To be deleted...

[nul-points]

Thanks for those tips, TK, i'll check out the specs

np

[itsu]

Back on track after some PC problems.


I still had my bat42 setup running from the first battery (750mAH) after 1 week, so i have interrupted it to
make a new setup using some 3V/3W leds powered by 2x  4 2000mAH NiMh in series providing 2x 5.5V.

I made a blackbox in which i have tested one of the 3W leds and calibrated it using DC from my bench PS.

So now i can crosscheck the measured power calculated by my scope with the equivalent DC power from a graph i have made.
It turns out to be fairly accurate.

Video here:  https://www.youtube.com/watch?v=NyOTWG0F914

I will do some run down tests with the filled up supercaps only.
The blackbox also enables me to find the max output on a given frequency at the lowest possible input.


Regards Itsu

[nul-points]

...
I still had my bat42 setup running from the first battery (750mAH) after 1 week, so i have interrupted it to
make a new setup using some 3V/3W leds powered by 2x  4 2000mAH NiMh in series providing 2x 5.5V.

I made a blackbox in which i have tested one of the 3W leds and calibrated it using DC from my bench PS.

So now i can crosscheck the measured power calculated by my scope with the equivalent DC power from a graph i have made.
It turns out to be fairly accurate.

Video here:  https://www.youtube.com/watch?v=NyOTWG0F914

I will do some run down tests with the filled up supercaps only.
The blackbox also enables me to find the max output on a given frequency at the lowest possible input.
....

Nice work Itsu - good progress!

If, as you say, you're getting good agreement between scope math value for LED power and the DC Equivalent reading from your photosensor graph, then my earlier comment appears to hold true about the mismatch you reported between measured values for LED power compared with i/p supply, that it was more likely that the scope LED power value was the more accurate

You mention that your 750mAh i/p cell has been running for a week - using 'ball-park' figures, i/p cell was approx discharged at end of run (1.18V), although these pulse circuits can extract almost the last 'drops' of energy from cells (and continue to operate down to approx 0.5V or less); the o/p cell is showing almost fully charged at 1.4:V (some of which could be 'phantom' charge, of course!)

If your mention of 'running for a week' is indeed 7 days non-stop operation, then as an order of magnitude the LED was running for 7×24h at approx 16mW average, ie. approx 2700mWh 

i think i calculated in an earlier post that a 750mAh NiMH cell would store approx 1000mWh of energy

So your circuit appears to have achieved some interesting efficiency, if these numbers hold true, to both run at that power and, in addition, to charge the o/p cell

Is that correct?

Thanks for sharing
np

[itsu]

Hi NP,

Yes, it seems that this way of checking on the led power consumption agrees with what my scope calculates.
Around 32mW with this new heavy duty setup, but still when measuring the individual battery/supercap stacks it does not add up.

So i still make some measurement error there (allthough i also used a csr in the battery lead which makes it even more confusing) or the signals (pulses) interfere somehow.
It annoys me that i do not get any logical data out of this setup.


Anyway, with the old setup it was running for a week, but i am not sure about the 16mW consumption of that 10mm led.
It might have been for a short while with fully charged batteries, but it will drop fast till about 5mW when the supply battery get lower.
Perhaps i could retry that setup with the blackbox attached (have to calibrate that led first) to see how fast the power drops off.

Thats the problem with batteries, they can decive you easily.


More to come,   regards Itsu

[nul-points]

...
Yes, it seems that this way of checking on the led power consumption agrees with what my scope calculates.
Around 32mW with this new heavy duty setup, but still when measuring the individual battery/supercap stacks it does not add up.

So i still make some measurement error there (allthough i also used a csr in the battery lead which makes it even more confusing) or the signals (pulses) interfere somehow.
...
Anyway, with the old setup it was running for a week, but i am not sure about the 16mW consumption of that 10mm led.

It might have been for a short while with fully charged batteries, but it will drop fast till about 5mW when the supply battery get lower.

Perhaps i could retry that setup with the blackbox attached (have to calibrate that led first) to see how fast the power drops off.

Thats the problem with batteries, they can decive you easily.
...

Lol, yes, tricky beasts, batteries!  But they can still give up some of their secrets to us - we just have to be trickier than them  ;-)

I wouldn't expect the LED current draw to change very much over a test run - we're only discharging  the input cell in that 1st test by approx 0.25V, equivalent to a drain from max voltage to around 80%

In my supercap rundown illumination tests you can see that the illumination level is almost constant for the first 40 seconds or so, while the supply volts drop from max to aprox 60% - this is a significantly greater difference in voltage to our tests with NiMH input supply

Hopefully, if you're able to repeat the 1 cell test with your photometer box, this will confirm the power profile - if the LED power is fairly constant (or just reasonably linear) then we can apply the maths in my previous post with an average value for the LED power and the run time to find the total energy converted by the LED

Finding the amount of energy in the output cell, due to charging, could be determined by an initial full charge, then time a discharge from max to 1.2V  into a measured resistor (say 20ohms, .5W or 1W). This enables a calc of an approx capacity of the cell.  The energy in subsequent charges of that cell can be compared by discharging the result through the same resistor and using the time to reach 1.2V as a proportion of the time to discharge the full cell

Obviously this is quite time consuming (but it works), so better to try & get the test equipment to agree with each other about measuring equivalent values !

I think you've made a very useful addition to your test bench with your blackbox - i'm inspired to use one of those solar cells like yours, in place of my photodiode, when comparing brighter sources

Hope you have a more productive weekend than last - PCs can be a real pain in the butt when Windoze goes AWOL !

All the best with your experimenting
np