Switch mode supply mod

[El_Steak]

I'm about to order my additional power supplies.

My pack is 24s Lipo, so I need 100V and 1400Watts would be cool.

I currently have one S-350-24 (clone). I could simply buy 3 more 24V models and be done with it, but I thought it may be more useful in the long run to buy different voltage models. Perhaps a 12V, a 15V and a 48V. This (plus my current 24V) would give me the same 100V / 1400W charger but could also be reconfigured for pretty much any possible voltage under 100V should I need it for other projects.

Am I looking for trouble putting 4 different voltage models in series ?

BTW, on my S-350-24 clone model, adding a 1K resistor in parallel with the existing thermistor did the trick for the "always on" fan mod.

 

[Gregory]

Aren't you going to be limited by the lowest current rated supply?
In this case the 48v 7a so max watts would be 100v x 7a =700w- other wise i'd say it sounds like a good idea to get a variety of supplies.

 

[El_Steak]

Aren't you going to be limited by the lowest current rated supply?

Yeah, I think you are right. Not a good idea in that case.

 

[MrKang]

Am I looking for trouble putting 4 different voltage models in series ?

BTW, on my S-350-24 clone model, adding a 1K resistor in parallel with the existing thermistor did the trick for the "always on" fan mod.

If you just keep one modified for regulating the amps it's ok. all the others only do the fan-mod

I would buy 3 24V for charging your 24s with 14A and maybe buy only one 12V for making different configs.
Then you an charge 24-36-48-60-72-84-100

 

[Arlo1]

I'm about to order my additional power supplies.

My pack is 24s Lipo, so I need 100V and 1400Watts would be cool.

I currently have one S-350-24 (clone). I could simply buy 3 more 24V models and be done with it, but I thought it may be more useful in the long run to buy different voltage models. Perhaps a 12V, a 15V and a 48V. This (plus my current 24V) would give me the same 100V / 1400W charger but could also be reconfigured for pretty much any possible voltage under 100V should I need it for other projects.

Am I looking for trouble putting 4 different voltage models in series ?

BTW, on my S-350-24 clone model, adding a 1K resistor in parallel with the existing thermistor did the trick for the "always on" fan mod.

You going to be limited by the amps of the weakest charger. I would just go with 2 48v supplies to keep things small and if you want 14 amps then 4 48v supplies 2 in seires then the two packs parlleled. But the 4 48v psu will pull ~20 amps from the wall so do you have a breaker rated that hi?

 

[number1cruncher]

US building code requires 20 Amp receptacles in the kitchen and bathrooms. Of course a home could have been built before this code was implemented. And El Steak is in Canada, but I think 20 Amps is reasonable. I'd go Mr. Kang's suggestion, as it seems more scalable.

 

[El_Steak]

It gets pretty cold around here during the winter so I had a semi-portable garage heater installed when I built the house. That means I have a nice 220V / 30Amp plug in my garage: 6000W+

That being said, with the proper R33 or shunt mod or Gary limiter, the 4 x 24V supplies should output a max of 1400 Watts. That's less than most hair dryers (1800 watts). So a standard 110V / 15 amps should be able to handle it if nothing else is live on the circuit. I don't know about power spikes though, mkeefer reported that he would pop the breaker on his high power setups.

My pack is 1332 watt/hour, so that charger would give me pretty much within-spec 1c charging 8)

 

[GGoodrum]

It gets pretty cold around here during the winter so I had a semi-portable garage heater installed when I built the house. That means I have a nice 220V / 30Amp plug in my garage: 6000W+

That being said, with the proper R33 or shunt mod or Gary limiter, the 4 x 24V supplies should output a max of 1400 Watts. That's less than most hair dryers (1800 watts). So a standard 110V / 15 amps should be able to handle it if nothing else is live on the circuit. I don't know about power spikes though, mkeefer reported that he would pop the breaker on his high power setups.

My pack is 1332 watt/hour, so that charger would give me pretty much within-spec 1c charging 8)

I use four S-350-24 Meanwells in series off the same circuit, and it hasn't yet been a problem. One of these units has the R33 mod, and the current starts out at 14A.

-- Gary

 

[mwkeefer]

Stick with multiple units of the same voltage for your primary charger - as someone else said, perhaps grab a 12v too for rigging odd voltages and one offs but things really are simpler and more reliable when running supplies in parallel or series to use the same supplies where possible

-Mike

 

[El_Steak]

Alright, 3 clones are on the way to my place. Damn this is an expensive hobby.

I bought 2 from this guy:
http://cgi.ebay.ca/ws/eBayISAPI.dll?ViewItem&item=160459987701

and another one for a couple of dollars less from this guy:
http://cgi.ebay.ca/New-24V-DC-14-6A-350W-Regulated-Switching-Power-Supply-/120589302236

I'm curious to see if they will all be the same inside...

 

[mwkeefer]

El,

Compared to hot rodding, or many other "movement" related hobbies - this is far less expensive (by factors of scale) and infinately more useful in the grand scheme of things

-Mike
PS: Have you ever seen a 10K set of carbon fiber golf clubs? Add to that the membership, clothes, etc... your looking at 15k startup costs and even worse transport costs getting to the course

 

[El_Steak]

Yeah, the problem is this e-bike addiction is just leading to harder stuff... like e-cars !

I even came close to buying a 5 speed 1997 Geo Metro for a conversion but the GF put an end to my madness by reminding me of all the remodelling I have to do in the house

With small kids, the problem is more time than money for me. Still a car conversion would be a great father-sons project when they get a bit older.

 

[mwkeefer]

A metro really, thank god for your GF! Why waste the time converting a metro? of all the choices
-Mike

 

[El_Steak]

I modded my S-350-24 tonight. I grinded one of the shunt until I got an output of 350W. Its working, unfortunately I get the annoying and fairly loud buzzing sound that others have reported.

From what I could gather from this thread this is "normal" (expected?). Any worries long term ?

BTW, If I put a load of a bit under 350W on it (iCharger 206B), I get no buzzing at all. It really only happens when the PS has to limit the current because of the load.

 

[amberwolf]

A metro really, thank god for your GF! Why waste the time converting a metro? of all the choices
-Mike

It's a popular conversion because it's easy and has been done a lot, so plenty of ways to copy if you like, cheap to find fixup parts, etc. Also, it's small and light; I could probably run one from a wheelchair motor (though it wouldn't go very fast. :lol

 

[El_Steak]

A metro really, thank god for your GF! Why waste the time converting a metro? of all the choices
-Mike

It's a popular conversion because it's easy and has been done a lot, so plenty of ways to copy if you like, cheap to find fixup parts, etc. Also, it's small and light; I could probably run one from a wheelchair motor (though it wouldn't go very fast. :lol

Yes, its a good starter car for us up north who don't have access to a lot of road worthy VW beetles. You can easily fetch a decent one with a 5 speed transmission and good body for under 500$. Being that light, I could run it at a relatively low voltage like 88-100v and reuse some of my ebike stuff (Lipo - if I tripple it to 24s9p), my upcoming Gary celllog-based bms (soon I hope) and this new 4 x S-350-24V power supply charger.

Plus I know I'll enjoy the look of contempt from some of my SUV-driving neighboors when they see an '97 metro in my driveway

But this is not going to happen in the short term so I'll continue to have fun with e-bikes in the mean time.

 

[Hyena]

Heath just posted some good info on how to repair these meanwells if you let the magic smoke out.

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=21156&p=310725#p308009

 

[fechter]

I've been working on a functional current limiter for these things and I think I have a workable solution. I still need to build and test more, but it looks promising.

The limiter board will go in line with the output and measures the current. When the current tries to go above the adjustable limit, the output voltage will drop enough to keep it right at the limit until the load decreases enough to go back to full voltage. This board will require tapping into a spot on the board, which varies by model variation, but so far the few samples I've looked at seem pretty easy. It's just one wire that taps into the voltage sensing signal line for the voltage regulator, which is one side of a resistor usually located near the voltage adjusting pot.

If this works out as planned, it will transform a Meanwell into a true CC-CV power supply, which will be ideal for charging batteries. If the CC limit is lowered to a reasonable level (substantially below the rated current), the heating and reliability will be much improved.

So far the circuit looks like it can be pretty small, possibly small enough to fit inside the case. It doesn't have a crapload of parts either, so shouldn't cost too much to build.

Stay tuned for further developments...

 

[fechter]

Testing is going along like usual. A careless wiring mistake put 24v into an op amp rated for 16v. With 13 amps, it didn't last long. It blew the top of it off with a loud snap and a flash of light. Like a mini version of the big gas explosion down the street.

It also fried pretty much everything else on that board, so I decided to just build from scratch the actual circuit I had in mind.

The prototype is pretty crude, but good enough for testing. A 'real' version of this could be quite a bit smaller, like about the size of a large postage stamp. Here's how it looks wired up for testing:


No fireworks this time. Everything seemed to work close to expectations.

One limitation of the SP-320-24 is the voltage can't go below about 17v. This would only be an issue if your pack voltage was less than this. I think for most applications it won't be a problem. It seems the PWM duty cycle can't go below a certain point. The S-350-24 does not seem to have this limitation and the output can go almost to zero.

The circuit is about as "cave man" simple as I can get it. A single op amp works as a comparator to compare the voltage across the 5 milliohm shunt resistor to a reference set by the adjustment pot. A voltage regulator is needed for the circuit, and a couple of resistors and a diode. I didn't install the voltage regulator caps and it seemed to work OK, but it would be a good idea to use them. Whenever the current gets up to the point where the op amp triggers, it pulls up on the voltage sensing line in the supply, causing the output voltage to drop.

For supplies with under 35v output, only the LM78L12 regulator is needed. For over 35v, a pre-regulator (Q1) is recommended. The circuit draws around 2ma.

Here's the schematic:

I think at some point I can make up some boards and offer them.

 

[Arlo1]

I love you guys!

 

[Tiberius]

......
Here's the schematic:

I think at some point I can make up some boards and offer them.

Richard,

remember this ? http://endless-sphere.com/forums/viewtopic.php?f=14&t=6749#p209081

What are you doing about stability, with two feedback loops controlling the PWM?

Nick

 

[fechter]

Dang! those brain cells are dying fast these days.

Funny, I think the SP-320-24 I'm testing with is the very same one Methods was using (he sent it to me).

I was worried about oscillation, but I think the loop gain is low enough that it seems to be behaving. I did test over a fairly wide range of loads. Since my feedback is going into the same spot as the voltage feedback, the same compensation network is getting used.

So far so good, but I want to try it with the S-350-24 I have.

The SP-320 has the 'hiccup' overcurrent protection so isn't good for a battery charger, but seems to work great with the circuit attached.

 

[GGoodrum]

This is a MUCH better solution, I think.

Below is what a small implementation of this would look like. It is 1.25" x 1.5", and would mount to the MeanWell's output terminal block, using longer screws and standoffs.

View attachment MW Current Limiter-v4.2.2.png

 

[peterperkins]

I've just done the mods on three Meanwell S-350w psu's. Thanks to everyone else for all their work on these.

In bought them from the below, came in about 10 days from china. Three chargers came to about $110 which is very cheap compared to a 'proper' EV battery charger. I paid over £500 for my last Zivan charger and it took two months to arrive

48v items - Get great deals on Tools, Prototype PCBs items on eBay.co.uk Shops!

I'm using them to charge my 50 x 20ah A123 cell pack. I have them running at 60v and 5.5A output.

They seem pretty easy to mod, with all the bits identified in the endless sphere forum threads in the right places on the pcb.

The three supplies all have the following mods.

1) Fan mod to activate fan as soon as psu is turned on. Simple 470R resistor inserted instead of thermistor.
2) Raised output voltage mod. Simple replacement of 2k resistor with a 1.2k device to push up the operating voltage range window.
2A) Replacement of over V voltage detection zener with a 68v device to enable 2) above.


One of the psu's (Let's call it the Master) has three further mods.

1) Current limit mod to lower output current so total supply power does not exceed 350w or so. I started by removing one of the three shunts on the pcb reducing output current by a third.
2) I then removed R33 and added a 500R pot into the free space on the pcb at SVR2 and a link at R27. Gives fully variable current limit.
3) I also brought out two connections which bridge the output voltage pot to use that to remotely control the operating voltage later. Perhaps via a digital pot.

All three psu's in series should run at the current as determined by the Master.

The charger is controlled in three ways.

1) The total output voltage of the combined psu's is adjusted to the pack full voltage limit or a tad below. So they limit themselves when output voltage rises to the preset maximum. So if the BMS fails and the pack is reasonably balanced it should be impossible for them to overcharge the pack.

2) The BMS controls the output voltage of the Master Psu. If a cell goes over V it can cutback the voltage of the the master PSU reducing charge current.

3) The BMS controls the solid state mains relay which powers up all three Psu's. If a cell goes over V and it can not be corrected by 2) above then the mains can be cut.

 

[fechter]

The charger is controlled in three ways.

1) The total output voltage of the combined psu's is adjusted to the pack full voltage limit or a tad below. So they limit themselves when output voltage rises to the preset maximum. So if the BMS fails and the pack is reasonably balanced it should be impossible for them to overcharge the pack.

2) The BMS controls the output voltage of the Master Psu. If a cell goes over V it can cutback the voltage of the the master PSU reducing charge current.

3) The BMS controls the solid state mains relay which powers up all three Psu's. If a cell goes over V and it can not be corrected by 2) above then the mains can be cut.

Nice work. That's really the right way for the BMS to limit current. What are you using for a BMS?

The S-350 has a nice current limiter section, so it makes sense to use it. The 320 and some other variations don't have a secondary side limiter and can only hiccup on overload. There also seems to be a steady stream of changes to the circuit layout, so it gets confusing to keep up with all the revisions.