drewjet
10 kW
Anybody else have trouble with the fans? I have had 2 die in the last week.
Switch mode supply mod
- Thread starter Jeremy Harris
- Start date
mwkeefer
1 MW
drewjet said:
Which fans the stock on the S-350 or the ones on the SP-320? What mods have you done for fans?
With regards to the SP-320 models, I have created a CC limiter which can be slapped on the front end... its running too hot right now and I need to make some serious adjustments before completing a charge cycle but I can now connect my 3 SP320s set for 20.75 in series through the limiter and connect to a pack even fully discharged to 3.0v per cell over 15S and it will prevent hiccup mode!
That said, I've run it about 2 minutes total (over 8 or 9 attempts) and the heat buildup is quite bad - I think the PWM needs adjustment lower (meaning less current) and also that I need to heatsink better...
I hope to come up with a bench version this week which can be used for full charging at up to 15A rate (for starters) and up to 75v (change the mosfets for higher voltage/current)... once I have successfully tested a few charge cycles... I will release the schematics here and also PCB artwork for those who want it.
-Mike
drewjet
10 kW
mwkeefer said:
Actually it is a s-400-12. I had one die while still hooked to the thermister and another one die after hooking up directly to the 12v output. I have the voltage set at 15v. Only mod I have done is cutting 2 shunts to limit amps to 25.
Hi,emzero said:
my modification (described a few posts before) doesn't work
Could anybody help me please!!!
First everthing seemed working fine:
The voltage could be set up to 66.8 V until the unit shuts down.
So i fixed the voltage to 66.0 V and connected it to my really empty!! (about 30 V - 40x 1,2 V NiCad) batteries.
In the beginning everthing seemed perfect: a constant load of 2.1 A ... and no problems ... even heat was acceptable
But as the voltage reached around 50 V the currrent slowly went down to only a few mA
Does anybody have any idea what is happening here????
mwkeefer
1 MW
emzero said:Hi,emzero said:
my modification (described a few posts before) doesn't work
Could anybody help me please!!!
First everthing seemed working fine:
The voltage could be set up to 66.8 V until the unit shuts down.
So i fixed the voltage to 66.0 V and connected it to my really empty!! (about 30 V - 40x 1,2 V NiCad) batteries.
In the beginning everthing seemed perfect: a constant load of 2.1 A ... and no problems ... even heat was acceptable
But as the voltage reached around 50 V the currrent slowly went down to only a few mA
Does anybody have any idea what is happening here????
I can take a stab or two at this one if you like...
First - you refer to the blown area in the first pic on the right being circled in red but I couldn't find it.
Second - you seem to have found an SP-100 model which has shunts, kudos.
Third - as part of your initial modification - in prep for 2 of these in series for your 15s pack (Lipo I assume right?) you removed a shunt. This supply the SP-100-27 is already designed for a maximum of 130% rated power or a peak of 4.815 A output current at 27v. By removing a shunt, you effectively cut that limit in half at 27v leaving you approx: 2.40/2.41 A possible output power (you are in effect running the unit at 65% of continuous rated power without the second shunt).
Fourth - You then upped the voltage (not sure if it's right or not, I didn't have a similar unit - smallest i have is the S-145 which I have taken to 83v with 100v output caps) - this is your biggest problem:
Now you have a supply rated at 100w continuous or 130 maximum but you have halved that to 65w maximum - at the same time you cut the constant current rate by 1/2.
I explained the effect at 27v (2.4/2.41 A CC to 27v) but that scales when moving to higher voltages, lowering current as the voltage increases....
The math says 65w / 66v = .98 A or just below 1A - add to that the inefficiencies and the inexact calibration and well.. that is the most you could possibly expect at 66 V, then it would taper down to MA after that as the CV phase did it's natural taper.
So.. you connected 30v (dead) pack to the charger and things were fine for a bit... (at 30v the 2.x A limiting was fine - 65w / 30v 2.1x A) - then you say at 50v it dropped to mere mv.... how many mv (500/900?)
At 50v the maximum you should see at 65w would be in order of 1.3A output at 65w...
To make this a CC/CV true charger you will need to do the following...
Assume 66V peak and put the other shunt back in place for a true 100w of power
@ 66V when the taper off will begin (it will begin before that due to under power, once the 4.1A can't be maintained it will taper) it needs to have been CC mode at no more than 130w / 66 = 1.96 A... if you can manage a 1.96A current limit from 30V up to 66V you will then see the normal taper off but you have a very low current charger.
Again... I think this will work but I cannot guarantee it...
Also - the idea of a small fan, not a bad idea.. yes the unit is passively cooled but if you coupled a simple variable pot with a 250ma 5v cell phone charger you could connect the charger guts inside and use them to power the fan... the resistor would be to adjust the fan speed to an ideal one.
One other issue.. with temperature changes come voltage changes, it's true of all the models... the best solution I have found so far is to relocate the precision resistors for voltage and current off board and onto daughtercards.
BTW: My 83v CC/CV charger required replacement caps (of large capacity to maintain level charge state in output stage) and also was based on an S-145-48 which I modified in much the way you did... My current from 30v to 83v is constant at 1.9*-2.0A where it begins to taper off... I am actually maxxing (with active cooling solution) to about 165w but that doesn't last too long and only for about 15 minutes of the final stages of the charge until the delta is reached.
I hope this helps, that I wasn't too complicated and that I wasn't wrong!
-Mike
Hello Mike,
first of all i would like to thank you very much, hoping i understood you right!
BTW: The blown area on my pic is unfortunately cut off within this forum ... don't know why :-(. If you rightclick on the pic it will be shown completely.
My charging-method is a little different:
My EV has NiCad-aircraft-cells which should be charged CC/CC.
First step is a CC-charge at 50 A - 60 A up to 63 V, which is done by another power-supply which shuts down at 63 V.
The second step shall be a CC-charge of 2.0 - 4.0 A for 2-3 hours for equalizing the cells (this shall be done by the Mean Well in parallel but stopped by time instead of voltage).
During this second step voltage usually rises up to appr. 64 V (maybe even 65 V).
So i was hoping that 66 V would be enough guarantee CV-Mode without dropping Current (or at least not to much)
The SP100-27 only has one single shunt inside. I flattened it and took off appr. half it's material by a kind of rasp so that current was limited at 2.1 A
So after all modifications the maximum-output should be (65 V x 2.1 A) 135 W - but for less than 1 hour. To avoid overheating i fastened a big aluminium-radiator (hope it's the right word!) to the case.
Did i get you right that reduction of the shunt not only causes lower current but even lower power at all even if voltage is double?
Now i did another test:
I replaced, as you said, the modified shunt with the original one from the second supply, discharged my batteries down to 31.5 V and then quickly connected the modified unit.
At the beginnig (unfortunately my computer only displays current in 0.5 A-steps) it startet with 5.0 A (most probably 4.8 A).
But just after seconds alike to the increasing of voltage the current went down, and at appr. 39 V it was less than 500 mA (when I checked it later with my multimeter there was a constant charge of 380 mA ... and it seems to hold that 380 mA)
That means about 15-20 W! :-(
I wonder if there was something else influencing the output-power besides the SVR1 + its resistor, the zener-diodes (voltage) and the shunt (current)
Disadvantage wrote about his "GuaLiLai"-modifications:
But Mike, if you are with me, i would like to ask another question:
The modified SP-100-27 would only be my second solution (of course i would like to solve this problem!!), but i would prefer to do the described equalizing-charge by pulsing with my SP-320-48 in Hiccup-Mode.
But therefore i would have to influence the hippcup-intervals because at the moment the medium charge is less than 2.0 A.
You wrote that by changing one restistor it was possible to keep the charge a little longer before the unit shuts down, which would increase the medium charge .
If you could help me finding this resistor, i would try to modify my SP-320 to do pulsing with a medium charge of 2.5 A - 3.0 A an i would have a great charger for lowest money
first of all i would like to thank you very much, hoping i understood you right!
BTW: The blown area on my pic is unfortunately cut off within this forum ... don't know why :-(. If you rightclick on the pic it will be shown completely.
My charging-method is a little different:
My EV has NiCad-aircraft-cells which should be charged CC/CC.
First step is a CC-charge at 50 A - 60 A up to 63 V, which is done by another power-supply which shuts down at 63 V.
The second step shall be a CC-charge of 2.0 - 4.0 A for 2-3 hours for equalizing the cells (this shall be done by the Mean Well in parallel but stopped by time instead of voltage).
During this second step voltage usually rises up to appr. 64 V (maybe even 65 V).
So i was hoping that 66 V would be enough guarantee CV-Mode without dropping Current (or at least not to much)
The SP100-27 only has one single shunt inside. I flattened it and took off appr. half it's material by a kind of rasp so that current was limited at 2.1 A
So after all modifications the maximum-output should be (65 V x 2.1 A) 135 W - but for less than 1 hour. To avoid overheating i fastened a big aluminium-radiator (hope it's the right word!) to the case.
Did i get you right that reduction of the shunt not only causes lower current but even lower power at all even if voltage is double?
Now i did another test:
I replaced, as you said, the modified shunt with the original one from the second supply, discharged my batteries down to 31.5 V and then quickly connected the modified unit.
At the beginnig (unfortunately my computer only displays current in 0.5 A-steps) it startet with 5.0 A (most probably 4.8 A).
But just after seconds alike to the increasing of voltage the current went down, and at appr. 39 V it was less than 500 mA (when I checked it later with my multimeter there was a constant charge of 380 mA ... and it seems to hold that 380 mA)
That means about 15-20 W! :-(
I wonder if there was something else influencing the output-power besides the SVR1 + its resistor, the zener-diodes (voltage) and the shunt (current)
Disadvantage wrote about his "GuaLiLai"-modifications:
But i didn't understand the meaning of this resistor and have now idea of how to figure it out :-(
But Mike, if you are with me, i would like to ask another question:
The modified SP-100-27 would only be my second solution (of course i would like to solve this problem!!), but i would prefer to do the described equalizing-charge by pulsing with my SP-320-48 in Hiccup-Mode.
But therefore i would have to influence the hippcup-intervals because at the moment the medium charge is less than 2.0 A.
You wrote that by changing one restistor it was possible to keep the charge a little longer before the unit shuts down, which would increase the medium charge .
If you could help me finding this resistor, i would try to modify my SP-320 to do pulsing with a medium charge of 2.5 A - 3.0 A an i would have a great charger for lowest money
mwkeefer
1 MW
Question - How long did you leave the charger connected in this 15w mode, did you pull it or leave it a while to see if it popped back up ?
I am working on the resistor to change duration and also which CAP to short to override the IC internal hiccup mode altogether (but think I was wrong about the CC limiting on these models)... I was cruising the Meanwell USA site looking for model differentiation and came across this in a FAQ:
How do we choose a power supply to charge a battery?
Ans:
MEAN WELL has ESC, SC, PA, PB series for battery charger purpose (120~360W). If these series does not meet customers' demands and they like to choose a power supply as a charger, we advice to pick up one which over load protection (OLP) mode is constant current limiting. The models in this mode provide constant current even when the protection circuit is triggered.
The second choice is fold-back current limiting or constant wattage model. In this model, when a battery is running low, the output current of the power supply will gently increase. The level of increase depends on battery's capacity and degree of exhaustion.
Hiccup or shut down model are not recommended because it will stop to generate current when OLP happens.
The SP-320 series has only hiccup mode according to their own specs, the next size up has CC limiting then a larger (I think 750w) unit provides both constant current mode and 3 second hiccup mode. All in all without some other front end to handle the issue of current limiting... I'm afraid it may not work.
I will double/triple check my hiccup mode timing mods on a single supply later tonight when my son goes to bed... I will log it with eagle tree... The issue with my current resistor mod, it takes longer to trigger because the current through the resistor (higher impedance) is lower but that is causing heat when I tested for 1-2 minutes straight... I need to figure out if I can increase the CAP on the other side and upgrade the resistor to a higher wattage one to handle the additional dissipated heat.
-Mike
I am working on the resistor to change duration and also which CAP to short to override the IC internal hiccup mode altogether (but think I was wrong about the CC limiting on these models)... I was cruising the Meanwell USA site looking for model differentiation and came across this in a FAQ:
How do we choose a power supply to charge a battery?
Ans:
MEAN WELL has ESC, SC, PA, PB series for battery charger purpose (120~360W). If these series does not meet customers' demands and they like to choose a power supply as a charger, we advice to pick up one which over load protection (OLP) mode is constant current limiting. The models in this mode provide constant current even when the protection circuit is triggered.
The second choice is fold-back current limiting or constant wattage model. In this model, when a battery is running low, the output current of the power supply will gently increase. The level of increase depends on battery's capacity and degree of exhaustion.
Hiccup or shut down model are not recommended because it will stop to generate current when OLP happens.
The SP-320 series has only hiccup mode according to their own specs, the next size up has CC limiting then a larger (I think 750w) unit provides both constant current mode and 3 second hiccup mode. All in all without some other front end to handle the issue of current limiting... I'm afraid it may not work.
I will double/triple check my hiccup mode timing mods on a single supply later tonight when my son goes to bed... I will log it with eagle tree... The issue with my current resistor mod, it takes longer to trigger because the current through the resistor (higher impedance) is lower but that is causing heat when I tested for 1-2 minutes straight... I need to figure out if I can increase the CAP on the other side and upgrade the resistor to a higher wattage one to handle the additional dissipated heat.
-Mike
Hello Mike,mwkeefer said:
i'm sorry, but i didn't look for the time, but it must have been something between 30 min and 1 hour till volltage was 50 V when i ended that test.
And i was surprised that caused by that 15 W (40Vx0.38A) -20 W (50Vx0.38A) the case and the radiator became warm.
The Mean Well-supplys are all different :-(
E.g. the SP-100, SP-150, SP-200 have Constant-Current-Limiting
SP-300, SP-320 have Hiccup
SP-480 has CCL
SP-500 has Foldback CL
RSP-600, RSP-1000 have CCL
RSP 2400 and RSP 3000 have adjustable CCL
I just got a MW SP-480-48 (48V, 10 A) and it's CL works somehow different to the e.g. SP-100, S-350:
It has no shunt inside and it's current is really limited to exactely 10 A (=100%).
On the other hand a friend of mine has the RSP-3000 (48V, 62.5 A) inside his EV, and he is able to adjust ist up to 75 A (=120%) again.
I just bought one DRP-480-48 at ebay which should arrive this week and i am very curious what this one will be like.
If you are interested i can take a pic of the SP-480 and the DRP-480 (as soon as i have it here) to post it here.
addition:
I've just made another test with my SP-320-48.
First result is that i don't need to modify the interval, because the medium charge is 3.1 A per hour, which would be perfect!!
Second is that there seems to be the same phenomenon i have with my SP-100-27 described above:
I modified the SP-320 in changing the voltage adjustment (poti+resistor) an the zener, so that i could set voltage to 66 V
But even at appr. 50 V the current drops and at 60 V there is only a charge left oft 1 A per hour (instead of 3.1 A before)
And i bet, if i would change the zener to a even higher value which would allow me to set voltage >70 V, this wouldn't change anything concerning the current-drop starting at appr. 50 V
Result:
Both the SP-100 and the SP-320 reduce their charge when the nominal voltage (27 V/ 48 V) is exceeded a few volts
I'm no whiz at this electronics stuff, but I understand the basic principles that have been laid out in this thread. Things aren't going quite according to plan though.
First off, I added a 1K resistor in parallel to the thermistor to keep the fan on at all times. This works fine.
Next I measured R37 to be 327ohms, (I'm using a S-350-24 Meanwell clone). I have a Turnigy watt meter and it shows 20.65A when I connect a 12V heater from a car to it. I think this is too much for the heater as after a few seconds it will cut off, and you can hear something trip inside. After it's cooled for a few seconds it clicks again and I can use it for testing. But the 20.65A is constant each time I try it.
I want 15A, so 15/20.65 x 327 = 237.5 ohms. Using the parallel resitor calculator gives me R2 as 867. I measured a 1K resistor at 1006 ohms, so putting this in parrallel with R37 should give 246 ohms, which when I measured it was 246. So 246/327 x 20.65 = 15.6A. However, when I hook my load back up it's drawing over 17 amps.
I'm going to pop in to town and pick up a 0-2K trimpot, but in the meantime can any of you guys see where I might be going wrong?
As an aside, I have come to realize that my soldering skills SUCK!
First off, I added a 1K resistor in parallel to the thermistor to keep the fan on at all times. This works fine.
Next I measured R37 to be 327ohms, (I'm using a S-350-24 Meanwell clone). I have a Turnigy watt meter and it shows 20.65A when I connect a 12V heater from a car to it. I think this is too much for the heater as after a few seconds it will cut off, and you can hear something trip inside. After it's cooled for a few seconds it clicks again and I can use it for testing. But the 20.65A is constant each time I try it.
I want 15A, so 15/20.65 x 327 = 237.5 ohms. Using the parallel resitor calculator gives me R2 as 867. I measured a 1K resistor at 1006 ohms, so putting this in parrallel with R37 should give 246 ohms, which when I measured it was 246. So 246/327 x 20.65 = 15.6A. However, when I hook my load back up it's drawing over 17 amps.
I'm going to pop in to town and pick up a 0-2K trimpot, but in the meantime can any of you guys see where I might be going wrong?
As an aside, I have come to realize that my soldering skills SUCK!
I've now got this running at a constant 15.5A under load using a 1K trimpot set at about 500 ohms. By my calcs that should give me about 12.5A, but as long as I've got 15A I'm happy. I opened up the heater, cut the lead to the blower motor and used an aligator clip to keep the bimetallic strip from opening, (the clicking I heard). Got it glowing nice and red. I was able to let it run for 30 seconds or so, checking that the PSU didn't over heat. Of course, I should have kept an eye on the heating element as left running continuously it melted the plastic case it was fixed to. 
So now all I need to do is do the fan mod to the other 3 supplies, hook 'em up in series and away we go!! I can't thank you guys enough for your posts. This solution has saved me well over $100, and will make installation so much easier as I will be able to mount the PSUs to make more efficient use of the space I have.
I was able to let it run for 30 seconds or so, checking that the PSU didn't over heat. Of course, I should have kept an eye on the heating element as left running continuously it melted the plastic case it was fixed to. So now all I need to do is do the fan mod to the other 3 supplies, hook 'em up in series and away we go!! I can't thank you guys enough for your posts. This solution has saved me well over $100, and will make installation so much easier as I will be able to mount the PSUs to make more efficient use of the space I have.
mwkeefer
1 MW
Dave,
Sorry I didn't get to you with the previous post in time... What your seeing in the first post is the effect of a lower voltage rail dropping (the TL494CN expects a 48v rail) under the load of the FAN... this happens with all of them... The faster the fan, the lower LOW VOLTAGE rail sags... you can see this when the fan hasn't been modified as a temporary spike in current from about 5.2A to 6.2A when the fan kicks in on the S-350-48 units (I still have one which I never even bothered to mod the fan on, it spikes still but I have found that doesn't hurt anything).
I see by the follow up that you have succeeded in the work around... did you note any buzzing, crackling sounds? I think your running voltage non-modified right? The fan covers the noise so once you do the mod you can't hear the sound anymore.. mine still makes the sound (without the fan mod) but it's minimal and I'm accustomed to it now.
The real test isn't at 12v against a 20.x v supply but rather at HVC or at perhaps in your case.. Basically I think you will see the taper begin much sooner than you expect (which should be at your exact HVC or voltage setting on the unit)... though with the FAN mod, you should be able to set each supply for a continuous 16.95A to the 20.65v for 350w continuous.
-Mike
Sorry I didn't get to you with the previous post in time... What your seeing in the first post is the effect of a lower voltage rail dropping (the TL494CN expects a 48v rail) under the load of the FAN... this happens with all of them... The faster the fan, the lower LOW VOLTAGE rail sags... you can see this when the fan hasn't been modified as a temporary spike in current from about 5.2A to 6.2A when the fan kicks in on the S-350-48 units (I still have one which I never even bothered to mod the fan on, it spikes still but I have found that doesn't hurt anything).
I see by the follow up that you have succeeded in the work around... did you note any buzzing, crackling sounds? I think your running voltage non-modified right? The fan covers the noise so once you do the mod you can't hear the sound anymore.. mine still makes the sound (without the fan mod) but it's minimal and I'm accustomed to it now.
The real test isn't at 12v against a 20.x v supply but rather at HVC or at perhaps in your case.. Basically I think you will see the taper begin much sooner than you expect (which should be at your exact HVC or voltage setting on the unit)... though with the FAN mod, you should be able to set each supply for a continuous 16.95A to the 20.65v for 350w continuous.
-Mike
Mike, I do get a buzzing sound under load that's clearly distinct from the fan noise. It's not loud though, but it is there. With no load there is no buzzing.
I am running the supplies at the rated 24V and I've just tweaked the current up from 14.6A to 15.5A. When I've got them all wired up in series I will set them all to my final voltage and recheck the 15.5A. With the test load I was using the voltage dropped to 16V with 15.5A. Does this mean that when I start to use it as a charger I'll need to "recalibrate" the current limit to the actual load of the batteries?
The one I modified for current limit only had R37 populated, while the other three had both R37 and R33 populated. As the current will be limited by the first one this shouldn't effect things right?
I am running the supplies at the rated 24V and I've just tweaked the current up from 14.6A to 15.5A. When I've got them all wired up in series I will set them all to my final voltage and recheck the 15.5A. With the test load I was using the voltage dropped to 16V with 15.5A. Does this mean that when I start to use it as a charger I'll need to "recalibrate" the current limit to the actual load of the batteries?
The one I modified for current limit only had R37 populated, while the other three had both R37 and R33 populated. As the current will be limited by the first one this shouldn't effect things right?
mwkeefer
1 MW
Normal when modding the resistors for current, I think modifying the shunt alone prevents this (based on others posts)...DaveAK said:
No the load in question is only on the low voltage side of the TL494CN and only because of the fan's draw. I should actually say, I don't know... when I started and in all my testing I have used a real lipo pack (or many in parallel) for 24v unit loading - I've never tried any other load... should be okay, just double check it = )DaveAK said:
DaveAK said:
Technically it shouldn't but.. Meanwell claims the S- series don't do parallel so if your using them in parallel at all, it basically means that they don't do proper duty cycle sync and that would allow one unit to be pushed harder than another - this means that for a safe parallel charger (more than one meanwell in parallel) you would need to current limit each of the parallel units to acheive the proper and balanced (or close to it) output... since I think your going to series these up, the CC limit of the lowest unit should force limit them all... but if you have the time to check with a DVM or your whats up with it connected inbetween the series units (I should actually test this here, I have the same S-350-24s) we could confirm with observation that indeed the CC factor cascades and balances in series..
Sorry if the answer is more complicated than the question - it's been a long day.
-Mike
The answer was great mike. Just one question about putting the Turnigy watt meter between the units when in series. How would I do that? First off, it's rated to 60V and I'll be at 96V. Secondly if the current is measure across say the +ve lead then I could connect that between two units, but I'll need to power the thing, and I don't have a suitable battery to hook up. If I did, would it read the current between two units?
mwkeefer
1 MW
Argh - had to hit me hard this late at night... I hadn't thought about how to do it really, beyond just thinking of doing it but here goes...
Your voltage is at 96v but I have seen alot of guys with packs beyond an EagleTree meter in the past simply hook up into series or parallel to part of the pack...
In your situation you have I assume 4 power supplies pushing roughly 15A current each in series to produce your close to 100v output total?
Honestly - the only thing I can think of (and will test here now that I thought about it) is to connect a normal DVM between the units at their series points... a DVM usually claims 10A max but the cheaper ones are unfused (and the better could be upgraded with larger fuse)... I've used 8.00 DVms from Home Depot to test up to 20Amps. I wouldn't suggest doing it for long and honestly you may want to use some better connections (3.5mm gold plugs from Hobby king fit nice)... then you can measure the current from each and every unit (with 3 or 4 total DVMs) in the series as it charges your pack.
I would only test this at LVC... Nominal and finally at the beginning of the taper and only for a moment or two at most and expect the DVMs to smoke up... but ive done this in the past in other situations (car audio with multiple batteries in series) and it works.
Do I have your configuration correct? What will the final # of meanwells in series and their voltages be? I have a pack I could charge to 83v and the data could be extrapolated... also I've been meaning to fast charge these in parallel (5.4 amps, 74v nominal A123 packs) for some time to see how it works out.
If I have it right, I will test this in the next few days - I guess I could test a 20S pack of 2P (I can whip one up from spare cells)...
-Mike
Your voltage is at 96v but I have seen alot of guys with packs beyond an EagleTree meter in the past simply hook up into series or parallel to part of the pack...
In your situation you have I assume 4 power supplies pushing roughly 15A current each in series to produce your close to 100v output total?
Honestly - the only thing I can think of (and will test here now that I thought about it) is to connect a normal DVM between the units at their series points... a DVM usually claims 10A max but the cheaper ones are unfused (and the better could be upgraded with larger fuse)... I've used 8.00 DVms from Home Depot to test up to 20Amps. I wouldn't suggest doing it for long and honestly you may want to use some better connections (3.5mm gold plugs from Hobby king fit nice)... then you can measure the current from each and every unit (with 3 or 4 total DVMs) in the series as it charges your pack.
I would only test this at LVC... Nominal and finally at the beginning of the taper and only for a moment or two at most and expect the DVMs to smoke up... but ive done this in the past in other situations (car audio with multiple batteries in series) and it works.
Do I have your configuration correct? What will the final # of meanwells in series and their voltages be? I have a pack I could charge to 83v and the data could be extrapolated... also I've been meaning to fast charge these in parallel (5.4 amps, 74v nominal A123 packs) for some time to see how it works out.
If I have it right, I will test this in the next few days - I guess I could test a 20S pack of 2P (I can whip one up from spare cells)...
-Mike
I have four 24V in series and they will be set for a little under 96V when it's complete. Still waiting on my batteries, (maybe tomorrow), and BMS, (waiting on G/F BMS V4.0 to be available), so I'm stuck for a load. Would a 1500W A/C hair dryer work, or does it have to be DC?
I have a cheap and expendable DVM that I can put in series to check the current, so I'll try that later. Should be good for a few seconds if it doesn't blow immediately.
I have a cheap and expendable DVM that I can put in series to check the current, so I'll try that later. Should be good for a few seconds if it doesn't blow immediately.
sjoerdvdd
10 mW
Hi guys,
i've red the last few pages, but didn't really get very far..
my plan is to use a meanwell rsp-1000 PSU to charge my 72V40Ah pack.
would this be possible? and what modifications do i need to make. i've red that this is a CCL type so it should be possible. i'm worried about the 72V though.
if 72V wouldn't work. charging up 36V80Ah would i think.
shortend up. what do i need to do to safely charge a 72V40Ah pack managed by a fechter/goodrum BMS (V4)
regards
i've red the last few pages, but didn't really get very far..
my plan is to use a meanwell rsp-1000 PSU to charge my 72V40Ah pack.
would this be possible? and what modifications do i need to make. i've red that this is a CCL type so it should be possible. i'm worried about the 72V though.
if 72V wouldn't work. charging up 36V80Ah would i think.
shortend up. what do i need to do to safely charge a 72V40Ah pack managed by a fechter/goodrum BMS (V4)
regards
mwkeefer
1 MW
RSP-1000 is a nice choice but at 300.00+ it may be a bit too much, granted it has current sharing and does all the CC/CV stuff you could want but I haven't used or opened one and so I can't attest to how easy it would be to modify the device to run at your higher voltage... I haven't pushed a single 48v past 64v or so maximum.
The other thing is you say 72V but I assume that's nominal so about 83v would be the true requirement...
Stock range on the 48v unit is from 43v to 55v adjustable so as I see it you have 2 options:
1.) Run 2 of the 48v units... modified for slightly lower voltage of 41.5 (for 83v CC->CV transition) and set the current to it's maximum of 20A (but really at 41.5 you could get closer to 24A I would think) and then run in Series
or
2.) Take the risk of configuring one of these to run at 83v cutout (resistor mod, ZD1 type mod) but I wouldn't risk it without consulting Meanwell, drop them an email and explain you need a CC/CV of 20A rate in a single unit and need it to go up to 83v which BTW in 100w is only a 12.04 A charger CC/C
Personally since running a single unit brings you down to 12A at 1000w (assuming with higher voltage comes higher efficiency as per their own docs/specs) I would just build up a quad set of slightly modified SP-480-24s modified to produce 41.5v and you would end up with CC/CV of about a 23A charge rate at 100% output power from each of the 4 units... Similar you could do this with S-350's (48v modified for slightly lower 41.5v each) and a 4 pack could produce an 83v CCCV charger of 16.87 A which is still higher than a single modified unit of the 1000w variety and would cost much less (I think you could get 4 of the 48v supplies for 30.00 ea + 40.00 s&H for a total of 160.00 and more power than your single unit.
Really the best bet would be send them an email asking about an 83v supply with CCL. if you want a single unit charger of 20A rate.
-Mike
The other thing is you say 72V but I assume that's nominal so about 83v would be the true requirement...
Stock range on the 48v unit is from 43v to 55v adjustable so as I see it you have 2 options:
1.) Run 2 of the 48v units... modified for slightly lower voltage of 41.5 (for 83v CC->CV transition) and set the current to it's maximum of 20A (but really at 41.5 you could get closer to 24A I would think) and then run in Series
or
2.) Take the risk of configuring one of these to run at 83v cutout (resistor mod, ZD1 type mod) but I wouldn't risk it without consulting Meanwell, drop them an email and explain you need a CC/CV of 20A rate in a single unit and need it to go up to 83v which BTW in 100w is only a 12.04 A charger CC/C
Personally since running a single unit brings you down to 12A at 1000w (assuming with higher voltage comes higher efficiency as per their own docs/specs) I would just build up a quad set of slightly modified SP-480-24s modified to produce 41.5v and you would end up with CC/CV of about a 23A charge rate at 100% output power from each of the 4 units... Similar you could do this with S-350's (48v modified for slightly lower 41.5v each) and a 4 pack could produce an 83v CCCV charger of 16.87 A which is still higher than a single modified unit of the 1000w variety and would cost much less (I think you could get 4 of the 48v supplies for 30.00 ea + 40.00 s&H for a total of 160.00 and more power than your single unit.
Really the best bet would be send them an email asking about an 83v supply with CCL. if you want a single unit charger of 20A rate.
-Mike
Hi guys I wat to use a SP-320-48 to charge a LIPO battery pack 10S3P with the balancers plugs connected to form a unique battery of 30Ah.
Can I use the SP-320-48 adjusting the output voltage at 42V? Do I need to make the mod to reduce the current?
Thank you.
Can I use the SP-320-48 adjusting the output voltage at 42V? Do I need to make the mod to reduce the current?
Thank you.
Kingfish
100 MW
BB, the spec for the SP-320-48 says you can tune it between 41V and 56V, so I would say yes you should be able to tune it down to 42V. Best way to tell is to go hook it up and test your unit as you trim it down.
I don't have an answer on how to reduce the current; not my forte.
Curious and Concerned: If you take your 10S3P and make a 30Ah battery the nominal voltage will be 18.5V. I would think charging at 42V is beyond the envelope. Not sure I'd want to stand too close while this is going. Are you are planning on an early 4th of July celebration? :wink:
~KF
I don't have an answer on how to reduce the current; not my forte.
Curious and Concerned: If you take your 10S3P and make a 30Ah battery the nominal voltage will be 18.5V. I would think charging at 42V is beyond the envelope. Not sure I'd want to stand too close while this is going. Are you are planning on an early 4th of July celebration? :wink:
~KF
Kingfish
100 MW
I think we're lost in translation:
I run 10S3P now on my P0 eBike. It's 37V nominal, and at max-V I charge to 41.1 +/- 0.5V using the High Power 36V/4A charger that came with my kit (de-rated from 43.6V to 41.1). Each of my batteries are ZIPPY 5S LiPo @ 5000mA, which means at 10S3P = 15Ah, so to get 30Ah I'd have to be at 5S/6P = 18.5V nominal/20.55V max. Let's review your quest:
At 10S3P, yes you can charge at 42V. I guess the part that's missing is what is the Ah rating for each cell. Sounds like quite a pack: I'm all ears and keen to learn. Who's the manufacturer/supplier?
I run 10S3P now on my P0 eBike. It's 37V nominal, and at max-V I charge to 41.1 +/- 0.5V using the High Power 36V/4A charger that came with my kit (de-rated from 43.6V to 41.1). Each of my batteries are ZIPPY 5S LiPo @ 5000mA, which means at 10S3P = 15Ah, so to get 30Ah I'd have to be at 5S/6P = 18.5V nominal/20.55V max. Let's review your quest:
At 10S3P, yes you can charge at 42V. I guess the part that's missing is what is the Ah rating for each cell. Sounds like quite a pack: I'm all ears and keen to learn. Who's the manufacturer/supplier?
mwkeefer
1 MW
Kingfish,
The balancers on your pack - what type are they? What is the discharge capacity?
With regards to your Meanwell - it should be fine and able to adjust out of box a bit more than they say from 39.1v to 59v on my stock units.
One other thing... I would suggest setting the Meanwell to 4.155v per cell or 41.55v - this will be the maximum voltage and when it's reached the meanwell would taper out to nothing...
So your current (for 100%) should be: 350 / 41.55 = 8.42A constant current... the efficiency is about 85% or so at that voltage so really, 7.16 A with losses accounted for...
The real issue is the SP series (the 350 not the 480) doesn't do CC limiting it does 3 second pulse reset, so unless you have a current limiter to fit on the front end to perform the CV phase taper you won't have the results you want...
I also don't suggest using a drain balance with a pulse supply, it will not balance properly.
So far as limiting the current, the SP series you would need has a shunt... does yours?
All,
I am working on the S-145 and S-150 models right now, looking to build a 15-18S charger with 2 current levels 100% and < 1A from 63v to 75v output...
These should be great for balance charging larger packs with smaller resistive bleed balancers...
Gary,
Ive noticed the accuracy of Battery Medic display is lacking (it's actually fairly off) and sometimes they even seem to try and balance the wrong cells (using 2x cell logs + a balancer connected for accurate monitoring and logging of balance process) - all that said, even 3 of them across 5s packs of 10AH will get my pack into a balance of +- 7mv so I assume that even with their inaccuracy overall (or atleast incorrect display voltage) that they do a nice job of balancing as long as the current is properly limited or in the CV phase....
-Mike
The balancers on your pack - what type are they? What is the discharge capacity?
With regards to your Meanwell - it should be fine and able to adjust out of box a bit more than they say from 39.1v to 59v on my stock units.
One other thing... I would suggest setting the Meanwell to 4.155v per cell or 41.55v - this will be the maximum voltage and when it's reached the meanwell would taper out to nothing...
So your current (for 100%) should be: 350 / 41.55 = 8.42A constant current... the efficiency is about 85% or so at that voltage so really, 7.16 A with losses accounted for...
The real issue is the SP series (the 350 not the 480) doesn't do CC limiting it does 3 second pulse reset, so unless you have a current limiter to fit on the front end to perform the CV phase taper you won't have the results you want...
I also don't suggest using a drain balance with a pulse supply, it will not balance properly.
So far as limiting the current, the SP series you would need has a shunt... does yours?
All,
I am working on the S-145 and S-150 models right now, looking to build a 15-18S charger with 2 current levels 100% and < 1A from 63v to 75v output...
These should be great for balance charging larger packs with smaller resistive bleed balancers...
Gary,
Ive noticed the accuracy of Battery Medic display is lacking (it's actually fairly off) and sometimes they even seem to try and balance the wrong cells (using 2x cell logs + a balancer connected for accurate monitoring and logging of balance process) - all that said, even 3 of them across 5s packs of 10AH will get my pack into a balance of +- 7mv so I assume that even with their inaccuracy overall (or atleast incorrect display voltage) that they do a nice job of balancing as long as the current is properly limited or in the CV phase....
-Mike
Kingfish
100 MW
Herro Mr. Mike
My Batts look exactly like these: ZIPPY Flightmax 5000mAh 5S1P 15C
EDIT <sorry - forgot this is what you really asked for...>
Balancers: Astro Blinky. I typically balance once a week; takes all afternoon with two. I am waiting for the good gentlemen to finish the BMS Version 4 <ahem... tap foot>
I use 10S3P for terrorizing the townsfolk, er... normal running around, and 10S9P for long rides. Got to say: I have been extremely pleased with them as batteries go; faithful lil' beasties
That being said, I haven't done a diddly-do with my Meanwell SP-320-48s. Been waiting on your findings. Hyena gave me some advice on fetching a S-150-12, S-240-12 or even an S-350-24 to limit current charging for my 15SXXP wicked long-arse distance battery array though I haven't made the plunge yet. Got bigger fish to fry (read hubs and controllers).
And now it's gone back to raining here in the PNW so I am twiddlin' inside.
Playin' nice... for the moment
~KF
My Batts look exactly like these: ZIPPY Flightmax 5000mAh 5S1P 15C
EDIT <sorry - forgot this is what you really asked for...>
Balancers: Astro Blinky. I typically balance once a week; takes all afternoon with two. I am waiting for the good gentlemen to finish the BMS Version 4 <ahem... tap foot>
I use 10S3P for terrorizing the townsfolk, er... normal running around, and 10S9P for long rides. Got to say: I have been extremely pleased with them as batteries go; faithful lil' beasties
That being said, I haven't done a diddly-do with my Meanwell SP-320-48s. Been waiting on your findings. Hyena gave me some advice on fetching a S-150-12, S-240-12 or even an S-350-24 to limit current charging for my 15SXXP wicked long-arse distance battery array though I haven't made the plunge yet. Got bigger fish to fry (read hubs and controllers).
And now it's gone back to raining here in the PNW so I am twiddlin' inside.
Playin' nice... for the moment
~KF
Kingfish said:I guess the part that's missing is what is the Ah rating for each cell. Sounds like quite a pack: I'm all ears and keen to learn. Who's the manufacturer/supplier?
My cells are 10Ah each one and the manufacturer is Trasmetic and are used on big RC airplanes.
http://cgi.ebay.it/LiPo-Akku-10000mAh-14-8V-25C-Transmetic-o-S-NEU_W0QQitemZ170405302098QQcmdZViewItemQQptZRC_Modellbau?hash=item27acf28f52
mwkeefer said:
I bought the SP-320-48 under suggestion of Gary because I would use it with his BMS V4 when available. Now I would try to charge my 10S3P LIPO battery supervising while charging with a LIPO volt meter that reads every cell voltage like Cellog 8S does.
Can you explain what should happen when the voltage of the battery is close to the voltage that I set on the SP-320-48? I have a stock unit with no mods.
