Bulk charging via digital DC-DC boost Minghe BST-900W

Similar setup I knocked together from a cheap "1200W" DC-DC boost converter from ebay:

http://www.ebay.co.uk/itm/1200W-DC-DC-Boost-Converter-Power-Supply-Adapter-Transfer-Step-up-Shield-Module-/302187923773?hash=item465bcdc93d:g:Ce4AAOSwZQRYbiCL

12-60V input, 12-83V output. Claims 20A input rating.

It's running at 74V 4A to slightly under-charge an 18S li-ion pack (BMS is limited to <5A charge current).

PSU is a cheapo 48V Meanwell clone. Volt/ammeter is another cheapo ebay unit and powered by its own tiny 230VAC-5VDC PSU.

Case is a generic equipment case with a handle added to the top.

20170316_212516.jpg

20170316_213152.jpg

20170316_213136.jpg
 
Overclocker said:
well it arrived ridiculously fast from china! it worked fine at the default 20v / 5a. when i upped the voltage to around 60v it stopped working and fried my PSU :evil:

upon closer inspection it uses a schottky diode that's only rated 45v! this is a module that's supposed to go up to 120v! i guess that's why it failed at the higher voltages, but then again i'm no switch-mode expert

here's a simplified diagram of a boost converter

Upon inspection, mine has a STMicroelectronics Branded "STPS20150" Schottky Rectifier : View attachment STMicroelectronic brand STPS20150CT 150V Schottky Rectifier.pdf
It has an IXYS Branded "IXFH150N17T" POWER MOSFET (175V V, 150 A, <12 mΩ) : View attachment IXYS brand IXFH150N17T POWER MOSFET.pdf

I got it from ebay seller "joytownoline" : http://www.ebay.co.uk/itm/BST-900W-DC-DC-CNC-Boost-Converter-8-60V-Step-up-10-120V-Solar-Charging-CVCC-/302213076941
Should be good to go up to 120V DC no problems.

In case anybody is looking for it...

Ming He BST900W MANUAL : View attachment BST900W Boost Converter Manual.pdf

Matador
 
So for a while I've been worrying about the lack of protection from small persons fingers on these bst400 dcdc s. I created an enclosure which channels the air over the heat sink and allows the user to get to the buttons and see the digital display. It's a bit crude and moving the interface panel to the side via two ribbon cables or on higher riser pins. Will post us photos later but I'm confident it is a lot safer than it was and much better protected.

View attachment 1
so this is printed upside down.

bstbox.png
Here you can see the slot to guide the air from the fan along the heatsink.

There are two pilot holes for the fan which need to be drifted a touch to line up with those in the heatsink but its close enough. everything else fits like a glove.

I may want to add another hole to hide the fan power wires too. Also mk2 might have allow the user panel to be flush with the case with printed buttons to further help from spills from above. Best off not actually drinking your brew while/near your charging .
 
markz said:
Meanwell Clone's
Are they sold on ebay?
What is the search term I use?

Want to do something similar.


https://endless-sphere.com/forums/viewtopic.php?f=14&t=82731&start=25

on page 2 are pix of this Smun meanwell clone. true clones have indentations on the aluminum case where the power transistors sit on. and have 99% identical PCB layout and mechanical design

you'll also find PSU's w/ flat sides (no indentation). they're not true clones. totally different PCB. often with some corners cut i.e. missing the AC input filtering stage and 220/110v switch
 
Gsk0nJq.jpg


just an update i built another PMS this time w/ a 48V Smun PSU to charge my 63v nominal batt pack. also printed out a cheat sheet

the reason for this is that the BST900W isn't very efficient at boosting 36v to 71.4v max
 
Has anyone tried connecting 2 boost converters in parallel? I'm trying to do that with the 1200w modules from ebay, but the result is that always one of the modules is taking all the amps, no matter how I set the pots. I'm running them on constant current (battery is not charged and therefore the set voltage is not yet reached)
Any idea how to avoid that? Maybe adding a diode somewhere on the trace?
 
Well, you could put a diode in series with each supply's positive output to the load, so the cathodes of the diodes are connected to the load, and the anode of one diode is on one supply's positive, and the anode of the other diode is on the other supply's positive.

That would isolate the outputs of the supplies from each other, but I don't know exactly what effects it would have on their monitoring capabilities (if any).

You'd want to measure the voltage drop across the diodes and then add that much to each supply's voltage, though.
 
cerberpeter said:
Has anyone tried connecting 2 boost converters in parallel? I'm trying to do that with the 1200w modules from ebay, but the result is that always one of the modules is taking all the amps, no matter how I set the pots

This was my experience with some lower-power ("800" I think) units as well. Even though I'd seen it written that they could be paralleled.
 
Paralleling is harder to do, obviously, because they werent meant to be paralleled, obviously.
Just thinking about that in regards to the Meanwell PSU's, because them MW's are specifically designed to share the current for certain models of MW's. Perhpas in this case the circtuitry does not accomodate paralleling, and maybe the impedence is off and the lower one takes it all. Maybe impedence is the wrong term to use, but over all resistance perhaps. Because you are changing the OVERALL circuit when you parallel two units.
 
Adding diodes didn't help unfortunately. They are acting the same way.
At first glance it looks fairly simple task, but in reality it isn't :(
It would've been great if we can parallel these boost converters, as this would make super cheap high power (fast) charger for our larger batteries.


pp.. I think I realize what is happening. The current shunt is placed on the negative side and the current there is always split equally through both of the boards, because the resistance path is almost identical. So no matter which board is boosting the voltage and pushing current through it's inductor, the current flow on the negative always splits between the 2 current shunts and makes them think they work the same when they are set to do that.
So I wonder if I try to brake the paths for reference voltage from the shunt and place shunt on the positive site, connecting its 2 ends where the reference voltage should go might make it work? :|
 
This is all over my head, but the basic concept is to use a cheap SMUN switching psu and then a dc-dc buck converter. Makes sense, efficiency is the concern as I read the thread.

What do I need to do to get 41.5V (10S) and 25A or higher charge doing the DCDC thang?

There are SMUN psu's that are 24V 25A/33A/50A @ $75/$151/$193(CAD) would have to split my pack in half and parallel.
I currently already have three Meanwell HRP-300-24 rated at 21.6V-28.8V and 15A, I use two in series for 10S at 15A.

Nice note to add - I got a shit ton of Dell 12V 62A psu's laying around. So DCDC looks like the way to go.

***Maybe I gotta start my own thread in Battery.
 
The DC-DC converter acts as a voltage & current limiter, allowing you to use any old PSU as a power source (not a current-limited on or one with exactly the right voltage).

You aren't going to do 41.5V @ 25A though with the common parts from ebay.

The largest common dc-dc boost converters are 1200W max (if you'd dare run anywhere close to a China rating lol). It's also limited to 20A max current, but this is INPUT current. So if you're using a 12V PSU then you'll get only 12 x 20A = 240W before it burns up.

You want a PSU voltage that's as close to your desired output voltage for best efficiency. So I'd be looking at a 36V PSU. That's still only going to give you 36V x 20A = 720W input. Output would be 720W / 41.5V = 17.3A x 0.9 (approx. efficiency) = ~15A at best.
 
I decided to use this setup for charging my 18s 32ah multistar pack. I ordered 2x of both of the models listed in this thread so I could test things out.

I already blew one of the BST-900's somehow, I had it at maybe 600w, hit the voltage up slightly, and it blew the fuse, and took out the fet and diode with it.

On the bright side, I ordered some higher quality fets, they max out at 100v, but are much lower resistance: https://www.infineon.com/dgdl/irfp4468pbf.pdf?fileId=5546d462533600a40153562c73472019 and diodes: https://www.onsemi.com/pub/Collateral/MBR20100CT-D.PDF

I'll attempt to repair the blown one, and upgrade the others after a bit more testing.

Also have an Eltek Flatpack2 (1800w) coming in in the next couple days, which should power these without issue.

I beefed up the traces on these boards, because they were pretty damn sad. Here are some pics: https://goo.gl/photos/N6Jqa1UxCX9CGwxWA
 
So 3 Dells 12Vx3=36V and 62A is 2232W, and the one dcdc I found puts out 30A. Yeah I need $pecial gear.
MW SE models or something. Then that current limiter board.
 
I have one from Drok. It worked nicely at 15A12v In to 24V out for hours, but blew up the IRFP4568 (150V fet by specs) at 9A 36V in and 120V out in under 1 minute. It overheated apparently, with additional fan blowing on the sink.
Is it a bad idea 3x+ multiplication or just a counterfeit Fet?
 
Quote I have from Doctorbass in a Personal Messages when we talking about the BST900 :

ORIGINAL QUOTE IN FRENCH :
"Wow, bon achats :wink: Fait gaffe pour le BST900 quand tu le branche. met toi une diode en serie avec l'entrée si tu peux, au moins 30A car l'entrée est sensible aux spark qui font des poiontes de courant élevé et momentané quand les condensateur d'entrée se rechargent."

ENGLISH TRANSLATION :

"Great buy :wink: Be carefull with the BST900 when you plug/connect it. Make sure to put in a diode in series with the input if you can, at least 30A value because the input is sensitive to spark that stems from transient current surges when the input condensers get charged up."

I havent burned up my BST900 yet (still sits in it's box)... What do you think ? Could that solve the issues I see you guy experiencing ?
 
Well i've been using my BST-900 with my meanwell s-350-27 for the past couple weeks with everything working fine.

I just got my Eltek Flatpack2 in, and I made a makeshift wiring harness to plug into it using some flattened bullet connectors and epoxy. Plugged that bad boy in and it worked first try. Putting out a stable 53.6v.

I plugged it into the BST-900, and it's powering it perfectly to charge my packs. The extra 20v on the flatpack2 compared to the meanwell really helps to not heat up the BST-900 when charging an 18s pack. I am not pushing it past 6a until I have my replacement fets in so i'm able to fix it if it does blow. I believe that once I replace the fet and diode with the better ones I ordered, I should be able to push this to ~1000w boosting from 53v to 75v. We'll see though, that's what testing is for.

Pictures: https://goo.gl/photos/N6Jqa1UxCX9CGwxWA
 
parabellum said:
tostino said:
I believe that once I replace the fet and diode with the better ones I ordered, I should be able to push this to ~1000w boosting from 53v to 75v. We'll see though, that's what testing is for.
What Fet and diode did you order?

Here is what I got:
tostino said:
On the bright side, I ordered some higher quality fets, they max out at 100v, but are much lower resistance: https://www.infineon.com/dgdl/irfp4468pbf.pdf?fileId=5546d462533600a40153562c73472019 and diodes: https://www.onsemi.com/pub/Collateral/MBR20100CT-D.PDF
 
I've been using mine with and esp120 54.6v power supply. And can get a full 15a at 58.4v

16x3.65v for lifepo4.
I only charge like this when I can be near and manually terminate the charge. It is handy for a quick charge and not far off the 18a that the 60ah 0.3c batteries can take.

Sent from my ALE-L02 using Tapatalk
 
15A @ ~60V is very very good.

I am not doing dcdc but I can charge two seperate batteries @ 15A and ~60V out of the same 110V outlet pulling 12A or so. Using 4 MW HRP-300-24's, all in I am at $50+$170 so $220cad plus I got a spare MW so 5 total for $270cad.

With your setup, I am sure it was much cheaper.
 
Esp120 was £50 off Sacko here on ES

Sent from my ALE-L02 using Tapatalk
 
Back
Top