Why doesn't this work? (self-oscillating circuit) / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Why doesn't this work? (self-oscillating circuit)

Move Thread

LAN_403

Dr. Dark Current

Sat Apr 11 2009, 11:09AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

I think you all know the "single transistor flyback driver" which produces more heat than output.

Based on the topology, I wanted to create an efficient Class-E (or at least near class-e) driver for the TV transformer. This is the circuit I came up with, note that the resonant capacitor is not shown as it killed the operation completely (see later):

With the circuit as shown, I tried both polarities of the feedback winding. In one polarity it just shorted the supply, in the other one nasty HF oscillations started appearing and the gate driver heated up (gate signal was anything else but square wave; there was no visible output from the xfmr). When I tried to add the resonant cap, the oscillation stopped completely.

Any ideas?

Hon1nbo

Sat Apr 11 2009, 02:26PM

Registered Member #902 Joined: Sun Jul 15 2007, 08:17PM
Location: Pacific Northwest USA
Posts: 1042

reason why one polarity of feedback shorts your powersupply, and forgive me if any of this is inaccurate, is that it must be coordinated with your primary - after current conducts via the feedback into the transistor stage, which turns on the primary, then it must be designed such that the current induced in the feedback from the primary must oppose the normal current flow, therefore shutting off the transistor, so if you improperly connect it it would keep the transistor on at all times, and current is being increased rather that being decreased... but again, what I stated might be wrong (I've only messed with the design from Powerlabs using a 2N3055, which I used for a lot of things just fine)... however, like you stated they more or less produce heat that output : - P

but also, is there anything else connected to that TC4422 Mosfet Driver? according the datasheet, it has it's own Vin, etc... I dunno if you just left that out of the schematic, but maybe something there is going on

it also seems that the frequencies used in the graphs on the TC4422 datasheet indicate a higher frequency that what I thought flybacks operate at... at least efficiently, maybe that is causing issues with the HF oscillations, but i'm not sure at the moment, am in a rush to get out of the house so this is just to toss a few things out there? - it can be found HERE

MRacerxdl

Sat Apr 11 2009, 02:56PM

Registered Member #989 Joined: Sat Sept 08 2007, 02:15AM
Location: SÃ£o Paulo, Brazil
Posts: 476

I did a similiar thing once, but I figured out that a startup oscillator helps, Here are my Cold Cathode Fluorescent Lamp Inverter Schematic:

Dr. Dark Current

Sat Apr 11 2009, 03:08PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

>DaJJHman: I know that only one polarity will work. But the problem is that the oscillations are in the MHz range and there is absolutely no output from the transformer.

And yes, the supply rails for the TC driver are omitted and of course connected, to +15V.

P.S. The windings should actually be polarised the other way so that the feedback is positive, when the transistor turns on it causes even more current to the base, holding it ON for a while, until the core saturates and the base current falls so low that it's no longer able to hold the ramping collector current, then a flyback pulse happens which is seen with negative polarity on the base, holding the transistor OFF during the flyback pulse.

Kolas

Sat Apr 11 2009, 10:48PM

Registered Member #102 Joined: Thu Feb 09 2006, 08:15PM
Location: Raleigh, NC
Posts: 169

I don't know, it almost seems like you'd need to be using inverting logic for it to set up real oscillations.

Dr. Dark Current

Sun Apr 12 2009, 11:00AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Kolas wrote ...

I don't know, it almost seems like you'd need to be using inverting logic for it to set up real oscillations.

But I can invert the feedback... ?

I think the fact that it works with bipolar transistors is because of their very nature and electrical characteristics, as they must de-saturate just before the flyback pulse happens. The gain of my circuit is practically infinite so this can not work...

Proud Mary

Sun Apr 12 2009, 11:44AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Have you calculated what the frequency of oscillation should be, lad?

Dr. Dark Current

Sun Apr 12 2009, 11:57AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Harry wrote ...

Have you calculated what the frequency of oscillation should be, lad?

In the original circuit it is probably mostly determined by the transistor de-saturating, or by the base current falling too low, probably a combination of both. The OFF period lasts as long as feedback voltage does not fall back to cause conduction of the transistor.

In any case, the voltage on the FB winding cannot reverse while the transistor is ON, so this is the reason why it can never turn off with my circuit.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob