Cool joule ss/ssg - Another view

[ion]

I would respectfully submit that interested parties run a test of a battery voltage vs. a change in room temperature with and without a slight load.

Remember, a battery of the type being tested is a chemical reaction which changes voltage slightly with temperature. Not as relatively temperature insensitive as a "standard cell".

[TinMan]

@ Itsu
Im not quite sure why it seems to only work with the 1n4004,as i have tried many different diode's now and only the 4004 gets this effect ??.
The other thing is,the run battery go's up then down-up then down in voltage-this may be related with what ION said about temp?.The charge battery dose not seem to be worried about temperature,as i have had it running non stop for 2 1/2 days now-and it charges both day and night.Days about 25 C ,nights about 4-6 C.Tonight i will do a rundown test on the second charge battery-as i did with the first.

[Itsu]

It could be the battery is recovering now, but none of the other diodes showed this.
So both batteries are increasing now, so lets see where this goes to.........
b2 start voltage when 1n4004 was inserted is 0.9823V, now after 15 minutes its 0.9872 and climbing.
B1 is steady on 1.064V

Regards Itsu

As expected, it seems the b2 battery was recovering from the bigger loads it had from the other diodes.
It went up till 0.9923 last night, but this morning it shows 0.9667 and after 10 minutes 0.9666, so slowly decreasing.

The B1 battery however steadily increases, now showing 1.075V

Temperature yesterday evening on the both batteries was 21.6 degrees C, this morning 19.7, so not much variation here.

Regards Itsu

[TinMan]

Here is an update video(day 3).It would seem ION was right about the temp making a difference,as during the day the run battery increases in voltage-and during the cold of the night it go's down again.I have lost .006 of a volt since i started this circuit 3 days ago.I am now going to build another,and try and return the charge back to the source battery.

Good result's with the 4001 Itsu,and im looking forward to hearing how slider is going with his setup.I have an idea for a self runner with this circuit,and will put it together as soon as my parts arrive from altronic's.

http://www.youtube.com/watch?v=kzu4-rTQOok

[Chet K]

Perhaps ION is playing here too? as in Solar "charge" during the daytime??
throw a Fresnel lens over one ?[maybe with a cooling fan to keep the temp the same?]
thx
Chet
PS Off topic [sorta]
I'm gonna get "johny" a new Pet to raise.[the next "silk worm"]

http://www.newscientist.com/article/dn23431-bacteria-churn-out-first-ever-petrollike-biofuel.html

[Slider2732]

Here's my upload from late last night:
http://www.youtube.com/watch?v=J7Fxr3OP1sU

Today, first circuit carries on:
Run - 1.205V
Charge - 0.838V

I'm not seeing the decrease at night, as would be expected from heat changes in a non air controlled house. But, I am sure ION is correct, it makes sense and is in keeping with hundreds of self made batteries. The tendency is to get more out of the battery when warmed, but the life of it goes down. Stick a flat AA on a hot radiator and it will charge a bit, that sort of thing.

The second circuit has shown some interesting changes, just by swapping transistors.
I was thinking about Negisters and how Dr. Stiffler uses a 2222A with no Base connection on his seeming self runner based on the SEC 18X.
MPS222A dropped voltage on the run battery at the comparatively dizzying rate of 20mV per half hour, which was no good, junked it.
But, 2N3904 (Negister related) not only fired up as soon as the charge battery went in, but gained 3 or 4mV on the charge battery over 2hrs, with LED on. Not a lot however.
S9014 looks good, needs more testing.
C3198 works, yet did drop 10mV over the 10hr run overnight on the run battery of this second circuit.

Did it drop or did the lowering ambient temperature make it appear to lose the voltage ?
The charge battery only went up by 2mV however, to 1.204V.

One focus is on the hFE of small signal transistors. I don't have any top hFE ratings transistors, like S9014 'D' etc.
Perhaps a Sziklai arrangement is an idea ?


@Chet - random factoid. I still have the Highest Score in Oklahoma for SilkWorm the arcade machine. Had one til about a year ago and score was 296,500. Loved that machine, but economics forced sale.
http://forum.okcoin-op.com/index.php?topic=862.0

[Slider2732]

New run of the second circuit.
I've now changed to exactly the same setup as circuit 1.
Diode is a RPG10J which is what is written on the other and why I could find no datasheet...apologies again.
Transistor is a C3198 'GR'
500ohm pot set to 100 ohms
Batteries rested for a day, these are the ones to do ION's ambient temperature experiment.
LED lit.

14.50 Run - 1.159V Charge - 0.919V
15.20 Run - 1.156V Charge - 0.925V

So far: 3mV used, 6mV charge with LED on.

Updates:
15.50 Run - 1.152V Charge - 0.927V
16.20 Run - 1.150V Charge - 0.929V
17.20 Run - 1.146V Charge - 0.932V
18.20 Run - 1.144V Charge - 0.934V
19.30 Run - 1.141V Charge - 0.937V

Right on the line !
Used 18mV, gained 18mV
But, all the while the LED has been on

[TinMan]

I have a brain teaser for you all coming up soon(video is loading).
It seems that it's not magnetic induction between the run and trigger coil that is fireing up the transistor???.

[TinMan]

Here is the cool joule 3 up and running.This one has to run coils in series,and no resistor on the base.I have removed the LED for a faster charge on the charge battery.We are now running on about 70 uA at 1.5 volt's.As far as the trigger coil running the system while inside a steel tube with large magnets surrounding it-well that might be one for ION to answer,as it's new to me.I suspect the emmiter is being pulled down below 0 volt's,wich is giving us a positive potential at the base of the transistor-??

http://www.youtube.com/watch?v=z7DlD8MIEes

[Slider2732]

I liked Giant Killer's explanation of the circuit - that the ubquitous storm over your shed is a source of white noise EM.
You know you have something there anyway when it rains  


Quick tip - starting troubles.
Put a 10uF electrolytic across the power rails.
Starts every time on one circuit without poking the legs of the transistor. Another one it didn't. Might improve very troubled circuits though.

Edited after trying on another.

[TinMan]

I liked Giant Killer's explanation of the circuit - that the ubquitous storm over your shed is a source of white noise EM.
You know you have something there anyway when it rains  


Quick tip - starting troubles.
Put a 10uF electrolytic across the power rails.
Starts every time on one circuit without poking the legs of the transistor. Another one it didn't. Might improve very troubled circuits though.

Edited after trying on another.

Hi Slider
I dint think the white noise thing is correct,as i have the trigger coil submersed in a strong magnetic field-this was to eliminate any magnetic interference with the trigger coil.

[TinMan]

@Itsu
How is your setup cpmeing along Itsu?

[Itsu]

@Itsu
How is your setup cpmeing along Itsu?

Hi Tinman,

well, i kind of stopped after trying several scenarios.
I used severall coils, severall diodes and severall transistors like 2n3055 and 2n2222a (metal case), but in all cases the outcome is the same:

  charging battery B1 will slowly increase
  running battery B2 will on average decrease more then the increase of B1, however it also show some increasing for a while (typical a few hours)
  when it was loaded heavily for some reason (typical battery recovery behaviour).

So nothing extraordinary to report :-(

Regards Itsu

[TinMan]

Hi Itsu
Im not sure why you arnt getting the result's like myself and slider are-infact i believe slider has surpassed me lol !where is slider anyway?!
I may have worked out a way to delete the trigger coil altogether-but still a work in progress.
Here is a video of Lidmotors replication.
http://www.youtube.com/watch?v=5Mbp1iuB7as

[Slider2732]

Sorry, yes, I wasn't sure what to do when hitting a 'certain situation'.

Without further a do, here is the Cool Joule GTi

http://www.youtube.com/watch?v=aYm-ldEkYkw


Update -
13.20 Run 1.352V, Charge 0.890V
14.05 Run 1.348V, Charge 1.191V
14.20 Run 1.347V, Charge 1.202V
14.50 Run 1.347V, Charge 1.211V
15.20 Run 1.347V, Charge 1.217V
15.50 Run 1.346V, Charge 1.221V
16.20 Run 1.346V, Charge 1.227V
17.20 Run 1.345V, Charge 1.233V
18.20 Run 1.343V, Charge 1.239V


Run -9mV
Charge +349mV
LED On

Looking at the figures, here's the analogy...people jumping into a pool. After so many, they have to find spots to place themselves as it fills up.

[Slider2732]

Another run, before heading back to single battery runnings (only got 1 pin board).

Same Run battery, different Charge battery.
A near 12hr run.

22.54 Run 1.343V Charge 1.190V
23.54 Run 1.338V Charge 1.242V
00.54 Run 1.336V Charge 1.252V
01.54 Run 1.334V Charge 1.256V
02.54 Run 1.333V Charge 1.261V
10.30 Run 1.320V Charge 1.287V

Run -23mV, Charge +97mV
LED On

S9014C
100ohm
2x1000uH
1N4148 on Collector
100uF cap across power inputs
2X 1N4004 equiv sharing with LED cathode connection

[ion]

Run -23mV, Charge +97mV

Big error to compare battery voltages at two different points on the charge / discharge curve and think that it is a meaningful comparison of energy used / recovered.

[Slider2732]

It was meant as a test of another battery at a different point in its discharge, but using the same Charge battery.
Mainly, what can be done with 1 charge from a charged battery and do the figures equate out of discharge vs charge with an LED lit as a 'load'.
Should another deadened battery have been used ? probably yes.

The Run battery was little used in terms of discharge. The Charge batteries were at different points.
With most charge having been shown in the earlier stages of the first run, would the second charge battery run a similar course from where the first battery had been at that point.
With no input on the thread the last couple of days, that's the way I rolled.

Point noted for the future

[TinMan]

What i have done this time.
I charged the second charge battery to full voltage(over 4 days)
I then ran the run battery(the one used from day one)in the garden light until it wouldnt run the garden light any longer.The run time was very close to 4 hour's-so that shows very little drain from the run battery in the cool joule for the last 6 day's.
I am now charging that run battery with the second charge battery.
The standing start voltage of the original run battery was(now being charged) .763 volt's.
Over night it has charged to 1.087 volt's (about 10 hours). So this battery is now being charged by a battery that it charged up in the first place.

[Slider2732]

I guess the main thing is whether it runs the solar light again, for more than a fade out.
What's full voltage on your batts ?
If I charge mine with an R/C balance charger, they get to nearly 1.4V ! If I charge with solar, they tend to be around 1.270V. A lot stick at 1.267V and I thought that was the top til the R/C charger was tried.

Over here, i'm running at 10 to 20uA.
If I disconnect the charge battery , it 'shoots up' to 40uA.

The method, is to wind a thin gauge of wire about 15 times around the Positive/Collector inductor. One end goes to the Negative rail and the other is free in the air. The free end can have it running down into the single digits depending on where it is placed next to, but the light is too dim to be of any use. Have to use 10uA or a little more to see the light...but the circuit remains running.
Doing such a thing doesn't do much for the charge battery lol
Did you ever find a way to omit the other inductor ?

First pic is of running at 20uA.
Second is when the charge battery is disconnected, running at 40uA.
Third is the light in a darkened room at 40uA.
Transistor is a C1815 'GR'.

[TinMan]

@ Mark
I have not yet worked out how to remove the trigger coil from the system-not yet anyway.
I have just finished processing a video of the cool joule circuit running a pulse motor.After reviewing the video,i have decided not to post it just yet-as i dont even believe what im seeing in the video ,! and i was there lol !So i could just imagin what others would say.
I will have to go through the circuit,and try different DMM's on the input-as something is amiss here.I cant see how or why the DMM would be reading wrong,as i am going from the battery(flat 9 volt)through the DMM and to a 9400uf cap bank as the run cap's.I have returned the output back to the caps as you did Mark,and the efficiency has gone through the roof-a too good to be true sort of rise.

Here are the p/in figures
7.8 volt's(near to) @ .6uA
Now when in solid state opperation,i see every second pulse is different. How can every second pulse be different?-and the difference is uniform through the oscillations.

[Slider2732]

 
I think I know what you've done...at least for the circuit, haven't tried the motor idea yet.
If you try to measure across the LED the LED will go out.
If you put a diode out from between the output diode and the LED, you CAN measure it. Now, if you take your reading as a Negative from the output diode that returns back to the input you'll see some really nice numbers.
It doesn't make sense (to me) that it would be a Negative going back to the Positive rail, but that's what it is. The Positive voltage reading is taken from that diode that is sitting with open leg from the output diode and LED junction.
Example would be 1.250V in, 2.791V out - the amperage is of course tiny though, so the LED is only 1/5th bright at best.
The 'trick' is to do something with that output and also keep the LED going...at least to keep the transformer type action going.
The 'secret', I think so far, is to use a cap and perhaps a 1:1 isolation transformer.
Losses have to be appreciated, while we're dealing on the micro size of all and that's been shying me away from a transformer. Might try a toroid with same winds on each side though.  
The output has to not see a draw, else the whole thing equates out, as we are familiar with.

I'm testing a circuit for Steve at the moment (a Cool Joule GTi), but do aim to try the motor/cap idea.
One question, being that the pulses of the collapse of the motor coil(s) will be out of phase with the input pulses, could that be a factor for the 2 pulses in your scope shot ?

Multi frequency circuits are an exciting field, as it were.

[PhysicsProf]

Thanks Slider and Tinman for your continued work on this...  exciting stuff.

Tinman: "I have returned the output back to the caps as you did Mark,and the efficiency has gone through the roof-a too good to be true sort of rise."

Would like to hear more about that when you're ready!

[TinMan]

edit-This is just an example.
Ok i made a couple of small mod's to the circuit,and getting result's that don't seem to fit the realms of normal.Below is a quick sketch-a question for you all.
Inside our theoretical circuit box,there is large smoothing cap's-so as the input and output are smooth DC.With scope set on highest voltage resolution(.5mv)there is no ripple at all across the two resistor's.We also have swapped the resistors with each other,and run test again.
My question is-is there any way this can be achieved using conventional method's of some sort that I'm missing ?.As i do over look the simplest of things some time's  

[Slider2732]

I think i've lost what you're doing, what are the tiny voltages ?
One of those times where a pic would bring an 'ohhh' moment


Last night saw the first surpass of the Run battery by the Charge battery.
It's all on Steve's set of components, so he can directly verify once I send them to him.
The biggest change has been to use a C2236 transistor.
Traditionally a mid performing NPN for other projects, this thing sent the GTi into a Turbo mode !
Normally, the top return diode (in circuit diagram below) will allow single battery uA running. When disconnected, the circuit may stop and normally does. Always does when running just 1 battery, because of no return path. However - get the right transistor and the circuit will carry on when it is removed. That's the trick here for Turbo mode.

The start of the day saw 2 very well rested batteries showing 1.306V and 0.954V.
During the day, several different ideas were tried, not all were good lol.
By 1am, I switched to the C2236 and pulled out the connecting pin for the top diode.
The circuit stayed on.
Here are the figures:

Start
Run 1.294V Charge 1.063V

After 3hrs
Run 1.267V Charge 1.269V

10.30am
Run 1.235V Charge 1.346V


I've swapped them over and will run it for 2hrs, then back again for 2hrs.
Then let both batteries thoroughly rest for final readings.
Throughout, the current used has been 1.9mA and the LED output very bright comparatively. About 1/2 full bright - useful light.