High Side Fet Driving

[ACharnley]

I need the efficiency.

And a high working voltage.

Attempting to have my cake and eat it!

 

[dknguyen]

What efficiency? Did you not understand the part where it just conducts through the MOSFET body diodes while the rest of the MOSFET does nothing? The circuit behaves the same if you removed the resistors and zeners and just left the gates floating. Hell, it's behaviour doesnt change if you removed the entire MOSFET as long as you left the parasitic diode there.

The circuit is not doing the equivalent function of a diode-bridge, It IS a diode bridge with a bunch of superfluous components. Those MOSFETs are not operating as ideal diodes.

EDIT: Oh wait, after staring at it a few more times, they are. nvm.

 

[ACharnley]

LTSpice says it'll work, so I'm going with it.

The PIC ADC monitors AC1, AC2, STORE and has three states, OFF, AC1 on, AC2 on.

I love the simplicity of it, plus no big resistors/zener to handle gate flow at 100V as per the earlier schematic.

As per earlier discussion I don't require it to work at 100V only to survive. It's only enabled for a 5V supply which is why the MIC5019 works. A higher working voltage would require an opto.

So long as the PIC is accurate and fast enough. I'll find out....

 

[ACharnley]

Hmm, question. Q43 has a negative charge on the source (Ac1) which would pull -100v below the gate. I suspect that'll be a problem, sorted if I use an opticoupler instead with a resistor to drain the gain.

If the opticoupler goes between pin 3 and 1 I'll need one with the collector-emitter breakdown to work at 100V?

I could keep the MIC5019's for the low side FET driving but it probably makes more sense economically to go with a 4 channel opticoupler.

Does this sound correct? My brain is fried!

 

[ACharnley]

Can anyone advise if this application is correct?

Since the transistor is isolated I don't need to consider the col-emi breakdown?

So all to consider is the strength from the opticoupler can't rise above the VGS rating?

 

[ACharnley]

And final question, regarding the 4N25, the LED creates a current at the base which allows the collector to flow. I'd expected the LED to create a voltage between pin 4 and 5 but the use of a transistor indicates otherwise?

 

[ACharnley]

OK I got it - VOM1271 is what I need here.

 

[ACharnley]

Can anyone clarify if any over-voltage resistors/zeners are required at the gate(s)?

VOM1271 are quite expensive, any recommended alternatives? I couldn't find anything else really, there's a Toshiba one but turn on/off is awful in comparison.

 

[ACharnley]

My PIC power source isn't tightly regulated, swinging from 2.4-5V and this has an effect on the opto's performance. I'm thinking about using the PIC's PWM's to keep the current constant. What do you think?

Picture shows the difference between 5V and 3.3V