Use for the Tesla Switch

[ldissing]

I did something right for a change. The Sg3524 circuit on pc board works better that expected. This morning I add a parallel cap to the value of 1.47uf and it still runs with fully adjustable duty cycle. Maybe I got hold of a good chip. I got it from a different supplier.
Now each side pulse every 6 second. That means a pulse every 3 seconds. The pulse width is around 500msec. Very difficult to measure on the scope.
With all this changes and different frequencies and duty cycles and different loads, the batteries still loose voltage over time.
With this slow pulsing you can see on the voltmeter that the voltage drops more every time when it goes in series than what it rises when receiving the 24V.
I’m currently using 12V 5W bulbs which under testing measure more like 3.8W.
Leroy suggested bulbs that does not draw more than the c20 rate. They light up much easier and at least it looks like something is happening at low duty cycles
Maybe someday I will pick up something and charging will happen..

I did suggest that to eliminate one of the variables, but I don't know that this is true for the TS system. But the C20 rate would be for the length of time on, so in your situation, Total time per cycle = 6 seconds, length of "on" time ~500ms, so the bulb could be as big as 12*your C20 for these particular lengths of time, you see what I'm saying.

I assume you are using roughly a 7 amp hour batter. So the C20 is 350ma. But, with a 7ah battery and your CURRENT "on" time compared with your "off" time, you could run 4.2 amps through it, during that 500 ms with no problem, you are not violating the C20, so to speak.

I believe that JB was using those really small bulbs when 3.2-3.8V when he was using those 7.2V batteries. 7.2 / 2 = 3.6V, so we should be able to use 6 V bulbs on 12V batteries as long as we are using some dead time in the system. A 50/50 duty cycle (each side on 50% or total on time = 100%) will blow those bulbs out, but a 25% duty cycle will not.

I think I know how to get both sides to charge and run the load, and as soon as I am positive and can show you a fool proof method of doing it and knowing that it is right, I'll post it.

Leroy

P.S. I almost for got to say....I did get rid of the digital meters.

[nvisser]

The voltage drop measured with a scope is about 6.5V pulses measured over the bulbs. I did measure the current going through them even when I know I should not use meters on JB circuits and found that about 250mA passed through to the parallel batteries.
Even with this 500msec pulse and even shorter pulses the batteries loose voltage. When using higher wattage bulbs or larger duty cycles to much current pass and the batteries run down fast.
So Leroy I wish you would tell us what you’re up to. I reached a dead end.
I did build a proper Joule thief charger that uses a 555 to detect max and minimum battery voltage that switch in a load to discharge the batteries at c20 rate till it reach 12.4v. Then it charge again to about 14v. At this stage I dont use a cap discharge circuit in it. Just straight spikes to battery. Can anybody tell me if that is bad for the batteries. Do I need a cap pulser in the circuit?
I was hoping that proper conditioned batteries will make the TS to charge.
Up to now I just hooked up a 14w solar panel up to them once a week to get them charged up to test again.

[Matthew Jones]

I have never seen where just spiking the batteries had accomplished very much. Neither battery capacity or over all charge performance went up in a Tesla switch. I always had to make sure with a pulse motor that they got a little amperage. That seemed to last the longest.

I am having the same problem you are Nvisser. To get the batteries on the charge side to go up to 14 volt I have to discharge the batteries on the primary side at such a rate that they end up loosing. Maybe some real big batts would help (100 amp hour plus).

I don't know though.

I have tried to no end. I am getting ready to just go back to my original plans and just try to balance out a good load like an inverter coupled with a small solar panel to maintain the charge in the batteries. Just use the thing for conservation.

Matt

[citfta]

Thanks Leroy

I did suggest that to eliminate one of the variables, but I don't know that this is true for the TS system. But the C20 rate would be for the length of time on, so in your situation, Total time per cycle = 6 seconds, length of "on" time ~500ms, so the bulb could be as big as 12*your C20 for these particular lengths of time, you see what I'm saying.

I assume you are using roughly a 7 amp hour batter. So the C20 is 350ma. But, with a 7ah battery and your CURRENT "on" time compared with your "off" time, you could run 4.2 amps through it, during that 500 ms with no problem, you are not violating the C20, so to speak.

I believe that JB was using those really small bulbs when 3.2-3.8V when he was using those 7.2V batteries. 7.2 / 2 = 3.6V, so we should be able to use 6 V bulbs on 12V batteries as long as we are using some dead time in the system. A 50/50 duty cycle (each side on 50% or total on time = 100%) will blow those bulbs out, but a 25% duty cycle will not.

I think I know how to get both sides to charge and run the load, and as soon as I am positive and can show you a fool proof method of doing it and knowing that it is right, I'll post it.

Leroy

P.S. I almost for got to say....I did get rid of the digital meters.

Thanks for the hopeful report Leroy. I have tried all kinds of duty cycles and loads and two different TS configurations. One with diodes and one without and although I have been able to run small loads for a long time without any apparent decrease in battery voltage I have not been able to get any real charging going. So I am eagerly awaiting your report. I am attaching the latest version of the TS with diodes that I have been testing. Do you think it could be used to get charging like you are trying to do? I like this one because you can connect a DC load without using a bridge.

Thanks again, Carroll

[ldissing]

Thanks for the hopeful report Leroy. I have tried all kinds of duty cycles and loads and two different TS configurations. One with diodes and one without and although I have been able to run small loads for a long time without any apparent decrease in battery voltage I have not been able to get any real charging going. So I am eagerly awaiting your report. I am attaching the latest version of the TS with diodes that I have been testing. Do you think it could be used to get charging like you are trying to do? I like this one because you can connect a DC load without using a bridge.

Thanks again, Carroll

@Carroll,

The load is in the emitters of the transistors, and I think it would be better if it were in the collector, but it should work fine. The setup that Bits has is interesting, and that could work without the extra diodes and no bridge either. I'm still working on the 6 transistor setup, right now, and no bridge, but two different loads.

Leroy

P.S. When I have a completed solution that is duplicatable, you should ask me questions, but until then, I'm still it is a crap shoot. I still have a few thoughts and solutions that nobody has talked about yet, so I still have hopes.

[citfta]

Sorry for the confusion

Hi again Leroy,

I just deleted that post because after looking at it in more detail it looked like it was going to give uneven current flow between the top and bottom batteries. I think I have been up too long. Do you want me to put it back?

Carroll

[ldissing]

Hi again Leroy,

I just deleted that post because after looking at it in more detail it looked like it was going to give uneven current flow between the top and bottom batteries. I think I have been up too long. Do you want me to put it back?

Carroll

@Carroll,

It actually is a very much like JBs 2 transistor version of the circuit. Woke up at 4:00 this morning, realizing what the deal was with your circuit too. I didn't pay that much attention to the whole circuit when you posted, just where the load was in relation to the collectors. I'd been working on the TS and some earth current stuff all day.

I did hook it up briefly, and you know, I was getting a reading of 14V over the "supposed" parallel batteries where JB indicated we should be getting 14V. I did get it peaking at almost 15V as I increased the "on" time. The light was as bright as it could possibly be, because there is actually 24V across it at least part of the time, but it did not burn out. The wave form was very interesting too, I know why JB needed differential equations, although, I do not remember much about differential equations even though I liked the class. There was some "going negative" going on, and the batteries did not seem to mind the "uneven current flow" as you called it. Once I got it turned in, the bottom batteries did not go down in voltage anymore. I did not look at the upper batteries AND I only ran it for 45 minutes or so, so not enough time to really tell anything, but the switching was really cool looking. I was using a 12V 10W quarts bulb for the load.

The waveform was different than anything I'd seen before with a 50/50 duty cycle.

The top transistors were dead cold, the middle transistors a little warmer, the bottom transistors the hottest, definitely above ambient but not hot.

Leroy

[nvisser]

Battery conditioning

This is the wrong thread for this question but I know there are a lot of knowledgeable people here with experience on battery conditioning.
How long resting time should I allow after the batteries were charged up to say 14v before I start discharging again?

[citfta]

Repost of circuit

Hi Leroy,

Well maybe I was too quick to delete that post. I hadn't actually tried it. It just looked like there was going to be a problem getting even current flow between the top and bottom batteries so I though I would pull it quick before anybody saw it. Maybe I am spending too much time thinking. I'll repost it in case someone else wants to play with it too. As soon as I get my batteries recharged I will try it too.

@ nvisser, I usually wait an hour or so before using my batteries after I have recharged them. I am using 7ah batts. I think the size of the batts may determine how long you should wait. I have noticed that after charging the voltage will drop when you take them off the charger. After a time they seem to settle on a resting voltage. At this time I start using them again. For my conditioned 7ah batts this is usually around an hour. If I am not in a hurry to get back to my testing I may let them rest longer.

Carroll

[nvisser]

Thank you for the information Carrol
That circuit you posted, according to me will do the same as Jb 2 transistor circuit but only in the bottom part. When the one side transistors switch your bottom transistor will switch the load in circuit with the bottom left battery through the series transistor and your bottom left battery will drive the lamp bright.
In Jb`s circuit it happens on top to and there are no loads. That’s why it fries diodes and transistors.
I really do not see how that circuit can operate correctly even with high wattage diodes and loads

[Matthew Jones]

I have had the 2 transistor version running a couple of times but it still didn't charge. I used 35 amp diodes and switched at or about 1 second. The load was 50 watt bulb.
It made alot of heat in the transistors, but it did work.

Matt

[nvisser]

Can you see in the circuit what I mean when I say that the transistors and diodes short over the battery and the same with the load in the bottom. That is why Leroy get the bulb to burn so bright.
But then he also said that his bottom batteries did not loose voltage in 45 min. It is possible that JB knows something the we can not see. Lets wait for Leroy`s results.

[nvisser]

@Carroll,
I did hook it up briefly, and you know, I was getting a reading of 14V over the "supposed" parallel batteries where JB indicated we should be getting 14V. I did get it peaking at almost 15V as I increased the "on" time.
Leroy

Leroy
How do you measure the 14 volt over the parr batteries . If the batteriy voltage are 12v it will pull that voltage before the 2 diodes down to their level.
I replaced one of the batteries with a large capacitor to see to what level it charge up to with one short say 100msec pulse and was quite amazed to see that it charge up to much more than 14v in that short time. That put me under the impression that there are enough voltage available for charging.
But still no charging.
Does anybody here tried flooded lead acid batteries to see if they got better results?

[ldissing]

Leroy
How do you measure the 14 volt over the parr batteries . If the batteriy voltage are 12v it will pull that voltage before the 2 diodes down to their level.
I replaced one of the batteries with a large capacitor to see to what level it charge up to with one short say 100msec pulse and was quite amazed to see that it charge up to much more than 14v in that short time. That put me under the impression that there are enough voltage available for charging.
But still no charging.
Does anybody here tried flooded lead acid batteries to see if they got better results?

Vissie,

I measured where JB indicated on his two transistor circuit, but I'm using six transistors. I'd think it should be bigger than 14V because of the diode drops to get the batteries to charge.

My voltage started out higher with shorter pulse width, but I increased the pulse width until I had 14V. The batteries actually have to start charging, but there is no point in having less than 14V at that point.

John K. is using flooded batteries.

Leroy

[ldissing]

Can you see in the circuit what I mean when I say that the transistors and diodes short over the battery and the same with the load in the bottom. That is why Leroy get the bulb to burn so bright.
But then he also said that his bottom batteries did not loose voltage in 45 min. It is possible that JB knows something the we can not see. Lets wait for Leroy`s results.

I'm not sure to whom you asked the question. If it was to me, then yes, I see it what you are talking about. I'm sure if we knew what differential equations to use (I guess I'm going to have to do some study of some old college material) then we could figure it out too?

Leroy