Mr. Al, Thanks for the equations. Especially the one for instability.
But C's circuit is running in discontinuous mode so I don't think that is his problem. Having said that, I don't know how to calculate the inductor size for discontinuous mode.
C, I think the current rating for your inductors is to low, but that is probably not why it is oscillating. Do you have the RC across the FET? It will ring like crazy without it when the inductor current goes to zero. The high frequency might be this. I'm surprised it works like this.

 

Mr. Al, Thanks for the equations. Especially the one for instability.
But C's circuit is running in discontinuous mode so I don't think that is his problem. Having said that, I don't know how to calculate the inductor size for discontinuous mode.
C, I think the current rating for your inductors is to low, but that is probably not why it is oscillating. Do you have the RC across the FET? It will ring like crazy without it when the inductor current goes to zero. The high frequency might be this. I'm surprised it works like this.

Yes, I did install the RC across the FET, with the exact 2k2 and 100pf values specified. LTspice shows a 39Khz waveform at the FET's gate, which is not what I'm observing in my scope. I have a slight suspicion that something might be going on with the way the 2N3904 is connected to the CV pin at the 555... perhaps it also needs an RC filter somewhere in there, since I'm reading a clean square wave output from the 555 when I disconnect it from the FET, but it's noisy as hell when it's connected as it should. I could be wrong, of course, but I think that means that whatever's wrong with the circuit, it has got something to do with its feedback into the 555...

 

I learned how to make spice do inductor saturation (finally). One of the problems with using the little inductor is that it rings like crazy when the inductor current goes to zero. You can see this in the simulation by scoping Vd. This is worse when the inductor saturates (as your will as it is to small). I think it is better with 1K and 500pf. You might try that. But it also kills the efficiency. It may be worse than it shows in spice due to layout etc.

 

I learned how to make spice do inductor saturation (finally). One of the problems with using the little inductor is that it rings like crazy when the inductor current goes to zero. You can see this in the simulation by scoping Vd. This is worse when the inductor saturates (as your will as it is to small). I think it is better with 1K and 500pf. You might try that. But it also kills the efficiency. It may be worse than it shows in spice due to layout etc.

Yes I can see that the ringing is more severe in this new model of yours, both in number of cycles and amplitude. Apparently it rings for at least 5 µs at an average current of around 100 ma... See attached image for comparison between your previous circuit (left) and this newer one (right)... I'll try the new values for RC that you're suggesting and then get back to you. Thanks.

 

My bad C. It is not that it is to small in value, but in current. That is why the inductor saturates and the current is so much higher in the second model. I couldn't model that before I learned how to enter the equation in the inductor.
It is the low value that makes the current higher than the circuit with the higher value inductor, but that's ok if you have an inductor good for the high current. The circuit where the inductor current goes to zero causes the ringing and requires the snubber. Chances are this high frequency noise is getting everywhere.

 

My uneducated guess is that the 555 is being overclocked at each burst, and that's why it's running too hot, since the datasheet says that 500 Khz is the top frequency for this device... but I could be wrong, of course.

View attachment 74899 View attachment 74900 View attachment 74901 View attachment 74902 View attachment 74904 View attachment 74905

Still trying to figure why the 555 is running hot...
Those bursts of high frequency be causing the 555 to get hot trying to charge and discharge the FET's gate capacitance. So why are there high frequency bursts? Others have made good suggestions. Here is another one (hopefully as good ):

It may be a coincidence but the burst rate looks like about 8 mS -- This would be the frequency of a full wave rectification of a 60 Hz power line. I wonder if your power supply voltage is sagging at the high peak currents while charging the inductor. If it is, then you need to add a large electrolytic cap on the output of the regulator. I would start with a 1000 uF cap and see what happens.

 

My bad C. It is not that it is to small in value, but in current.

Well... remember that I'm using three 100 µh inductors in parallel, so as to get the 33 µh I need ... the rated current of each is around 1 amp, according to their datasheet. Shouldn't that be enough?
Maybe Richard's right, my power supply might be too small... and that's what I'm going to look into tomorrow.
Thank you both for your insight.

 

When I look at the inductor current it looks like a little over 3 amps and the inductors are 1 amp each.

 

When I look at the inductor current it looks like a little over 3 amps and the inductors are 1 amp each.

I'll check into that, too... thanks

Update: I just checked your last circuit in LTspice, and although the inductor's current has an RMS of 1.78 Amps, it runs at a frequency of 3.58 KHz with peaks of 6.5 Amps! So maybe the three inductors in the circuit are saturating and ringing horribly... I'll see if I can get myself a bigger inductor and do more tests... so stay tuned, it might take me a week or so, but I'll get back to you all with the results. Thanks again.

 

it runs at a frequency of 3.58 KHz with peaks of 6.5 Amps!

I thought the frequency was more like 30 KHz. Is the 3.58 KHz a typo?

 

Sorry, yes, it was definitely a typo, the entire circuit runs at 47 Khz (don't know what I was thinking when I wrote 3.58 KHz), but the peak amps is definitely 6.5 at the inductor....

 

You can change the 3 (amps) in the inductor equation to see the effects of it saturating. At start up the current is higher so take a look there as well. I think the one I linked earlier will work ok. I don't know as this is the problem but it is not to good for the FET either.