ZK-SJ30 Buck Boost 30A Converter uses LT8705

[@davgregeric]

Had one of these new SJ30s in service today as a solar charge controller. Small string of panels on an east facing roof at 65V MPP, feeding peak 420W into my fork lift truck's 10kWh nominal 24V flooded lead acid, or around 15-16A at ~27V output. The SJ20 would have hit the OPP limit and turned off. I do like the flexibility these buck/boost converters afford... my SJ20 is currently creaming off power up to 300W from the fork lift's battery into 3kWh LiFePO4s, twiddling the MPPT pot to act as UVLO to shut off power draw below 26.5V set point. The SJ30 will be put to work in such surplus boost mode when further panels are installed on the roof, when a dedicated high power & voltage true MPPT buck converter will be needed to charge the big 24V FLT lead acid..

[@ecospider5]

Nice to know the SJ30 didn’t just instantly blow up in that situation. I just purchased 3 to charge some Lifepo4 batteries

[@aspendell209]

In boost mode the current limit of the module would be measured on the input side. In Buck mode the modules current limit would be on the output. So on your 20 amp module you measured >10 amps on the output while boosting to like 18v, while the input was almost certainly pegged to the 20 amp input limit.

[@JulianIlett]

Yes, the SJ20 seemed to have an input current limit of about 19A.

[@SpeccyMan]

I'm thinking these units might be more a subject for Big Clive to take a look at.

[@MrSummitville]

Don't look too closely, otherwise you will be disappointed.

[@ecospider5]

You definitely get what you pay for. And I really like playing with cheap crap. But don’t run your C-pap with this. 🤣🤣🤪😜😆

[@TheDefpom]

I would say limit on all units should be 10A.

[@JulianIlett]

Very sensible. But then there's no reason to buy the SJ30.

[@TheDefpom]

@@JulianIlett build quality doesn’t support the claimed rating anyway. It is common for ratings to be exaggerated on products from AliExpress, my general rule is to halve whatever it says, and if it can actually do it (rare) then it’s a bonus.

[@mikko3d]

Thank you for the testing! Pretty disappointed with the construction of the SJ30, I'll stick to my SJ20. I 3d printed a fan holder for a side-mounted 60mm fan. That helps with cooling the caps and the inductor as well.

[@user--PM]

not sure if you have a thick enough thermal pad, but you could try using one of them to transfer some heat from the board to the heatsink, otherwise i would add my own heatsinks to the top of the board with some thermal adhesive to aid in the cooling.

[@ecospider5]

I was thinking the same thing.

[@simosagimain]

Considering the already very dubious design, I wonder also if in addition to the terminals also some of the tracks are not really sized to carry 30A.

[@DustinWatts]

I keep making the mistake in my head that 7805 is a 5V regulator 😅

[@JulianIlett]

Yeah, I keep calling it a 7805 - it's an 8705.

[@DustinWatts]

@@JulianIlett lol

[@simontay4851]

But a 7805 IS a 5V regulator. Its not a mistake.

[@DustinWatts]

@@simontay4851 There are LDO's, with 7805 in it's name nut the 7805 is a house hold name for a 5V regulator. Easily mistaken if you mean 8705. I'd probably would make the same mistake over and over if I would make a video. I'm not going to, Julian beat me to it 😜

[@ecospider5]

I used to be able to keep track of numbers like that. Not anymore. I still know my brothers phone number so I’m not doing to bad. :)

[@power4one2go83]

Im an electrician and i have to say that the connector is made to work with 10amp at 300volt max peak voltage isolation 😏 maybe! But like most connectors that comes some time with an high voltage AC and a low voltage DC Amp difference, the high voltage may be 300V 10A and the low voltage +/-30V may be 30A 😯 I kant give equation for because the difference vary depending on the conductor metal type and grade and resistance and more 😔 It is always best to ask the manufacturer or check the datasheet if there is a ID 😅 on it.

[@g2tana]

I use 1 of the 1500w ones for charging my 18s scooter lipo and the connectors melted, charging at 20a. I removed them and soldered directly to the board and it works so much better it doesn't even get as hot now

[@SuperBrainAK]

Ooh I love the LTC8705! I've been looking for high current ones but nothing good over 10A. Hopefully the SJ models are popular enough that someone is going to make their own, see the downfalls of their design and do it again better.

[@frankowalker4662]

With that circular fan mount on top, it looks like a camping stove. LOL.

[@JulianIlett]

Without the fan you could probably cook an egg on it :)

[@frankowalker4662]

@@JulianIlett Ha ha.

[@TWforage]

I've been looking for this for the past two days. Before I ordered it, they advertised it as having MPPT capability. But now, I discovered it's based on the LT8705 chip, which only functions as a buck-boost controller, not a real mppt?

[@JulianIlett]

It's not real MPPT. You set a target input voltage and the unit pushes the input voltage down by increasing the load (increasing the switching pulse width) until that target is reached.

[@TWforage]

@@JulianIlett Thanks for explaining that!

[@absolute___zero]

you have felt for the marketing trap, they have brain-washed you that you have to buy the top-notch product because it is super efficient so you buy MPPTs for 500 bucks, but then when your batteries are full you throw away all the solar energy to the trash because you can't absorb it anymore. A smarter choice would be buy a PWM converter or a device like this one, at the end you spend less but get the same result. It is kind of like buying a sports car with 300HP when you live in a city where you only can drive at 60km/h

[@MrSummitville]

​@@absolute___zero PWM is ok for small / simple systems. But when you have 4 strings each, of 10 PV Panels connected in series, you *must* have MPPT. Nobody is brain-washed. Clearly, you have no clue how to design a large PV array.

[@absolute___zero]

@@MrSummitville my pwm buck converter bought in china (using latest modern IC) for 30 dollars 50A on output gives me 95% efficiency. Are you saying that 3% extra efficiency of MPPT(to have 98%) worth it? No way! Yes, I must put a diode to prevent backflow, but 10 pack of 80A diodes costs me 9 dollars + heatsink + assemble it yourself. Still way cheaper than 500USD mppt. Marketing propaganda is what you're buying.

[@ddistrbd1]

I have not has a good experience with boost converters, burnt a few, for me (as a total novice) the only "work around " in order to avoid destroying a boost converter is , to know exactly the amperage of the input before going inthe boost converter, in other words I avoid adjusting the current on the boost converter because I just don't trust them doing it right.

[@ecospider5]

Careful putting 15v to a car headlight I have been using one for years as a load but destroyed it this year by putting a little to much voltage to it. I don’t remember the exact voltage but it was more than 14.5v and less than 16v.

[@sebadyoda]

Awesome like always

[@antibrevity]

That's a stupid heatsink design and the heatsink likely hurts more than helps as it blocks airflow that might otherwise cool the PCB. They spent the money on a nice heatsink and then used it improperly. The SJ20 is the better option.

[@panospapadimitriou3498]

new toys .. and i needed one for the spare i burned. about terminals. btw i always break those plastic walls of those terminals to fit at least 2x1.5 mm copper cables on one crimped ringlug on each terminal and reach more than 10 amps cooler.. but only for some momments !!

[@iowahank1]

300V at 10A is 3000 watts. 3000 watts at 12V is 250A. The connector should be able to handle the higher current at 12V.

[@FireballXL55]

No that is the maximum current they are made to handle, the voltage is irrelevant. The voltage is given as the maximum rated voltage between the pair and also I would guess to the mounting surface.

[@julzes]

Where can I purchase the H7 bulb load testing pcbs?

[@JulianIlett]

You can't - I made them.

[@DavidBTB]

Are the current shunts on the SJ30 touching each other?

[@JulianIlett]

Not quite :)

[@rilosvideos877]

These cheap Pots often do fail and lead to malfunction in electronics. I bought quite a few 10xpots in several values but i later recognized that they tend to fail also! It is worth to spend some more money on reliable components that do work for years without trouble. The overall design seems to be questionable with heatsinks on the plastic part of the Mosfets! They simply don't care much about safety and stability.

[@alisoncarney1745]

15A/20A 900W/1200W DC Power Supply CNC DC Voltage Regulator XY6015L/XY6020L UK please have a look at this enjoy your youtube alison😂😂😂😂😂

[@ecospider5]

Do you know what the CNC stands for?

[@harrygood4783]

Well i think maybe you could increase the size of the output cables, the reduction of output power could compromised back the back flow of power returning back to the unit, this also could have an effect on the input power causing it to limit, but they do look really thin for a 30A flow, i would expect them to be at least twice the diameter, i most cable's of a good quality normally have a printed rating on the cables, just an idea Julian and trouble with china made components they are substandard, and made cheap, but that's mast production.

[@JulianIlett]

Oh yeah, the cables were grossly undersized and getting quite warm. No sense using properly spec'd cables just for a bit of testing.

[@Mr1Spring]

Bad design on all fronts. Including dodgy components: One potmeter blew up an IC Do not buy. And Julian, send them back - reason DOA

[@ecospider5]

I just ordered 3 of the ZK-SJ30 for $27 each. I will get what I paid for :). We will see how many I send back. I plan on only using them rarely so we will see if I can get them to limp along I’m definitely wondering if I should bypass the 10amp connectors.

[@davadoff]

7:47 PCB bending can crack and short those ceramic capacitors, causing the magic smoke to be released or worse. Bad design.

[@davadoff]

You really need to get a light for filming. I’m 3-4 mins in. The video is way too dark.

[@ecospider5]

That being said. Thanks Julian for the great free information.

[@cmyanmar13]

The current ratings on every Chinese power converter seem to be, at best, peak current, not sustained.

[@321tryagain]

Terrible thermal management and potential for self-annihilation due to cheap potentiometers. The SJ30 seems to have terrible design, would you agree?

[@JulianIlett]

The SJ20 is a better design.

[@pault6533]

The heat needs to be spread properly before air flow can do its job. Completely drop the bottom side heat sink and attach smaller heat sinks to the exposed metal patches on the top of the board. Only then will the top side fan be effective. But the fan will block access to the potentiometers.

[@davadoff]

If the PCB layout was done properly, then the pot going open circuit would be like turning it to its maximum resistance. As opposed to an infinite resistance.

[@simontay4851]

I think i will buy the SJ20. The SJ30 is stupid design.