This threads smps is a discontinuous flyback - the diode needs a fast turn-on, and although the turn-off (reverse recovery) should be as fast as possible, that is alleviated by the diode current naturally passing through zero at turn-off as well as the reverse voltage just resonantly increasing. The secondary winding is not 'push pull' in the context of general use of that description - both diodes conduct at the same time - one to bump the positive rail up, and the other to bump the negative rail more negative.
The scope plot in post #79 shows a 13.5us period (75kHz), and diode on time of about 6us (about 44% duty cycle). When the diode stops conducting, the diode voltage swings resonantly, as does the FET drain voltage during that portion of the period. The FET is conducting during about the 4us that the diode waveform is flat, and then the fet primary energy flies back to the secondary winding for the diode to pass.
This smps was the topic of another thread here as well: YH11068A 10-32V to 45-390V dc/dc - guess the control ICs
The other smps linked in post #74 is an unregulated isolated pp style with 37kHz ripple, that has the advantage of allowing many fixed secondary voltage levels to be set up (assuming a fixed nominal 12V input level) by either using a full-bridge or doubler rectifier output, and using the range of secondary taps.
The scope plot in post #79 shows a 13.5us period (75kHz), and diode on time of about 6us (about 44% duty cycle). When the diode stops conducting, the diode voltage swings resonantly, as does the FET drain voltage during that portion of the period. The FET is conducting during about the 4us that the diode waveform is flat, and then the fet primary energy flies back to the secondary winding for the diode to pass.
This smps was the topic of another thread here as well: YH11068A 10-32V to 45-390V dc/dc - guess the control ICs
The other smps linked in post #74 is an unregulated isolated pp style with 37kHz ripple, that has the advantage of allowing many fixed secondary voltage levels to be set up (assuming a fixed nominal 12V input level) by either using a full-bridge or doubler rectifier output, and using the range of secondary taps.
Last edited:
Thanks trobbins
Given how simple the circuit is for these PSUs, I am almost tempted to build a heavier duty version (I have some 600V IGBTs somewhere that may suit)
However the whole fly back transformer/coupled inductor subject is vast in itself! I found some pot cores (3C85 ferrite, 0.5" diameter by about an inch E cores) and I spent the last 2 hours trying to find reference to the cores, which I find are discontinued Phillips.
Anyway I think the core material is vaguely suitable, but perhaps I am better off trying to make a simple boost inductor using them?
It would seem far simpler a design exercise for a simple booster (and I probably have some design theory in my uni folders to help out)
Given how simple the circuit is for these PSUs, I am almost tempted to build a heavier duty version (I have some 600V IGBTs somewhere that may suit)
However the whole fly back transformer/coupled inductor subject is vast in itself! I found some pot cores (3C85 ferrite, 0.5" diameter by about an inch E cores) and I spent the last 2 hours trying to find reference to the cores, which I find are discontinued Phillips.
Anyway I think the core material is vaguely suitable, but perhaps I am better off trying to make a simple boost inductor using them?
It would seem far simpler a design exercise for a simple booster (and I probably have some design theory in my uni folders to help out)
faulty converted of faulty me?
Is the converter from the link above supposed to show DC output with no load?
I tried one with the jumper set for 110V without any load and with a 2K resistor across DC output (which should draw 55 mA) but the output voltage in both cases seems to jump over 600 VDC (the end of the range for my cheap meter).
I powered it with a 12V 1.5A adapter so perhaps it doesn't have enough current on the input to regulate correctly (though 1.5A should 'mathematically' be sufficient for only 55mA @ 110V output).
Or am I just testing it the wrong way?
Is the converter from the link above supposed to show DC output with no load?
I tried one with the jumper set for 110V without any load and with a 2K resistor across DC output (which should draw 55 mA) but the output voltage in both cases seems to jump over 600 VDC (the end of the range for my cheap meter).
I powered it with a 12V 1.5A adapter so perhaps it doesn't have enough current on the input to regulate correctly (though 1.5A should 'mathematically' be sufficient for only 55mA @ 110V output).
Or am I just testing it the wrong way?
Last edited:
HAve somebody already used this 500W Inverter Boost Board Transformer Power DC 12V TO AC 220V 380V Car Converter | eBay , any goods? thanks
DC input AC output.
I was looking at the PCB and the centre tapped VCC and thinking - that's a half bridge or a dual inverter alternating synchronous boost so the output is very likely AC. Then re-read the description.
If you have a look it's easy to just rectify - the rectification will be at the switching rate so choose your diodes carefully. Then use a cap that matches your current Vripple..
However other than a possible sense on the primary side, I couldn't see any voltage regulation feedback for the output - it literally is straight out of the transformer.
In the brief look I couldn't see if it was isolated so I would treat it at as non-isolated (just use an isolated 24V source).
It's not clear then if the AC voltage is peak or RMS..
TI and LT do nice sync boost chips that allow active rectification (virtually no voltage drop as no diodes) using mosfet switching.
I've used them. They work well.
Get the one with the heatsink attached - the fets are in a stupid place and it's almost impossible to screw heatsinks to them. I actually had to remove the heatsink and FETs as one unit to replace them when I blew one of them.
Use UF5408 or HER308 etc. Standard diodes will fail short and take out the FETs if the supply will dump enough current into them.
Get the one with the heatsink attached - the fets are in a stupid place and it's almost impossible to screw heatsinks to them. I actually had to remove the heatsink and FETs as one unit to replace them when I blew one of them.
Use UF5408 or HER308 etc. Standard diodes will fail short and take out the FETs if the supply will dump enough current into them.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.