material used to control RFI/EMI emissions. "Good" or low loss ferrite is great for transformers/inductors because they have no losses. But using this material to control EMI--The low-loss ferrite + wiring can "ring like a bell". Lossy ferrite will dissapate energy and reduce the chances of ring (high Q vs low Q materials).

Next, the way you pass the wires through the (typically) donut ferrite core.

You can pass one wire through the donut (or even wrap a single wire a couple times through the donut hole). This will not usually work well for DC currents. The high current flowing one direction will tend to saturate the ferrite and reduce any EMI control effects. Also, this add inductance the cable and could cause problems with the FETs.

Another is to take two wires (+/-) and pass it through the donut (and possibly wrap a turn or two through the donut). This is a "Common Mode" choke. This is good a blocking signals traveling down both wires at the same time (in phase). This also should not create problems for the FETs (the "differential" inductance cancel each other out). This also works well with differential DC current flow because the +/- current flows cancel each other out and will not saturate the ferrite.

-Bill

[boB]

Hello:

I have come to the conclusion that it is a really bad idea to put ferrites on the wires going into or out of the charge controller. These ferrites increase the inductance of the wire and if they are large enough they can stop the FET transistors from turning on and off correctly. It would take fairly large inductors to work with a high current system but even small ferrites are going to make it harder for the FET transistors to switch.

Daniel

Daniel, how much inductance do you think would be a problem ?


What happens to the FETs turning on and off with too much inductance do you think ?

Thanks...
boB

[PNjunction]

Ah, great information all! I really appreciate it since I understand how hard it can be to provide the most accurate answers without just blasting out a blanket-statement - every install / situation / usage pattern is different. In the end, it is still my responsibility to apply the answers received and see if it fits my own scenario. Fortunately, you guys are right on the ball for the most part!

Guess what was on my CC leads? Yep, HF-spec ferrites. I'll take them off and do some more testing later. I'm very familiar with common-mode rejection since some of my antennas have the common mode filtered, whereas some actually make use of the common mode - it all depends on what you want to do. However, I really wasn't thinking DC, and will investigate the additional inductance / saturation aspect with regards to the CC operations. Nice tip!

I love experimenting, and sometime later I'll probably get an mppt just to play with it, even though it isn't really necessary for me right now. No badmouthing taken at all. And nobody is on the hook so to speak for their responses since one size doesn't always fit all. :)

[BB.]

If you can run the electrical wiring in metal conduit (grounded to the source of the noise, such as the charge controller) will help too... About 10-15' of conduit (or shielded cable) is about the maximum "helpful" length. Any longer shielding will not make much difference (most RF energy has been "shorted" from the cable to shield by this point--And the natural impedence of the cable does not carry RF signals past this point).

-Bill