AussieJester
1 TW
mwkeefer have you tried any of the sp320 meanwels yet? I haven't moddded mine at all and it happily charges my pack
without even getting slightly warm.
KiM
without even getting slightly warm.
KiM
Switch mode supply mod
- Thread starter Jeremy Harris
- Start date
hydro-one
10 kW
Upon further inspection the transistor is a 2sc3320 by fuji. http://pdf1.alldatasheet.com/datasheet-pdf/view/88110/FUJI/2SC3320.html
both at end of case are burnt.
A question. could i use the zenar out of this 27v meanwell to add to my other 27v meanwell and get a 60v unit? I heard the zeners under a transformer. I would really like a 60v unit!!
thanks in advance
mikeg
both at end of case are burnt.
A question. could i use the zenar out of this 27v meanwell to add to my other 27v meanwell and get a 60v unit? I heard the zeners under a transformer. I would really like a 60v unit!!
thanks in advance
mikeg
baerfoot
1 mW
Hyena, thanks for your reply. I have a kill-a-watt. So I thought I would hook up the PSUs without any mods and see how much power they would draw. When I was hooking up the Negative lead to the pack there was a loud pop and spark. I took the lead away because I thought there may be a problem. I checked the PSUs and they still seemed fine, but I would like more advise before hooking up again.Hyena said:
Is it normal to have that type of arcing when hooking up the leads? Is there a way to prevent it?
I find that if I plug my power supply into the wall first it won't arc but I unplug it from the wall before un hooking it from the batteries! The last time I used mine I uhnhooked it from the batties in the middle of charging and fried 2 fets! It's all a matter of having the voltages as close as possible before conecting wires. I also. Use the power supply to charge up the controler before hooking up the batteries and it prevents arcing!baerfoot said:
hydro-one
10 kW
Just noticed that my new genny meanwell 27v will charge at 9amps unmodified . then i noticed that switch is set to 240v input!!! didnt get warm at all. changed to 120v now its at 15amps unmodified. guess i will snip the one shunt for 12.5 amps , mabye 15 would be ok, 12.5 barely gets the fan going,,,,,,,Im guessing six amps with the input voltage set 240 and one cut shunt!! new mod??
mike
mike
mwkeefer
1 MW
Hello all,
I've been playing with combinations of S-350-48s and SP-320-24s and have come to some observations...
1.) I attempted to use the SP-320-24s alone to charge the following:
5S2P (10AH)
5S4P (20AH)
5S6P (20AH)
6S2P (10AH)
The result with the S320 was consistent HICCUP mode...
2.) I then tested charging 15S and 18S2P packs using 3 x S-320-24 in series and again entered hiccup mode on all tests...
3.) Finally in desperation to atleast use the SP320 I hit the 15 and 18S packs with a S-350-48 in series with the SP-320-24... I had the 350-48 at about 40v (right at it's lowest) and the 24v set to make the difference... I was seeing 10-11A charge rate (about what I would expect) and it was flawless - though the 350 is not modified, I must atleast do the fan mod and replace the heat sink paste.
4.) Something dawned upon me so I quickly added 18G taps into 36v LiFePo4 pack (10AH) similar to the ones sold by e-bikekit.com... I tapped the first 6S cells from ground for a total 3.6v charge of 21.6v.
I then configured the 24v SP-320 for 22.2 (3.7v per cell) and connected a celllog8 for monitoring, charge was run via Watts Up and was pushing 14A at 22.2v - to my surprise, IT WORKED and didn't enter hiccup mode?
My best guess here is that due to the IR (Internal Resistance) of the LiFePo4 10AH cells, the supply was outputting less power than with lipo pack < 10 mohm but I am as always no engineer!
I do believe your success with the SP-320-48 is in part due to 2 factors:
1.) You are operating within the stock voltage range for charging 12S LiPo - the 48 must be modified to 60v+ for LiPo 15S or LiFePo4 16S.
2.) Your pack (or charge leads) may have a rather high impedance (internal resistance) which would in theory cause the supply to limit it's own power output.
I will continue to play/research, I am also looking at Gary's current limiting portion of the Charge Controller... yes there are some issues with too much power through the controller but... I think it could be perfected and plan to devote some time and effort to just that goal - I will not be sharing my mods with anyone but Gary and Fechter as they worked hard to bring us all these products and they have always shown me the curtosey of keeping privlidged info to themselves and not trying to step on my toes, I owe them that same respect.
-Mike
I've been playing with combinations of S-350-48s and SP-320-24s and have come to some observations...
1.) I attempted to use the SP-320-24s alone to charge the following:
5S2P (10AH)
5S4P (20AH)
5S6P (20AH)
6S2P (10AH)
The result with the S320 was consistent HICCUP mode...
2.) I then tested charging 15S and 18S2P packs using 3 x S-320-24 in series and again entered hiccup mode on all tests...
3.) Finally in desperation to atleast use the SP320 I hit the 15 and 18S packs with a S-350-48 in series with the SP-320-24... I had the 350-48 at about 40v (right at it's lowest) and the 24v set to make the difference... I was seeing 10-11A charge rate (about what I would expect) and it was flawless - though the 350 is not modified, I must atleast do the fan mod and replace the heat sink paste.
4.) Something dawned upon me so I quickly added 18G taps into 36v LiFePo4 pack (10AH) similar to the ones sold by e-bikekit.com... I tapped the first 6S cells from ground for a total 3.6v charge of 21.6v.
I then configured the 24v SP-320 for 22.2 (3.7v per cell) and connected a celllog8 for monitoring, charge was run via Watts Up and was pushing 14A at 22.2v - to my surprise, IT WORKED and didn't enter hiccup mode?
My best guess here is that due to the IR (Internal Resistance) of the LiFePo4 10AH cells, the supply was outputting less power than with lipo pack < 10 mohm but I am as always no engineer!
I do believe your success with the SP-320-48 is in part due to 2 factors:
1.) You are operating within the stock voltage range for charging 12S LiPo - the 48 must be modified to 60v+ for LiPo 15S or LiFePo4 16S.
2.) Your pack (or charge leads) may have a rather high impedance (internal resistance) which would in theory cause the supply to limit it's own power output.
I will continue to play/research, I am also looking at Gary's current limiting portion of the Charge Controller... yes there are some issues with too much power through the controller but... I think it could be perfected and plan to devote some time and effort to just that goal - I will not be sharing my mods with anyone but Gary and Fechter as they worked hard to bring us all these products and they have always shown me the curtosey of keeping privlidged info to themselves and not trying to step on my toes, I owe them that same respect.
-Mike
baerfoot
1 mW
GGoodrum said:
After a couple of hours of reading posts as of yet I have not found the information on how to lower the current on my s-350-48 supplies. Can anyone point me in the right direction? Strange thing about my 2 PSUs. One has the fan on all the time the other one started to get pretty warm and the fan never cycled on. Any simple mods to just keep the fan on?
mwkeefer
1 MW
The info is in this thread and the CC/CV and meanwell and... = )
Mostly kidding... Seriously though there is just a resistor R33 maybe ( I cant remember its 3am) you want to either increase or lower depending on your intent...
Before doing that, you will first want to perform the thyristor mod - replace the green thing near the coil with a resistor, again the value escapes me off top of head but that's what search is for... you just need the buzzwords...
In either case, the mods take about 15-30minutes for someone with little experience or about 10 mintues for a pro like Gary... me, I'm stupid I wait till 3AM to get started so it takes me 2 days!
Hope that helps a bit... I suppose a howto should be written up if doesn't already exist.
-Mike
Mostly kidding... Seriously though there is just a resistor R33 maybe ( I cant remember its 3am) you want to either increase or lower depending on your intent...
Before doing that, you will first want to perform the thyristor mod - replace the green thing near the coil with a resistor, again the value escapes me off top of head but that's what search is for... you just need the buzzwords...
In either case, the mods take about 15-30minutes for someone with little experience or about 10 mintues for a pro like Gary... me, I'm stupid I wait till 3AM to get started so it takes me 2 days!
Hope that helps a bit... I suppose a howto should be written up if doesn't already exist.
-Mike
Hyena
10 GW
500k resistor for the thermistor, if you don't want to mod r33 you can snip through one of the shunts and replace it with a few fine stands of wire. Connect a battery and look at the current then cut through the strands one by one if necessary to drop the current as required
Thud
1 MW
Hyena,
Thanks. I just picked up a couple S-350-48 chineese clones. I seem to only have 1 shunt.
View attachment 1
Thanks. I just picked up a couple S-350-48 chineese clones. I seem to only have 1 shunt.
View attachment 1
Hyena
10 GW
you can still snip through the 1 shunt and do as I said above, or do the r33 mod - its up to you. I cant remember the verdict with r33/37 mods but I think it was still causing buzzing in alot of cases
baerfoot
1 mW
Thanks for all your replies and patience with a newbie like me. I opened up my cases and it does seem like the mods are pretty straightforward. I am very new to electronics.
One of the s-350s I have looked the same as Thud's. The other one looks quite different. It does not have a resistor at r33. it does have something adjustable at svr2
Any idea what this adjusts? Maybe I can adjust current limit here 
Another difference is that it has no component at Q5
Maybe that is why the fan runs all the time.
Just to be sure I understand correctly: Will changing the thermistor to a resistor make the fan run all the time?
One of the s-350s I have looked the same as Thud's. The other one looks quite different. It does not have a resistor at r33. it does have something adjustable at svr2
Another difference is that it has no component at Q5
Just to be sure I understand correctly: Will changing the thermistor to a resistor make the fan run all the time?
So one of my new 7.3 48v meenwells came in but the other is in limbo so.... I managed to fix one of my old ones that poped a cap but my batteries are only charging at 3 amps and 4.5 if I crank the unloaded voltage to 104v total. Is there anthing else I should look at that might have fried when one of the 63 volt caps poped?
mwkeefer
1 MW
svr2 is for current limiting... i've been waiting for someone to bust out with pics of that model with the SVR2 populated, none of the models we have gotten so far were populated...
I would be very intersted in decent macro shots of the component side of that S-350 and if possible a scan of the trace side.
svr2 may be fine adjustment but locating fine adjustment internally doesn't make alot of sense and in my S350 without that portion populated there is a transistor and SVR2 and they run back to one side of the error amp (which manages duty cycle and thus current limiting)
-Mike
I would be very intersted in decent macro shots of the component side of that S-350 and if possible a scan of the trace side.
svr2 may be fine adjustment but locating fine adjustment internally doesn't make alot of sense and in my S350 without that portion populated there is a transistor and SVR2 and they run back to one side of the error amp (which manages duty cycle and thus current limiting)
-Mike
Hyena
10 GW
Yes very interesting! So R33 is missing and instead SVR2 is fitted.
As Mike said if you can post a high res pic of the whole board that'd be good.
What brand is it and where did you get it from ? Post an ebay link if its from there
As Mike said if you can post a high res pic of the whole board that'd be good.
What brand is it and where did you get it from ? Post an ebay link if its from there
mwkeefer
1 MW
Arlo,
The switch needs to be set for your AC mains voltage - 110 in the states (unless 3 prong 220) and 220 abroad... if it was set to 220 and you used it on 110 it will produce about 1/2 the output power and high end voltage range - also if you hook to 220 with 110 set you will blow the unit out without a doubt.
Now 2 x S-350-48 tuned to a no load (CC/CV crossover point) voltage of 104 and an assumed even split of 52v each supply...
700 / 104 = 6.73 A - this is the rate you should see the charger running at without any further mods (assuming they sustain).
If one of the units was set for 220 it would produce 1/2 (roughly) it's potential or 175w so... 525w / 104v = 5.048 A
If both units were configured for 220 it would be 350w total and at 104v that is = to 3.365 A - sounds like what your seeing though it could be damage to the flyback diode...
Check that both units are set for your AC current.
One further thing - what are you charging with these? Chemistry I mean... LiFePo4 would have higher Internal Resistance and thus may draw less power... have you done some current limiting and have you modded the units to replace the thermistor with a resistor so the fan stays on?
-Mike
The switch needs to be set for your AC mains voltage - 110 in the states (unless 3 prong 220) and 220 abroad... if it was set to 220 and you used it on 110 it will produce about 1/2 the output power and high end voltage range - also if you hook to 220 with 110 set you will blow the unit out without a doubt.
Now 2 x S-350-48 tuned to a no load (CC/CV crossover point) voltage of 104 and an assumed even split of 52v each supply...
700 / 104 = 6.73 A - this is the rate you should see the charger running at without any further mods (assuming they sustain).
If one of the units was set for 220 it would produce 1/2 (roughly) it's potential or 175w so... 525w / 104v = 5.048 A
If both units were configured for 220 it would be 350w total and at 104v that is = to 3.365 A - sounds like what your seeing though it could be damage to the flyback diode...
Check that both units are set for your AC current.
One further thing - what are you charging with these? Chemistry I mean... LiFePo4 would have higher Internal Resistance and thus may draw less power... have you done some current limiting and have you modded the units to replace the thermistor with a resistor so the fan stays on?
-Mike
baerfoot
1 mW
I took several images of the board. Here is a link to one with a higher resolution, and some other images of the board:http://picasaweb.google.com/surfinbaer/S350#5471708590960051922
I bought 2 units on ebayhttp://cgi.ebay.com/ws/eBayISAPI.dll?ViewItemVersion&item=270461332964&view=all&tid=526058272017 when I received them one was DOA. I returned it for a replacment. The board I am showing in the images is from the the first shipment. The second unit does not have sr3 populated. So if you bought from this seller I don't know which type you would get.
I bought 2 units on ebayhttp://cgi.ebay.com/ws/eBayISAPI.dll?ViewItemVersion&item=270461332964&view=all&tid=526058272017 when I received them one was DOA. I returned it for a replacment. The board I am showing in the images is from the the first shipment. The second unit does not have sr3 populated. So if you bought from this seller I don't know which type you would get.
mwkeefer
1 MW
If we can isolate that your PCB is identical to most of ours the defacto std design - ours have locations for SVR2 but it's not populated... if we can lift the components used in yours and the surrounding circuit we can adapt it to a module - easier to construct and would allow dial in of the current on these models as designed (well in yours atleast)...
This is actually how Fecther figured out the mods in the first place we posted pics of the PCB around the TL494CN chip and then I scanned some and finally I sent him a unit for testing and modification...
Now I don't know - 22 pages later, well...
I should give credit to Tiberius and Jermey Harris too... they did start this whole bit off in another similar thread.
-Mike
This is actually how Fecther figured out the mods in the first place we posted pics of the PCB around the TL494CN chip and then I scanned some and finally I sent him a unit for testing and modification...
Now I don't know - 22 pages later, well...
I should give credit to Tiberius and Jermey Harris too... they did start this whole bit off in another similar thread.
-Mike
baerfoot
1 mW
Let me know how I can help, although I don't have much experience. For instance I would not be sure how to reassemble the heat sinks correctlyif I removed the PCB.mwkeefer said:
Hey man thanks but what I was trying to say is my last S-350-48 does not have a 110/220 switch where it should be! its missing! I do have a S-350-48 that I blew up real good so I will look at pulling the switch out of it.mwkeefer said:
mwkeefer
1 MW
It may be automatic...
May not need the switch
-Mike
May not need the switch
-Mike
baerfoot
1 mW
I hooked up my 2 s-350 48s up to my pack and experimented with the pot at svr2.
I am using a kill-a-watt to measure incoming current. I do not know what the output current is.
When I started, the current was 990 w
I turned the pot all the way CCW and there was no change.
I turned the pot all the way CW and the current was 575 w there was also a hissing noise.
I then played around with the adustment and found that the area of adjustment is very fine, so it is tricky to fine tune.
I left it at 640+- w. I figure that is probably a safe level to not burn out the PSUs. I measured each PSU seperatly also at that setting.
I will call A the unit with the pot.
A: 291 w with an output of 37.2v
B: 324 w with an output of 43.8 v
The pack was at 80 v before starting the experiment
The fan started up in unit B as it got warm, so I may not need to mod the thermistor.
A friend recomended hooking up a seperate wire with a resistor as a lead first to avoid a big spark. I did this, and there was only a small spark. I then hooked up the normal lead.
I am using a kill-a-watt to measure incoming current. I do not know what the output current is.
When I started, the current was 990 w
I turned the pot all the way CCW and there was no change.
I turned the pot all the way CW and the current was 575 w there was also a hissing noise.
I then played around with the adustment and found that the area of adjustment is very fine, so it is tricky to fine tune.
I left it at 640+- w. I figure that is probably a safe level to not burn out the PSUs. I measured each PSU seperatly also at that setting.
I will call A the unit with the pot.
A: 291 w with an output of 37.2v
B: 324 w with an output of 43.8 v
The pack was at 80 v before starting the experiment
The fan started up in unit B as it got warm, so I may not need to mod the thermistor.
A friend recomended hooking up a seperate wire with a resistor as a lead first to avoid a big spark. I did this, and there was only a small spark. I then hooked up the normal lead.
Hyena
10 GW
Baerfoot you really need to measure the current between the psu output and the battery.
Your wattage tests confirm that svr2 does actually appear to change the current though, as expected.
Also that ebay link you posted has expired because its too old, can you look through your feedback and post the sellers name ?
From the pics it does appear to be a genuine meanwell, or atleast it claims to be unlike alot of the clones that have some other brand sticker on them.
Your wattage tests confirm that svr2 does actually appear to change the current though, as expected.
Also that ebay link you posted has expired because its too old, can you look through your feedback and post the sellers name ?
From the pics it does appear to be a genuine meanwell, or atleast it claims to be unlike alot of the clones that have some other brand sticker on them.
