AMX Schematics & Layout

[Carl-NC]

OK, thanks.

[Tinkerer]

For higher inductance coils it would be an advantage to have a higher voltage (for example VTXSW ) at Q7.
The impedance of high turn, high inductance coils is higher.
Higher voltage will reduce the compensation time, important with short cycle times.

[Carl-NC]

I don't understand. The compensation "kick" is fixed at about 1V, set by D9. VTXSW just has to be high enough to ensure Q7 turns on, and that is relative to VTX. A high inductance TX coil will also have a higher resistance, and this will drive VTX higher (for a given current) but the AS2312 only needs 4-5V to be completely turned on. Since VTXSW can be increased to 20V, there should be plenty of margin. Let me know if I'm missing something here.

However, it brings up another issue: the AbsMax rating for VGS on the AS212 is 10V so I need to either add some gate protection or find a different device.

[Tinkerer]

I don't understand. The compensation "kick" is fixed at about 1V, set by D9. VTXSW just has to be high enough to ensure Q7 turns on, and that is relative to VTX. A high inductance TX coil will also have a higher resistance, and this will drive VTX higher (for a given current) but the AS2312 only needs 4-5V to be completely turned on. Since VTXSW can be increased to 20V, there should be plenty of margin. Let me know if I'm missing something here.

However, it brings up another issue: the AbsMax rating for VGS on the AS212 is 10V so I need to either add some gate protection or find a different device.

Sorry, I did not explain well.
I mean to use higher voltage for the compensation "kick" instead of only 1V above VTX. D9 then serves to block the higher voltage from going backwards.
In other words: disconnect the Drain of Q7 from where it is and connect it to VTXW.

[Carl-NC]

I think that would provide too much kick and make the PWM control more difficult. With just the 1V kick, sims are showing that it takes a very short pulse width to flatten the current. If it turns out that it needs more kick, it will be a simple matter to lift the drain pin and hard wire it to something else.

Addendum: As it turns out, IC11-pin1 is right next to Q7-D and is connected to VTXSW. So I added some bridge copper to make connecting easier; cut the Q7-D trace to D9, and bridge the copper. With that change, and Mick's suggestion, I'm gonna pull the trigger and order some boards. Fingers crossed.

Late Addendum: I forgot to post the final layout that was ordered. I changed the Rev to 1.0:

AMX_GerberRev1.0.zip

There are errors in this layout (see later posts), the most serious being that the ADC footprint is wrong.

[KingJL]

Here is Rev 0.3 of the schematics...

Carl, What is the P/N for Q3 & Q4? P/N SIHB600N60E doesn't seem to be valid. I am trying to verify the footprint and pin connections.

[Carl-NC]

I think I meant SIHB100N60E. It's a TO263 package.

[KingJL]

Minor error (I think)... C51 wrong orientation/polarity.

[PICON]

Carl, The TXDAC from the micro is connected directly to the output of Opamp IC7b pin 7, I believe it should be summed via its own resistor to IC9b pin 6.

[Carl-NC]

Minor error (I think)... C51 wrong orientation/polarity.

Correct, it is backwards.

Carl, The TXDAC from the micro is connected directly to the output of Opamp IC7b pin 7, I believe it should be summed via its own resistor to IC9b pin 6.

The idea is that you would use either the DAC, or the integrator (IC7b), but not both at the same time. However, IC7a is required for both, so you can't just omit IC7b. I should have put a 0Ω resistor on the output of IC7b, instead I'll probably just lift pin7, assuming I get that far.

Thanks for both feedback.

[Dean Sarelius]

Layout is progressing, though slowly. I've made it to the hard part, where I need to figure out the power planes. Total board size is currently 100x150mm.

Carl,
Just curious to know what stackup you are using..?

[Carl-NC]

Top
Ground
Power
Bottom

[KingJL]

Carl, What package is Q3a & Q4a? I can't tell from the schematic as it is overlayed. In fact from the schematic, I did not see a Q3a or Q4a... only when I was examining the gerbers did I discover Q3a & Q4a.

[Carl-NC]

Q3 & Q4 includes TO220, TO247, TO247-4, TO263, and TO263-7 packages, all in parallel. So you have lots of MOSFET options.

[ivconic]

Top
Ground
Power
Bottom

Oh sure!
You could add several layers more; just to annoy and harass ivconic more!