toroid for car amp smps

[tennisballg]

I would say that they are ferrite, Iron powder usually comes with a color code, it's a shame though that you don't know the permeability of the material...maybe someone on the site could give you some pointers on finding it out.

I guess this is as good a place as any to ask about another power supply I am trying to fix. people tend to respond more here anyway. My friend brought me his amp. It works great on my bench after I replace the mosfets, but any time my friend pops the amp back in his car, it only lasts an hour or so before the mosfets explode.

This has confused me, because it works so nicely on the bench, checked the waveforms, everything looks great.

Digging deeper, though, I noticed that the inductor filtering the input from the +12V battery is shorted out on itself, so it's basically just a jumper. Is this minor, or is it the cause of the problem?

My thinking is that the inductor calms the instant inrush when the amp is powered on, and no inductor means blown mosfets. Sound reasonable?

[sss]

Originally posted by tennisballg
My thinking is that the inductor calms the instant inrush when the amp is powered on, and no inductor means blown mosfets. Sound reasonable?

imho i think that this inductor got nothing to do with it , this inductor is eliminating the switching noise so there wount be interference with other equipment .

[subwo1]

.

This has confused me, because it works so nicely on the bench, checked the waveforms, everything looks great.

It could have to do with things like dash board temperature, speaker load, and 13.8v power in the car vs. 12 on the bench, among other things.

[Eva]

sss:

This looks like a common mode filter so the toroid is probably made from high permeability ferrite

Anyway, It's easy to find out the optimum volts/(turn*Hz) value for any core by aplying constant voltage pulses of controlled lenght to a known number of turns and looking for saturation in the current waveform while progressively increasing pulse width [allowing the core to reset between pulses]. That kind of test is easier to perform with the help of some PWM IC like SG3525A, a switching mosfet and a clamping R and D network

[sss]

thanks again Eva
i found a way to know if its ferrite or not by trying to brake it , most of the time ferrites are fragile
so here it is

[BeanZ]

I am a power engineer and I wanted to get in on this forum. So posting this message will subscribe me. Also, I have some soft magnetic materials in my own stash that could be available if someone wants. I have some toroid ferrites, toroid MPP cores, Toroid High Flux MPP, and some Kool Mu. I have a boatload of the ferrites. If anyone is interested send and email and I will give you the information you need on the cores or on SMPS design. That's what I do! Check out my Class-T amplifier section with off-line SMPS for everyone to see.



Class-T Integrated Amp?

[Eva]

sss:

Poor little ferrite core. You are evil, did you know?

[tennisballg]

Umm....you DO have more where that came from, right?

[sss]

i found a junk store where i can get those for 1$

now i'm trying to wind the primarys and its kind of hard to make only 4 turns on that big core , i'll post some pics later

[wrenchone]

Some comments on earlier posts - 1.5W in a mosfet or two is ok, since it is easy to heat sink the things. 1.5W in a transformer winding of only a few turns is definitely not ok, as toroidal transformers have difficulty getting rid of heat. With 1.5W dissipation, the primariy windings will be really hot. If all one has is 22 gauge wire, the thing to do is to use enough parallel strands to get the dissipation down. With only a few turns on the primaries, this will help by increasing the core coverage, thus reducing leakage.

Also, it is not a sin to use a few more primary turns, even if the core is large. In fact, this may be a preferred approach with an unknown ferrite core, especially if it came from a common mode filter. Sometimes these filters are wound using power-grade ferrites, but they are also wound using high permeability ferrites. These ferrites aren't the best as far as core loss or high flux capability is concerned. However, using more primary turns will lower the flux density, which will reduce core loss and keep the core from saturation.

Someone a few posts earlier inquired about putting an inductor in series with the primary center tap. This is not a good idea at all unless there is some overlap between switching phases. 2-phase switcher chips(3524, 3525, 494, etc.) incorporate some minimum amount of dead time between switching phases to avoid overlap, which in a conventional push-pull converter will result in "shoot-through current". This happens if both phases conduct simultaneously, and it causes the equivalent of a short circuit across the input supply. This will overheat the switches, and could damage them if severe enough - hence the use of dead time. During the dead time both phases are off. If there is an inductor in the transformer center leg, there will be a large inductive kick during the dead time that could cause the switches to break down.

If one is dead set on using an inductor in the transformer center tap, an interesting approach may be to use a TL494 switcher chip and invert the outputs. The dead time now becomes a region where both switches always conduct. One could then set the chip for minimum duty cycle (inverted to maximum duty cycle), and modulate the overlap by varying the amount of dead time. Since the TL494 has a dead time control pin, this is easily done. What you have then is a 2-phase boost converter. The overlap between phases charges the inductor, and it discharges through the transformer into the load when only one of the primary switches is on.

For approaches other than the one above, the place to put an inductor is between the output diode and the filter capacitors. If the inductor is relatively small, it will serve to reduce the peak current into the capacitors, hence through the diodes and switches. The output will still peak charge to the maximum value of the secondary voltage. If the inductor is sufficiently large, it will serve as a low-pass filter and allow one to control the secondary voltage as a function of duty cycle. This is described exhaustively in application notes.
For the adventurous, one can take the aforementioned TL494 chip, use it as the controller for a push-pull converter with an LC output filter, and keep the output voltage relatively constant by modulating the dead time as a function of input voltage. This open loop approach is a way to get some output regulation without having to worry about the potential stability problems of a closed-loop system.