Testing the TK Tar Baby

[TinselKoala]

TK,

So, now what's the problem?

If channel 1 is the shunt voltage, I agree with your off screen measurements and it looks like about 320ma is flowing when the FG output is positive.  320ma across the 11R load resistor means that the load resistor is dissipating around 1.13 watts.  If Vbatt is 60 volts, there is about 56.5 volts across Q1 so Q1 is getting hot as it dissipates approx 18 watts.  Total dissipation (ignoring the small dissipation in the csr) is just under 20 watts.

PW

And that agrees with what I said, which is that 20 Watts are being dissipated _somewhere_ in the circuit.
And also there is some dissipation in the FG's internal 50R, I think. No problem as far as I can see... but Ainslie says that we don't know how to calculate the power, that I am wrong, and therefore you are too ... but she won't tell us why or how. Therefore, I conclude that SHE once again is simply wrong.

This is of course the screenshot that goes along with the "boiling water with no power drawn from battery" blog post that I've linked. Not only wasn't the water boiling, but substantial power is in fact being drawn and dissipated, and it ain't coming from superluminal zipons, neither.

[happyfunball]

http://www.youtube.com/watch?v=t9Y3mWDkB6o

[picowatt]

And that agrees with what I said, which is that 20 Watts are being dissipated _somewhere_ in the circuit.
And also there is some dissipation in the FG's internal 50R, I think. No problem as far as I can see... but Ainslie says that we don't know how to calculate the power, that I am wrong, and therefore you are too ... but she won't tell us why or how. Therefore, I conclude that SHE once again is simply wrong.

This is of course the screenshot that goes along with the "boiling water with no power drawn from battery" blog post that I've linked. Not only wasn't the water boiling, but substantial power is in fact being drawn and dissipated, and it ain't coming from superluminal zipons, neither.

TK,

During this part of the cycle the FG need only provide Q1 gate leakage current once it charges the gate capacitance, so current thru the FG in this mode (FG output positive, Q1 turned on) is very minimal (not significant).

PW

[picowatt]

TK,

Are you going to try .99's "Altoids" oscillator?

PW

[TinselKoala]

@PW: Yes, I see that now, or again, thanks. I've been concentrating on the other mode, negative only, for so long I forgot about how the positive mode worked. The FG is only in the series circuit during the negative gate drive portion and then the overall current is low due to the mosfets only being biased partially on and oscillating. And of course in the positive mode it just needs to charge the gate capacitance and then it's basically done, as you say. So maybe Rosemary is confusing the positive gate drive mode FG role with what we've been discussing lately, the negative case, where the role seems very different.

So really we seem to have a third mode of operation now: a partially on mosfet Q1 due to the low level positive "ON" drive, combined with the negative "OFF" drive giving the oscillations and passing some current through the Q2s. This is distinct from the case of the +10 or +12 volt drive that produces a fully on Q1 and a current of 3 or 4 amps, with a negative going portion for oscs, and of course it's different from the strictly negative drive that leaves Q1 out of the picture entirely. So we have 1) mosfet Q1 fully on with high current during "on" time and Q2s in linear region with oscillations during "off" time (but really negative drive), and lots of power dissipating in the load and mosfets; 2) mosfet Q1 partly on with low positive gate drive "ON" and Q2s partly on due to oscillations "off" really negative drive, with most of the power dissipating in the Q1 mosfet and a little in the load; and 3) strictly negative drive, ignoring Q1, oscillating and biasing Q2s to conduct some current, with small load heat and Q2 mosfets warming.
Have I left anything out?


Sure, I'll build the Altoids box. I intended to do it today but I got lazy and didn't feel like driving to the store in the thunderstorms. I need to get the right Zener and some misc. inductors, so I'll probably do it tomorrow. I like the idea of having the math done on-board but developing the circuit to do it is above my pay grade. If you or .99 can sketch a schematic using a comparator or opamp or the AD834 or similar I'll be happy to build it, if my source has them in stock. Or maybe I'll place a DigiKey order. I need some other stuff too and they usually deliver in two days, from placing the order, to my mailbox.