separating smps transformer cores
Avalanche, Fri Aug 18 2006, 01:48PMHi,
I've broken about 3 cores now, and I have one left here
I have a few smps transformers, and I have salvaged the former from one (by breaking the ferrite), but now I need the two ferrite halves to make a complete transformer again. I have managed to cut the former and wire off some others, leaving just the two halves of the core stuck together with some kind of white resin. The problem is, the resin is pretty strong. I have tried boiling one for about 20 minutes, but still the ferrite breaks when I try to separate the two halves... Any ideas? Is it even possible to get this glue to soften, maybe soaking it in petrol or something?
I have a good supply of these transformers, so I can experiment. If all else fails, I'll have to just wind my transformer on one of the salvaged cores, leaving the former out... that would be a shame though.
Re: separating smps transformer cores
Wilson, Fri Aug 18 2006, 02:01PM
Perhaps acetone?
btw, you can glue the broken ferrite back together with CA glue, and it works as new :P
Wilson, Fri Aug 18 2006, 02:01PM
Perhaps acetone?
btw, you can glue the broken ferrite back together with CA glue, and it works as new :P
Re: separating smps transformer cores
Marko, Fri Aug 18 2006, 02:07PM
ATX transformers are dipped in epoxy and getting them apart is usually very frustrating and messy job.
I got one core out only once by heating the transformer with 500ºC hot air fan, I made terrible mess, stunk out whole room and mutilated the former.
Temperature required for epoxy to melt is simply way higher than ferrite is 'rated' for.
Core formed microfractures, and when I took the core later it just fell in two pieces.
I fixed it with superglue but I'm not happy at all with it.
Most cores and formers I tried to pull apart that way were trashed in beggining.
Cooking in water usually helps only with very small transformers, temperature isn't too high and core usually survives if you are careful.
No way..
Marko, Fri Aug 18 2006, 02:07PM
ATX transformers are dipped in epoxy and getting them apart is usually very frustrating and messy job.
I got one core out only once by heating the transformer with 500ºC hot air fan, I made terrible mess, stunk out whole room and mutilated the former.
Temperature required for epoxy to melt is simply way higher than ferrite is 'rated' for.
Core formed microfractures, and when I took the core later it just fell in two pieces.
I fixed it with superglue but I'm not happy at all with it.
Most cores and formers I tried to pull apart that way were trashed in beggining.
Cooking in water usually helps only with very small transformers, temperature isn't too high and core usually survives if you are careful.
Perhaps acetone?
No way..
Re: separating smps transformer cores
Avalanche, Fri Aug 18 2006, 02:59PM
Thanks for the replies
I've changed my mind about these transformers now, they are so difficult to separate I might as well just use them as they are. By modifying the way the windings come out, I should be able to use them as a GDT for a halfbridge.
Something I did want to do though was make a mains inverter, I managed to separate a larger monitor transformer by boiling it. Just for the record, it appears that the brown glue (whatever it is) melts a lot easier than the white epoxy.
Avalanche, Fri Aug 18 2006, 02:59PM
Thanks for the replies
I've changed my mind about these transformers now, they are so difficult to separate I might as well just use them as they are. By modifying the way the windings come out, I should be able to use them as a GDT for a halfbridge.
Something I did want to do though was make a mains inverter, I managed to separate a larger monitor transformer by boiling it. Just for the record, it appears that the brown glue (whatever it is) melts a lot easier than the white epoxy.
Re: separating smps transformer cores
Marko, Fri Aug 18 2006, 03:11PM
If you re-use an ATX transformer you can make yourself a nice halfbridge SMPS.
You can use small drive transformer from ATX supply as GDT if main transistors were mosfet's.
You can use another small transformer for a mini flyback SMPS (I failed to get results from that, witohut regulation voltage spikes are real pain).
TV and monitor transformers tend to use flyback topology due to their low power, but I think you could run them with a fullbridge for some really fierce output.
Marko, Fri Aug 18 2006, 03:11PM
If you re-use an ATX transformer you can make yourself a nice halfbridge SMPS.
You can use small drive transformer from ATX supply as GDT if main transistors were mosfet's.
You can use another small transformer for a mini flyback SMPS (I failed to get results from that, witohut regulation voltage spikes are real pain).
TV and monitor transformers tend to use flyback topology due to their low power, but I think you could run them with a fullbridge for some really fierce output.
Re: separating smps transformer cores
Hazmatt_(The Underdog), Fri Aug 18 2006, 06:20PM
didn't someone sometime mention microwaving the transformers to separate the cores?
Hazmatt_(The Underdog), Fri Aug 18 2006, 06:20PM
didn't someone sometime mention microwaving the transformers to separate the cores?
Re: separating smps transformer cores
Marko, Fri Aug 18 2006, 06:28PM
I doubt ferrite would get to several hundred deg to actually melt the epoxy, since it's a high frequency electromagnetic core afterall.
It would get heated from the former and windings, but I believe think it would be much better than hot air or water (??)
Somebody tried it..?
Marko, Fri Aug 18 2006, 06:28PM
I doubt ferrite would get to several hundred deg to actually melt the epoxy, since it's a high frequency electromagnetic core afterall.
It would get heated from the former and windings, but I believe think it would be much better than hot air or water (??)
Somebody tried it..?
Re: separating smps transformer cores
ragnar, Fri Aug 18 2006, 08:51PM
With the white epoxy, you generally need to butcher it with a screwdriver to 'crack' off all the blobs of glue around the join, then sand until the only glue left is actually between the ferrite halves. If you then boil the transformer for a bit, and use gloves (stinking hot) you should be able to lever the top out, but I've found it _essential_ that any forces are applied on the round centre axis, otherwise you of course snap the ferrite wings off.
If there are largeish (0.3mm+) spacers in there, you can also rest the ferrite such that everything but the spacer is supported, and then with a knife etc 'crack' it out sideways, but it's difficult to get enough force.
If you've done it enough, it becomes a science and whenever you see some good ferrite, you'll know what to do with it. =P
ragnar, Fri Aug 18 2006, 08:51PM
With the white epoxy, you generally need to butcher it with a screwdriver to 'crack' off all the blobs of glue around the join, then sand until the only glue left is actually between the ferrite halves. If you then boil the transformer for a bit, and use gloves (stinking hot) you should be able to lever the top out, but I've found it _essential_ that any forces are applied on the round centre axis, otherwise you of course snap the ferrite wings off.
If there are largeish (0.3mm+) spacers in there, you can also rest the ferrite such that everything but the spacer is supported, and then with a knife etc 'crack' it out sideways, but it's difficult to get enough force.
If you've done it enough, it becomes a science and whenever you see some good ferrite, you'll know what to do with it. =P
Re: separating smps transformer cores
Marko, Fri Aug 18 2006, 09:19PM
Epoxy ''around edges'' isn't the problem. Problem is that space betweent he core and former is completely fiilled with epoxy. It is impossible to soften that amount of epoxy by boiling, or physically remove it somehow.
Some transformers are much easier to get apart, smaller ones are cooked easily but ATX are really vain.
Marko, Fri Aug 18 2006, 09:19PM
Epoxy ''around edges'' isn't the problem. Problem is that space betweent he core and former is completely fiilled with epoxy. It is impossible to soften that amount of epoxy by boiling, or physically remove it somehow.
Some transformers are much easier to get apart, smaller ones are cooked easily but ATX are really vain.
Re: separating smps transformer cores
Self Defenestrate, Sat Aug 19 2006, 12:19AM
I've had some luck with those ultra thin diamond dremel cut off disks, if you hold the core in something soft to ease the vibrations. Once the outside is done, I can usually boil and twist the cores apart. The inner rod part of the core is sanded a bit to flush everything up, and there you have it.
Self Defenestrate, Sat Aug 19 2006, 12:19AM
I've had some luck with those ultra thin diamond dremel cut off disks, if you hold the core in something soft to ease the vibrations. Once the outside is done, I can usually boil and twist the cores apart. The inner rod part of the core is sanded a bit to flush everything up, and there you have it.
Re: separating smps transformer cores
thom, Sun Aug 20 2006, 03:04AM
I've had no luck with boiling or solvents.
thom, Sun Aug 20 2006, 03:04AM
Self Defenestrate wrote ...
I've had some luck with those ultra thin diamond dremel cut off disks, if you hold the core in something soft to ease the vibrations. Once the outside is done, I can usually boil and twist the cores apart. The inner rod part of the core is sanded a bit to flush everything up, and there you have it.
hm thats a good idea as removing a small portion of the core is much better than a huge crack.I've had some luck with those ultra thin diamond dremel cut off disks, if you hold the core in something soft to ease the vibrations. Once the outside is done, I can usually boil and twist the cores apart. The inner rod part of the core is sanded a bit to flush everything up, and there you have it.
I've had no luck with boiling or solvents.
Re: separating smps transformer cores
Part Scavenger, Sun Aug 20 2006, 03:49AM
Hmm. I don't seem to have that much trouble. I usually just slide a utility knife down in the cracks and it pops apart. In fact, I rewound two today, then proceeded to drop one on the concrete floor.
Part Scavenger, Sun Aug 20 2006, 03:49AM
Hmm. I don't seem to have that much trouble. I usually just slide a utility knife down in the cracks and it pops apart. In fact, I rewound two today, then proceeded to drop one on the concrete floor.
Re: separating smps transformer cores
williamn, Sun Aug 20 2006, 05:07AM
I have alwaysed boiled mine as well with great success. It took a few times but after awhile you get it down. I have boiled some large ferrite transformers successfully. I use hi temp silicone gloves, boil, wiggle, boil wiggle, boil wiggle and eventually it comes loose. All the transformers I ever boiled came with yellow or blue tape around them, I dont know if that clues you into the type of epoxy used or not. Now I find it easier to just buy the dang ferrite/forms, no more ferrite soup.
williamn, Sun Aug 20 2006, 05:07AM
I have alwaysed boiled mine as well with great success. It took a few times but after awhile you get it down. I have boiled some large ferrite transformers successfully. I use hi temp silicone gloves, boil, wiggle, boil wiggle, boil wiggle and eventually it comes loose. All the transformers I ever boiled came with yellow or blue tape around them, I dont know if that clues you into the type of epoxy used or not. Now I find it easier to just buy the dang ferrite/forms, no more ferrite soup.
Re: separating smps transformer cores
Marko, Sun Aug 20 2006, 07:31PM
You disassemble big ATX supply transformers that way
(poorly epoxy-ed transformer?)
I decided boiling is'nt worth effort after several trials, core ended with tny fractures (epoxy wasn't bothered at all.
Thing I didn't try is sawing the edge of former and then making soup as some from here recommend, but seems like quite a bit of work, and you need 2 transformers dead to make one new..
Marko, Sun Aug 20 2006, 07:31PM
Hmm. I don't seem to have that much trouble. I usually just slide a utility knife down in the cracks and it pops apart. In fact, I rewound two today, then proceeded to drop one on the concrete floor.
You disassemble big ATX supply transformers that way
(poorly epoxy-ed transformer?)I decided boiling is'nt worth effort after several trials, core ended with tny fractures (epoxy wasn't bothered at all.
Thing I didn't try is sawing the edge of former and then making soup as some from here recommend, but seems like quite a bit of work, and you need 2 transformers dead to make one new..
Re: separating smps transformer cores
Part Scavenger, Mon Aug 21 2006, 01:58AM
Uh, no. My apologies for not knowing what I'm talking about, misunderstanding on my part...
Part Scavenger, Mon Aug 21 2006, 01:58AM
Uh, no. My apologies for not knowing what I'm talking about, misunderstanding on my part...
Re: separating smps transformer cores
Avalanche, Mon Aug 21 2006, 09:39PM
There's definitely 2 different kinds of glue they use...
If the transformer has the white epoxy, I've come to the conclusion that it isn't worth bothering. I guess you could unwind and rewind the transformer without dismantling it at all, which I might have a go at. The other kind of resin is brown and opaque, and seems to melt into liquid at <100 degC. I used to dismantle ordinary mains transformers as well, to get the steel laminates - they make great packing for lathe tools. I guess you could boil those transformers as well as they use the brown stuff!
Avalanche, Mon Aug 21 2006, 09:39PM
There's definitely 2 different kinds of glue they use...
If the transformer has the white epoxy, I've come to the conclusion that it isn't worth bothering. I guess you could unwind and rewind the transformer without dismantling it at all, which I might have a go at. The other kind of resin is brown and opaque, and seems to melt into liquid at <100 degC. I used to dismantle ordinary mains transformers as well, to get the steel laminates - they make great packing for lathe tools. I guess you could boil those transformers as well as they use the brown stuff!
Re: separating smps transformer cores
Marko, Mon Aug 21 2006, 09:56PM
hey, nothing to apologise, really
Regarding glue type all seemed equally difficult to me
Sometimes core isnt heavily glued to former but has just few drops of glue, adn that gets apart easily..
Marko, Mon Aug 21 2006, 09:56PM
Part Scavenger wrote ...
Uh, no. My apologies for not knowing what I'm talking about, misunderstanding on my part...
Uh, no. My apologies for not knowing what I'm talking about, misunderstanding on my part...
hey, nothing to apologise, really
Regarding glue type all seemed equally difficult to me
Sometimes core isnt heavily glued to former but has just few drops of glue, adn that gets apart easily..
Re: separating smps transformer cores
thom, Fri Aug 25 2006, 10:50PM
Price was not bad, $11 shipped for 10 disks. Ebay seller lau*** (he had lots of diamond tools for cheap).
thom, Fri Aug 25 2006, 10:50PM
Self Defenestrate wrote ...
I've had some luck with those ultra thin diamond dremel cut off disks, if you hold the core in something soft to ease the vibrations. Once the outside is done, I can usually boil and twist the cores apart. The inner rod part of the core is sanded a bit to flush everything up, and there you have it.
Quoting this again because I just received some diamond disks and they are very good at cutting ferrite. The normal cutoff disks are slow and crack the ferrite easily.I've had some luck with those ultra thin diamond dremel cut off disks, if you hold the core in something soft to ease the vibrations. Once the outside is done, I can usually boil and twist the cores apart. The inner rod part of the core is sanded a bit to flush everything up, and there you have it.
Price was not bad, $11 shipped for 10 disks. Ebay seller lau*** (he had lots of diamond tools for cheap).
Print this page