New bobbins needed for non isolated transformer based converters.

[Faringdon]

Hi,
The attached  (LTspice and PNG schem) shows a Two Transistor Forward and also a Buck converter. Both are converting 390Vin to 100Vout at 500W, and at 100kHz.
It is seen that the 500W Buck is not really useable, as it is likely to be so noisy that even its own control would be compromised. The 2 Tran forward has a leakage inductance in the transformer, and this reduces the di/dt in the switchs and diodes. Also, it reduces the peaks of the turn on spike currents. They both need snubbing, but the two tran forward needs far less. Also, if an artificial “stray” inductance is added to the Buck, then it simply causes way too high losses in the Buck’s semiconductors.
As such, we see that transformer isolated converters have another very important use. That is, Noise reduction when power is high. 
However, virtually all the offTheShelf bobbins for SMPS transformers are designed with mains isolation clearance/creepages in mind. This makes them too big, for those cases when the isolation spec is not strict. Also, a 1 layer coil will start and terminate on the same side of the bobbin….(for isolation purposes). This is not necessary for “less strict” isolation. As such, the new bobbin range should have bobbin pins placed such that a 1 layer coil would start at one end of the bobbin tube, and terminate at the other…no straggly awkward bit of wire being straggled across the bobbin for termination. Also, the bobbins should have entrance and exit “flaps” such that a coil , eg primary coil, can come on and off the bobbin without taking up room needed for the secondary  coil round the bobbin edges.
Also, bobbin pins should not be so weakly joined to the bobbin plastic. At the moment, any multi strand coil (for higher current) needs much heat when terminating  to the bobbin pin, and inevitably the heat melts the pin out of the bobbin plastic. As such, simple bobbin clasps are needed, and the transformers can then be soldered to the PCB directly. Or alternatively, soldered to a separate, optional  “Plinth”. The plinth would have metal pins very well embedded into the plinth plastic such that they don’t loosen when termination soldering is done.
So why do we not have these bobbins available to us?

[jbb]

I suggest you have another go on the buck converter with a Silicon Carbide (SiC) diode for D7. Could make a big difference to the turn on switching transient.

On bobbins… I dunno. I guess you get what you can get? Which is probably a function of what big customers have ordered in the past?

[Faringdon]

On bobbins… I dunno. I guess you get what you can get? Which is probably a function of what big customers have ordered in the past?

Thanks, i dont want to make this political, but big customers order big volumes from China, and get what China gives them.
I think the Western world has simply "turned off" to any kind of innovation in  this sector. As such, the bobbins spoken of dont exist.

I suggest you have another go on the buck converter with a Silicon Carbide (SiC) diode for D7. Could make a big difference to the turn on switching transient.

Thanks, yes it will improve things, but still, as i am sure you appreciate, the other benefits of the leakage inductance in the 2 Tran forward are still badly missing in the buck.

[tggzzz]

On bobbins… I dunno. I guess you get what you can get? Which is probably a function of what big customers have ordered in the past?

Thanks, i dont want to make this political, but big customers order big volumes from China, and get what China gives them.
I think the Western world has simply "turned off" to any kind of innovation in  this sector. As such, the bobbins spoken of dont exist.

And yet the point you make is completely political, without any other content.

[gnif]

"Thanks, i dont want to make this political, but"... proceeds to make it political.

Faringdon has been given a 7 day break for again breaking the forum rules with regards to politics.

[jonpaul]

RE  Farrington.

I hope that EEVBLOG reviews the  1000s?) of   posts,  they will see some other issues.

j

[tggzzz]

"Thanks, i dont want to make this political, but"... proceeds to make it political.

Faringdon has been given a 7 day break for again breaking the forum rules with regards to politics.

Thank you.

[Gyro]

I feel for the edaboard forum members! Not that they aren't already fielding most of the vaguely technical questions anyway. He mostly seems to be reserving eevblog for general troll threads and political assertions these days.

[Simon]

RE  Farrington.

I hope that EEVBLOG reviews the  1000s?) of   posts,  they will see some other issues.

j

We are well aware thanks! no one is perfect on here!

[EEVblog]

RE  Farrington.
I hope that EEVBLOG reviews the  1000s?) of   posts,  they will see some other issues.

We are well aware thanks! no one is perfect on here!

We all post something that can be deemed "political" from time to time, that's just the nature of forums, you can't micro-police it. But some users are just so darn persistant at it thta they ultimately have to be temp banned to send a message, or just outright banned because they will never stop.

[tggzzz]

Well, gnif suggested Farmington should take a hint. I doubted that would happen and, uncharacteristically, kept quiet.
https://www.eevblog.com/forum/chat/your-best-piece-of-electronics-engineering-equipment/msg5395682/#msg5395682

Sometimes it is a shame to be right.