Use for the Tesla Switch

[ldissing]

[...]

I am going to try and put this in simplistic terms, so to illustrate, if we close Q1, we apply 24V into batts 2 and 4. One would say that we are building pressure against batts 2 and 4, By design, Q9 provides the path to batts 1 and 3 when we reverse the process AND FEELS THIS PRESSURE. Most of us learned early on (when diodes were used) the pressure became to great and blew the diodes. Q9 overcomes this by becoming a "pressure release valve" and can be adjusted t0 insure the right pressure is maintained against batts 2 and 4.

[...]

Bit's

Bit's,

I see this still differently than you do. I have not put a scope on it, because I think the probes are fried, although, I may try it again today.

Q9 in your situation is exactly like a diode except that you can turn it off. So you are running a modified 2 transistor design for the most part. I see Q9 not as pressure release valve, but a pressure increasing valve. When you open the valve, more charge goes to batteries 2 and 4 and NONE goes back to battery 3. So you are closing off this path by opening the transistor. Now, you really have no time delay between firing Q1 and Q2. You could code it that way, but it was not there in the code you posted the last time I saw code from you.

I thought JB said we could run this from 10% to 50% duty cycle? He did not say 100%. You would need 1/2 dead time on each side to run at 50%. Hopefully you are still gaining charge, I will go check mine out this morning. He did however mention a path for current in the top and in the bottom, but we are actually creating AC on the two bottom legs in which AC was not meant to be created. AC was meant to be created in the battery 3 to battery 4 negatives. JB said that no charging took place on the AC side.

So, I can definitely duplicate what you have done, which is GREAT. I'm just not sure this is exactly where we want to be, but it is a repeatable experiment and everyone can do it if they have a digital means for control.

[***EDIT my batteries lost potential, but I coded it significantly different than Bits, so this has nothing to do with what Bits is doing. I did not believe that Bits was doing coding it correctly, but he may have been, because my batteries are significantly lower this morning than last night...about 7 hours run time. I was using delays between switching Q1 and Q2 and less off time for Q9, so I still need some fine tuning on the coding, the hardware seems to be fine. Should have just done what Bits did and see if batteries gained charge! ***]

Leroy

[nvisser]

Bit's,
I thought JB said we could run this from 10% to 50% duty cycle? He did not say 100%.
Leroy

100% duty cycle will be dc. High all the time. I think with 100% you mean 50% duty cycle . That is half the time on and half off.

[ldissing]

100% duty cycle will be dc. High all the time. I think with 100% you mean 50% duty cycle . That is half the time on and half off.

Well, with a PWM signal, which we are essentially generating, one side is on 50% and the other 50%. That is 100% on, but when you introduce a delay in the on time, you would be decreasing the duty cycle to say 50%, meaning 25% on one side and 25% on the other side with 50% being dead time. Does that make sense?

Of course, you could have a duty cycle where one side was on 30% and the other on 20% and that could still be termed 50% duty cycle I guess, but one side has less load than the other.

I do not know what JB means when he says 50% duty cycle. Is that 50% on each side (in turn), or 25% on each side with the remaining being dead time.

We (I) need to get our (MY) definition defined....here I go again, not knowing what I'm talking about! Par for the course!

Leroy

[ldissing]

What happens when you do this?

Bits,

So, what happens in your circuit if you turn Q10 and Q9 off all the time (normally on, closed for you)? What happens with the bulbs in this scenario?

Hate to ask you as you are doing some incredible work, but I need to confirm what I'm seeing. No problem to change the code and test it, right?


Leroy

[Bit's-n-Bytes]

Bits,

So, what happens in your circuit if you turn Q10 and Q9 off all the time (normally on, closed for you)? What happens with the bulbs in this scenario?

Hate to ask you as you are doing some incredible work, but I need to confirm what I'm seeing. No problem to change the code and test it, right?


Leroy

I have done that by pulling the Opto's. I think this will give you a false indication that it is working when in fact it is not. You need the "Snap". Put your DMM across batts1 and 3. Fire Q1. Are you able to drive the potential up to say 18 - 20V? If not something is not right. Repeat for batts 2 and 4.

Bit's

[ldissing]

I have done that by pulling the Opto's. I think this will give you a false indication that it is working when in fact it is not. You need the "Snap". Put your DMM across batts1 and 3. Fire Q1. Are you able to drive the potential up to say 18 - 20V? If not something is not right. Repeat for batts 2 and 4.

Bit's

You didn't say what happens to the bulbs? I do not think you could pull the optos in your setup, they have to be turned "on" to get the PNPs to be "off" or "open".

If you are seeing what I'm seeing, then the question is why? There would be no connection between the 24 positive side and the 12 positive side, yet if you are seeing what I'm seeing, then charge is still flowing...both ways? What is up with that?

Leroy

[Bit's-n-Bytes]

You didn't say what happens to the bulbs?
Leroy

Didn't I answer that in post 1731?

[ldissing]

Didn't I answer that in post 1731?

No, but it doesn't matter. No worries, I'll just go back to my testing and see what I can see.

Leroy

[***EDIT BTW, when it is coded the way you are doing it with PNPs (mine with NPNs) it does definitely create a voltage in excess of 18V. ***]

[nvisser]

Question to Electronic boffins

What is the best way to connect 3 optocouplers to switch at the same time from one of the outputs of a SG3524.
I think the voltage over each one is 1.2v with a typical current of 20mA.
Do I limit the current with a 470ohm and connect them all in series or do I drive each one separate with its own limiting resistor (560ohm)?
On the diagram of the scaler charger that Mr Bedini posted he connect two of them in series.

[blackchisel97]

What is the best way to connect 3 optocouplers to switch at the same time from one of the outputs of a SG3524.
I think the voltage over each one is 1.2v with a typical current of 20mA.
Do I limit the current with a 470ohm and connect them all in series or do I drive each one separate with its own limiting resistor (560ohm)?
On the diagram of the scaler charger that Mr Bedini posted he connect two of them in series.

Vissie, I tried and had no problem with three H11's in series plus LED and 330Ohm. I had two connected just as J.B did and LED as well. Working fine.

EDIT: Just noticed, your question wasn't addressed to me (Electronic boffins) Ooops! hope you don't mind.

Vtech

[mlindeblom]

What is the best way to connect 3 optocouplers to switch at the same time from one of the outputs of a SG3524.
I think the voltage over each one is 1.2v with a typical current of 20mA.
Do I limit the current with a 470ohm and connect them all in series or do I drive each one separate with its own limiting resistor (560ohm)?
On the diagram of the scaler charger that Mr Bedini posted he connect two of them in series.

There is less wasted energy using the optical couplers in series; hence: lower power dissipation in the SG3524 and series resistor(s).

The original resistor 560 ohms at 20 ma generated 11.2 v at 0.224 watts when the optical coupler is on.

A single series resistor needs to generate on 8.8 v at 20 ma, thus 0.176 watts. The new size is 440 ohms using ohm's law.

The power saving is 3 x 0.224 - 0.176 = 0.496 watts. The parts savings is 2 resistors.

Both your solutions work.

[hherby]

I have done that by pulling the Opto's. I think this will give you a false indication that it is working when in fact it is not. You need the "Snap". Put your DMM across batts1 and 3. Fire Q1. Are you able to drive the potential up to say 18 - 20V? If not something is not right. Repeat for batts 2 and 4.

Bit's

Bit's,
Is this 18-20v from neg of batt 3 to pos of batt 1 while Q1,Q9 and Q10 are all closed? If so, what is the voltage when you open Q9? If so, what is the voltage when you open Q9?




Thanks
Alex

[Bit's-n-Bytes]

Bit's,
Is this 18-20v from neg of batt 3 to pos of batt 1 while Q1,Q9 and Q10 are all closed? If so, what is the voltage when you open Q9? If so, what is the voltage when you open Q9?




Thanks
Alex

Alex you are correct in the hookup. Depending on the load voltage will vary with Q9 open. Obiviously with a very small load I have seen the sum of batt1 and 3. You could actually "short" where Q9 is (and we did this when using the diodes) when we fire Q1. The trick is to be able remove that "Short" (hence the function of Q9) to act as a relif valve and maintain our voltage. Currently I have 2 Rat Shack 6.3 Bulbs,1 at J5 and 1 at J6. My DMM shows about 18-19V when I open Q9 (Q10 in the oppisite side). I had to adjust the code with the following to prevent the bulbs from blowing out;

high 4
pause 800
high 7
low 4, 7
;pause 50
high 6
pause 800
high 5
low 6, 5

This code is totally adjustable.

Hope this helps.
Bit's

[Bit's-n-Bytes]

Alex you are correct in the hookup. Depending on the load voltage will vary with Q9 open. Obiviously with a very small load I have seen the sum of batt1 and 3. You could actually "short" where Q9 is (and we did this when using the diodes) when we fire Q1. The trick is to be able remove that "Short" (hence the function of Q9) to act as a relif valve and maintain our voltage. Currently I have 2 Rat Shack 6.3 Bulbs,1 at J5 and 1 at J6. My DMM shows about 18-19V when I open Q9 (Q10 in the oppisite side). I had to adjust the code with the following to prevent the bulbs from blowing out;

high 4
pause 800
high 7
low 4, 7
;pause 50
high 6
pause 800
high 5
low 6, 5

This code is totally adjustable.

Hope this helps.
Bit's

Here's more;

YouTube - Workig D-TS-1.MPG

Thanks.
Bit's

[Mark]

Nice demo Bits. It seems like it would work better when you made the delay longer. Potential was climbing up to 26 volts or so. Do you think the frequency of the "potential dump" is one of the more important points that causes the batteries to charge? I know the "snap" is important too.

I found using when charging with solid state charges the lower the frequency the faster the charging.

Great work!

Mark