tagan dead, caps to blame

[starfury1]

Re: tagan dead, caps to blame

put in KBU1010 (Thanks Gonzo )

In RS components got the following

(have to buy 2 thought, plus shipping))


well you may have to scratch the below I see it mentions surface mount ?
I thought the part number is suppose to be what it is... package wise
Maybe call them could just be a website thing

--------------------------------------
RS Stock No. 2509471076
Availability N Y Goods will be delivered in 6-8 working days
RoHS Status
RoHS certificate of compliance
Manufacturer HVCA
Manufacturer's part number KBU1010

Add to compare

Quantity Price
(ex. GST): Each
2+ $5.64
25+ $5.03
50+ $4.50
100+ $4.11
250+ $3.39
Quantity
min:2
Name
ER - Bridge Rectifier; 1000V; 250A; 420V; 10A;Surface Mount

[Brian C]

Re: tagan dead, caps to blame

Hi Paul_h... is the fuse blown ? In this case, you need to check the Switching Transistors as well. There are 2 pcs, mounted on the heatsink next to the 1200uF(200V) capacitors. TAGAN uses bipolar 2SC2625 (Fuji Electric) as the push pull switches.

I ever repaired few TAGAN PSU with shorted junction 2SC2625.

[paul_h]

Re: tagan dead, caps to blame

put in KBU1010 (Thanks Gonzo )

In RS components got the following

(have to buy 2 thought, plus shipping))


well you may have to scratch the below I see it mentions surface mount ?
I thought the part number is suppose to be what it is... package wise
Maybe call them could just be a website thing

--------------------------------------
RS Stock No. 2509471076
Availability N Y Goods will be delivered in 6-8 working days
RoHS Status
RoHS certificate of compliance
Manufacturer HVCA
Manufacturer's part number KBU1010

Add to compare

Quantity Price
(ex. GST): Each
2+ $5.64
25+ $5.03
50+ $4.50
100+ $4.11
250+ $3.39
Quantity
min:2
Name
ER - Bridge Rectifier; 1000V; 250A; 420V; 10A;Surface Mount

Yeah I found them on rsaustralia earlier this afternoon, but you have to actually buy 5 minimum when you try to order them, so $25 just to get some
farnell has nothing close to the rectifier I need.
Someone on ocau says jaycar has a 8A rectifier, but my old catalogue and there online one doesn't list it, the highest they list is 6A. But I might go there or phone to find out tomorrow.

Yeah that chip on the small PCB is a BTA08 triac.

This just isn't about economically repairing this PSU, but learning this stuff.
As I mentioned before, I'm an a/c mech. Most people in my field just tell people to buy new fridges or a/cs if they break. Or they replace whole circuit boards that can cost up to $1k.
I'm pretty much the only person that even bothered to get down to replacing blown MOVs. So I'm trying to extend my knowledge beyond that now that I'm trying to start my own business. Repairing rather than replacing is going to be my mantra, hence I don't care what it costs me in time or replacement parts to work on this PSU, since all refrigerators and a/cs are electronic now

[paul_h]

Re: tagan dead, caps to blame

There is a bit of circuitry next to the fuse holder and I can see a chip but no idea what it is

So if you can post that number it might help me understand it
and also the number of that large device on the little pcb

So thats why I am having a hard time cause I've never seen the guts of the tagan PSU up close and personal and its a mystery to me.

Cheers

BTW does you camera have macro?

The chip near the fuse is an 8 pin UC3842B chip. And yeah I've got macro, just uploading them at a low res at the moment to not annoy websites and hosting services (most of them have a tight limit on kb )

[linuxguru]

Re: tagan dead, caps to blame

Those 2sc2625s fail fairly frequently, also seen in Sparkle/FSP PSUs. I suspect the Vceo rating is not high enough, or the SOA is marginal and they're undergoing 2nd breakdown. I once used 2sd2498 horizontal deflection transistors to replace them, they worked for a while but failed within six months too. The base drive and/or pulse shaper may be marginal in the original design - it needs further analysis.

[paul_h]

Re: tagan dead, caps to blame

D

Hi Paul_h... is the fuse blown ? In this case, you need to check the Switching Transistors as well. There are 2 pcs, mounted on the heatsink next to the 1200uF(200V) capacitors. TAGAN uses bipolar 2SC2625 (Fuji Electric) as the push pull switches.

I ever repaired few TAGAN PSU with shorted junction 2SC2625.

On this PSU I've got a basic main 10a fuse blowing, blown primary cap and evidently shorted bridge rectifier. Does the fault you are talking about fit in with those symptoms, ie cause or result of this failure?
I'm just trying to learn here as a shorted cap or rectifier will definately cause the fuse to blow, where does this transistor fit in to it all?
* Just asking because this is a fairly busy, chock full of stuff PSU, with craploads of components and heatsinks, so it's hard to read any chips without major disassembly

[paul_h]

Re: tagan dead, caps to blame

BTW, theres three transistors mounted behind the primary capacitors, (on the other side of the large heatsink though ). I can't see one of them (the smallest one), but the other two are 2sc3320 chips.

[starfury1]

Re: tagan dead, caps to blame

yeah your not wrong on that...reading chip etc numbers

Humm yeah thats sometimes the prob you have to buy in a min amount
(same with the caps too)

Having a few spare bridges wouldn't hurt thought if you are going to repairs things
but its up to you

There are heaps of stuff out there on switchers, also look through the posts here as well...I tend to post links all over the place and mostly to do with more basic stuff learning wise.

Yes these days most repairs are PCB jockey, more so now I think since SMC has entered the scene you need a bit more stuff and skill set to fix them
The other side is as you mention cost of replacement PCB verses time and cost of tech to repair them....plus circuit complexity these days..not to mention getting parts can be a real pain (as you finding out with this one)


Gonzo, linuxguru and Brain C (plus others here) advice I would follow as well cause stuff that goes like this did are very liable to have other faults...you need to check.
( they work in the field and would have had more experience, knowledge of them then me
I used to fix electronics stuff but not switchers so don't have great knowledge of them)

here is the a PDF on that chip HIGH PERFORMANCE CURRENT MODE CONTROLLER
UC3842B

Unfortunately schematics seem to be somewhat hard to get your hands on lest wise around the net...don't know what the story is with companies asking for them
you use to be able to get service manuals schematics for a price from the manufacturers but don't know how well this holds true today...

So you tend to find your self reverse engineering out things....
although if you fix them often enough you get know the topology etc of how some are put together.

Well as you know sometimes you can fix something and not have to worry about the sequence of events that lead to the failure but other times you do or else it will happen again.

Anyway if you know ohms law etc know what a cap inductor and resistor and how they function in regard to DC and AC (frequency) then you got the basics in place you only really need to learn about active type components Transistor fets etc
(if you haven't all ready)
IC's like op amps, and IC's more as system blocks, like the one above

by the sounds of it your (line of work)
would be more along the lines of
Power control systems I guess

With the stuff you are working on it would most likely be a dedicated uP controller and probably in house code programs...some repair might be a bit limited but some stuff just has basic probs like the psu etc

Cheers

[gonzo0815]

Re: tagan dead, caps to blame

I like to fix things too, and i am still curious if i see a dead SMPS anywhere.
Only those TV SMPS are a little unhandy, but it makes fun to fix em too.

If the 2sc2625s are prone to failure, then i thin it is best to check them too.
The other device is most likely a Topswitcher, but AFAIR they do rarely fail. Their secondary side caps on the other side do fail regularly if from questionable quality.

[paul_h]

Re: tagan dead, caps to blame

I can't find any 2sc2652 chips, the one's near my main capacitors are 2sc3320 chips.
As far as the bridge rectifier, the closest I've been able to find is a 8A 800V which I ordered from the same place as the 1000uF capacitors.
Wouldn't 10A be overkill anyway? If these PSUs are sold for 110V, would they rate a higher amp rating just to cover that lower voltage? So running it at only 240V means I could get away with a lower amp rating of 8A with no problems?

[linuxguru]

Re: tagan dead, caps to blame

The 2sc3320 is a bigger brother of the 2sc2625, rated at Ic=15A vs. 10A, and Vcbo=500V vs. 450 on the c2625 - other specs, including the packaging, are nearly identical. It's possible that Tagan upgraded the transistors after seeing c2625 transistor failures on the original design.

This also points to the c3320 as a reasonable upgrade for the c2625 on FSP/Sparkle designs with failed primary switching transistors.

[starfury1]

Re: tagan dead, caps to blame

pdf on both

2SC 3320

2SC 2625

links may die but the others were china japan ? english..
nothing wrong with that if you can read them.

generally speaking
You have to be careful with component substitution, this is where it really helps to know how something operates and what changes in parameters you can get away with without upsetting the circuit.

The other side (as Above) maybe what was there really wasn't good enough to start with
(whether thats because of cost cutting or availability or a bad design)

(yes linuxguru sounds like an idea to have some to hand )

sometimes its just down to the part you cant get or is obsolete

Ideally circuit design should dictate how the transistor etc operates
but sometimes a change in some other aspect or parameter can throw the circuit out and it wont operate as it should.

bottom line, mostly you can and a lot of the time you don't have much of a choice in the matter

really just a word of caution, if in doubt post a query

just a few thoughts on it really

If you dont know how to test transistors its a bit like 2 diodes back to back
(bipolar transistors that is mosfets are different but mostly go short.)

here a few links, just in case

http://www.kpsec.freeuk.com/components/tran.htm

http://www.electronics-radio.com/art...istor-test.php


http://www.elexp.com/t_test.htm

this covers a few components

http://www.repairfaq.org/sam/semitest.htm

http://www.uoguelph.ca/~antoon/gadgets/mostest.htm

http://www.4qdtec.com/mostest.html

you will have to sort through them and see whats useful but hope it helps.

Another good piece of test equipment to have is an ESR meter

see willawake's post on Bob's pre build EVB version

https://www.badcaps.net/forum/showthread.php?t=1287

Doesn't have to be his of course
but for the money, his is good value for what you get
(Unfortunately DS no longer make the kit but it is available from over sea's source..or buy the pre built)

you can get away without one just handy to have if you are going to do power supplies of switcher types.

"As above" the secondary caps (DC output voltages) are a major part that
does die

anyway like I said HTH

Cheers

[Brian C]

Re: tagan dead, caps to blame

I can't find any 2sc2652 chips, the one's near my main capacitors are 2sc3320 chips.
As far as the bridge rectifier, the closest I've been able to find is a 8A 800V which I ordered from the same place as the 1000uF capacitors.
Wouldn't 10A be overkill anyway? If these PSUs are sold for 110V, would they rate a higher amp rating just to cover that lower voltage? So running it at only 240V means I could get away with a lower amp rating of 8A with no problems?

The 2SC3320 is the big brother of 2SC2625, better specification. Some TAGAN PSU use 2SC3320, especially higher wattage type. The 2SC2625 is found in the 330W TAGAN PSU.
TAGAN uses a pair of bipolar power transistors as the switches, operates in push-pull (upper transistor ON, lower transistor OFF and vice versa). The PWM controller for this operation is TL494 or the KA7500. In TAGAN PSU, it will be TL494. When both upper and lower transistor had a shorted ce junction, it will cause fuse blown (that is what happened to my friend's PSU). Supply supervisory circuit is found on the TL494 controller board. It uses the LM339 quad comparator IC to monitor the output voltage level. Its job is to shutdiown the TL494 when output voltages are out-of-the-range. You can adjust the trip point by changing few resistor value.
The current mode controller (UC3842) is the controller for the Standby Circuit, i.e. The Standbby 5V supply and on-board auxilliary supply. The most common failure of this section is the open-circuit of the supply line resistor ~62k (connect to the Power Supply Zener Voltage pin). Sometimes it could be the fault of MOSFET switch as well.
I attached some pictures of TAGAN PSU for ur reference... the location of Active PFC, PWM controller, etc...

[The Active PFC Controller Board]


[The Push-Pull Transistor Switches]


[The PWM Controller and Supply Supervisory Board]


[The UC3842 Auxilliary and Standby 5V section]

[starfury1]

Re: tagan dead, caps to blame

Thanks brain C for the info and pictures

Paul h

Wouldn't 10A be overkill anyway? If these PSUs are sold for 110V, would they rate a higher amp rating just to cover that lower voltage? So running it at only 240V means I could get away with a lower amp rating of 8A with no problems?

I dont know if strictly speaking thats correct cause not sure how the front end is operating (old ones 2 different modes of operation, as above)
even then id have to think about it

But yeah you might get away with it, although id rather go same or above to be sure.

There should be in rush limiting for the caps and I suppose how much gets pulled via the bridge diodes is Dependant on how heavily loaded the supply is
it is at that point more or less a linear supply.

So I am guessing it will be ok under light moderate loading
Heavy loading "might" kill it

But again guess work on my part

I can just see that there is 2 X 200V 680 uF (I think caps) on the psu of brains photos.

Also the wires are "Red" blue black white
The front ends look similar

based on the fact that the UC 3842 is doing what brain said and that section looks the same.

I guess you don't have Active PFC and that little PCB is a Voltage Detecting Auto switch (Triac on it is 8A)

If what I am guessing here is correct.
The triac is only on if the mains is 115V (Voltage doubler mode)
but it really depends on where those 3 wires go
They sort of look in the places they should be for it to be doing that

So if the above follows...The front end is basically the schematic link (atx) I posted
(not yanz's)

anyone comment on if the above is correct?
Or am I totally barking up the wrong tree here.

From Brains info, if you have the TL494 then the supply is "old school" with a few changes

I tend errr on the side of caution so you will probably get away with it

Its the peak current drawn though the diodes, if that gets exceeded then good bye bridge

if anything you have probably just reduce your safety margin a bit

Just make sure you check the transistors and I would throw the light bulb in series when you first fire it up
It may save its bacon if you have missed something.

But not the guru here so anybody else in the know please comment

Cheers mate and good luck with it...keep us posted

[gonzo0815]

Re: tagan dead, caps to blame

I think you are right, an TL494 is not a PFC controller. So it must be the classic design with voltage doubler.

The rating of the rectifier bridge probably depends on the capcitors inrush current, not on the continuous current.
But an 8A one should handle this.

[starfury1]

Re: tagan dead, caps to blame

yes thats my guess its the "classic design" (well said Gonzo)
updated a bit.

If you have a look at the 530W review pics it appears to use the same Vertical secondary PCB as used in Brain C's pics,
from the little bit and number I can see
(as in the TL494, LM 393)

But the mains side is different again, although 2 caps are there

Seems they were just revamping the Mains side, possibly to meet new specs
(although the review says "No PFC" on the last page for this one)
and boosting the secondary outputs if JG's is anything to go by
(err yeah that might be more on paper then reality )

Interesting

[paul_h]

Re: tagan dead, caps to blame

Sorry to disappear, I had heaps of trouble getting the parts on order. Problems with the courier etc, they came in two lots so had to repeat all the problems with the courier. Anyway got the rest of them today.

I stuffed up, I did heaps of research into finding a capacitor that will fit. I went and ordered what I thought would fit (a panasonic "c' series ecos2), I ended up ordering and getting 2 of ecos2dp102Ds by my mistake, which of course like every other bloody capacitor that rating, is 30mm dia, so they don't fit.
Doesn't look like they have any c series 25mm ones either.
Ah well, stuff it. Tagan made the parts too hard to get, by the sounds of it having capacitors rated at 1200uF isn't common.
Thanks for all the help anyway.

[gonzo0815]

Re: tagan dead, caps to blame

Sad, that it is so hard to get the parts.
Rectifiers are cheap here, at least up to GBU8M it is easily available here for merely pennies.
I have a quantity of GBU8J laying around, as i need this for some control PCB`s of a coffee maschine.
I bought them for about 0,30€ each. But sure, large can caps are very expensive here, if not bought from surplus dealer or ebay.

[Big Pope]

Re: tagan dead, caps to blame

Tagan made the parts too hard to get, by the sounds of it having capacitors rated at 1200uF isn't common.

What is the diameter of this pair of Fuhjyyu 200V 1200uF? May be i can offer you Samxon in the right size.

[starfury1]

Re: tagan dead, caps to blame

to quote paul_h from his first post

It's a 1200uF 200v cap, 25mm dia, 45mm high, oh yeah, it's a fuhjyyu capacitor

seems the 25mm is an odd ball size to get hold of here
(although RS etc carry stuff like this in OZ they don't tend to have a vast range of it)

I do remember you mention something a while back about some 200 or 250 Volts types mains ones Big pope

HTH

yeah bummer on that paul_h
but big pope might be able to help you out here