Been waiting on a higher power buck boost module to come out. The previous one that you could buy was a 10a with no display. I think they had issues with it and took it off the market. This module looks good especially for solar. Great video! Thanks for letting us know this is available!
I bought both of them (SJ20 and SJ30) version. So, here is what I noted after half-hour of playing with these things: - The manufacturer that makes these modules also makes regulated power supplies, they have lots of other products that start with "ZK" prefix, for example: zk-6522c . All their stuff uses microcontrollers, so they are probably good at programming them and hence a unique module design idea to control the buck converter with a microcontroller - So the board (with display) is composed of a buck board (or "motherboard"), and a microcontroller board (the display). Now why do you think the display is detachable? Well my hypothesis is: because these guys also manufacture power supplies (regulated, laboratory-style), they made a generic board based on the LT 8705 chip (very good IC, it even has sync pins to sync many of these in parallel) which they are using on their other power supply products. But how do you sell a buck converter to the solar guys? Exactly, you label it as "MPPT". A regular buck converter has a potentiometer called "low voltage", but for the solar guys they relabeled it as "MPPT". This way you manufacture the same board, but for solar guys you sell the PCBs with "MPPT" label, and for the electronics guys you sell the same board but with label "Low voltage protection" (and you don't have to print it, because we all know how it works). If you sell a power supply, then you insert the display board in the already existing slot. Result of this design: same "motherboard" is reused everywere it can be. - The reason you have to short the outputs to set the overcurrent protection is because the overcurrent protection goes on the display (the microcontroller board). But when you already have the PCB designed and used in other products, how do you control it with external board without modifications? Exactly you control it by detecting a short-circuit. On short-circuit the voltage drops to 0 immediately so there is no mistake. They found this easy solution for them, but for the User it is not very pleasant. Imagine a 30A current flowing on the output and you have to short-circuit it (in order to set overcurrent protection value), you will get a lot of sparks and probably smoke. While you are turning the potentiometer the short-circuit must be kept active, otherwise the value won't be set, you could possibly heat the screwdriver a lot if you aren't very quick. They could do it with another button, but because they didn't want to modify the buck-board (and this is my speculation because I don't know the real reason) they decided to go for "short-circuit way" to tell the microcontroller board that it has to set the output current value in its memory. Only engineers from that country can come up with such idea. - The fan can't be turned off, doesn't work on temperature-sensing. Too bad Will keep testing and maybe find another issues.
I bought 2 of these. 1 with and the other without the display. The mainboard works exactly the same just transfer the display to the other. The display also has a power limit function. If the output exceeds 300W, it turns off. You have to push the on/off button to resume. If you don't have the display module, there is no power limit until something blows up.
@@JulianIlett there's another "mppt" module available I'm not sure if you have seen it. It is boost only but based on LTC1871 and is advertised as 10 to 60v in, 10 to 85v out. 600w max. I bought a few and have one used for two 18v panels in series to the 48v home battery bank. It's been running for months, gets warm but quite reliable after setting up. Google LTC1871 600w mppt and it should come up.
Hey there, which buck-boost would you recommend today, with the following properties (from higher to lower priority): - efficiency, no or very little heat dissipation, - protected with a resetable fuse - voltage and current display - size of module preferable small - solid input and output Terminals - temperature controlled fan, best without one needed - voltage 5V in / out - cuttent 10A in / out - power 100W to 200W in / out
check out the blue buck-bust converter with LTC3780 chip, costs 6 dollars ,synchronous rectification, 10A . The best thing I have ever found on Ali. You can parallel them up and they won't burn. Ceramic fuse included!
The output capacitors most likely get hot because of the high ripple current they carry when in boost mode. (In buck mode the input capacitors have the high ripple current). The worst case output ripple current occurs when the output is twice the input voltage - which is the case in your test - 12V in to 24V out. Making a few assumptions about the switching frequency (200kHz) and inductor (22uH), with 10A output the output capacitors have approx 10Arms ripple current. Those 2 x 330uF 100V won't be rated for that (probably 2.5A max each, probably less), and they likely won't share the ripple current well (because the hotter cap of the two will have lower ESR and thus hog the ripple current, getting hotter still). The designers should have provided 10uF or more (40uF would be good) of very low ESR, high ripple current ceramic capacitors (or expensive polymer/hybrid electrolytics) paralleled with the input and ouput electrolytics. Problem there is that multilayer ceramic caps suffer drastic loss of capacitance with DC bias - eg. a 1210 10uF MLCC 100V X7R might be less than 2uF with 80V of DC bias. Thus they would need at least 5 of them for 10uF net capacitance. I don't see many MLCCs on the supply so the electrolytics are likely overstressed and will fail early and may even explode if operated at 1:2 voltage ratio, 20:10 amps for any length of time. This is a common problem on cheap convertors beacuse high ripple capacitors are expensive, especially for higher voltages. Note that the capacitor ripple current is independent of the input and ouput voltages - only the currents and the voltage ratio.
Yeah, I believe this is what's happening. The capacitor temperature isn't always directly related to the heatsink temperature. This unit will be tested long-term in my shed. I'll let you know if the caps fail.
Got an idea for you! You should "analyze" the RFI from these things, would help people in buying one and would also lengthen your videos to get you to the magic 10+ minutes as you test them under various loads/etc.
@@JulianIlett Haha! You are too polite. Love the videos. Thanks. Thought I'd make a brilliant suggestion since you love doing videos on buck convertors.
The title should mention mppt tracker cause I missed to watch it and found it only by the most recent video which showed a flexible 100W panel . Aliexpress calls it also MPPT and it is a quite affordable mppt one for roughly 30€. I like this one a lot for a 300W, but there are also for a bit more units that come with a case and fit far better into a RV.
Hello Julian, can you help me with what size stud fork connectors I should get for this unit's In/Out screw terminals? #6, #8, or #10? I am new to DIY. Thanks a lot!
I think i will buy this but i'll replace the caps with higher quality panasonic FR ones. The input will never be more than 30V so i'll replace the 100V input caps with 35V ones and the output caps with 63V ones. I won't ever need more than 60V. I'll power it from a 24V AC toroidial transformer.
Nice little module, but the MPPT seems a bit misleading. On the aliexpress site they really talk about MPPT of solar panels, but you have to set it manually which is not MPPT. Would be very interesting if it really could handle 300W as stated. I doubt it can handle 20 A for longer time than a few seconds without fan. Also for a buck-boost the efficiency would be interesting. Maybe you can add this to your test, for different load situations would be optimal 🙂
True, it's not 'Maximum Power Point Tracking'. I call it maximum power point targeting. Apparently the unit cuts out at 300W with an error message. Although how that works when the display unit is not attached, I don't know.
@@biezitis I have bought 1500W (50A) buck converter on Ali for 33 bucks (with shipping included), it manages 2 solar panels of 550W , so I am getting 94% efficiency (measured it myself so nobody can fool me), why the heck do I need an MPPT? To gain 4% more? Come on.... 4% doesn't worth paying 10x the price of a buck converter.
When you say single inductor but also buck-boost, on the basis it's described as synchronous, and "like the 3780," I would hazard a guess what you're actually looking at is a SEPIC converter. I think buck-boost has just become a shorthand for a DC converter that can go up or down. Possibly I'm splitting hairs, but it's an interesting development to see more of these.
No, it's not a SEPIC converter. Both the LT8705 and LTC3780 are described as Synchronous 4-Switch Buck-Boost Controllers. The data sheets explain it. But I have just bought a super little SEPIC buck-boost converter, fully digitally controlled (using the FP5139) for a very specific purpose. Details soon :)
@@JulianIlett thanks, interesting. I'm always on the lookout for a DC converter I can control with some sort of data input, something with a decent enough response time to implement a PID for LED current control. Never found one yet... Be interested if you've come across anything that might work.
When are we going to see the next shed extension to house all these bucking boosters? Are you taking a leaf out of SDG's book but replacing soldering iron stations with DC/DC converters? 🙂 I'd hate to see your AliExpress wish list. ;-)
This seems ideal for charging an 36V or 48V ebike battery from a small ~18V solar panel, right? Without the "MPPT targeting", traditional boost converters will collapse the voltage as they try to max the current. One should only charge a battery with a BMS so it can cut off the charging when full.
Prolong float charging at full-charge voltage of lithium batteries shouldn't be used but it is fine with lead-acid batteries as long as the regulation is good. All switchmode converters have a negative input resistance characteristic, so they will all collapse the voltage of a PV array if the array can't meet the power requirement and there is nothing to prevent collapse.
hi @JulianIlett do you know what type of fet transistors are installed there? i am wondering if I can to change them to more powerful ones (with less resistance) if this module will be used not for 80 but with 40 volt panels. don't want to install additional cooling )
I know the OPA1612 as I have used them as audio opamp. They are pretty low noise and have low THD. Pretty good for audio use is my experience. They are also marketed as an audio opamp (SoundPlus is the line of opamps from TI). But their high precision makes also sense as a feedback opamp to give a very precise output voltage which I believe they are using it for in this case. So surprised to see it and also not surprised to see it :D Edit: It is not a cheap OpAmp btw. €3.31 per unit if you buy 1000 on Mouser. A single unit is €5.88.
Yes, there's a brief overshoot of the output current limiter system. Not a problem if charging a battery (or even a supercapacitor), but it could kill a fragile load.
Always enjoy watching your videos. Out of interest, I would like to give something back to you and this community, would you be interested in a free PSU for your projects and testing modules.
@@JulianIlett I have a Mean Well and TDK-Lambda switch mode power supply's. I have a list of them on the eevblog forum. Please feel free to pick what one would be best suited to your requirements.
Thanks so much for the generous offer. I've had a look at your post on EEVblog and I notice these are mains PSUs. I now only use battery power supplies (power banks / power stations) for my experiments.
Great video as always. When you speak about 10A,.... Input ? Output ? I will like to know if you can test it by charging a 24v battery from a 12v battery. 10A? 15A with fans? So 240w at output is possible? I think you also have an infrared camera so it will be a good test !!!
Been waiting on a higher power buck boost module to come out. The previous one that you could buy was a 10a with no display. I think they had issues with it and took it off the market. This module looks good especially for solar. Great video! Thanks for letting us know this is available!
I like the forks on the XT-60 connector. Nice review.
I bought both of them (SJ20 and SJ30) version. So, here is what I noted after half-hour of playing with these things:
- The manufacturer that makes these modules also makes regulated power supplies, they have lots of other products that start with "ZK" prefix, for example: zk-6522c . All their stuff uses microcontrollers, so they are probably good at programming them and hence a unique module design idea to control the buck converter with a microcontroller
- So the board (with display) is composed of a buck board (or "motherboard"), and a microcontroller board (the display). Now why do you think the display is detachable? Well my hypothesis is: because these guys also manufacture power supplies (regulated, laboratory-style), they made a generic board based on the LT 8705 chip (very good IC, it even has sync pins to sync many of these in parallel) which they are using on their other power supply products. But how do you sell a buck converter to the solar guys? Exactly, you label it as "MPPT". A regular buck converter has a potentiometer called "low voltage", but for the solar guys they relabeled it as "MPPT". This way you manufacture the same board, but for solar guys you sell the PCBs with "MPPT" label, and for the electronics guys you sell the same board but with label "Low voltage protection" (and you don't have to print it, because we all know how it works). If you sell a power supply, then you insert the display board in the already existing slot. Result of this design: same "motherboard" is reused everywere it can be.
- The reason you have to short the outputs to set the overcurrent protection is because the overcurrent protection goes on the display (the microcontroller board). But when you already have the PCB designed and used in other products, how do you control it with external board without modifications? Exactly you control it by detecting a short-circuit. On short-circuit the voltage drops to 0 immediately so there is no mistake. They found this easy solution for them, but for the User it is not very pleasant. Imagine a 30A current flowing on the output and you have to short-circuit it (in order to set overcurrent protection value), you will get a lot of sparks and probably smoke. While you are turning the potentiometer the short-circuit must be kept active, otherwise the value won't be set, you could possibly heat the screwdriver a lot if you aren't very quick. They could do it with another button, but because they didn't want to modify the buck-board (and this is my speculation because I don't know the real reason) they decided to go for "short-circuit way" to tell the microcontroller board that it has to set the output current value in its memory. Only engineers from that country can come up with such idea.
- The fan can't be turned off, doesn't work on temperature-sensing. Too bad
Will keep testing and maybe find another issues.
I bought 2 of these. 1 with and the other without the display. The mainboard works exactly the same just transfer the display to the other. The display also has a power limit function. If the output exceeds 300W, it turns off. You have to push the on/off button to resume. If you don't have the display module, there is no power limit until something blows up.
The power limit is a sensible idea :)
@@JulianIlett there's another "mppt" module available I'm not sure if you have seen it. It is boost only but based on LTC1871 and is advertised as 10 to 60v in, 10 to 85v out. 600w max. I bought a few and have one used for two 18v panels in series to the 48v home battery bank. It's been running for months, gets warm but quite reliable after setting up. Google LTC1871 600w mppt and it should come up.
Hey there,
which buck-boost would you recommend today, with the following properties (from higher to lower priority):
- efficiency, no or very little heat dissipation,
- protected with a resetable fuse
- voltage and current display
- size of module preferable small
- solid input and output Terminals
- temperature controlled fan, best without one needed
- voltage 5V in / out
- cuttent 10A in / out
- power 100W to 200W in / out
check out the blue buck-bust converter with LTC3780 chip, costs 6 dollars ,synchronous rectification, 10A . The best thing I have ever found on Ali. You can parallel them up and they won't burn. Ceramic fuse included!
great little module. Very informative video
The output capacitors most likely get hot because of the high ripple current they carry when in boost mode. (In buck mode the input capacitors have the high ripple current). The worst case output ripple current occurs when the output is twice the input voltage - which is the case in your test - 12V in to 24V out.
Making a few assumptions about the switching frequency (200kHz) and inductor (22uH), with 10A output the output capacitors have approx 10Arms ripple current. Those 2 x 330uF 100V won't be rated for that (probably 2.5A max each, probably less), and they likely won't share the ripple current well (because the hotter cap of the two will have lower ESR and thus hog the ripple current, getting hotter still).
The designers should have provided 10uF or more (40uF would be good) of very low ESR, high ripple current ceramic capacitors (or expensive polymer/hybrid electrolytics) paralleled with the input and ouput electrolytics. Problem there is that multilayer ceramic caps suffer drastic loss of capacitance with DC bias - eg. a 1210 10uF MLCC 100V X7R might be less than 2uF with 80V of DC bias.
Thus they would need at least 5 of them for 10uF net capacitance. I don't see many MLCCs on the supply so the electrolytics are likely overstressed and will fail early and may even explode if operated at 1:2 voltage ratio, 20:10 amps for any length of time. This is a common problem on cheap convertors beacuse high ripple capacitors are expensive, especially for higher voltages.
Note that the capacitor ripple current is independent of the input and ouput voltages - only the currents and the voltage ratio.
Yeah, I believe this is what's happening. The capacitor temperature isn't always directly related to the heatsink temperature. This unit will be tested long-term in my shed. I'll let you know if the caps fail.
Got an idea for you! You should "analyze" the RFI from these things, would help people in buying one and would also lengthen your videos to get you to the magic 10+ minutes as you test them under various loads/etc.
I'd be interested in knowing how noisy these are too.
Excellent idea. The basis for a successful and 'lucrative' TH-cam channel I'm sure. And because I'm nice, I'm not even going to steal it.
@@JulianIlett
Haha! You are too polite. Love the videos. Thanks. Thought I'd make a brilliant suggestion since you love doing videos on buck convertors.
I use XT60 a lot and never thought to put some connectors on it like that. I'd even cover them in solder if you want to try and push a bit of amps.
LT is one of the best maybe the best. Sometimes it's worth the price.
Nice for a first look at yet another buck boost converter 🙂
Squeak!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
@@andymouse CHEESE!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Looks like now there is a beefier 30 amp version of this available (ZK-SJ30).
I'll get one, thanks :)
@@JulianIlett Nice, looking forward to a video about it 🙂
The title should mention mppt tracker cause I missed to watch it and found it only by the most recent video which showed a flexible 100W panel .
Aliexpress calls it also MPPT and it is a quite affordable mppt one for roughly 30€.
I like this one a lot for a 300W, but there are also for a bit more units that come with a case and fit far better into a RV.
5.09 - how to set the power-up behaviour. Been trying to figure that out, many thanks :-)
Hello Julian, can you help me with what size stud fork connectors I should get for this unit's In/Out screw terminals? #6, #8, or #10? I am new to DIY. Thanks a lot!
Julian, when are you doing the next video on this, the one driving the unit from solar panels?
Well, well, well. I hadn't seen you in a while. I thought you might be off to have your bits done. Good as new i guess?
Awesome review, Julian!! 😃🤩
nice module Julian
I think i will buy this but i'll replace the caps with higher quality panasonic FR ones. The input will never be more than 30V so i'll replace the 100V input caps with 35V ones and the output caps with 63V ones. I won't ever need more than 60V. I'll power it from a 24V AC toroidial transformer.
Nice little module, but the MPPT seems a bit misleading. On the aliexpress site they really talk about MPPT of solar panels, but you have to set it manually which is not MPPT. Would be very interesting if it really could handle 300W as stated. I doubt it can handle 20 A for longer time than a few seconds without fan. Also for a buck-boost the efficiency would be interesting. Maybe you can add this to your test, for different load situations would be optimal 🙂
True, it's not 'Maximum Power Point Tracking'. I call it maximum power point targeting. Apparently the unit cuts out at 300W with an error message. Although how that works when the display unit is not attached, I don't know.
If you don't have a display, the module is always on. So no power limit until something blows. I tested it to just 160w and it already needs a fan
Good point. It is not a true mppt device.
@@biezitis I have bought 1500W (50A) buck converter on Ali for 33 bucks (with shipping included), it manages 2 solar panels of 550W , so I am getting 94% efficiency (measured it myself so nobody can fool me), why the heck do I need an MPPT? To gain 4% more? Come on.... 4% doesn't worth paying 10x the price of a buck converter.
When you say single inductor but also buck-boost, on the basis it's described as synchronous, and "like the 3780," I would hazard a guess what you're actually looking at is a SEPIC converter. I think buck-boost has just become a shorthand for a DC converter that can go up or down. Possibly I'm splitting hairs, but it's an interesting development to see more of these.
No, it's not a SEPIC converter. Both the LT8705 and LTC3780 are described as Synchronous 4-Switch Buck-Boost Controllers. The data sheets explain it. But I have just bought a super little SEPIC buck-boost converter, fully digitally controlled (using the FP5139) for a very specific purpose. Details soon :)
@@JulianIlett thanks, interesting. I'm always on the lookout for a DC converter I can control with some sort of data input, something with a decent enough response time to implement a PID for LED current control. Never found one yet... Be interested if you've come across anything that might work.
When are we going to see the next shed extension to house all these bucking boosters? Are you taking a leaf out of SDG's book but replacing soldering iron stations with DC/DC converters? 🙂
I'd hate to see your AliExpress wish list. ;-)
Yes, I can see the shed getting several of these SJ20 converters. I want a conveyor belt of energy transfer between batteries :)
This seems ideal for charging an 36V or 48V ebike battery from a small ~18V solar panel, right? Without the "MPPT targeting", traditional boost converters will collapse the voltage as they try to max the current.
One should only charge a battery with a BMS so it can cut off the charging when full.
Prolong float charging at full-charge voltage of lithium batteries shouldn't be used but it is fine with lead-acid batteries as long as the regulation is good.
All switchmode converters have a negative input resistance characteristic, so they will all collapse the voltage of a PV array if the array can't meet the power requirement and there is nothing to prevent collapse.
Hi. I'm curious what is the module's passive current draw from the output when the input is unpowered?
can I use this module to power an AMD motherboard (around 165w MB plus 65w processor)?
hi @JulianIlett do you know what type of fet transistors are installed there? i am wondering if I can to change them to more powerful ones (with less resistance) if this module will be used not for 80 but with 40 volt panels. don't want to install additional cooling )
NCE82H140 originally installed
I know the OPA1612 as I have used them as audio opamp. They are pretty low noise and have low THD. Pretty good for audio use is my experience. They are also marketed as an audio opamp (SoundPlus is the line of opamps from TI). But their high precision makes also sense as a feedback opamp to give a very precise output voltage which I believe they are using it for in this case. So surprised to see it and also not surprised to see it :D
Edit: It is not a cheap OpAmp btw. €3.31 per unit if you buy 1000 on Mouser. A single unit is €5.88.
At 5:09 it seems to have exceeded the current limiter? That may be hazardous depending on what (was) on the output 😅
Yes, there's a brief overshoot of the output current limiter system. Not a problem if charging a battery (or even a supercapacitor), but it could kill a fragile load.
Do you actually get that much efficiency (98%)? I thought this possible only with GaN?
What is that chip on the left in a TO-263 (5 lead?) housing?
Looks like a switching regulator on its own?
It's an XL7015
@@JulianIlettso what is that switching buck regulator doing here?
Hi Julian, awesome video! We'd love to partner with you! How can we get in touch?
There's an email address in my 'about' page.
Always enjoy watching your videos.
Out of interest, I would like to give something back to you and this community, would you be interested in a free PSU for your projects and testing modules.
Thanks for the offer. What kind of PSU is it?
@@JulianIlett I have a Mean Well and TDK-Lambda switch mode power supply's. I have a list of them on the eevblog forum. Please feel free to pick what one would be best suited to your requirements.
Thanks so much for the generous offer. I've had a look at your post on EEVblog and I notice these are mains PSUs. I now only use battery power supplies (power banks / power stations) for my experiments.
@madprofessor9423 I could take one off your hands for the community 😂
Great video as always. When you speak about 10A,.... Input ? Output ?
I will like to know if you can test it by charging a 24v battery from a 12v battery.
10A? 15A with fans? So 240w at output is possible?
I think you also have an infrared camera so it will be a good test !!!
🎉
good module but not worth at this price