most unreliable power converter circuit?

[coppercone2]

out of all the DC DC converters what is the most unreliable circuit topology that is always giving you agita?

have you noticed any trends? i think its flyback

[dustooff]

flyback converters running free are the worst 

Trying to persuade a multi winding output to behave as its creator intended, has contributed to the grey hair count.

One thing I have always wondered, is it possible that small transformers particularly multi output that are manufactured down to a price do some aging where they no longer match their original impedance and magnetic performance?

[T3sl4co1l]

In what respect? Design? Repair? Production?

I guess it's free reign to post relevant memes then,

Tim

[coppercone2]

flyback converters running free are the worst 

Trying to persuade a multi winding output to behave as its creator intended, has contributed to the grey hair count.

One thing I have always wondered, is it possible that small transformers particularly multi output that are manufactured down to a price do some aging where they no longer match their original impedance and magnetic performance?

yo I swear there is something not in the blue prints for this bull shit

The circuits are so low cost, and required test equipment is so high cost, no one will ever find it. And if they do the product count is so high every measure will be taken to hide it under the rug


actually who am i kidding, they probobly can't find it.

I feel like thinking about transformers is unhealthy lol, to me it represents balding

[ganfetsic]

The half bridge is a swine, because of the problem of current mode where it goes unbalanced.
Any voltage mode converter is a PITA because if you add load capacitance, it can go well unstable.
The pushpull is a pain because vdrain goes up to 2xvin when the opposite fet turns on, and so youre snubbering will often dissipate a lot due to the high voltage it sees. (unless a complex snubber is done)
SMPS with sync rects is also challenging when there is much secondary side leakage L to ring with the sync FET cds.
Any 400V approx input SMPS which uses bootstrapping (gate drive)  is a pain as they require beautiful layout, which often cant be achieved due to winding tracks around big heatsinks etc.

[coppercone2]

the synchronous rectifier is like disguising a uranium centrifuge as a merry go round for young engineers looking for a cool project

I think that one threw 'treez' into orbit


Look at this neat way to reduce heat sink requirements, your boss will love you for it 

I had someone propose that one to me before. I immediately said its under my skill level, and the senior engineer said hell no (and I usually accept challenging projects). I remembered those circuits where they put capacitors around a rectifier on some test equipment, I also renembered some ranting I was hearing about EMC. I figured it was likely to result in desk banging and deranged psychos scribbling on white board.

[ganfetsic]

yes, whenever doing sync rects, always bring up with the sync rects disabled at first, is a good idea.

[coppercone2]

Do you think it could do class B EMC on a medium power product (~1kW)?


That was the test bed, then it was supposed to move up to ~30kW, But I think that one had reduced EMC requirements, maybe.But it had to be functional for a while on the smaller product, even if it barely did anything.

[jbb]

Not the converter per se, but I think you’ll appreciate it.

As part of an undergraduate university project the class had to make a buck-boost converter. They were provided with a 2-sided board and a library of parts.

One group approached the Teaching Assistant (TA) and said their converter was t behaving right; it would cut in and out of experience wild voltage transients at random.

The TA turned it over and half the components fell out. Turns out they had just put the wires through the holes, bent them over and cut them off. Almost no solder in sight.

[coppercone2]

sounds like that was a future boeing leader. they figured out how to save money on the process

but its almost accurate. if you can't solder, that means you can get like 15 meetings in before someone gets a contractor to do it. That should be at least 2-3 weeks of busy calendar. That is like counting how many logs are on your raft, for where those people end up.

[gnif]

ganfetsic was banned, this is farringdon again

[coppercone2]

lmfao

i remembered it too that he was talking about them a while back

say synchronous rectifier 3 times while looking at your bathroom mirror and he shows up




the circuit diagram looks like a kind of odd trap, when you step on the green pressure plate it grabs your foot

[ArdWar]

Geez, what's with the flyback slander in this thread? It's so simple hardly anything could go wrong. It's just unfortunate that its simplicity making it the prime target for cheap crap that corners got one cut too many.

[coppercone2]

I think their terrible, even the ones inside of CRT monitors, they totally got me off making any of the HV projects a long time ago when it was popular. They all use flyback and I hate all of them

I thought I could make a lifter and a singing arc and that spending all that time taking CRT apart on my lunch was worth it. NOPE 

[coppercone2]

Tim

the data sheet drawing for your chip is rather strange

[Kim Christensen]

ganfetsic was banned, this is farringdon again

Date Registered: Today at 09:23:58 am
Last Active: Today at 03:47:08 pm

That was fast!   
Your Farringdon detector is top-notch!