Switch mode supply mod

[GGoodrum]

I agree with you, but up to a point. Lowering the CC limit by changing a resistor is a better method than one I came up with first - cutting a shunt. It has to be worth someone tracing out the various circuits round that current comparison area, because there do seem to be quite a few variations. I've even had nominally identical PSUs with different R values in places.

I agree, but with these new SP-320s, the layouts are identical, and so are the value of the parts that I can read, at least.

Where I disagree slightly is in the need for hacking the voltage around. True, turning a 24 V PSU into a 48 V one isn't needed; its better to get a 48 V one to start with. But the native tuning range on each model is only +/-10%; we need to increase that or move the centre around to cover all the voltages needed.

I agree again, that a greater voltage range will be good, but first and foremost, in my mind at least, is coming up with as simplest a mod possible to get the basic units usable as reliable chargers.

 

[AussieJester]

To that end, Kim, you still haven't supplied a picture that has enough resolution, and is from the right angle,

I left them at 1080 rez and took it from same angle Methy did :? have you tried saving them to your PC and zooming in? ...i shall pull the thing apart again, now take more pics for you...

OK..10 pics of front and back of the board left in the rez they come off the camera...21mb worth

http://members.ii.net/~aussiejester/index/pictures/Custom%20RC%20Cruiser/Hi-Rez%20Meanwel%20SP320-48v%20PSU/

I can't get a shot of the resistor in the R33 spot i cant bend the component hiding it i cant even see it myself no matter how i look at it and im NOT de-soldering anything sorry guyz. If there's any others you can't see and need to know ask and ill see if i can read them for you on the board...

Best i can do hope it helps

KiM

p.s Mike...the shots of the back of the board are there for you also mate...

 

[wookey]

I had 2 of the 48V models arrive last week. I've just got as far as testing the no-load voltages, taking them to bits and trying them in series (all good). I'll take some pics forthwith. I might even try and do a bit of schematic tracing.

<fx: faffs about for q>

OK. I've taken a photo of top and bottom, and merged two bottom pics to remove the flash-overexposed bit.
They are uploaded here:
http://wookware.org/pics/meanwell

I've also drawn a schematic of the control area round the controller chip and put it there too. It's a very scrawly bit of paper so far laid out like the PCB rather than anything comprehensible, but it may help a bit. I can tidy it up and measure a load of values if that'll help.

I realise this is a poor substiture for having one in your hands, but I can't help there (being in the UK).

I note that my ZD1 (under the big transformer) is actually two diodes with the tails wrapped together, but not soldered! Tres crufty. Think I'll improve that a bit.

 

[Velocipede]

Hi, just read this entire post with interest and this looks like an interesting way to go.

Are we aware of a model that have SVR2 already populated so we can adjust amps without any mod?
My local shop lists: SE-600, SE-1000, SP-500, SP-320, S-320.

If I buy a lower power rated unit like S-240, the amps will be naturally limited lower, is that right? Or I will blow it or throw it in protection mode by plugging my battery to it?

If the details are clear about a particular model, I may try the mod.

Something like this is awesome with the display and all, but at 3x the cost...
http://www.gwinstek.com/en/product/productdetail.aspx?pid=38&mid=76&id=178

 

[Hyena]

My new meanwell arrived yesterday, sure electronics eventually agreed to replace it under warranty. (this was the completely unmodded one that blew up 2 minutes after I took it out of the box)

I tested it briefly with a lipo pack connected and it was putting out 10a (~500w) straight out of the box.
I opened it up and tried the resistor mod but instantly got the buzzing sound. After blowing up 2 SMPSs I wasn't willing to risk a 3rd so I decided to mod the shunts instead. Mine is the model that has 3 large shunts and by cutting one it drops the current to 6.9a - pretty much exactly their rated output.
So, if anyone is looking for a quick and easy fix to the over-current problem, snip through one of the shunts.
I decided to bump the current back up just a little to 8a, so around 400w (one of my other 350w units has been used daily for 6 months at 450w without any problems) I soldered 2 strands of fine wire across the gap - it's a good way to tune the exact current, get a piece of 20ga wire or similar, untwist the strands and fan them out with both ends soldered across the shunt you've just cut. Fire up the power supply with a max load connected and snip through the strands until you get the desired current.
Yeah it's a bit ghetto, but that's how I roll
Replacing R33 with an appropriate pot would probably be the better way to go but it still results in the buzzing and it requires messing around taking the board out which involves removing the glued in fet mounts and reapplying thermal paste afterwards (if you bother)

Also, re: the earlier topic of making the fan stay on, I measured the thermistor and it was reading about 1500 ohms at ambient temperature (about 26 celcius at the time) and when heated with a heat gun turned on the fan when the the resistance dropped down to about 700 ohms. I replaced it with a 500 ohm resistor and now the fan starts up immediately when you power up. If you short it it shuts down, so there needs to be some resistance there. I actually tried putting another resistor in parallel with the thermistor, so it'd just kick in earlier but bizzarely when I put a 500 ohm resistor in parallel with the 1500ohm themistor the resistance actually increased to 2200 ohms. Curious... I'm far from an electronics genius but I know that putting 2 resistors in parallel lowers the resistance. I guess it must be a property of thermistors that their resistance increases at a fixed temperature if you put a resistor across them ??

 

[dozentrio]

I liked your approach to having the fan on constantly. I was able to do this by putting a 1k resistor in parallel with the thermistor though. My thermistor is labelled 502. I am sure if you're curious you can discover what that means. I was not curious enough, being happy that the 1k resistor did the job.

I have some questions about the current limit. First of all, where are these shunts and what do they look like?

Also, what is a good way of measuring the max current output of my supply? The multi-meters i have access to are limited to 10 amps. Which is about what the output will be at 130%. So, if I am willing to risk the fuse on my multi-meter, do I simply connect the two probes of the multi meter across the powersupply + and - and then plug it in? Something tells me this is not the best or proper way of doing it. Something about inrush current... ? I also have a watt-meter I could use, which has a higher current rating.

 

[Hyena]

I liked your approach to having the fan on constantly. I was able to do this by putting a 1k resistor in parallel with the thermistor though.

Yeah that should work, as it is doing for you, but for some bizzare reason my resistance actually increased...

I have some questions about the current limit. First of all, where are these shunts and what do they look like?
Also, what is a good way of measuring the max current output of my supply?

These are the shunts - there's 3 of them in my unit (as for most 48v 350w models I believe) and cutting one will result in it safely operating at its rated power output of 7 amps


I use a TURNIGY WATT METER FROM HOBBYCITY
They're $24 and measure voltage, current, wattage, ah etc.
I'm in the process of building it into the charger permanantly so I can accurately and easily adjust the voltage to charge a range of packs without connecting up a multimeter. Seeing the current is a bonus as it lets you know once it starts tapering off that the charge is nearly done.
You could also use a cheap analog panel meter, that's what I have one one of my other power supplies. A 10a unit should be sufficient, any more than that and the meanwell will have a short life.

 

[MrKang]

I have a 5A lifePo4 charger for 48V.
If i measure the output it's 60.1V which is normal. But i want to modify the chargers voltage output to charge my 12S lipo.
I allready found the pot.meter to adjust the output voltage. what is the safest voltage i have to adjust it to?
is it 12 x 4.15V ? or 4.20 volts? so that leaves me a total output to 49.8 or 51V

This charger i could use at work.

 

[dozentrio]

Well, I connected my watt-meter and did some messing around. There is only one shunt on my power supply, and I didn't want to cut it because it would have been difficult to replace. The resistor (R37 on my 48V 350W clone supply) says it is a 390 ohm, and I modified its value to 200 ohms. Now the power supply outputs around 5 amps, which is just what I would like for charging my pack. The supply supply buzzes/crackles under load, though, and I wonder if it is safe to allow it to do that for hours and hours on end.

 

[wookey]

SO. My R33 measures 740R in circuit, which matches the colour code. The layout is such that you can fit R37 and SVR2 instead of R33 in order to have an adjustable resitance here. Probably the neatest mod. I'll buy bits for say 360R plus 500R variable and try it out.

Personally I'm interested in mwkeefer's flavour-conversion scheme as well as the simple current-limit mod.

No-one seems to have answered his questions about labels so here's what my board says (same layout as Ggoodrum's, I think mwkeefer actually wanted to know details for the other layout aussiejester has pics of, but nevermind, might as well write it down):
U1: 98AZTLM TL494CN and a TI logo, so that makes it a TL494CN SMPS chip: http://www.alldatasheet.com/datasheet-pdf/pdf/28822/TI/TL494CN.html
U2 is HA17358 (hitachi dual op-amp)
Heatsunk devices: Q2 2 x 2SC3320 (fat transistor) at the end, and D13 is D92M-03 (low loss high speed rectifier) at the side.

RG1 is 7812CT (even I know what that is: 12V regulator
BD1 is D15XB60 (600V 15A bridge rec)

(I've just discovered that LED strip-lights are really good for reading part numbers with)

Now I've got that SMPS datasheet it ought to be relatively siple to make sense of the circuit. I'll do some more documenting and schematic-decoding over next few days. Then hopefully someone who actually understands this stuff can say something sensible about it all.

 

[mwkeefer]

Wookey and all,

Thanks for the info and updates, as you may have heard I've been a bit busy lately with family medical issues.

That said, I have done a few more mods and tests and it seems as if nearly any change can make these supplies buzz or crackle (sounds like bacon doesn't it)... On my S-145-24 and 48 models, simply changing the limiting Zener will cause buzzing too (without any other mods) - I made another discovery, one I don't yet understand but someone here may so:

If I take the stock 58v zener from a S-350-48 which runs silent in stock mode and within the adjustable voltage range (39.1 - 58.4 on this particular unit out of box) and replace it with 2 zeners of 29v (not actually the values I used, tried a few combinations but this will illustrate the point) in series I get instant buzzing and / or crackling within the same stock voltage range. Just to confirm no damage, I removed the 2 in series and replace with the original diode and buzzing is gone (or inaudible).

If anyone has ideas on why this mod alone would introduce noise, I'm all ears = )

With regard to limiting these units... I've been running a S-350-48 with a tiny 5v (actually used an old iPhone USB cube) AC/DC converter embedded to power a 5v fan which runs all the time and prevents the internal fan from ever coming on. This unit is setup to deliver a maximum of 400 watts in the range between 62.20 and 62.60 (fine tune with smaller 100 ohm pot in series with SVR1) which works out to a consistent 6.3-6.4 A across the whole CC phase of the charge cycle. I have used this 15S charger on 5,10 and 20AH packs and it works flawlessly on all these. Though I haven't tracked each charge (I use it in the trunk for opportunistic charging) cycle, I have had this one working for about 2 months and I do about 3-5 charge cycles to a 10AH pack (to about 50% SOC before cutout and recharge), this supply has seen not only daily use for 2 months (as hyena has posted earlier I think his lasted 6 months so far) but ABUSE of road travel and commute via Train on the folding bike.

Still on topic but just some observations:

1.) It seems that everyone is saying modifying the shunts limits the current without causing buzzing or crackling noises? How could this be any different than changing the impedance of the error amplifier (TL494CN) to acheive the same limit of current? In other words how could this method be acheiving limiting without introduction of noise.

The only theory I have (it's a theory I will test when I have the time) is that modifying the shunt directly keeps the mod on the output side of things while the R37 (or akin) resistor mod is actually within the isolated bus surrounding the TL494CN which could be how the noise is being injected and then reproduced audibly?

2.) Swapping the single ZD1 (or equiv) for a series pair of even the same voltage (combined) also introduces noise, even in an otherwise unmodified unit... replacing the original ZD1 back into the PCB eliminates the noise issue.

I suppose all Zener Diodes have some forward voltage drop (.7v right?) so 2 in series would cause a 1.4v drop (yes?)... Could it be that simple, the circuit is calibrated for a .7v drop in the forward bias and by adding the second and causing a 1.4v forward drop it causes the buzzing and crackling sounds? I will test this tonight with a zener diode from a 48v unit that is fried and transplant it to a 24v unit without modification - in theory the high end of my range should increase slightly (to within the range of the adjustment circuit but without the zener limitation) and should cause ZERO abnormal noise. I have a few "must do" items tonight but since I have managed to postpone my move (home and office) for another month, I can slide some things around and get some tests done tonight.

Don't know if anyone cares but... I have been working on modifying the S-150/S-145-48 units to function as a 15-20S charger (62-84v) with a charge rate of 2.4 A to 1.78 A with success... I have noise issues in these too (as mentioned earlier in this post) but they are smaller, lighter and lower powered but perfect for slower charging of large packs (approx 4hr for 10AH 15S) and may even extend the lifespan of the pack due to the reduced charge rate of 1/4C (also appropriate for LiPo). These units have only a single shunt and as of yet I have not modified it to regulate the current and I see about 130% of rated output (consistent with all the other meanwells I've tested). The final version will wait until I see if I can get this buzzing sound out... I'm not worried about it persay, if the things (any of the models) last 6 months with 2-3 charges a day I will consider that my monies worth (*I've toasted several iChargers to date, 3 and in far less time) but in the end, it would be nice as Gary stated earlier... to have these for use as a "DEPENDABLE" and "RELIABLE" charging solution (oh yea, and low cost).

-Mike

PS:

I should have 3 of the newer S-320-24 models here on hand early next week and will begin immediate testing, 2 will be modified only slightly to produce a 10.2A 62.25v charger for my 10 and 20AH 15S packs but the other is to be fully reverse engineered and I will attempt to convert it first to 10S (41.5) then 12S (49.8) finally 15S (62.25). There are still some inconsistencies I have found between S-350-24 units which hopefully the UL/AU certification requirement will have eliminated in these newer listed units...

 

[amberwolf]

For the zener issue, what is the *actual* value that the original zener conducts at, vs the pair (or other set) of smaller series zeners?

If it is conducting at a voltage less than the others, perhaps it is absorbing little tiny spikes at a lower voltage than the others do (if they do at all) and those spikes induce voltages or currents in something else in the supply that causes the noise?

 

[Hyena]

That's interesting, the replacement I was sent for the meanwell clone that blew up for no good reason buzzed out of the box and I'm pretty sure it came from the factory with 2 zeners in series. I made it for a friend and it continues to buzz intermittently during the charge but hasn't blown up yet. I told him that's normal... :lol:

 

[Hyena]

hah, boy did this asshole pick the wrong forum to spam :lol:

 

[wookey]

I just fitted R37 and SVR2 and removed R33, setting the adjutable resistance to the same as R33 (746ohm). It still works as expected. In order to test the current limiting I need a suitable load. Using my battery pack isn't really convenient (or safe for the pack). What do people use for a 350W load for testing these things?

 

[Hyena]

Using 10ah pack is perfectly safe for the batteries as the max charge rate will only be about 1.5C, and what could be more convenient than using the actual batteries you're going to charge anyway ?

 

[regmeister]

deleted

 

[Hyena]

The maryma one is the one I posted pics of earlier in the thread. They have a MW symbol on the front sticker but only the genuine meanwells have Meanwell and the logo laser engraved into the top of the case.

 

[regmeister]

deleted

 

[wookey]

The problem with using a pack for charger-testing is that I only have 24V-worth assembled so far due to a shortage of mounts. Also I've carefully filled my cells up individually so don't really want to charge them any more until they've been used.

 

[regmeister]

deleted

 

[cell_man]

I need to get some of these. They are very readily available nearby me. There is also a manufacturer in Nanjing which is not so far from me, not Meanwell which is a HK company AFAIK but they look very similar. I'll see what they can do on a quantity and if they can supply them to spec with a bigger voltage adjustment and maybe adjustable current limit.

I'm especially interested in getting some small PSUs that can be re configured to give 3.6-3.7V with 10A plus and using these in series to charge a bigger pack. I picked up up a 35W Meanwell PSU from a shop for about 6USD and they do these in 5V versions too. I've even seen a small 3.3V with +-10% adjustment that might do the job but current is a bit low.

There are lots of companies that do a massive range of PSU in pretty much any voltage and current combination you require. I'll try to pick up some samples and go from there.

What are the popular sizes of PSU used and what sort of price do they go for on Ebay?

 

[Velocipede]

I understand that an adjustable PSU without the buzzing sound is the goal right now.
We can get them on ebay for around $50USD.

 

[Hyena]

Most of us are using the S-350 meanwells or clones in 24 or 48v forms.
They're rated at 350w but put out well over 400 when used for lipo charging and will self destruct if the current isn't pulled back a bit. Then tend to buzz if you do this by modding resistors which would be the preferred method for making the current adjustable, but I've found modding the shunts will lower the current to a fixed level silently.

They go for around $40-50 delivered from china or methods posted a link a few posts up where they can be had a bit cheaper from a seller in the US

 

[regmeister]

deleted