Why Meanwells?

x88x

10 kW
Joined
Oct 11, 2010
Messages
527
Location
MD, USA
So, this is something that's been bugging me for a while and I was wondering if anyone could point out what I'm missing.

I've seen around here several people driving custom and/or high power battery chargers using Meanwell switching power supplies. Unless I'm missing something, the desired characteristics in a power supply in this case are steady, clean power (thus switching, not transformer) and high current levels. So, here's what's been bugging me. Why Meanwells? Sure, they're high quality and have easy hookup right out of the box, but I feel like the easy hookup probably isn't so much a need for the people using them, and they're quite expensive for what they do, imo.

Why not use PC power supplies?

They're a lot cheaper. If you get a high quality one, you'll have just as clean and reliable power. I mean, let's face it, they're designed to power a lot more sensitive electronics than battery chargers. Then, if you pick the right high-powered one, you can pull quite high current levels at 12V.

Take the power supply currently in my PC, the Corsair TX750. 750W PSU, it can put out a constant 60A on the single 12V rail, 24/7, and it costs $110. Compare that to the nearest Meanwell supply I could find, the SP-750-12. Same 750W rating, puts out a steady 62.5A @12V, and it costs $272.50! For that money, you could get a Corsair AX1200, 1200W supply pushing a steady 100.4A on the single 12V rail. :shock:

So...am I missing something?
 
PC power supplies lie (not matter what brand/model etc).

Server power supplies don't lie, and many folks use them for powering RC chargers etc.

Meanwell's have the big advantage of enabling CC/CV mode and voltage/current adjustability.
 
liveforphysics said:
PC power supplies lie (not matter what brand/model etc).
This definitely used to be true, but a lot has changed in the world of consumer PC power supplies in the last 5 years or so. This sounds like an interesting experiment..see how close to actual specs I can get them to go. :twisted:
 
You're confused....were not running high powered chargers off of meanwells, we are using the meanwells as high powered chargers :)

I'm using 2x 48v models in series to charge 24s lipo CC/CV style. They are tuned to around 49.5v each for 99v total. Cost about $35 each, IIRC
 
x88x said:
So...am I missing something?

People like me running higher voltages who bulk charge using a BMS dont wan't the over complication of 4+ 12v block PC PSU modules + the wiring mess/hassle of all the unused +/-3.3V/5V rails to deal with. This all assumes PC PSU's can provide a CC mode as well....
 
AmpEater said:
Cost about $35 each, IIRC
grindz145 said:
Constant Current and Cheap.
I must be looking in the wrong places then...everywhere I've looked, they're hideously expensive.

EDIT:
boostjuice said:
the wiring mess/hassle of all the unused +/-3.3V/5V rails to deal with.
There's always wire cutters. ;)
 
I didn't want to post this until I had links to back me up, so here we go. Regarding the constant current capabilities of modern PC power supplies, this is a commonly held misconception that has the fairly unique merit of actually being true at one point. However, with the massive rise in power requirements of high-end components in the last 5-6 years, this has changed dramatically. The website hardocp.com does a lot of computer hardware reviews, including power supplies, and I chose them to back me up on this because they keep (and post) very detailed records of their tests. With power supplies in particular, they have some pretty intensive processes that they put them though, detailed here. For the purposes of this topic, one of those steps involves running them at 25, 50, 75, and 100% rated load for 30 minutes at each level. Granted, this is not a horribly long time, but it gives a snapshot. I've included links to their load test pages for a few high quality units, but the general point is, none of these dropped below 11.87V at 100% load, and all of them handled 100% load (or damn close) just fine.
Thermaltake Tough Power 750W : 60A @ 12.08V
NZXT HALE90 750W : 60A @ 12.06V
PC Power & Cooling Silencer MkII 950W : 80A @ 11.97V
Corsair AX1200 1200W : 98A @ 11.87V
Corsair AX850 850W : 68A @ 11.98V
Seasonic X-560 560W : 44A @ 11.98V

Now, in all fairness, not every PC power supply can put out 100% rated load without breaking a sweat...
Ultra Talent Atomic Juice PS-700 700W : 4x 14A @ between 11.93V and 12.06V ..high voltage variability and it came out the other end at a toasty 74C :shock: (note: having multiple lower-power 12V rails is a fairly common way that mid-to-low-end PSU manufacturers try and boost their numbers)
..or these...
FSP Everest 1010W : ...well, it couldn't actually even start the test at 100% load...
FSP Everest 700W : 4x 14A @ between 11.56V and 11.80V

Like most things, it depends what you buy.

This has made me curious though. I have a couple good (and some not so good) PSUs that I'm not using any more....methinks a marathon torture test is in order. :twisted: I just need to get a nice big shunt and off I go. Time to see how long these units can hold 100% load. :twisted: *looks around for fire extinguisher*

As for getting more than 12V, that is a good point. If my torture tests go well, a custom variable DC/DC converter may be in my future. :D
 
They pop up cheap on Ebay alot, and thats where most people get them. Good luck getting 80V CC out of a PC power supply..
 
x88x said:
what I'm missing.

I've seen around here several people driving custom and/or high power battery chargers using Meanwell switching power supplies.

Your missing we are using Meanwels as chargers not to power chargers. I for example use a 48v Meanwel to charge my 44v 20ah lipo setup... The majority aren't using them to power rc chargers....yet.

KiM
 
AussieJester said:
Your missing we are using Meanwels as chargers not to power chargers.
AmpEater said:
You're confused....were not running high powered chargers off of meanwells, we are using the meanwells as high powered chargers :)
Yeah, I think that was it. I also did a little more looking, and the price curve for higher current Meanwells isn't anywhere near linear...a new 1000W 24V one wasn't a whole lot more than the 750W 12V one I linked in my OP. ..now I'm curious though. :D

grindz145 said:
Good luck getting 80V CC out of a PC power supply..
Heheheh, we shall see...we shall see. :twisted:
 
What will you charge with 12v? Its not really close to anything - 3S charge at around 12.5v, or 11V if they are LFP.
Would you run several in series?
If so, how independent are the 'rails' on the PSU's that have 2,3,4 rails? Could you series connect different rails from the same PSU and get yourself 48v from 4 X 12v rails?
 
Samba said:
Could you series connect different rails from the same PSU and get yourself 48v from 4 X 12v rails?
Probably not since they all share a common ground. I was thinking more along the lines of a custom variable DC/DC converter circuit. :D

Looking at the prices of the high powered Meanwells (1kW), I can see the appeal, price-wise..not much more than a similarly powered PC power supply, and no modifications are needed. But if you're like me and already have a bunch of PC power supplies lying around, there's no reason why they wouldn't work just as well with a little modification. :mrgreen:
 
A couple of things...

Most people here are using multiple MeanWells, either in series, to get higher voltages, or in parallel, to get higher current. The reason is that this is a much cheaper way to get higher power than trying to buy a single 700-1000W unit. You can get two 350W supplies for a lot less than $100. I use two S-350-24 MWs in parallel to drive a 550W Hyperion EOS 1420i 14s/20A charger, and I also use two S-350-24 MW, but in series, to bulk charge a 12s3p LiPo-based pack. The MW outputs are completely isolated so you can put them in parallel and/or in series, to achieve pretty much any voltage and current output.

Most higher output RC-type chargers, like the Hyperion I mentioned or several of the iCharger models, need to have 20V+ to achieve their max power ranges. I don't care how many amps a 12V supply is capable of putting out, it is not going to drive any of these high power chargers to their max output.

-- Gary
 
x88x said:
So, this is something that's been bugging me for a while and I was wondering if anyone could point out what I'm missing.

I've seen around here several people driving custom and/or high power battery chargers using Meanwell switching power supplies. Unless I'm missing something, the desired characteristics in a power supply in this case are steady, clean power (thus switching, not transformer) and high current levels. So, here's what's been bugging me. Why Meanwells? Sure, they're high quality and have easy hookup right out of the box, but I feel like the easy hookup probably isn't so much a need for the people using them, and they're quite expensive for what they do, imo.

Why not use PC power supplies?

They're a lot cheaper. If you get a high quality one, you'll have just as clean and reliable power. I mean, let's face it, they're designed to power a lot more sensitive electronics than battery chargers. Then, if you pick the right high-powered one, you can pull quite high current levels at 12V.

Take the power supply currently in my PC, the Corsair TX750. 750W PSU, it can put out a constant 60A on the single 12V rail, 24/7, and it costs $110. Compare that to the nearest Meanwell supply I could find, the SP-750-12. Same 750W rating, puts out a steady 62.5A @12V, and it costs $272.50! For that money, you could get a Corsair AX1200, 1200W supply pushing a steady 100.4A on the single 12V rail. :shock:

So...am I missing something?

Mea culpa............

I'm probably to blame, because a couple of years ago I bought one on ebay, as an impulse buy, and modded it to charge a 16S, 4P pack of Headways. It worked OK, so I rather innocently started a thread: http://endless-sphere.com/forums/viewtopic.php?f=14&t=4125 Little did I realise that it'd spawn a rash of developments, all using these cheap power supplies. Since then others have massively improved on this idea, adding current limiting and other control features. Overall, these power supplies are popular because we're now familiar with them and know they work.

Jeremy
 
Jeremy Harris said:
Overall, these power supplies are popular because we're now familiar with them and know they work.
Haha, fair enough. I'll be sure and post the torture testing and anything I do with boosting the 12V output with DC/DC converters...open up a different route for those that might want it. One nice (for these purposes) thing about people who build high-end PCs is that they tend to replace the PSU with each major upgrade...which translates into a fair number of high powered PSUs available second hand for a good bit less than retail. It might not be that good for lower power (say, 500-600W or less), but if you want, say, a multi-kW charger...
 
x88x said:
Samba said:
But if you're like me and already have a bunch of PC power supplies lying around, there's no reason why they wouldn't work just as well with a little modification. :mrgreen:


With the exclusion being that they don't produce the rated power, 12v isn't the voltage we want, they don't have a CC mode, they have a under-voltage trip instead, and meanwell's are so cheap/light/easy that it's tough to complain.

I think my cheapest set of meanwells was like $32 each shipped (buying 4) for 350w versions. If you want to charge at a continuous ~1350w (about all you can get out of a standard 15amp 110vac breaker) it's costing you like $128 in meanwell supplies, and that's it. You make the needed series or parallel charge harness (for whatever type you got and the pack voltage etc), adjust the voltage trim pots to dial things in where you want them, and the charger is complete. Depending on which supplies you buy, you have the flexibility of a charging range from a 200v pack or a 12v pack all with no in-between electronics, just the meanwells.

If you want to make a 12vdc DC/DC to step up to some other pack voltage, you're in for a lot more size/weight/volume/parts cost than just starting from scratch and stepping down from 110vac. If you want to see what a 12vdc to 30-60vdc DC/DC power supply looks like, the first half of any car audio amplifier is exactly what you need to build.
 
You have a point about the voltage problem. Though, now I'm curious...I wonder if I could make multiple 'virtual' isolated rails to stack..I never really thought about stacking DC power supplies before, but from what people are saying they've done with Meanwells it's apparently possible..just requires the proper filtering, I would guess. *adds to research list*

What do you mean by 'CC mode'? I had assumed constant/controlled current, ie, constantly providing whatever current the load requires...which is exactly what these do..but given the comments on this thread I wonder if I may be missing a nuance.

liveforphysics said:
they don't produce the rated power
http://endless-sphere.com/forums/viewtopic.php?f=14&t=23604#p343507
Any PC power supply worth anything is at least 80% efficient (ie, outputs 80% of max rated current at max load with the other 20% dissipated as heat), with the good ones sitting around at least 85-90% efficient at load. For more info, see the 80-plus rating. From the datasheets on Mouser, that 85-90% range is where the Meanwells I checked are too.

Don't get me wrong, Meanwell supplies are very good power sources. I'm not trying to say one is better than the other, just trying to open up options.
 
x88x said:
What do you mean by 'CC mode'? I had assumed constant/controlled current, ie, constantly providing whatever current the load requires...which is exactly what these do..but given the comments on this thread I wonder if I may be missing a nuance.

Batteries will take whatever current the source can supply. If the current is not controlled (Constant Current), the battery appears near enough to a short circuit.
PC power supplies are Constant Voltage.

The meanwells we use, have constant current limiting. Or we modify them to be so. When they hit the output current limit, they lower the output voltage to keep the current supplied at the limit point.

Hope this helps.

Amanda
 
x88x said:
You have a point about the voltage problem. Though, now I'm curious...I wonder if I could make multiple 'virtual' isolated rails to stack..I never really thought about stacking DC power supplies before, but from what people are saying they've done with Meanwells it's apparently possible..just requires the proper filtering, I would guess. *adds to research list*

What do you mean by 'CC mode'? I had assumed constant/controlled current, ie, constantly providing whatever current the load requires...which is exactly what these do..but given the comments on this thread I wonder if I may be missing a nuance.

liveforphysics said:
they don't produce the rated power
http://endless-sphere.com/forums/viewtopic.php?f=14&t=23604#p343507
Any PC power supply worth anything is at least 80% efficient (ie, outputs 80% of max rated current at max load with the other 20% dissipated as heat), with the good ones sitting around at least 85-90% efficient at load. For more info, see the 80-plus rating. From the datasheets on Mouser, that 85-90% range is where the Meanwells I checked are too.

Don't get me wrong, Meanwell supplies are very good power sources. I'm not trying to say one is better than the other, just trying to open up options.

There is nothing to research about paralleling power supplies, and no procedure to it. If they are isolated (like a meanwell), you just wire them up as many in series or parallel as you like, no different than connecting up a battery (also an isolated power source).

By, "don't produce the rated power", I mean, when I bought a number of PC power supplies for running my 12v HID ballasts for aquarium lighting, the "38amp 12v" rail would fall on it face after 5minutes of drawing only 20amps. Then later when I needed to power RC chargers, the damn things would shut down every time the charger pulled the voltage below ~11.8v or so, and ended up burning up. In the computer it was in, it worked fine never having a problem. I'm sure there are some better than others, server power supplies being quite good.

Constant current (CC) means, the power supply stops trying to regulate the voltage at a fixed point, and lets it drop down to whatever it needs to be, while just holding the current it supplies constant. Most lab power supplies can even have the output leads just shorted together, and they will just steadily output whatever current they are regulated to provide through the short, like 10amps or whatever. If you remove the short, it leaves CC mode, and jumps up to CV mode, where it holds voltage at exactly the set voltage point until something exceeds that 10amps CC current, and it voltage drops down and it just supplies the set 10amp current.

In your last paragraph, you're talking about efficiency (power in vs % power out), yet you're proposing using something with 4-5 different voltage rails in the same package (all causing no-load currents), and then running this lower voltage into a DC/DC converter where it can then be transformed again to some useful level through another lossy converter (which would be larger than the meanwell in the first place)...
 
liveforphysics said:
By, "don't produce the rated power", I mean, when I bought a number of PC power supplies for running my 12v HID ballasts for aquarium lighting, the "38amp 12v" rail would fall on it face after 5minutes of drawing only 20amps. Then later when I needed to power RC chargers, the damn things would shut down every time the charger pulled the voltage below ~11.8v or so, and ended up burning up. In the computer it was in, it worked fine never having a problem. I'm sure there are some better than others, server power supplies being quite good.
Sounds like you just got some shit power supplies. :p OOC, do you remember the manufacturer? There are a lot of really bad manufacturers that sell a lot of units just because they're just so cheap..like the FSPs I linked reviews of. Just like in any industry, there are good and there are bad (yes, even among those made for servers ;) ). The thing about computers is that they really don't consume all that much power, which is why the computer you put it in was fine. I have a pretty beefy system, but I've never seen it pull more than about 175W even under heavy use. One of the servers I use at work (dual quad-core Opterons), I've never seen top 450W, even when it was at near 100% resource use.

The point I was trying to make with the efficiency figures was about producing rated power, so yeah, power in vs power out, ie, a 1000W PSU at full load, working at 90% efficiency will put out 900W with most of the remaining 100W lost as waste heat. For example, I'm guessing the power supplies you got for your fish tank were sold as 500W units. If they crapped out at 20A, they were, at most, operating at 53% efficiency at 100% load on the 12V rail...possibly lower considering how quickly they failed. Cheapo heatsinks are not happy dealing with 216W of waste heat. :lol: PC PSUs actually hit peak efficiency around 50% load, rising a bit from low-load to there and dropping a bit from there to 100%. Nice little bell curve, usually only a few percentage points high with high quality units.

I'll admit, I did forget that most AC/DC supplies are isolated from the AC line...there's the circuit completion I was looking for that explains the series config working.. :oops:

I hadn't thought of the voltage wanting to drop, but now that I think back on my reading on battery/cap charging, that is what happens, isn't it... :oops: There should be a way to make that work though. I've used a big potentiometer to manually adjust a 12V rail from 0-12V, I'd just have to automate it.

IDK, it's sure not the easiest way (you have convinced me of that), but it should be interesting. :D
 
Doesnt mean its 50% efficient... might be 97%. Just means its continous current rateing is BS.

If you're going to build a custom DC/DC, just start from 110vac. The current being switched can be 1/10th on the primary side, and everything gets much much easier, smaller, more efficiency, etc.
 
Just so I'm clear on Meanwells used as chargers. I could series up 3 of the S-60-24 Meanwells 3x24v=72v. Adjust the output voltages so the sum of the stack is proper for the Lipo pack, and I have a 2.5amp charger? Then connect the balance taps to a bms or battery meds or something to keep the cells in balance. I need to learn this because the next time HK has a lipo sale, I'm in.

The firewal here at work blocks hobby kings site :evil:
 
yopappamon said:
Just so I'm clear on Meanwells used as chargers. I could series up 3 of the S-60-24 Meanwells 3x24v=72v. Adjust the output voltages so the sum of the stack is proper for the Lipo pack.

Yes...but you would be better off using the S-350-24 meanwels they are ~14amp

KiM
 
Back
Top