How does the MPPT potentiometer on the ZK-SJ20 buck boost converter work?
Power Queen 100Ah LiFePO4 Battery:
https://uk.ipowerqueen.com/?ref=eqsfmkg1
ZK-SJ20 Buck Boost MPPT Converter 300W
https://www.aliexpress.com/item/1005004966703221.html

Hello Today I'm charging the power Queen 12.8 volt, 100 amp hour life Pro 4 battery using the Sj20 power supply fed from a solar panel. Now this is a flexible solar panel. It's an odd one. actually.

it's um, got 32 cells. now. Normally these panels have 36 cells, but it does claim that the maximum PowerPoint is 17 volts. Let's have a look.

Yeah, so this is a 100 watt panel. Allegedly, 17 volts is V Max power IMAX power is 5.9 amps and then you've got a short circuit current of 6.1 amps because we won't see that because we're trying to get maximum power and I'm running the gauntlet a little bit between Sun and clouds. but let's see how we get on. Oh pretty good.

78 Watts by setting the Mppt voltage to 17. let's take that down to 16. at 74 Watts back up to 17 volts, 76 watts and if I take it up to 18 volts, it falls off again. So you can see we're trying to find this elusive maximum PowerPoint by setting the voltage at which this power supply puts on the solar panel in order to get that voltage.

So it seems that the maximum PowerPoint is somewhere around 17 volts. I Think we have a cloud. Of course, just after the cloud disappears, we get the maximum power and then that drops down gradually as the solar panel warms up. Because as panels warm up.

Oh another. Cloud I'm afraid. Okay, yes, as solar panels warm up their potential difference, their voltage drops. The current isn't affected in much, the sun's come out again.

But it does mean that the power drops off because of that drop in voltage. Can we get 80? Watts Not quite like I Say when the Sun goes behind the cloud, the panel cools down a little bit. Let me just get a little burst of extra power before it settles down. Unfortunately I Can't do anything about the flickering of the display.

It's a cloudy moment now. but I Just wanted to look at the other values on here. We've got the output voltage. Now that's essentially the voltage of this battery when current is Flowing It will look a little bit higher here because there's a bit of Vault drop in these cables, but the battery is at 13.3 volts the input voltage, the voltage coming from the solar panel 16.2 and the set point.

The solar panel voltage always seems to be a little bit below the set point, but the set point is set to 17 volts. The sun's coming out from behind the cloud does seem to take a little bit of a while for this to get up to maximum power as it allows the solar panels voltage to creep up. Yep, there we are. That's maximum power.

about 78. Watts Why is it dropping off? There must be some hazy cloud. Yeah, the sun's just sitting on top of a cloud, so it's uh, kind of half behind the cloud and half knot. I'll wait until it we get a bit of clear sky.

now. When I had this panel connected directly to the battery I was seeing six amps. but you mustn't be deceived because although I was getting a bit more current, I was getting a lot less voltage because the panel would course be pulled immediately down to the battery's voltage of 13.6 and not allowed to float at this higher voltage of about 17 volts. So although I was getting more amps and I'm getting less amps in this scenario, if I press and hold this, you can see I'm getting more watts.
So this is doing its job. It's allowing the solar panel to float up to its maximum power point, passing that through the efficient in this case Buck converter because I've got a high voltage on the input than the output and pushing current into the battery on the output. this Watts figure is the output Watts Actually, so the input Watts will be slightly higher because there is a bit lost in the electronics I just nudged this Mppt point up to 17.8 I Don't think that's helping, so let's take it back down to 17. Yeah, I'm getting more watts.

Um, Interestingly, if I take that all the way down to about 13 volts, which is what the battery voltage is, we don't lose that much power. 63 Watts at 13 volts of solar panel voltage and at 17 volts of solar panel voltage, 70 Watts it I was getting 80 Watts before why is that? Is it not a completely clear sky? Well, it's clear where it's clear. I can't look directly to the sun I Don't want to put my camera directly at the sun either. So at 70 Watts Um, the heatsink is warm, but it's not.

um, uncomfortably warm. Press and hold that to get amps. 5 amps is quite a low current. This allegedly can take 20 amps, but you do need air cooling.

These capacitors were getting hot when I was testing it at 10 amps. That's interesting because they're not warm at all now. So I wonder if they're just getting air cooled because there is a bit of a breeze? Now can we simulate uh, the sun coming up in the morning? Well, not directly because I'm just going to shove this connector in and um, see what the power supply does except boots. It should turn on the output that red LED there is on and yes, it immediately goes up to the Mppt point.

The Mppt lead is also on and we'll play with trying to make that go off in a minute. Um, but yeah, that shot up to the 73 watts. Can I improve on that 73 Watts Down a bit on the CV point? not really up a bit on the CV point. No, that's definitely gone in the wrong direction.

Definitely seems to work best at around 17 17 and a half volts. Oops, There's a bit of pots a little bit um, not very smooth. That seems to be the maximum. Now if I turn down the constant voltage on the output pot, you can very quickly get it to just collapse away to nothing.

So we're probably limiting the voltage on the output to a couple of volts. Now, raise that back up and that feedback loop doesn't interfere with the Mppt feedback loop. In terms of the current part. If I turn that down.

this is another one where it's not very smooth actually. I've got a screwdriver here. Let's use that. Uh, turn that down.

Should see the CV LED Come on. that's the one directly above. Yes, it's Come on. Now you just see it glowing red directly above the red.
LED here. CV LED has Come on. and we've lost power output. So you really don't want to be current limiting on the output either.

So we'll raise that up so that we're not current limiting and we're back up to 70 watts. Can we get a bit more by raising up the output? Con: The output voltage limit? Not really. No, so that looks like it's about the maximum we can get.

By Julian

Youtuber, shed dweller, solar charge controller aficionado

11 thoughts on “Part1: testing zk-sj20 buck boost mppt function”
  1. Avataaar/Circle Created with python_avatars Nginmu Mbetse says:

    I'm finding my panels (old secondhand 230W panels ex solar farm, roughly 40V Voc, listed Vmp 29V) are giving me the best output at 26V on the MPPT twiddle pot, today. Really cloudy October day late afternoon, panels in shade, getting a whole 2 Watts! I noticed that as I creep up on 26V on the MPPT pot, the output rises very steadily until just above 26V, there's a very sudden collapse of the output. I guess this makes sense when you look at the shape of the power curve. I got to thinking – is it better when setting the MPP 'target', to choose a level a little lower than the one you think the panel's going to operate at, most of the time – so that although you'd be harvesting a little less power overall, you might avoid 'falling over the cliff' on sunnier days, and stay on the more gently sloped side of the power curve? I don't know how much the MPP shifts around but I might try say 24,25V instead of my 26V and see if it gives more 'consistent' operation

  2. Avataaar/Circle Created with python_avatars Simon Andrew Crane says:

    Good demo but would be interested it's efficiency (not watched part 2 yet).

  3. Avataaar/Circle Created with python_avatars tudor nedelcu says:

    There are many constant charging variables influencing the MPP . Maybe this can be the case for controllers with a small AI implementation to track the maximum power of solar cells. Potentially, later loading the discovered data into a multi dimensional mapping array; similar to what we have on car ECU(s).

  4. Avataaar/Circle Created with python_avatars ratgreen says:

    I would call this a MPP in that it is able to be adjusted to its maxiumum power point. But no T as it does not track, ie automatically adjust itself (to try obtain the maximum wattage from the panels available amps X volts)

  5. Avataaar/Circle Created with python_avatars Graham Gillett says:

    Not much a MPPT controller when you have to turn pots for maximin smoke!!!

  6. Avataaar/Circle Created with python_avatars jlucasound says:

    ๐ŸŽผ๐ŸŽต๐ŸŽถ"Here Comes the Sun, Doo-N-Doo-Doot…"๐ŸŽถ๐ŸŽต๐ŸŒž๐Ÿ˜โญโญโญโญโญ

  7. Avataaar/Circle Created with python_avatars Krzysztof Cygan says:

    Do you get much power boost if you attach garden hose to the back of the panel to keep it water cooled?

  8. Avataaar/Circle Created with python_avatars lez briddon says:

    resurrect the muppet!

  9. Avataaar/Circle Created with python_avatars deslomeslager says:

    The idea of MPPT (t stands for tracking) is that if a cloud passes by (not a huge one of course), it still manages to find the sweat spot (max Power). Your device does not track. And I see many comments saying exactly that. You can get better MPPT devices, and they are (or should be) worth their money if they do their job well. If there is a good working ESP32 version anywhere, I would build it.

  10. Avataaar/Circle Created with python_avatars electronzapdotcom says:

    Nice test.

  11. Avataaar/Circle Created with python_avatars Charles Manning says:

    These are not really MPPT chargers because they don't do tracking. They operate according to a setpoint.
    However it should be quite easy to modify something like this into a true MPPT by adding a microcontroller to monitor the voltage/current and "twiddle the pot". A simple Arduino circuit should do it.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.