Downsides to using a meanwell type power supply instead of benchtop psu?

[schroederjd]

Er no. Maybe their cheapie LED drivers might but their real power supplies do not. They have excellent voltage regulation. I've used them for years to charge battery packs and cells. Right now on my bench I have a PSP-600-27, a HRP, a HRPG, a LRS (not current limiting) and 8 RSP-100s. I've got a stack of others on the shelf.

@Maast, do you typically use a protection diode of some kind to prevent backfeeding upon charge completion/PSU shut-down? If so, could you provide any diode recommendations for top balancing 272Ah cells with an HRPG-300-3.3?

Thanks in advance.

 

[Maast]

@Maast, do you typically use a protection diode of some kind to prevent backfeeding upon charge completion/PSU shut-down? If so, could you provide any diode recommendations for top balancing 272Ah cells with an HRPG-300-3.3?

Thanks in advance.

With my 20A power supplies I'm using two parallel axial 20A shottkey diodes thoroughly embedded (especially including the leads as they're the most thermally conductive to the diode core) in thermal epoxy to a 1.5x3 inch aluminum plate with heatsinks epoxied to it and I have a fan blowing over it. The diodes got 'holy crap' hot (thermocouple measured temp 180F) without the thermal epoxy/heatsinks and now I can touch one without discomfort @ 22A.

I just used what I had laying around from other project residue, if you're going to buy new I'd highly recommend you use a flat diode with a through hole mount like this one: https://www.digikey.com/en/products/detail/stmicro/STPS20SM60D/2827130 since you'll get MUCH better thermal transfer to your heatsink (and you'll definitely need a heatsink) and you wont end up using 10x the value of the diodes in thermal epoxy.
I should have gone with a flat plate diode but I got stubborn and didnt want to wait.

I really like this stuff as a thermal epoxy https://www.amazon.com/gp/product/B00HKJ6QHK it's cheap (for what it is), it works awesome and it's super strong. I've stuck a lot of aluminum heatsinks onto things with with it over the years.
With a through hole mount you won't even need epoxy, just some sort of heatsink grease. Whatever you mount the diodes to will need to be electrically isolated as the through hole mount is electrically connected to one of the leads.

I've got a HRP-300 that I use as a charger and I've seen 64A to a depleted battery thats at 2.9ish volts, you'll need 4 or maybe 5 diodes in parallel. You never want to max out a diodes capacity and ideally you want to keep it at 50% or lower.

I like using shotty diodes as they self level when in parallel because their forward voltage drop increases as the current goes up.

 

[schroederjd]

With my 20A power supplies I'm using two parallel axial 20A shottkey diodes thoroughly embedded (especially including the leads as they're the most thermally conductive to the diode core) in thermal epoxy to a 1.5x3 inch aluminum plate with heatsinks epoxied to it and I have a fan blowing over it. The diodes got 'holy crap' hot (thermocouple measured temp 180F) without the thermal epoxy/heatsinks and now I can touch one without discomfort @ 22A.

I just used what I had laying around from other project residue, if you're going to buy new I'd highly recommend you use a flat diode with a through hole mount like this one: https://www.digikey.com/en/products/detail/stmicro/STPS20SM60D/2827130 since you'll get MUCH better thermal transfer to your heatsink (and you'll definitely need a heatsink) and you wont end up using 10x the value of the diodes in thermal epoxy.
I should have gone with a flat plate diode but I got stubborn and didnt want to wait.

I really like this stuff as a thermal epoxy https://www.amazon.com/gp/product/B00HKJ6QHK it's cheap (for what it is), it works awesome and it's super strong. I've stuck a lot of aluminum heatsinks onto things with with it over the years.
With a through hole mount you won't even need epoxy, just some sort of heatsink grease. Whatever you mount the diodes to will need to be electrically isolated as the through hole mount is electrically connected to one of the leads.

I've got a HRP-300 that I use as a charger and I've seen 64A to a depleted battery thats at 2.9ish volts, you'll need 4 or maybe 5 diodes in parallel. You never want to max out a diodes capacity and ideally you want to keep it at 50% or lower.

I like using shotty diodes as they self level when in parallel because their forward voltage drop increases as the current goes up.

Thank you so much for all the info and recommendations! This exactly what I needed. I think I’ve got a clear path forward now.

 

[Ampster]

With my 20A power supplies I'm using two parallel axial 20A shottkey diodes

Before I read read this thread I got a great deal on EBay for two Meanwell 55Amp 3.3 volt (adjustable to 3.65 volts) power supplies. After I read the thread I checked the specs and realized these had no current limit and if I connected them to a bunch of 280 Ahr cells they would just go into hiccup mode.
I don't have much invested but intellectually am challenged to find a simple solution. Also this might be an informative discussion for others, which is why I am bumping this thread. One solution might be to continuously tweak the voltage on the power supply so the current is limited. Another thought is to use some power diodes in some way, which is why I quoted the post at the top. I do realize that the main purpose of a blocking diode is to prevent backflow but it does have some current limiting orooerties but my physics is rusty.

Any body with more knowledge have any ideas for this thought experiment?

 

[Maast]

Before I read read this thread I got a great deal on EBay for two Meanwell 55Amp 3.3 volt (adjustable to 3.65 volts) power supplies. After I read the thread I checked the specs and realized these had no current limit and if I connected them to a bunch of 280 Ahr cells they would just go into hiccup mode.
I don't have much invested but intellectually am challenged to find a simple solution. Also this might be an informative discussion for others, which is why I am bumping this thread. One solution might be to continuously tweak the voltage on the power supply so the current is limited. Another thought is to use some power diodes in some way, which is why I quoted the post at the top. I do realize that the main purpose of a blocking diode is to prevent backflow but it does have some current limiting orooerties but my physics is rusty.

Any body with more knowledge have any ideas for this thought experiment?

The simplest would be as you said: turn the voltage down until it quits hiccuping and then turn it up as the cell voltage rises. Diodes wouldnt work so well - you could do it but it'd be a PITA and then you'd run into problems on the high end. A better way would be to use a variable resistor to limit the current and then dial the resistance down to finish off the top end - but if you're going to be futzing with it that way you might as well just adjust the power supply voltage in the first place.

 

[Ampster]

The simplest would be as you said: turn the voltage down until it quits hiccuping and then turn it up as the cell voltage rises.

I was hoping you would chime in because your earlier posts seem to understand the issue I was trying to deal with. I have two of those power supplies so after I get my Ammeter and voltage display dialed in using one I might run them in parallel from 240 volts. I am putting the parts together and will fabricate some cables to handle the current. Those power supplies only cost $15 each, delivered so it will be a fun project and will save me some time when I get more cells and decide to parallel top balance 48 cells. I can use this system to step charge to maybe 3.5 volts and then use my other power supply with current limiting at 30 Amps to take the paralleled cells to 3.6 volts.
Part of my measurement meters include a voltage controlled relay that I can use with a contactor to cut of charging at a selected voltage so the back feed will not be an issue. I can calibrate that to my fluke and run sense wires to the cells to get better accuracy.
Thanks again for the feedback and reality check.

 

[DJSmiley]

You also can add a big resistor in series. That will limit the maximum current.

Once the cells gets filled up, current slowly drops, and the voltage drop over the resistor will decrease, eventually to almost zero.

If you add a 0.06 ohm resistor, that will limit the current to 55A (@ 3.65V). I think a smaller value will already do. I think if you just simply use a long wire it will probably already be enough to keep the current below 55a, preventing the power supply to enter hickup mode.

When paralleling 2 supplies, just 1/2 the resistor. I think its best to do some testing. Note you need to use a nice wire gauge to handle the currents. Not too thick, otherwise the required current limit won't be reached. Only pay attention to heat, cables will get hot, but I think if you just don't keep them on a spool (just lay them eg outdoors) I think that would work. (Or just throw a spool of (insulated) wire into a pool for cooling )

 

[Maast]

If you add a 0.06 ohm resistor, that will limit the current to 55A (@ 3.65V). I think a smaller value will already do. I think if you just simply use a long wire it will probably already be enough to keep the current below 55a, preventing the power supply to enter hickup mode.

The 'long wire' would work, but thats 100 feet of 8 awg which'll be a little spendy. I like the idea of putting the wire in a bucket though - its what I do when I use .032 stainless safety wire as a resistor for a high amperage load.

 

[Pontipine]

A bit late to the party but someone else might find this useful. This is a tab mounted schottky 100A 15V blocking. 0.45 forward drop. There are higher voltage versions. They could be bolted to Busbar for conduction and cooling.

https://www.mouser.ca/ProductDetail/Vishay-Semiconductors/VS-100BGQ015HF4?qs=iMG30BJ9VGgoY%2FK0JFxGuA%3D%3D

 

[schroederjd]

A bit late to the party but someone else might find this useful. This is a tab mounted schottky 100A 15V blocking. 0.45 forward drop. There are higher voltage versions. They could be bolted to Busbar for conduction and cooling.

I attempted to diode-protect my Mean Well HRPG-300 using a series of 30A schottky diodes. I just couldn't get the power supply to work when the diodes were in line. To overcome the voltage drop (the ones I used had 0.5 V forward drop), the supply would have to increase the voltage above the over-voltage protection limit, and fault out. In the end, I ran the supply without diode protection and it worked fine in that state. That's not to suggest that anyone should run these without diode protection...just sharing my experience.

Also against all advice, I ran the HRPG at max current (60A) for days on end and it performed flawlessly. With the sensing leads, and my arduino-based control/monitoring setup, I had a lot of confidence that this setup wasn't going to overcharge my cells.

I documented the top balancing process with my Mean Well power supply on my build thread starting here, if anyone's interested.

 

[Just John]

With my 20A power supplies I'm using two parallel axial 20A shottkey diodes thoroughly embedded (especially including the leads as they're the most thermally conductive to the diode core) in thermal epoxy to a 1.5x3 inch aluminum plate with heatsinks epoxied to it and I have a fan blowing over it. The diodes got 'holy crap' hot (thermocouple measured temp 180F) without the thermal epoxy/heatsinks and now I can touch one without discomfort @ 22A.

I just used what I had laying around from other project residue, if you're going to buy new I'd highly recommend you use a flat diode with a through hole mount like this one: https://www.digikey.com/en/products/detail/stmicro/STPS20SM60D/2827130 since you'll get MUCH better thermal transfer to your heatsink (and you'll definitely need a heatsink) and you wont end up using 10x the value of the diodes in thermal epoxy.
I should have gone with a flat plate diode but I got stubborn and didnt want to wait.

I really like this stuff as a thermal epoxy https://www.amazon.com/gp/product/B00HKJ6QHK it's cheap (for what it is), it works awesome and it's super strong. I've stuck a lot of aluminum heatsinks onto things with with it over the years.
With a through hole mount you won't even need epoxy, just some sort of heatsink grease. Whatever you mount the diodes to will need to be electrically isolated as the through hole mount is electrically connected to one of the leads.

I've got a HRP-300 that I use as a charger and I've seen 64A to a depleted battery thats at 2.9ish volts, you'll need 4 or maybe 5 diodes in parallel. You never want to max out a diodes capacity and ideally you want to keep it at 50% or lower.

I like using shotty diodes as they self level when in parallel because their forward voltage drop increases as the current goes up.

Thanks, very helpful for those of us that don't have a clue.
Sure, I understand the basics, but that's much more advanced. The purpose of the diodes is to prevent the cells from feeding (pretty much) unlimited amps back into the supply when switched off, correct? Certainly the diodes you linked are pretty cheap, add thermal compound and a heatsink (I'd add a fan). Put it all in a case with the 60 amp at 3.3v meanwell and you have a nice balance charger.

I wonder if it is cheaper (or better) than the 15v at 60 amp supply here:

148.98US $ 30% OFF|Bench Source Laboratory Dc Stabilized Power Supply Adjustable Variable Digital Regulated Power Supply 30v 20a 60v 10a Wanptek - Switching Power Supply - AliExpress

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

To me, it seems as if it will be almost the same (ok, slightly cheaper) price.
I have the 30v at 40 amp version of this supply, and seems to run quite a bit warmer then my Tekpower 15v 40a supply (which was $190).

 

[enocorecom]

I looked into the HRPG
I don't think its current is configurable.
It has voltage sense leads which will be handy to allow it to compensate for voltage drop.
It also has pins for external control.
From memory the external control pins are shorted to terminate the charge.
So with a suitable voltage sensitive relay that allows to automatically terminate the charge on voltage at the cell terminals I think it would make a good way to top balance cells.
I'm reasonably confident that it can adjust its voltage low enough to maintain constant current for a battery that is discharged to 2.5 volts.
For 280 amp hour cells in parralel using the 450 watt version it would be...
280ah * 8 cells / 90 amps * .5 SOC = ~12.4 hours
For a single cell it would be ~1.5 hours although the rubber band effect will be more pronounced.

In another thread I was looking at using a battery pack with bms and an SSR + meanwell psu on a single cell within the pack to top balance.

First things last...
I believe charging individual cells with a charge voltage of ~3.66 volts at ~.2c until the cells reach ~3.65 volts is all that is required to prepare cells "close enough" for usage in a serial pack.
I think its more important to get the cells deterministically to the same point on the high knee than it is is to get them all perfectly full.
Further to this I believe that the cells diverge less and converge better if the pack will not be drawn down too far and/or too fast.

tl;dr
Voltage is a pretty good indicator of state of charge at the very top(3.65 volts) and the very bottom(2.5 volts).

I think you are right.