Showed a neodymium magnet to ferrite transformer, now no workie?

Thread Starter

Hamlet

Joined Jun 10, 2015
313
I built this thing, was watching the current and voltage, when I laid a hard drive magnet
on top of the "39" ferrite transformer to see what it would do... Amp went up, voltage down,
and now, with the magnet removed, my circuit no longer works...

I checked the pins of my transformer for ohm/continuity, all looks good, diodes good,
transistor replaced, but still no worky.

Did I screw up the ferrite somehow?
 

ronsimpson

Joined Oct 7, 2019
684
While the magnet was on the transformer, it saturated the transformer core. This is a non working condition.
With the magnet removed, the transformer should return to a good state.
You may have killed the MOSFET or diodes.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
Thanks, I'll check again.

This is the second ferrite transformer that is giving me trouble. The first one was fine, but after
an extended soak test, the transistor burned up, and after replacing it and the diodes, it never worked
again. In placing the magnet on my second build, I was attempting to see if the 1st transformer
somehow became goofed up with a surge of current thru the fused transistor via the thick primary winding.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
I replaced everything, no cross continuity on transformer, no voltage, but about 300kHz,
and the current looks about right... Replace bjt with mosfet, shorted 2k2, and no volts,
but 350kHz (I used a mosfet before the troubles, and I know it works with it.)

I'm going to tear down the transformer and see if the secondary is shorted.

schematicferrite trans problem.JPG
 
Last edited:

dendad

Joined Feb 20, 2016
3,535
The 1N4007 diodes are not the type to use in the secondary side here.
They are a bit slow, so it may be worth looking at some high speed diodes for a start.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
Yes, the diodes are slow, replaced with some fast ones, no difference. I just tore down the
transformer, no indication of heating. I just now rewound it with a few dozen turns on
the secondary, 20t + 20turns on the primary side, installed it, and nothing. Used a smaller
cap, and saw a few millivolts, with current at 0.5A. Tested freq. and saw about 40kHz.

My summation is that it is oscillating at the primary side, but no magnetic connection with
the secondary, which leads be to believe whatever mystery ferrite this transformer is made from,
I've somehow scrambled it's structure, and it is no longer magnetically conductive.
I tested the core with a leaf of soft iron, and it doesn't have any residual magnetizm that I can detect.... Baffling.
 

dendad

Joined Feb 20, 2016
3,535
Can you post a few pictures of you setup please?
How are you measuring things, do you have an oscilloscope?
If the primary is oscillating, at what level is it running?
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
Can you post a few pictures of you setup please?
How are you measuring things, do you have an oscilloscope?
If the primary is oscillating, at what level is it running?
My phone is tied to the wall with string; I don't have a 'scope, just a DVMM.

I don't know what you mean in the last part, "about what level it is running"?
 

dendad

Joined Feb 20, 2016
3,535
Is the primary switching between 0 and 12V, or only changing by a volt or so.
It is pretty hard to debug this sort of thing without a 'scope. They are often referred to a "Truth Meters", and it is a great help to see what is going on.
I would suspect the switching transistor. Putting a FET there could have killed the FET in question too as there is no over voltage protection on the FET gate.
How do you come up with the 350Khz figure?
And I do not understand your....
My phone is tied to the wall with string;
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
I don't have a smart phone with camera handy like nearly everyone else. I already have
6 computers, have to draw the line somewhere...

I used my digital volt meter, on the Hz setting. It might not be right, but I get relative readings,
for what it's worth.

I think the ferrite is borked, but that seems unlikely.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
Curious Hamlet; are you planning to repeat again the trick of the neodymium magnet with a new transformer?
I'm tempted, if it helps me get to the bottom of this mystery...
I just wanted to build this damn project, almost there, soak testing,
and then all this weirdness started happening.

Ferrite isn't supposed to saturate and stay that way, is it?
I even ran it over my electromagnet/demag machine, but
didn't seem to help convince the core to straighten-out.

I know if you over-amp a PM motor, you can demag the magnets,
maybe something like that happened to my first core?
 

Alec_t

Joined Sep 17, 2013
11,302
I used my digital volt meter, on the Hz setting. It might not be right, but I get relative readings,
Most DMMs assume a sinusoidal waveform. The circuit shown has a more or less sawtooth waveform, so your frequency measurement could be way off.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
Most DMMs assume a sinusoidal waveform. The circuit shown has a more or less sawtooth waveform, so your frequency measurement could be way off.
Thanks, I'll make a note. Someday I need to find a scope. I'm like a missionary without a bible.

Stayed up late and noodled around, discovered ferrite has not gone sour,
but maybe a lousy bunch of transistors, huh. Any suggestions for a replacement?
 

Alec_t

Joined Sep 17, 2013
11,302
maybe a lousy bunch of transistors, huh. Any suggestions for a replacement?
With only ~20 turns of 1mm diameter wire (as per your circuit) for the primary coil, if the oscillator fails to oscillate for any reason the transistor/FET current is limited only by the power supply, the fusing current limit of the 1N4007 and the very low resistance (a few milliohms) of the coil, so could be tens of Amps (briefly). That's way beyond the capability of a TIP31 and many FETs. The higher the current rating the better for any replacement.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
313
I finally got my circuit working, then I was blowing the triac section (omitted from schematic).
Installed some freewheel/snapback/flyback diodes, and all is well. Still haven't got to the bottom
of the ferrite problem, maybe tonight. Thanks for all your help.
 
Top