Fechter's Mini Meanwell Limiter Board

[auraslip]

Thanks neil - you are the boss and have practically taught me everything I've needed to know about these damn things.

[NeilP]

Hey, thanks for that, but all I have really done is collated the info in my mind from all the multitude of sources, then spewed it back out when asked.

I have built my own twin Meanwell unit recently, so have all the info immediately to hand, I can't take credit for any of the work

Here is a link to recumbents page with all the info set out

http://www.recumbents.com/wisil/e-bent/ ... efault.htm

go bak to his home page, he has loads of other interesting useful stuff there too

[auraslip]

That site is legit - his black and yellow recumbent trike is probably the most bad ass electric vehicle I've ever seen!

[fechter]

There's a better picture of where the sensing line attaches on page 4 of this thread.

When using the limiter board, you can leave the limiter resistors in the Meanwell in their original configuration. If you modded the resistor in the Meanwell, it will limit at whichever one is set lower.

[NeilP]

Thanks Richard...I knew I had seen a better one somewhere, I just posted a re link to the first one I came across

[Gregory]

That site is legit - his black and yellow recumbent trike is probably the most bad ass electric vehicle I've ever seen!

I think you are confusing Mr bad ass Matt Schumacher Recumpence with this website guy who is on ES but I don't remember his name.

[TylerDurden]

Website guy is Warren.

memberlist.php?mode=viewprofile&u=6456


Yellow trike is Matt.

memberlist.php?mode=viewprofile&u=6262

[hillzofvalp]

Here's my situation.. I built a 24V supply from two esp114 server supplies for my 1420i. However, for my 14s5p pack I would like faster charging. Could I split the pack into 7s10p for a quick charge? My supplies max out at around 50A. If necessary I could put in the original server fans for better throughput.

There are some pots inside that may adjust the voltage.. but I've never messed with them. Is 24 volts or a little less okay to charge a pack that is 23.1 nominal? I guess fully charged would be 3.42V/cell... but that's probably over 90% charged. I could then balance on the 1420i

If I need a current limiter, which one would be best from a member of ES? OR-- how could I build one from radioshack parts?

[fechter]

I think we'll have to see a picture of the board on that one to make any guesses. I have no experience with that model. There should be some way to do it.

When charging at high rates like that, be careful that no cell group goes over voltage toward end of charge. Bad things could happen fast. 3.65v is the nominal target, but some commercial systems go even higher. 3.42v should be pretty safe and near 90% as you estimated.

[hillzofvalp]

I will take it apart tonight and post some pics. I'm hesitant not having a cut-off because I would like to walk away from the charger every once in while...

WHat guage wire do you guys use over 6 foot to charge at 35-45A? I've got some 10 gauge I thought I might try.

Also, when the battery reaches psu voltage, does the current go to 0?

[hillzofvalp]

Before I was to disassemble, I tested it. I hooked up on 7s5p pack... psu stayed on as if nothing happened... but lights flickered a little bit. It blew lukewarm air for 5 minutes. since the pack is at 23.3V I don't think it's pulling very much, no? I see a .4-.5V jump when it is connected to psu.

I am browsing around to find the voltage control pins on this esp114 server supply, because a lot of people have had success adjusting voltage this way on other server supplies.

[fechter]

I will take it apart tonight and post some pics. I'm hesitant not having a cut-off because I would like to walk away from the charger every once in while...

WHat guage wire do you guys use over 6 foot to charge at 35-45A? I've got some 10 gauge I thought I might try.

Also, when the battery reaches psu voltage, does the current go to 0?

10ga should be OK. Check it when it's running to see how warm it gets.
Yes, as the battery voltage approches the supply voltage, the current will taper down to near zero.
If you try discharging the the pack a little first, you should see higher currents at first.

[hillzofvalp]

how does the charging current curve look?

[fechter]

It depends on your batteries, but here is a typical charge profile:

A123 Charge Curves.jpg (62.15 KiB) Viewed 633 times

[steveo]

Hey Everyone

i just purchased 2 1000w meanwell 48v psu

i am having the issue where i connect the psu to a load (my batteries) and the psu powers down.. I've had the same issue with a small 250w psu which i fixed with one of fechters mini meanwell limiter boards....


I was wondering if anyone could help me on where to connect the sense wire for the limiter board on the 1000w meanwell

here is a couple photos of both sides of the psu board

thanks
-steveo

[NeilP]

What is the white JST style multipin connector, looks like a LiPo balance lead socket?

Are you running the supplies in series or parallel


Try connrecting to a pack that is only a little discharged and see if the supply still drops out

[steveo]

I tried a pack that approx 5v difference between supply and battery and it cutout...


I gues if its 2-3v difference it wouLd not

However when charging a batter i can not keep tweaking voltage to keep charger going , it just has to work!

Hopefully i can get the limiter board wired in there

Steveo

[NeilP]

or try the other type of current limit mods, if you can find the control rrsistor on this new board type

What exact model is this ?

[fechter]

Here's the datasheet for the switching controller chip:
http://www.onsemi.com/pub_link/Collateral/SG3525A-D.PDF

Looks like the error amp inputs are pins 1 and 2.

If I stare at the pictures long enough, I can probably trace it out.
I also see an optocoupler. These usually tie into the feedback,

[steveo]

Here's the datasheet for the switching controller chip:
http://www.onsemi.com/pub_link/Collateral/SG3525A-D.PDF

Looks like the error amp inputs are pins 1 and 2.

If I stare at the pictures long enough, I can probably trace it out.
I also see an optocoupler. These usually tie into the feedback,

I will give pin 1&2 a try

I measured 2.5v on each, which is jusy my old psu that has limiter board

[fechter]

It's pretty hard to trace from the pictures. Do you see anything that looks like a shunt?

My guess is raising the voltage on pin 2 will reduce the output. You could possibly try using a resistor like 100k from pin to the positive output and see what it does to the output voltage. You really want to feed just a little over 2.5v to pin 2 so you need to be careful with jumpers during testing. I think the chip fries if you put a little to much into there. There is probably another part of the regulator circuit that uses the LM358 and that feeds into pin 2.

Near the adjustment pot, there is a spot marked OVP. That might be the 2.5v reference spot.
On most meanwells, the sensing line is connected to one side of the adjusment pot.

The connector has pins marked S+ and S-. These might be remote voltage sensing inputs. You might see what the voltages on those pins look like.
The pins marked RC seem to go to the optocoupler and must be a remote control of some sort. Feeding voltage into the RC input might lower the output voltage. I suspect it's a 5v signal there.

[steveo]

It's pretty hard to trace from the pictures. Do you see anything that looks like a shunt?

My guess is raising the voltage on pin 2 will reduce the output. You could possibly try using a resistor like 100k from pin to the positive output and see what it does to the output voltage. You really want to feed just a little over 2.5v to pin 2 so you need to be careful with jumpers during testing. I think the chip fries if you put a little to much into there. There is probably another part of the regulator circuit that uses the LM358 and that feeds into pin 2.

Near the adjustment pot, there is a spot marked OVP. That might be the 2.5v reference spot.
On most meanwells, the sensing line is connected to one side of the adjusment pot.

The connector has pins marked S+ and S-. These might be remote voltage sensing inputs. You might see what the voltages on those pins look like.
The pins marked RC seem to go to the optocoupler and must be a remote control of some sort. Feeding voltage into the RC input might lower the output voltage. I suspect it's a 5v signal there.

hey fechter

I see the shunt wires on the opposite side of the capasitors, 3 shunt wires on this psu

do you mean use use a 100k resistor connected to pin 2 on that chip & connect it to the 48v dc output of the power supply?

sounds like that will make things crispy...

-steveo

[fechter]

hey fechter

I see the shunt wires on the opposite side of the capasitors, 3 shunt wires on this psu

do you mean use use a 100k resistor connected to pin 2 on that chip & connect it to the 48v dc output of the power supply?

sounds like that will make things crispy...

-steveo

Yes, that's probably a risky way to test it.

Measure the voltage on that 'OVP' pad near the voltage adjusting pot. I think it goes to the pot wiper. Otherwise try to get on the pot wiper somehow. That's really where you want to tie in the limiter. On most supplies it's around 2.5v, but some run 5v or more.

What you really want is to feed the reference point with just a little more than 2.5v. Up to 5v should be safe for the input. Depending on the resistances, 100k to the 48v might be in the safe range, but better to not risk that. I use a resistor to feed test points to limit the current.

Take a pic of the shunts. I saw something that sort of looked like one, but I think it was a jumper.

[steveo]

Hey Everyone

i just purchased 2 1000w meanwell 48v psu

i am having the issue where i connect the psu to a load (my batteries) and the psu powers down.. I've had the same issue with a small 250w psu which i fixed with one of fechters mini meanwell limiter boards....


I was wondering if anyone could help me on where to connect the sense wire for the limiter board on the 1000w meanwell

here is a couple photos of both sides of the psu board

thanks
-steveo

Hey Everyone,

I just want to say that i have gotten the 1000w meanwell to work with fechter's limiter board, The sense wire must tap into the "OVP" spot on the board, right next to the voltage turn pot..

I will post a video on my youtube of it working shortly, stay tuned..

....

Fetcher does the limiter board limit the current at 14amps?.... thats what i maxed out at!

-steveo

[fechter]

It should be close to that. It depends on what value resistor was used above the pot. The shunt should be good for 20A in theory, but will get hotter than hell. I think it will be OK at 15A. It can run pretty hot without damage.