MC34063 spice sim issues

[reportingsjr]

Hello,
I tried designing a 24VDC to 3.3VDC converter using the MC34063 and I can not figure out what I have done wrong. After assembling the circuit and blowing up 3 chips I decided to try and simulate it in LTSpice (first time really using spice), but I can't seem to get the simulation working.

I am using the model found in the second post of this thread: https://www.eevblog.com/forum/beginners/simulate-mc34063-on-ltspice/ (it is ON Semi's model. I am actually using STM's chip currently.)

When I load the demo that comes in that zip I find it works ok, but when I tried building my circuit I get strange results. On the timing capacitor I am getting ~24V and no waveform with similar results on pretty much every other node in the path of the power. My schematic can be found here: https://dl.dropbox.com/u/9135021/psuSim.asc . Looking at the demo schematic and mine side by side, besides some output differences and value differences I can't see anything wrong with mine.

I do realize that the values for my schematic may not be exactly optimal, but it still looks like it should work even if the output is very choppy/noisy. The best I got when I actually built the circuit was ~2V on the output, which I don't really understand.

Thank you!

[c4757p]

I think it's the model. Your circuit works (not well, though) for me with this model.

[c4757p]

Looks like you have ground and feedback swapped. I switched them around and now I do get 3.3V out (though it oscillates horribly).

[c4757p]

As a starting point - 47pF across R2 cleans it up significantly, though it's still not perfectly stable.

Adding some external series resistance to C2 (68 milliohms) seems to take care of the rest of the oscillation. I just put in "perfect" values though, with no attention to variation, so it's probably right on the verge of oscillating again. Just take that as a starting point.

[Bored@Work]

I think it's the model. Your circuit works (not well, though) for me with this model.

As he wrote, it doesn't in real live. And with turning to SPICE he now doubled his problems. Fixing the real live circuit and fixing the simulation. I know on what I would focus instead of trying to fix two problems. And it ain't the simulation.

[reportingsjr]

Okay, I think I am understanding now.

So the output transistors are limited to switching below 40KHz to 50KHz. I saw that rating under f,osc on the datasheet and was confused as to what it meant. The chip is clearly rated to run up to 100KHz. The actual oscillator can run that fast and I suppose that could be considered the speed at which the MC34063 cycles, but the output transistors can switch faster than 40-50Khz. So if I increase the inductor in the buck setup to ~4x what is is now (around 120uH) then the time to peak current through the inductor should increase by the same about and the frequency that the transistors will need to switch at will be closer to 20KHz, a much better value for them. Excellent! Hopefully this fixes my problems, ordering a new part right now!

Thank you for all of your guys' help for now, except bored@work . haha

[reportingsjr]

Hmmmmm, alright. I am going to try and increase the inductor size for now. I do not believe their are any mistakes with the wiring since it is an exact copy of what is in the datasheets. Most of the blown up chips were caused by my own stupid mistakes (put chip in backwards, shorted output and gnd, placed 100uF cap across output without thinking, etc).

Unfortunately I'd rather not work in another more modern controller, because I already created this board and paid $20 for three of them. I'll try a couple more fixes before I do that. Too much money invested in this one already!

[darko31]

I don't want to make a whole new thread, so a quick question about MC34063.

Is it possible to make variable voltage supply (and output would be of course higher than input voltage) by changing the value of R1 and R2 resistors replacing them with trimmers, in step-up operation, given that inductor is large enough?

[kripton2035]

did you try to calculate your design online here : http://dics.voicecontrol.ro/tutorials/mc34063/
or here : http://www.nomad.ee/micros/mc34063a/index.shtml
*before* burning chip(s) or making pcbs ?