EPCOS SMPS Transformer substitute

[dkneyle]

Trying to help a mate who services industrial equipment get this SMPS going again, otherwise the whole machine is a write off. Replacement boards not available from supplier.
This is an industrial power supply from a commercial machine; I think it's a dishwasher. Lightning strike. Mt mate had a first go and assumed it was diodes or caps that had failed. He said the coupling capacitor had actually melted the solder pads.
Measurements with meter showed nothing seemingly failed on the primary side other than the big electro which seems to be reading low. Very suspicious about the transformer, which when you look very closely (not visible in pictures) looks like there could be a burn mark through the yellow plastic insulation. I'm making an assumption of shorted turns. Which caused the primary side to overheat as it was driving into a shorted load.
Anyway, would like to replace the transformer with either a direct replacement or substitute transformer. It has a part number (?) EPCOS 4310 P8910. Google search doesn't throw up a direct match anywhere.
Was thinking I could maybe substitute a small transformer from the 5V supply of TV SMPS I have hanging around. Any clues as to the specs I should be looking at trying to emulate?

[redwire]

Re: EPCOS SMPS Transformer substitute

It would be really hard to find an off the shelf part to replace this transformer.

Only chance would be to work backwards from IC100 design notes and datasheet, to find the reference design that may be used in this dishwasher.
Since a custom transformer is such a hassle, many manufacturers use a stock transformer. Some SMPS transformers here:
http://www.coilcraft.com/transsel_ac.cfm
Wurth has some offerings too.

Mainly you have to match IC100 and the output power/voltage requirements.
Dual output- mainly 12VDC for the relay and another output? I estimate 5-10W. It doesn't look like it is providing isolation though.

IC100 might be LNK363
Design sheet shows how complex the transformer can be:
https://ac-dc.power.com/design-suppo...-30-w-charger/

You can carefully disassemble the transformer and write down the # turns etc. and rewind it, if surgery is your thing. Put it in the oven at 250F and the cores slides apart. Just to not melt the bobbin. Kinda last resort.

[dkneyle]

Re: EPCOS SMPS Transformer substitute

Thanks Redwire for getting back to me with a wealth of good ideas. I'm more accustomed to TV SMPS where hot and cold sides are easy to identify. Without having drawn out the circuit but following the tracks it looked to me like the 240VAC supply was given a simple half wave rectification with a power diode and the filter cap, which supplied the IC100, an driver oscillator which runs of 400 volts! I found it with a Google search, and yes, it is an LNK363PN. I too estimate it is only about 5 -10W driving the relays. I might do a bit more reverse engineering of the tracks and transformer pins to see how it is configured. From memory the primary was just two outer connections used (centre tap unused). The secondary had four pins and all appeared to be used.
Anyway, will let you know how I go. I too would be surprised if it was a unique tranny just for this machine. I'm sure it must be an off-shelf component of some type.

[dkneyle]

Re: EPCOS SMPS Transformer substitute

Looking at design notes it refers to an EE16 flyback transformer. Gee, they look close from images I can see! I think you have put me on the right track at least.

[Longbow]

Re: EPCOS SMPS Transformer substitute

SMPS design is not in the same category as a linear supply. Lots of engineering goes into SMPS design. The circuits may look the same, but you can bet that the parts are not interchangeable. The Coilcraft web site lists a lot of transformers that are available for the purpose, but the parts become the basis for a new design and are not considered replacements for existing supplies. I suppose it is possible to get lucky and identify a part number from a generic supplier, but this possibility is very unlikely. In short, playing with replacement transformers will not provide you with a reliable repair.

All that aside, statistically speaking the switching transformer is the LEAST likely part to fail in a SMPS, simply because the parts connected to it are far more sensitive to abuse. In this case, you can't argue with lightning. When it comes to a lightning strike, you are wasting your time.

If you have an expensive unit that warrants repair, by far the best solution is to purchase a supply that provides the required voltages. Remember that your dishwasher might also have electronics associated with an operation panel and other sensor functions. Those other pieces could easily be ruined along with the main supply. Just my 2 shillings worth.

[dkneyle]

Re: EPCOS SMPS Transformer substitute

Thanks Longbow for your 20c worth. (Two shillings before we went to decimal currency in 1966 here in Oz!)
I've been on a fast learning curve and have come to the same conclusions you have provided here. The dishwasher is a very expensive industrial model and my mate is keen to see if he can't get it operational again.
I've dealt with lightning strike boards before and usually you can see physical damage. In this case there is soot present, yet the components seem to read good. Except for the 10UF cap which seems to read low we can't see a lot of issues. Under a mag glass there is this slight imperfection in the yellow shrink wrap coating on the transformer, and the windings are exposed, but I can't honestly say I can see any signs of burning. I too am wondering if the transformer is in fact OK?
I think I need to look real carefully for any semis further down the board that could be defective.
When I get home tonight I'll post a picture of the entire board. Might prompt some thoughts?

[starling22]

Re: EPCOS SMPS Transformer substitute

The post is getting a bit on in time but I have got involved in the similar board out of a washing machine. The evidence is similar but a bit more pronounced with the LNK chip cracked and some dark dust around.

Working around the components and tracing the diagram in this area, a bit laboriously but I am curious.

Just wondering what did happen her in the thread, just getting interesting then no more posts??

[R_J]

Re: EPCOS SMPS Transformer substitute

It looks like a fairly simple smps, should not be that hard to repair, If the ic is cracked it is bad, post a clear straight on picture of the top and the bottom so we can see the traces.
I would really doubt the transformer was bad.

[starling22]

Re: EPCOS SMPS Transformer substitute

Yes I agree it looks straight forward but .... At this stage I do not need help apart from I would like to identify that transformer, no matter what.

For one to put in my ECAD schematic and second to see if it is one of those specially designed ones with extra screening or whatever the talk is in the LNK363 pdf file.

I have found out that the LinkSwitch-XT family has been superseded by the LinkSwitch-XT2, where there is an very obvious rearrangement of the die making it not direct replacement. Roughly described a 90 degree rotation so all 4 Source connections now is on one side instead of 2 on each side.

My guess the company found there is a problem with the chip. So I am preparing for redesign in that area of the board. Partly for the fun of it as it is a family affair being my sons washing machine and he trying to have a go at electronics with limited knowledge in trouble shooting and I am located far away.

He is ordering obvious parts around the area and can replace them but identifying or testing if more parts are gone may be hard. He can take good photographs when he knows what is needed.

I have a number of photos but you would at quick glance say there is nothing wrong with the board my son described his findings to me and I imagined the chip was blown to bits and charred area but when the photos arrived it was not immediately obvious the chip was cracked.

What I did notice I could not read the markings on the chip, that was unusual unless deliberate grinding off had been carried out but I think the part is legal or real enough.

I will attach couple of pictures

As I said identifying that transformer or the series and get some technical details is my aim. I thought that TDK might have them but not sure yet the EE16 is the best so far.

I will also attach what I have decoded so far, some body may be interested in checking my part schematic or perhaps add as bit 'in pencil' in such case I will add it in the ECAD and post back.

[budm]

Re: EPCOS SMPS Transformer substitute

I cannot find any info about that transformer either, but the original IC is still available at MOUSER. I do not see EOL (End Of Life) about this LNK363PN either.
https://www.mouser.com/ProductDetail...9Uk%252bMuQ%3d

[starling22]

Re: EPCOS SMPS Transformer substitute

The details of being superceded I got from the manufacurers site when looking up details on the XT and found it was now XT2, I did not take notice of the time the new line was in sales.

[stj]

Re: EPCOS SMPS Transformer substitute

how much current at 5v can it supply??

[starling22]

Re: EPCOS SMPS Transformer substitute

Go to this website and look for some reference designs with the LNK363:

https://www.power.com/search/site/linkswitch-xt/


Depends on what you need and surroundings, in this case enough I should think but I am yet to find out what voltage it is running at or should be running at.

Enough as there is not really any evidence of long term overheating.

[starling22]

Re: EPCOS SMPS Transformer substitute

After some sweating and headache, the power supply section around the LNK363 I believe have been decoded, a schmatic and overlay produced.glk. This is now being used to check out and see what will happen next when the chip gets replaced.

Will attach schematic and overlay for the power supply section only at this stage. There is a jpeg and a windows xps file with good resolution of the schematic

xps file not on allowed list, will have to see what I can do, pdf version is not too god.

[budm]

Re: EPCOS SMPS Transformer substitute

So that SMPS is for two negative power supplies? That is what it looks like.

[starling22]

Re: EPCOS SMPS Transformer substitute

Yes the negative supplies I find a bit confusing. I should not be confused but am so used to just seeing positive.

As the switching runs at a nominel 120 kHz, I think I got that from the specsheets of the LNK363.

The need for some fast/ultra fast switching diodes, identified as

2DD101 UF1006 0519 800V / 75ns
https://au.mouser.com/ProductDetail/...btsdVN7mcwI%3d

2DD102 SB140 0369
https://au.mouser.com/Semiconductors...=SB140&FS=True


2DD103 SF24G B10 (B1zero) 200V / 35ns --> ES1D 200V / 25ns Diodes Incorporated
https://au.mouser.com/Search/Refine.aspx?Keyword=SF24G

While these diodes seem to be available they are hard to find in single quantity. Normaly RS Online Australia can provide but this time it seems only in quantity for mass production. Need some more checking with them.

Mouser got none in stock either it seems.

However if anybody has suggestions as replacements please let us know. At the moment I am aiming for a trr of 35ns but I have wondered if 75 ns is close enough. Say out of pc switchmode power supplies (got plenty of those under the house)

[starling22]

Re: EPCOS SMPS Transformer substitute

Ah ha, just stumbed over an .xps to pdf converter that seems to do a nice job. So here is the schmeatic devloped so far, around the power switcher LNK363.

I also found older top view of the complete power board (there is another nasty complicated microprocessor board associated, hope not having to deal with this)

Icludes the X-ray of the tracks side where I have been working at annotating.

Warning: very likely not correct as I was struggling more with my ECAD software at the time. Check it yourself if you are going to use it

I do not have a full size board completely annotated yet, I am hoping that I do not need to as it is considerable work.

[petehall347]

Re: EPCOS SMPS Transformer substitute

new ic might be a good start along with fresh capacitors

[dkneyle]

Re: EPCOS SMPS Transformer substitute

Yep, even though I don't have the board any longer, pretty sure it is the same as the one I was working on.