Meanwell RSP-2000-48 Current Adjust

[DaleKramer]

I have an RSP-2000-48 and I love it when I charge at home with my heavy duty AC circuits to charge from but I would like to be able to use it on less 'capable' circuits away from home.

Has anyone figured out how to current adjust the RSP series?

Thanks,
Dale

 

[Doctorbass]

the 1500W version i have is able to output his full CC-CV without tripping the breaker ( normal 15A)

i see you want to limit it to be able to not trip your 15A breaker right? :wink:

You have to find some shunt resistor like 0.05 to 0.5 ohm smt resistor iner or on the top of the pcb. than change that value by removing some of these parallel resistor and it should work. At least it worked on my RCP-1000-48.. but i never tried on my RSP-1000-48 or my rsp 1500-48

Doc

 

[DaleKramer]

Actually I want an adjustable CC stage output current by a pot/knob or something so I can use it on both 'capable' and 'crappy' circuits.

Are you suggesting a knob that adds parallel resistors as needed to the shunt reading circuitry or just different hi power shunt resistors?

Dale

 

[DaleKramer]

So I broke down and took the cover off and found an SVR2 pot set at the factory at 1.0502 kohm that seems to adjust current from about about 36 amps at pot max resistance (turn to left) of 1.8 kohm to about 44 amps at pot min resistance (turn to right) of 0 ohm.

The pot seems to be in series with a 4.4 kohm surface mount resistor.

I am thinking of replacing the 4.4 kohm surface mount resistor with a 10 kohm one to see if I get more range. I would be hoping that the supply can actually work at lower amps and not blow up.

Any takers on 'blow up' or just 'reduce current further'?

Dale

 

[ZOMGVTEK]

Generally, you need to be worried about low resistance. If you want to get a wider current range, or lower currents, it should be safe. The SVR2 pot probably has a series resistor to limit the minimum resistance, since near 0 ohms will probably be a bad thing on that. The supply might get unstable if that value starts to get exceedingly high, or goes open, but replacing a 2K pot to a 10K should be perfectly fine. Definitely swing for a multi-turn pot.

As long as the supply isnt making any odd noises, it's probably fine. But of course, this is simply my opinion.

 

[DaleKramer]

It didn't blow up.

Those resistance values I mentioned earlier were 'in circuit' measured so they were not absolutely correct for part values.

There are actually 2 series resistors with the 2k SVR2 pot. One whose real value is 5.6 kohm and it goes to GND and one whose real value is 2.7 kohm that is on the other side of the pot and it goes to the output of an op amp and the CS testpoint.

Anyway I have achieved what I wanted by replacing the 2k pot with a 10 kohm pot and the 5.6 kohm resistor with a 4.7 kohm resistor.

The 10 k pot now adjusts the current from 20 amps to 40 amps.

 

[adralien]

I have tried this with an RSP1000-48...

I opened up case, found trim pot that is 20K and labeled SVR1... it forms a non-inverting amplifier out of an LM324 nearby... There is a 20K resistor from -ve pin of op-amp to GND and the 20K trim pot is in series with a 90.9K resistor from the op-amp, typical non-inverting setup...

Changed 90.9K to 33K and trim pot to 100K.

And guess what I found?

I can adjust the -fan speed-. The current limit stays the same. So I think the RSP-1000-48 does not have adjustable current limit like the RSP-2000-48.

Continuing my examination to see where current is regulated. The fan speed does respond to overall power/current level, but that may just be with a signal from the HV side of the switcher. It could be that current limiting is done on the HV side, and there is no feedback from the LV side.

 

[Doctorbass]

I have tried this with an RSP1000-48...

I opened up case, found trim pot that is 20K and labeled SVR1... it forms a non-inverting amplifier out of an LM324 nearby... There is a 20K resistor from -ve pin of op-amp to GND and the 20K trim pot is in series with a 90.9K resistor from the op-amp, typical non-inverting setup...

Changed 90.9K to 33K and trim pot to 100K.

And guess what I found?

I can adjust the -fan speed-. The current limit stays the same. So I think the RSP-1000-48 does not have adjustable current limit like the RSP-2000-48.

Continuing my examination to see where current is regulated. The fan speed does respond to overall power/current level, but that may just be with a signal from the HV side of the switcher. It could be that current limiting is done on the HV side, and there is no feedback from the LV side.

:lol: The fan speed.. ok.. so it might be usefull to tune up the air flow in case you boost the output current :lol:

No kidding, let me know even if you find the way to adjust that current value.. maybe following the current sense path on the pcb near the shunt resistor should help to locate the circuit that regulate that current...

These shunt resistors are near the negative output under the pcb.

Doc

 

[Nuts&Volts]

It didn't blow up.

Those resistance values I mentioned earlier were 'in circuit' measured so they were not absolutely correct for part values.

There are actually 2 series resistors with the 2k SVR2 pot. One whose real value is 5.6 kohm and it goes to GND and one whose real value is 2.7 kohm that is on the other side of the pot and it goes to the output of an op amp and the CS testpoint.

Anyway I have achieved what I wanted by replacing the 2k pot with a 10 kohm pot and the 5.6 kohm resistor with a 4.7 kohm resistor.

The 10 k pot now adjusts the current from 20 amps to 40 amps.

Thanks for the great information. Think it would be possible to turn it just a bit lower? I may try series another power supply with a RSP to charge my 21s pack both on a 220V50A circuit at home, but it would be nice to easily turn it down to work on a 110V15A outlet when riding around for an opportunity charge.

 

[DaleKramer]

Think it would be possible to turn it just a bit lower?

Yes, I am pretty sure you could expand the range futher but I can't try this right now.

 

[DaleKramer]

I have tried this with an RSP1000-48...

I opened up case, found trim pot that is 20K and labeled SVR1... it forms a non-inverting amplifier out of an LM324 nearby... There is a 20K resistor from -ve pin of op-amp to GND and the 20K trim pot is in series with a 90.9K resistor from the op-amp, typical non-inverting setup...

Changed 90.9K to 33K and trim pot to 100K.

And guess what I found?

I can adjust the -fan speed-. The current limit stays the same. So I think the RSP-1000-48 does not have adjustable current limit like the RSP-2000-48.

Continuing my examination to see where current is regulated. The fan speed does respond to overall power/current level, but that may just be with a signal from the HV side of the switcher. It could be that current limiting is done on the HV side, and there is no feedback from the LV side.

Too bad, I was hoping this was an RSP series mod. I changed the thread topic to reflect your info.

 

[adralien]

Hopefully I'll have an answer for the 1000-48... here's my findings from today:

Donned nitrile gloves and took the unit apart... lots of thermal goo, overall nice design where components don't have to be unbolted from the case.

R130 to 133 are current sense resistors, only R130 and 132 are populated. Current sense from the resistors enters the vertical daughter PCB at pins 9 and 10.

For a permanent adjust of current sense, it is possible to change the sense resistors. Right now they are 0.005R each, for total of 0.0025R. So removing one would half the output current.

That is tempting for my system. I'm trying to get the max charging power out of a 1000W Honda generator. I have electrified a Cal20 sailboat, and am now finishing up the "hybrid" portion of it by having our on-board generator charge the Li-ion batteries through the RSP1000.

Today's trial: Removed 1 sense resistor to double the signal for the same current. This limited current to ~7.5A. Interestingly with both sense resistors in place the current limit is ~24A. I suspect that the reason the we're not closer to half is that the sense leads from the sense resistors are not even close to proper "Kelvin" connections. So a lot of stray IxR voltage is seen on the sense.

This is good news and means that current limiting is from those sense resistors. Maybe there's still hope for a simple variable current fix.

More later, got to do actual work!

 

[Kingfish]

I am definitely keen on the RSP series but haven’t made up my mind on the 1500 or 2000 because I’ll need two in series to reach my target voltage. The part that I am interested in is how to adjust the current without opening up the box.

Also, could we use Fechter's Mini Meanwell Limiter V.3 instead?
~KF

 

[adralien]

Kingfish, Fetcher's project is good stuff. I have been contemplating a similar idea, except using a PIC to monitor current and run an LCD (to report delivered power).

My concern with a solution like that is that the voltage adjust is only down to 40% of the nominal. So if your batteries (or whatever) can pull more than your max current below 40% of rated output, then you will see a higher current.

I believe the built-in limiting of the Meanwell will run the current limiting down lower than 40% of nominal voltage, but haven't tested this yet.

For my PIC idea, I was going to control on/off of the Meanwell , so if current stops being regulated at low voltages, the output would just be turned off. That could probably be worked into an analog circuit too...

Just an idea at this point though.

 

[Doctorbass]

I have done the mod for my RSP-1000-48 with the current sense resistor and it work well.

I have installed a 20A switch switch in serie with the second sense resistor to get full or half current. I get exactly Half the current by ising only oe resistor.

I also have investigate on the vertical board,and tried to parallel to add some 100kohm resistor with few resistor that i thought could be part of the voltage divider on the reference voltage near the Op amp..to change the value while i had a full load on the meanwell but it did not worked.. maybe it was the wrong resistor set.. But i'm convinced that there is a way to find that...

maybe with all us searching for a solution we will finally acheive that :wink:

Doc

 

[Doctorbass]

Yesterday i built my 1500-2500W charger for 30s lipo ( 126Vdc) from two meanwell RSP-1000-48 connected in serie and i also used 4x DC-dc converter BXB150 connected in 2s2p to boost the voltage of each meanwell to 63V.

I have installed a switch on the second shunt resistor to allow full power from a 120V 15A outlet or from a 220V 15-30A outlet. It work great!

The overall size is the same as the RSP 1500 model but is having a heatsink and 4 DC-Dc converter on it to boost the voltage.

The measured efficiency between the AC input POWER and the DC output POWER is 84% while the Power factor remain 0.99.

The efficiency of the DC-DC is in the best optimum area of the input voltage vs efficiency curve since i am supplying them at 50Vdc from the adjusted meanwell so the lost of the dc-dc is minimal.

I made a thread about that:

Doc

 

[adralien]

Doctorbass, sounds like you've been busy! 30s lipo is a big stack.

I played around some more with my RSP1000-48 but have not found a magic fix to adjust current limit. I traced out some of the current sense lines, but after playing with some things I got nervous and decided that it wasn't worth burning out my unit for.

I have simply modified my sense resistors, which seems to work well.

 

[Doctorbass]

Doctorbass, sounds like you've been busy! 30s lipo is a big stack.

I played around some more with my RSP1000-48 but have not found a magic fix to adjust current limit. I traced out some of the current sense lines, but after playing with some things I got nervous and decided that it wasn't worth burning out my unit for.

I have simply modified my sense resistors, which seems to work well.

yeah this is what i tried too .. well awith the sense resistor i had that can match the power they require.. Also I had succes dividing by 2 the output by removing one of the current sense resistor in parallel.

Doc

 

[Doctorbass]

Guys, I did not checked and always assumed that the 2000W was the same as the 1000 and 1500W model... lol.. i was stupid!...

There is a pot on the RSP-2000 that adjust the max current limit !!!

It is located on the top of the vertical pcb and is calles RVS2.

 

[Doctorbass]

Oh and btw if you want to add a protection when connecting many in serie, just add some shottky diode in parallel to the output in NON CONDUCTING mode

I have ordered some 60A shottky diode from ebay. These are in the popular To-220 case and there is two diode in each case that are connected with common anode. I just parallel them together. I have added a copper block between both anode to solder them together and soldered the case directly to the meanwell terminal.

 

[Doctorbass]

Thanks for the update Doc. Does the pot on this daughterboard limit current too? This is from the RSP 1000-48 I got from you.

No unfortunatly on the RSP-1000 there is no current adjust. the pot on the daughter board does not adjust the current limit. it is only on the RSP-2000)

But i began to investigate the daughter baord on the RSP-750 and i'm trying to reverse engineer it it look very similar to the rsp-1000

Doc

 

[dnmun]

if you can take pictures of both sides of that daughter board and where the traces go from the wiper and terminals of that cermet trimpot over to those two ICs on the board and the labels on the ICs so we can try to deduce the curcuit. then maybe guess where to add feedback to adjust the output current. on the kingpans the trimpot was in parallel with the shunts but that cermet is not the normal Bournes ten turn type trimpot either.

 

[Doctorbass]

if you can take pictures of both sides of that daughter board and where the traces go from the wiper and terminals of that cermet trimpot over to those two ICs on the board and the labels on the ICs so we can try to deduce the curcuit. then maybe guess where to add feedback to adjust the output current. on the kingpans the trimpot was in parallel with the shunts but that cermet is not the normal Bournes ten turn type trimpot either.

Sure i will !

I have a SE-750
I have a RSP-750 ( deffective and got it for free for rev engineer)
I have a RSP-1000
I have a RSP-1500
I have 3 RSP-2000

I love the meanwell!! 8)

Doc

 

[circuit]

RSP-750 has some nasty behavior... Sometimes shows error when adjusted with external voltage.