EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: asgard on July 16, 2015, 02:30:29 pm
-
Here is an update of a project I have kicking around for a while, the BLiVIT. The point is to have a smart battery manager, with galvanic isolation on output so you don't fry your perfectly good scope (if it works I might even submit it to the FDA for medical approval :).
Enclosed are the schemat and a 3D render of the board. What most concerns me, is that whether it is a good idea to have the isolation after the regulator, or would be advisable to have it after the isolation. Not the use of a USB-compliant current limit switch directly before the output. Previous version placed regulation after the isolation transformer, and used a (pricey, and decreasingly available) Maxim part as the power manager. Too often I have released the blue smoke in the $5 Maxim device when accidentally throwing 19V into the DC jack, hence the linear regulator just for that. Is it safe to assume the USB will be sufficiently constrained and therefore not need any protection regulation?. Prior versions had electrolytic SMD cans at C3 and C4, but I hope now the ceramic 0603 caps will do o.k. there. Advice?
I hope the board layout looks good, at least.
-
Normaly you would feed the isolated output back to the regulator at the primary side via an optocoupler (driven by a TL431 for example). Or you could place a regulator at the secondary side, if the DC DC converter provides a fairly stable voltage. But this approach would decrease the efficiency of the voltage regulation.
-
What about the notion of completely relocating the final DC-DC converter/regulator after the isolation stage?
-
The normal way would be having voltage regulation and isolation in one stage, ususally with an optocoupler for feedback. For low power this usually is a flyback converter, but a forward converter with an extra induktor is possible as well.
At least the ratio for the isolation could be choosen that way that a buck or boost regulator is enought. Even a low drop linear regulator mitgh be an option if the voltage does nor vary that much - e.g. having a flyback converter without feedback from the secondary.
If there are up to 19 V at the input, the poor 7805 at the input might overheat.
-
The point is to have a smart battery manager, with galvanic isolation on output so you don't fry your perfectly good scope
Why would I fry my scope when charging batteries? I don't get it... *confused*
-
What about the notion of completely relocating the final DC-DC converter/regulator after the isolation stage?
Why do you want to add an additional regulator, when the isolating flyback can do that job? You would need a flyback controller with a relatively stable output voltage without having any feedback, because the second regulator will need some usable input voltage (not too low, not too high). The only case where a second regulator would make sense, is a linear regulator for low ripple. But to limit the losses of the linear regulator, you would use a regulated flyback with feedback. Since your circuit includes a battery you should make everything as efficient as feasable. And that's the isolating flyback with feedback.
-
How about this, for example?
As to why would a charger fry your scope - I think Dave has several times mentioned the problem of earth ground reference with bench instruments. If I have a DUT connected to the output of the BLiVIT with no isolation, for example, and the incoming power is from a wall-wart (or even a badly designed hub on the USB) then a ground loop exists from the scope (or that poor old multimeter :) probing the DUT, back through the power supply rail to the ground reference of the supply. There is potentially (pun intended) high current loading that can be presented to the instrument. Bang.
J.R. Stoner
BDG
http://www.bifrost-group.com (http://www.bifrost-group.com)
-
How about this, for example?
Much better :-+