Quote from: poynt99 on August 01, 2012, 11:36:44 PM
I'm still checking in, yep.

In what configuration are you thinking of TK? Series I suspect?

Yes, I think this would pretty much kill the oscillation during what used to be the oscillation phase.

Yes, I meant in series, one at each battery terminal.
This would kill the oscillations throughout the system?

And also... if the NERD circuit is in "resonance".... what are the L and C values that are resonating at 1.3 MHz, and why isn't the resulting oscillation a pure sine wave?

Sigh.

Sure. To measure the internal resistance of a battery, you just set your DMM to "ohms" and hook it across the battery terminals. Right?





It's a good thing that Fluke makes their meters to be smarter than their users.

Quote from: TinselKoala on August 02, 2012, 12:29:23 AM
Yes, I meant in series, one at each battery terminal.
This would kill the oscillations throughout the system?

I just tried a single MUR160 forward-biased on the BAT+ terminal in the simulation, and it kills the oscillation.

Quote
And also... if the NERD circuit is in "resonance".... what are the L and C values that are resonating at 1.3 MHz, and why isn't the resulting oscillation a pure sine wave?

If we knew the exact L, we could find C. The total estimation of L is as per my diagram in the detailed analysis document.

The oscillation is not sinusoidal because of the body diode in Q1. Pull Q1 and Fo will increase, and the wave form should become more sinusoidal.

TK:

It would be amazing if people got to "know" their batteries and measured the internal resistance themselves.  I think somewhere on this thread we spoke about doing spot checks of the internal resistance.

You mentioned that Rosemary's multimeter will not blow up if she puts the probes across the battery terminals.  I hope so for her.

It feels to me like we are really in the wind-down phase.  Even with Poynt's help, there is only so much he can do and he can only offer up a certain amount of his precious time.

I love the smell of burning wire insulation in the morning.  lol

MileHigh

Quote from: poynt99 on August 02, 2012, 11:24:33 AM
I just tried a single MUR160 forward-biased on the BAT+ terminal in the simulation, and it kills the oscillation.

And presumably this would also be true if you put the diode between cells in the battery, like between #3 and #4 of the stack of 6, 12 V batts.
So...does this then indicate, as RA believes, that the oscillations and the current associated with them must pass "through" the batteries? Why, then, if the current is more negative than positive, and is flowing through the batteries along with the oscillations.... why do the batteries still discharge?
Why does putting caps _across_ the batteries filter the oscillations from the measurement, but still allows them to continue... but putting a diode in _series_ kills the oscs completely?

Quote
If we knew the exact L, we could find C. The total estimation of L is as per my diagram in the detailed analysis document.

OK, so I just read all the way through, again, your detailed analysis 06 .pdf. Using the last schematic in the paper and adding up the inductances by hand, since I can't find a statement of the total inductance you used, I get 11.23 microHenry. For this circuit to be in resonance at 1.3 MHz it will need a capacitance of  1335 picoFarad, or about 1 and a third nanoFarad, and will have total characteristic impedance of around 92 Ohms. The input capacitance of a single IRFPG50 is more than twice that at 2800 pF.

Quote
The oscillation is not sinusoidal because of the body diode in Q1. Pull Q1 and Fo will increase, and the wave form should become more sinusoidal.

Well, it's certainly true that if you pull Q1 the f0 increases and the waveform becomes cleaner, a bit. But does this really indicate that the circuit as a whole is in true resonance?





Now, isn't the internal resistance of a battery measured by the current it will deliver to a known, low resistance load?

Surely this could be done with an ordinary DMM and a bunch of power resistors. You take your 1 Ohm, 1000W resistance, hook it across the battery terminals with some heavy wire, and use the DMM to measure the voltage drop across the load. This will give you the current, from which the internal resistance can be found by Ohm's Law.

Vbatt, Rload given.

I = V/R
Iload = Vdrop/Rload

R = V/I
Rint+Rload =Vbatt/Iload

Rint=(Vbatt/Iload)-Rload