diy solar

diy solar

Battery cell capacity testing.

Here’s one I purchased on Amazon. Curious to see how it compares. Mine is rated for 20a as well. Haven’t used it yet.
 

Attachments

  • A480A205-E916-4166-9395-D69F028139D7.jpeg
    A480A205-E916-4166-9395-D69F028139D7.jpeg
    124.6 KB · Views: 13
  • 0F2C3F3D-FDA8-4B7D-A9CA-9E207AF9182E.jpeg
    0F2C3F3D-FDA8-4B7D-A9CA-9E207AF9182E.jpeg
    84.4 KB · Views: 13
Here’s one I purchased on Amazon. Curious to see how it compares. Mine is rated for 20a as well. Haven’t used it yet.
It has two diodes. If you run it at 20 amps, the PCB will discolor and turn brown, same design, same problem. I have 3 different versions, they are all the same basic design. Be careful and don't burn your fingers.
 
Its the junction temperature that yoiu have to be concerned about, typical max junction tempearture is around 150c, you will have to calculate the size of the heatsink based on thermal resistance of the device spec,, etc., thermal resitance will be shown degree C per Watt.

I tend to design assuming PCB and air will be at 85C. A power device often has spec 150C maximum junction temperature and I would design to keep its junction no higher than 125C. Cooler of course would be more reliable and might be a target for some systems.

With his thermocouple on bottom PCB, if the plane there isn't electrically connected to the device, may read much lower than case.
Top side where package tab solders to PCB would give a better reading. From there I would calculate junction temperature using power dissipation and theta J-C

I can cool a part quite effectively without a heatsink (like this diode) using FR4 to carry heat from electrically connected copper to ground plane, in the event it doesn't connect to the plane. But to do that I specify layer thicknesses and layout area.

This board could be reworked by desoldering the diode and adding a large piece of thick copper underneath as a heatsink.
 
I tend to design assuming PCB and air will be at 85C. A power device often has spec 150C maximum junction temperature and I would design to keep its junction no higher than 125C. Cooler of course would be more reliable and might be a target for some systems.

With his thermocouple on bottom PCB, if the plane there isn't electrically connected to the device, may read much lower than case.
Top side where package tab solders to PCB would give a better reading. From there I would calculate junction temperature using power dissipation and theta J-C

I can cool a part quite effectively without a heatsink (like this diode) using FR4 to carry heat from electrically connected copper to ground plane, in the event it doesn't connect to the plane. But to do that I specify layer thicknesses and layout area.

This board could be reworked by desoldering the diode and adding a large piece of thick copper underneath as a heatsink.
I also stich Thermal VIAS between top top layer and bottom layer of PCB beside using proper heatsink to manage the thermal issue. The thermal probes also need to be place at the right locations to get more accurate reading
 
Its the junction temperature that yoiu have to be concerned about, typical max junction tempearture is around 150c, you will have to calculate the size of the heatsink based on thermal resistance of the device spec,, etc., thermal resitance will be shown degree C per Watt.
Oddly enough, the last diode I replaced (MUR3060PT, not on this tester) has a junction temperature rating of 175 degrees Celsius, and routinely dies after 30 to 60 minutes with the heatsink measuring 45 degrees Celsius. Like I said, it runs too hot. I'm sure it will last a while, how long is the question, anything that will burn your fingers is running too hot. They also like to die as a short (which may or may not ruin your day and cell). Spec sheets and cheap Chinese parts are different things.

Here is what I'm talking about on the "180 watt" version of the tester (two heatpipes and two fans, same design).
PCB_discolored.jpg
Diodes.jpg
 
Last edited:
Very interesting!

I've read about the Riden units. What do you think about the switching power supply itself? Not too noisy?
The step down/lcd bit is quite interesting. There is a review on eevblog.

I see you talk about automotive fuse holders I'm not familiar yet with those and see many models online.
Are you referring to this kind?
1615111467217.png

Do you know of a programmable electronic load at the price and features of the ET5410 with a 4 wire system for remote sensing?
If the voltage from the load isn't accurate, how do you handle LVD properly? Do you just offset it by the difference seen betwen it and your multilmeter reading?

I've also seen an interesting youtube review from Adam Walch on youtube about the ZKE Tech EBC-A05+. Have you heard of it?
It features a charger and load in the same enclosure. With a pc software to setup and see results of charge cycle.
Looked like a nice (from my beginner perspective) all-in-on setup.
1615112884784.png


Considering the advices of "Limiting to ~75% of rated power or 8A @ 12V far safer."
This one is the 180W : https://fr.aliexpress.com/item/1005001458325206.html
So if I wanted to test a 272AH 3.2V lifepo4 cell individually, could I still go up to 25A@3.2v as it's less than 75% of 180W?
 
Very interesting!

I've read about the Riden units. What do you think about the switching power supply itself? Not too noisy?
The step down/lcd bit is quite interesting. There is a review on eevblog.

I see you talk about automotive fuse holders I'm not familiar yet with those and see many models online.
Are you referring to this kind?
View attachment 39929

Do you know of a programmable electronic load at the price and features of the ET5410 with a 4 wire system for remote sensing?
If the voltage from the load isn't accurate, how do you handle LVD properly? Do you just offset it by the difference seen betwen it and your multilmeter reading?

I've also seen an interesting youtube review from Adam Walch on youtube about the ZKE Tech EBC-A05+. Have you heard of it?
It features a charger and load in the same enclosure. With a pc software to setup and see results of charge cycle.
Looked like a nice (from my beginner perspective) all-in-on setup.
View attachment 39930


Considering the advices of "Limiting to ~75% of rated power or 8A @ 12V far safer."
This one is the 180W : https://fr.aliexpress.com/item/1005001458325206.html
So if I wanted to test a 272AH 3.2V lifepo4 cell individually, could I still go up to 25A@3.2v as it's less than 75% of 180W?
Just purchased the EBC-A40L Zketech. Based on what I have seen on off-grid-garage, its the simplest and coolest way to test high capacity cells, as it also can charge the cell before it discharges, and graphs the test on your PC, and can runto 40amps. Andy has done it on his channel multiple times, with no issues.
 
Very interesting!

I've read about the Riden units. What do you think about the switching power supply itself? Not too noisy?
The step down/lcd bit is quite interesting. There is a review on eevblog.

I think the 800 watt version of the switching power supply needs $5 worth of parts and a little work. After that I've tested it at the full 720 watts output for hours. Jerry Walker did a great video series reviewing the RD6006 and RD6012. The 1000 watt supply they sell with the RD6018 isn't quite up to putting out the necessary power for the full 18 amps at 60 volts range, Riden recommends you limit it to 950 watts if using 120 volts like we do in the USA. But if you have a 48v system, the RD6012 properly modified makes a great charger, and you can always parallel them if you need more power. I tried a 1200 watt power supply from Aliexpress, and while it does output the necessary power for the RD6018, it has a fan that screams like a 1u server, and runs much hotter than I'd like. I have a couple brand new in the box for sale if anyone is interested.


I see you talk about automotive fuse holders I'm not familiar yet with those and see many models online.
Are you referring to this kind?
View attachment 39929

Yes, don't bother with those above 20 amps.

Do you know of a programmable electronic load at the price and features of the ET5410 with a 4 wire system for remote sensing?
If the voltage from the load isn't accurate, how do you handle LVD properly? Do you just offset it by the difference seen betwen it and your multilmeter reading?

Yes, the Kunkin KP184. The front panel interface however is a pain. You can modify the ET5410 and install your own remote sense leads, search on EEVBLOG for details.

Kunkin KP184 is quite popular.



I've also seen an interesting youtube review from Adam Walch on youtube about the ZKE Tech EBC-A05+. Have you heard of it?
It features a charger and load in the same enclosure. With a pc software to setup and see results of charge cycle.
Looked like a nice (from my beginner perspective) all-in-on setup.
View attachment 39930


Considering the advices of "Limiting to ~75% of rated power or 8A @ 12V far safer."
This one is the 180W : https://fr.aliexpress.com/item/1005001458325206.html
So if I wanted to test a 272AH 3.2V lifepo4 cell individually, could I still go up to 25A@3.2v as it's less than 75% of 180W?

I have (and quite like and recommend) the Zketech ECB-A40L.


The cheaper testers, I've made some recommendations on their limits, but so far they are quite accurate and I've not had one die yet.

 
Back
Top