Building a substitution miniature flyback transformer

[richard1991]

Hi. The miniature flyback transformer burned out. I've taken it apart and I know the primary and secondary wire gauge and length*. Hopefully I can get an E core, bobbin and wire and make myself a substitute transformer. But, I'm not experienced, so I may need help.

The original transformer was made by VOGT (Now taken over by Sumida). The following was marked on the transformer core: "VOGT R1" The bobbin was marked "VOGT 32555".  Examining a VOGT catalogue "E cores and Kits" section, I have determined that the product Part Number must have been either 254 03 325 88 or 254 03 328 88.  These are described as E12.6/3.7 core shape.  325 relates to Fi 325 core material, 328 as  Fi 328 core material.  88 in the number I added taken from a table and is the standard AL value, 88 being 560 nH (the actual nH for the cores is 600 nH).

I've now got to try to source a replacement E core set and bobbin.

I attach a schematic.

* I may get my primary length wrong.  We will see.

P.S. The primary wire was in bits, and some has gone missing. So, I probably do not know the primary length, but could make an estimation (a matter of estimating what length of wire I've lost and adding that to the wire I have). It's about 44 SWG. Secondary wire gauge is about 32 SWG and is 952mm in length.

[IanMacdonald]

The main question would be whether it is a diode-split type or uses a separate rectifier.

Also be aware that on the separate rectifier type the interwinding capacitance of the secondary is somewhat critical because is typically tuned to 3x or 5x the input frequency. (it works like a Tesla coil, producing a greater output voltage than the turns ratio would suggest) Less critical on the diode split type, but then you need the diodes between layers.

[T3sl4co1l]

The main question would be whether it is a diode-split type or uses a separate rectifier.

Also be aware that on the separate rectifier type the interwinding capacitance of the secondary is somewhat critical because is typically tuned to 3x or 5x the input frequency. (it works like a Tesla coil, producing a greater output voltage than the turns ratio would suggest) Less critical on the diode split type, but then you need the diodes between layers.

Ehh...

Ohh wait, you mean like a CRT flyback transformer?  For high voltages?  Nononono, he's just got a little DC-DC supply.  Different beast.

Tim

[T3sl4co1l]

Hi. The miniature flyback transformer burned out. I've taken it apart and I know the primary and secondary wire gauge and length*. Hopefully I can get an E core, bobbin and wire and make myself a substitute transformer. But, I'm not experienced, so I may need help.

Did you break the core and bobbin in the process?  If not, they are perfectly reusable.

What burned on it?  Wire?  Bobbin?  Everything?

Unfortunately, length is not relevant: the number of turns (loops) around the bobbin is what matters here.

At least since you know the A_L of the replacement core, you can calculate turns based on the TOPSwitch appnote.  Look it up and give it a read.  It'll tell you everything you need to know about this sort of circuit.

If you just need it fixed, it's much easier to simply get a replacement module.  AC-DC and DC-DC modules are cheap and plentiful: $10 for something this size.  You may not be able to find an exact, drop-in replacement, but a little hack wiring will allow almost anything in its place.

Tim

[richard1991]

Further: Water must have ingressed the machine and the board the transformer is on, resulting in the primary winding being broken just before the connection pin on the former.  I set about carefully dismantling the transformer. Well, it was a bit of a disaster.  I broke the core, and I could not unwind the primary. So, I cut right through the primary which created lots of small bits of wire. I was measuring the wire bits to get a total wire length (so I could cut some wire the same length and rewind).  but for a week I did some other work and when I got back to this task, I found some of the wires missing from the bench. They probably got lost when my woolly hat landed on them at some point.

What I did manage to do was record the primary turns, there were 34, wound on the outside, 34 SWG.

I probably blew the TOP209P, so bought a new one.  I did go to Power Integration's site to see if there was anything there that I could use. I downloaded some application, but I'm not sure if it will help or not.

My reasoning has been that if I can get a similar transformer core set and former, then knowing the wire lengths I could just wind on the wire and I'll be done. But, as I say, I don't think I can get to know the primary wire length.  Well, I might be out by say 20% if I estimate the wire that's gone missing.   I have some TopSwitch app note. I'm uncertain whether it will help or not.

As has been mentioned, maybe I can hack the board a bit and put in a replacement circuit.  The item that I'm effectively repairing is a Bosch washing machine.

[DBecker]

This is a simple non-isolated white goods supply.  It's designed to work with a wide input voltage range, as cheaply as possible.

Read the datasheet.  The transformer isn't critical, you just shouldn't get it grossly wrong.  Especially if you only need it to work with a single input voltage.

Before you bother getting a replacement transformer, verify the rest of the machine works by using a cheap wall wart 12VDC supply.  That way you won't be debugging a power supply issue when the problem is further in.  And you may find that it's an acceptable solution.  Even if just for the short term need of wearing clean clothes.

[richard1991]

If I took the "hacking option" especially as an interim measure, would I not buy a miniature embedded AC/DC PSU, like for instance something like RAC03-12SC or equivalent?  I'm guessing about the wattage being 3W  (12V @  250mA).  £11.50 at RadioSpares.

https://www.recom-power.com/pdf/Powerline-AC-DC/RAC03-SC.pdf

[richard1991]

I am tempted to do a sort of hack to get the appliance working (and may do that), but still feel I should try to make a substitute transformer.  I have all the relevant Power Integrations application notes, but they intimidate me a bit.

First of all, let's deal with this on page 3 of AN-18. "In order to meet international safety regulations, a transformer for use in an off line power supply must have adequate insulation between the primary and secondary windings. For transformers using standard cores and bobbins, there are two basic transformer insulation methods: margin wound construction and triple insulated wire construction".

Okay, "off line" I believe means taking power directly off the ac/mains power line, without an (isolating) transformer between the IC/smtp transformer.

If you look at the circuit I posted, power to the flyback transformer and to the IC is taken directly from ac supply. Now, there were no margins on the transformer I took apart. The margins I take to be tape wound on the transformer bobbin adjacent to the inner side of the bobbin flanges.  Is margin wound construction inferred to meet off-line safety regulations? I think today it is.

Perhaps in the appliance design there was a reason not to use margin construction, or maybe the regulations in Europe were different 20 years ago (when my appliance was purchased). There is a reason possibly why there was no reason for a margin wound transformer.  Something I think to do with low isolation requirements. Probably identifiable in the circuit I posted. Probably, if the line voltage got onto the secondary, it would short circuit and blow a fuse.

[T3sl4co1l]

The circuit is common ground: a DC-DC converter.  It could be buck format instead, but that might give poorer performance because of the very low duty cycle.  (There are ICs that do this function, all in one, just like the TOPSwitch does flyback, but they're even more expensive IIRC.)

So, only functional insulation is needed.

Tim

[richard1991]

My sitation is this:  Looking at Power Integrations design applications (I have two downloaded: PIXIs Designer 9 & PI Expert) the specific IC used in the circuit (TOP209P) is not listed as a choosable component.  Does that mean then that PI have have no available software design aid to help me design the substitute transformer that I'm seeking to make? The answer I guess will be a straight yes or no.

If the answer is a no, and I think it is (From a PI forum - "While we still supply TOPSwitch-II devices we recommend starting new designs with TOPSwitch-JX which we have spreadsheet support for. We no longer support TOP 2 in the software." ) - then is there any other software design application that I can use to help design the transformer?

The PI application note dealing with flyback transformer design covering TOP 2 was AN-17 and one could get "AN 17 spreadsheet". Looks like that is now not available.  Therefore I'm wondering whether I am unable to use any other design software to help. I do have most of the transformer characteristics of the original transformer, such as core shape,size, AL, Ae, Ve,  Le etc.  I have the original transformer primary SWG (but not turns) & secondary gauge of 32 SWG with 34 turns.

P.S. Actually PIXIs 9 might be the AN 17 spreadsheet. But, in the latest version (9)  it only caters for products TOP252PN/GN to TOP262LN  and TOP264E to TOP271K.

A substitute for 209P might be  TOP221, TOP222 & TOP242.  Not covered in the list.  Probably covered in some version of PIXIs that is not now available for download.

[richard1991]

Hi. I finally decided to just substitute the original SMPS circuit with a cheap SMPS module: 12V @ 450mA. Has >1000V DC isolation. Washer now works. I think it's safe.