Switch mode supply mod

[mwkeefer]

So what are the chances of being able to parallel the output on the SP-320s? I'd like to take the four I ordered and set them up as a 96V/13A charger. Maybe even get two more and make a 20A charger, although I don't know if I have the space for that.

Dave what core voltage are your SP-320s, if they are 48v... turning then down to 48v each and running series/parallel you end up with 2S2P in supply terms and have about a 13a charger (limit the current circuit, that's covered in the thread). The reason I ask if they are 48s, the 48s have 100v caps in them and would hold up better (imho) over the long run.

-Mike

 

[DaveAK]

Dave what core voltage are your SP-320s, if they are 48v... turning then down to 48v each and running series/parallel you end up with 2S2P in supply terms and have about a 13a charger (limit the current circuit, that's covered in the thread). The reason I ask if they are 48s, the 48s have 100v caps in them and would hold up better (imho) over the long run.

-Mike

Yup, they're the 48V ones. I'll actually need them slightly under 48V a piece I think, but well within the normal adjustability. (26 cells @ 3.60V per cell giving 93.6V, plus a little extra. OK, so 96V as near as damn it. ) I was thinking 2s2p like you say, but I've heard words of caution that they might not like running in parallel. What's your take on this?

 

[mwkeefer]

Wired properly the work fine in parallel (I've done it, not to 90+v) - one other caution, all my old units S-350 are stamped on the circuit board with 0-94v which could simply mean that individual possible range points are 0-94v each (don't try it).

You will need to limit the current on 2 of the units so that they only provide approx 6.8A then, you will series each of those modified units to the non-modified units (the lower limit will be the regulating limit)... and you will have your 90+v 13AH charger.

I would also suggest that you look at the posts in this link about Hyena's resistor mod (replace thermistor with resistor) which will keep the fans on all the time (this is good) and I would perform this mod to each of the 4 units prior to calibrating the current on 2 of them.

Sounds like your running LiFePo4 cells which have a max charge above what you spec (not much above) so I will assume it's intentional and correct (26 Series @ 3.60v per cell = 93.6v) for the 93.6v... each unit (S-3X0-48) must be configured for exactly 46.9 (well within their range) ... though I suggest replacing the SVR1 with a 10 turn 1K (check the existing value first) precision so that you can get the voltage as accurate as possible.

Finally... hook up the two modified for 6.8A units in parallel for 46.9v @ 13.6 A Charger. Now hook the remaining two supplies in parallel (Negative to Negative, Positive to Positive) and last, connect the limited and non limited 2P parts into a 2S (Positive to Negative) and you have your charger.

How many AH capacity is your pack and what is the actual chemistry... 13.6A seems alot for LiFePo4 which seem to like 1/4C which means 13A would be best suited to 40-45A LiFePo4 packs and that would be rapid charge.

Hope it helps!

-Mike

PS:

Feel free to PM with questions, I'm happy to help.

 

[DaveAK]

Thanks Mike. Excellent instructions which are a huge help. My pack will be LiFePO4 60AH cells, and I believe 1/4C charge rate is well within their specs. I got the 3.60V per cell from the latest specs of the Goodrum/Fechter BMS. I must have missed the thermistor to resistor post, but I fully intend to reread this whole thread before attempting the mods!

 

[Eric]

...all my old units S-350 are stamped on the circuit board with 0-94v which could simply mean that individual possible range points are 0-94v each (don't try it)....

You might want to double check the stamping on the PC Board. 94V-0 is a flammability rating that's stamped on most bare boards.
Eric S.

 

[mwkeefer]

Eric,

I stand corrected, I made the assumption long ago that it was the voltage range but since it was chinese it was backwards (seriously, no jokes intended).

Thanks for the real explaination!

Dave,

60AH Cells? Cell_Man stuff? Get the max charge rate (although this may shorten the life of the pack) and divide 60 by that rate to get the proper charge current... then limit 2 of the 48v units at that current and do the 2P2S bit you intended. At 1/4C that's 15A charge rate... which will be approx the charge rate (you can play with the current limits to get what you want, which is why I like a pot resistor on my current limit circuit until I have it tweaked just right).

The thermistor to resistor mod was courtesy of Hyena (I think) and is a much better method than I came up with (as always I overcomplicated it).

-Mike

 

[Hyena]

94V-0 is a flammability rating that's stamped on most bare boards.

94V-0 must be a good rating then, I've had a few go up in smoke but no flames yet :lol:

 

[mwkeefer]

Hyena - I've never had a flame out Meanwell (the iCharger was like a roman candle) nor a white puffer... the ones which went bad, just did.

-Mike

 

[DaveAK]

60AH Cells? Cell_Man stuff? Get the max charge rate (although this may shorten the life of the pack) and divide 60 by that rate to get the proper charge current... then limit 2 of the 48v units at that current and do the 2P2S bit you intended. At 1/4C that's 15A charge rate... which will be approx the charge rate (you can play with the current limits to get what you want, which is why I like a pot resistor on my current limit circuit until I have it tweaked just right).

The thermistor to resistor mod was courtesy of Hyena (I think) and is a much better method than I came up with (as always I overcomplicated it).

-Mike

Oh how I wish they were cell-man cells!! Unfortunately that's out of my budget. I got Thundersky cells, which to be quite honest were also outside of my budget! I think they'll charge at 1/2C but I can't remember exactly. I do know that they recommend an initial conditioning charge at 1/3C. If I had bought a couple more supplies I could have done this, but money and space are tight.

Do you think upping the current to 15A will be OK with the fans permanently running? I think $100 for a 96V/15A charger represents great value for money, and while a faster charger would be nice, I think this will be quite adequate for my needs.

 

[mwkeefer]

Dave,

I would appologize but for gods sake man, 60AH of capacity? Im lucky to cobble 5p15S into 25AH packs of LiPo at best (1556.25wh, 34A/69A load).

It's sad but my DD is slower, less efficient and less powerful than my geared hub (peaks of 5.6hp so far on 9FET controller) they are 26wh/40wh mi at 25mph which limits my range to approx 60 miles on a charge.

Your pack is 96v hot and 60AH of capacity, with my system I could get up to 222 miles range at 25mph.

It should be very cool either way.

-Mike

 

[DaveAK]

Dave,

I would appologize but for gods sake man, 60AH of capacity? Im lucky to cobble 5p15S into 25AH packs of LiPo at best (1556.25wh, 34A/69A load).

It's sad but my DD is slower, less efficient and less powerful than my geared hub (peaks of 5.6hp so far on 9FET controller) they are 26wh/40wh mi at 25mph which limits my range to approx 60 miles on a charge.

Your pack is 96v hot and 60AH of capacity, with my system I could get up to 222 miles range at 25mph.

It should be very cool either way.

-Mike

Mike, I neglected to say anywhere that this is for a motorcycle project. I'm aiming for 40-50 miles at 60-70mph.

I'm just glad that Gary pointed me in the direction of this thread because I think I've probably saved myself $200-$300 dollars by going this route. Your help, and that of everyone who has contributed to this thread is very much appreciated I can tell you!

 

[Hyena]

Do you think upping the current to 15A will be OK with the fans permanently running?

It should hopefully be ok, but you're running it right at the limit. I'd get the 350w units if you want 15a though. I have a 350w unit that has reliably put out 17 A for the last 6 months but the 320w units haven't been as widely tested, plus they're rated at a lower wattage to start with.

You could test it and see, if you can hold your hand on the back of the unit after a few minutes you'll probably be OK. In my experience (with blowing these up :lol: ) they'll all get warm which is normal but once they get so hot at the back that you can't keep your hand there that's when they'll pop.

 

[DaveAK]

It should hopefully be ok, but you're running it right at the limit. I'd get the 350w units if you want 15a though. I have a 350w unit that has reliably put out 17 A for the last 6 months but the 320w units haven't been as widely tested, plus they're rated at a lower wattage to start with.

You could test it and see, if you can hold your hand on the back of the unit after a few minutes you'll probably be OK. In my experience (with blowing these up :lol: ) they'll all get warm which is normal but once they get so hot at the back that you can't keep your hand there that's when they'll pop.

I got these because they were only $25 a piece, which seemed like a bargain to me. I think I'm too chicken to run them at 15A straight away, so I'll limit them to 13.3A which is still pretty good and see how it goes. If I use a pot like Mike suggested I can always turn them up a little bit when I'm more confident.

 

[mwkeefer]

Hyena,

Do you have a temp logger? or even a DVM with a temp probe and max hold?

If you could be so kind as to connect this to the back... then we will get a baseline for normal vs blowout pending, and blown values - with that we can add a self protect circuit (thermistor) to cut the output.

Just an idea.

-Mike

 

[EV4LIFE]

Hi guys,
maybe its just me, probably is, but I'm having a hell of a time trying to find where the thermistor is.
got pics of what one looks like, but still can't find it on my meanwell 350. can you point out to me where it is?

Thanks guys.

Martin

 

[Hyena]

Mike all I have is a turnigy temperature sensor. It seems to read ambient accurately and goes up to 160 oC so should be accurate enough in the range that we'd expect to see for the case temperature. I'm not willing to risk blowing up another one (I've killed 3 experimenting now @ $50 ea) but I can measure mine at normal operating temperature as a reference before i permanantly mount the sensor inside my hub.

Dave $25 each is cheap, I'd risk blowing 1 up to see how it runs at 15a - if it happily runs for 20 minutes it'll probably keep running. When mine have blown they've all been within minutes of turning on. Thermal runaway perhaps ? See what current a single unit runs out of the box - it may well be around the 7.5 amps you want anyway.
If it starts to get too hot then power it down and back the current off a bit (either resistor or shunt mod). It won't blow up instantly so just keep a close eye on the temp and you have little to fear.

ev4life look a few pages back, I posted a picture of it.
edit: Look at this picture
http://img20.imageshack.us/img20/5914/mw350.jpg

It's the little green component that looks a bit like a capacitor off to the bottom left hand corner of the large transformer

 

[DaveAK]

Maybe a dumb question, but what kind of load can I put on this to determine the current draw while I'm setting up/testing? And what can I use to measure the amps? I have a DVM, but I seem to recall that these shouldn't be used to test current for more than a few seconds. I'm not into RC stuff, but I could go to the local store and see if they have a WattsUp or similar I suppose, but what about the load? Preferably something common as I'm not going to be needing a piece of equipment that I'm unlikely to use again.

48V @ 7.5A = 360W. Four 100W household bulbs maybe?

ETA: I just bought the last one of these SP-320-48s that were being offered so now I have a spare for testing or if one should crap out on me at any time.

 

[heathyoung]

Mike all I have is a turnigy temperature sensor. It seems to read ambient accurately and goes up to 160 oC so should be accurate enough in the range that we'd expect to see for the case temperature. I'm not willing to risk blowing up another one (I've killed 3 experimenting now @ $50 ea) but I can measure mine at normal operating temperature as a reference before i permanantly mount the sensor inside my hub.

Dave $25 each is cheap, I'd risk blowing 1 up to see how it runs at 15a - if it happily runs for 20 minutes it'll probably keep running. When mine have blown they've all been within minutes of turning on. Thermal runaway perhaps ? See what current a single unit runs out of the box - it may well be around the 7.5 amps you want anyway.
If it starts to get too hot then power it down and back the current off a bit (either resistor or shunt mod). It won't blow up instantly so just keep a close eye on the temp and you have little to fear.

ev4life look a few pages back, I posted a picture of it.
edit: Look at this picture
http://img20.imageshack.us/img20/5914/mw350.jpg

It's the little green component that looks a bit like a capacitor off to the bottom left hand corner of the large transformer

That green component is a PTC thermister (sometimes called a posister) that is used to limit inrush current.

Do you have any dead ones still? I'm looking for a 24V one, would be interesting to know the failure mode of these...

 

[Hyena]

Heath I have a few dead ones but they're all 48v units (my 24v one is still going strong)
Heres the pic I posted a few months back showing where it let go.

Also check out these larger images for overlays of both sides of the board..

http://img109.imageshack.us/img109/2338/mwoverlay2.jpg

http://img25.imageshack.us/img25/2143/mwoverlay3.jpg

 

[EV4LIFE]

It's the little green component that looks a bit like a capacitor off to the bottom left hand corner of the large transformer

Thanks Hyena

 

[cell_man]

I've picked up a 100W, 3.3V, 30A PSU today. I checked the voltage range and it can give 3.67V just by turning the voltage adjustment pot to max. I'm gonna look at reducing the current limit to maybe 20-25A and see how it performs as a cell charger. If it works out ok I'll get a load more of them for my test setup. They come in at just under 10USD retail each so I reckon if you can get a few of these strung together it would make a great high power balance charger. If anyone's interested I'll get some pics of it and if there's enough interest can look at getting some direct from the factory which would mean they could be a bit custom too (possibly).

 

[cell_man]

I've got the PSU connected to a few discharged cells in parallel and it's delivering 25A without any mods (current limited @ 3.48V on the ouput terminals). It's a bit less than the 30A claimed (likely at 3.3V) but still not bad. I haven't modified it in any way, just turned the volts up to the max which is about 3.64V (no load) when checked with my DMM. I wonder if it's power limited rather than current limited It's connected to about 90Ah of cells so I'll leave it going for a few hours and see how it does. If it blows, it's only 10$ so I'll know to reduce the current limit next time. There appears to be 3 parallel shunt resistors so snipping 1 should do the trick.

 

[dozentrio]

I have a clone of the 48V 350W supply.

I was just playing around with it, and noticed that when I lower the resistance of the thermistor, the current output increases significantly. This seems like it may have been a design flaw? Here's what happens:

I turn it on, and it is outputting about 5.8 amps. This is less than the rated 7.3 amps because of the resistor mod I did. I added some resistors in parallel with the R31, until I got the desired output current. Now I point my heat-gun at the thermistor, and the fan turns on. When the fan turns on, the current jumps to over 8 amps. The same thing happens when I add a resistor in parallel with the thermistor, or if I short the thermistor out. Now it seems to me that if your powersupply is overheating, you want the fan to turn on. This is all well and good, and we can assume that it will work this way. I haven't tried it myself, and I'm afraid to, since I heard of someone's powersupply exploding out of box because the fan turned on too late. Anyway, even if the thermistor value is properly tuned to detect the heat before it gets too high, I foresee one big problem. As soon as the thermistor resistance reaches the threshold, the fan turns on and the current output... doubles?!. I think something is backwards here, and perhaps they wanted it to decrease current. If this is the case, perhaps replacing the thermistor with ... the other kind, the funky kind where the resistance increase with temp... maybe that would be a good idea. Who knows. Hopefully someone will find this useful.

 

[Indubitably]

Copied from my post in the Goodrum's BMS thread--->

This is actually pretty darn cool. I hadn't realized that a straight power supply could be used as a charger. This, combined with the fact that the BMS is so much more affordable than others, makes my project a lot more economically feasible than I had previously expected.

One question though, I'm going to be using a 16s array of 40ah (thundersky LiFeYPO4 cells) batteries that have a max charge rating of 3C and 4.25V, what would probably be the largest current I would realistically want to charge them at using this BMS? Would I want more than the 7amps at 48v that I could get out of an SP-350-48? If I did want more, could I just hook a bunch of 350s in parallel at 48, or would I be better off hooking them up in series at 15amps and 24v, etc? Any advice would be appreciated.

PS: This is my first EV project, so please bear with me if I am missing some of the basics.

 

[dozentrio]

With regard to my earlier post, I may have jumped the gun a little. So, to correct: When I short out the thermistor, power output drops to 0 and fan turns on weakly. When I parallel the thermistor with 100 ohms, no output, and fan turns on a little more strongly. When I increase the resistor I use to parallel with the thermistor to 1k ohm, the fan turns on strongly and the current output is large. (8 amps
).