Use for the Tesla Switch

[John_K]

OK Lero,

Since you asked...

Reference the attached pic from JB. You'll also need to just have one side on all the time, not alternating (i.e just Q1 & Q10) to do this test.

The idea is to get the most voltage over the 12V bank (at least 14V), whilst having the most voltage on the 24V bank (at least 18V). As JB says, you need at least 2V over the battery to get it to charge.

Set you meters up as per the picture and start with a small load like an LED, but make sure you have the same load on each side even though only one is lit up. Measure the voltages and record them. For example, when I used an LED (with 330R to protect) I got 13.08V for the 12V bank, but 23.8V for the 24V bank. The point is that I don't have enough to charge the 12V bank, but I have heaps of potential to use on the 24V bank.

OK, next try the other extreme like a car headlight halogen bulb. When I did this I measured 13.29 (after 15 seconds) on the 12V side but under 18V on the 24V side. The other problem was that the halogen took too long to light up - which means you can't switch fast enough to get any charge into the 12V side.

The best result I had was 8 x #47 bulbs. These were arranged so I had 4 pairs paralelled of 2 bulbs in series. I had the same load setup for both sides. With this arrangement it gave me 13.28V on the 12V side and still 23.16V on the 24V side. My batteries were all 130Ah big Trojans.

All 4 batteries did charge up slowly - about 0.01V every few minutes, but then they all stopped gaining after about 30 minutes.

When you do this test, it's also a good idea to have a meter on the bottom battery on the 12V side to see how fast (or if) the voltage increases.

I don't think I still had the right load, but I was starting to run out of #47 bulbs.

But then I put the same load (8 x #47) on a set of 7Ah gel-cells and it was too much for the 24V side - so that's why I said you have to pick the load for the battery size you are using.

Anyway, that's my theory which I still have to test out properly on the bench. I'd like to get some 1157's and four motorcycle batteries as the only decent fooded LABs I have are the big Trojans and it takes too long to see any impact. (And the 7Ah gel-cells are too unreliable)

Let me know how you go...


John K.

John K.,

You will not be wasting my time. I'm in the midst of it all, so please post or PM me with your thoughts. I ran 12V 10W bulbs (17ah) batteries, about the right amperage for the batteries, but I fear, too big a load in voltage. The batteries lost .2 V over 9 hours, with the batteries basically dead already. Pretty impressive. Let me know.

Lero

[ldissing]

OK Lero,

Since you asked...

Reference the attached pic from JB. You'll also need to just have one side on all the time, not alternating (i.e just Q1 & Q10) to do this test.

The idea is to get the most voltage over the 12V bank (at least 14V), whilst having the most voltage on the 24V bank (at least 18V). As JB says, you need at least 2V over the battery to get it to charge.

Set you meters up as per the picture and start with a small load like an LED, but make sure you have the same load on each side even though only one is lit up. Measure the voltages and record them. For example, when I used an LED (with 330R to protect) I got 13.08V for the 12V bank, but 23.8V for the 24V bank. The point is that I don't have enough to charge the 12V bank, but I have heaps of potential to use on the 24V bank.

OK, next try the other extreme like a car headlight halogen bulb. When I did this I measured 13.29 (after 15 seconds) on the 12V side but under 18V on the 24V side. The other problem was that the halogen took too long to light up - which means you can't switch fast enough to get any charge into the 12V side.

The best result I had was 8 x #47 bulbs. These were arranged so I had 4 pairs paralelled of 2 bulbs in series. I had the same load setup for both sides. With this arrangement it gave me 13.28V on the 12V side and still 23.16V on the 24V side. My batteries were all 130Ah big Trojans.

All 4 batteries did charge up slowly - about 0.01V every few minutes, but then they all stopped gaining after about 30 minutes.

When you do this test, it's also a good idea to have a meter on the bottom battery on the 12V side to see how fast (or if) the voltage increases.

I don't think I still had the right load, but I was starting to run out of #47 bulbs.

But then I put the same load (8 x #47) on a set of 7Ah gel-cells and it was too much for the 24V side - so that's why I said you have to pick the load for the battery size you are using.

Anyway, that's my theory which I still have to test out properly on the bench. I'd like to get some 1157's and four motorcycle batteries as the only decent fooded LABs I have are the big Trojans and it takes too long to see any impact. (And the 7Ah gel-cells are too unreliable)

Let me know how you go...


John K.

Will do. I've been thinking about what JB and Bits said....put the bulbs in parallel, but I've burned up all I had, so back to the store. I'd like to find some 4V bulbs at about 850ma or so, probably the perfect load, but time will tell. I'll give it a shot after I get some more bulbs.

Just so I get what you are saying, you had 8 bulbs on one side. 4 sets paralleled of 2 in series. I probably only need 4 on a side for the 17ah batteries, though. I'll buy as many as I can and test out different situations.

Thanks,

Lero

[John_K]

Tesla Switch

Lero,

Yes, I had to place 2 bulbs in series because the #47 is rated at 6.3V and I started to blow them if I didn't series them up.

If it were me I'd buy a bunch of different sized 12V bulbs to give me something to play with.


John K.

Will do. I've been thinking about what JB and Bits said....put the bulbs in parallel, but I've burned up all I had, so back to the store. I'd like to find some 4V bulbs at about 850ma or so, probably the perfect load, but time will tell. I'll give it a shot after I get some more bulbs.

Just so I get what you are saying, you had 8 bulbs on one side. 4 sets paralleled of 2 in series. I probably only need 4 on a side for the 17ah batteries, though. I'll buy as many as I can and test out different situations.

Thanks,

Lero

[nvisser]

The seventh opto

I was concerned about it...JB said it was for the capacitor...to run the 3524 and associated circuitry from the cap...if I'm not mistaken. Did you find something else I missed.
Leroy

You can see in the Bill and Ray scaler charger how the opto was used to only charge a large cap during the same time that the batteries charge the two charger caps during the charge cycle. During that pulse the 12v return from the battery is fed through the opto and a 27ohm 2w resistor to that large cap like the one on JB`s TS photo. They than use that cap to supply the 5V to drive the TTL chips and the scaler energy that was generated did not interfere with the TTL. The 5V return is taken between the 27ohm and the cap.
I repost the schematic that was posted by JB

[vallentin]

I was concerned about it...JB said it was for the capacitor...to run the 3524 and associated circuitry from the cap...if I'm not mistaken. Did you find something else I missed.

Leroy

I know you did, Leroy. I was pointed that to you along the way. The part with master is speaking...
Nothing more that it did not work as expected and it was disconnected.

Thank you nvisser. That's it!

[ldissing]

You can see in the Bill and Ray scaler charger how the opto was used to only charge a large cap during the same time that the batteries charge the two charger caps during the charge cycle. During that pulse the 12v return from the battery is fed through the opto and a 27ohm 2w resistor to that large cap like the one on JB`s TS photo. They than use that cap to supply the 5V to drive the TTL chips and the scaler energy that was generated did not interfere with the TTL. The 5V return is taken between the 27ohm and the cap.
I repost the schematic that was posted by JB

nvisser,

Thank you so much for that explanation. I could not figure out how that was going to work, and still don't quite understand it, but it makes some more sense to me. I did not pay attention...attention deficit disorder I guess, I was not working on the scalar charger, so I did not pay close enough attention. Thank you valentin, and nvisser for pointing me to this.

In your opinion, the 12V point is connected to the positive of the TOP capacitor, and the bottom is connected to the negative of the BOTTOM capacitor, right? When the opto is off, the capacitor is discharging to the controller and when the opto is on it is being charged and being utilized to run the controller. I'm thinking that those parts are for a 12V system, though, not a 5V one, but I could be mistaken. I do not think they were using TTL chips.

Leroy

[nvisser]

They mentioned on the interview that they used TTL. It was 1984.
I do not know if this will work on the SG chip as you need at least 8V to run it. JB said that his did not work. He used the words "goofed out".
That is why he supplied his with a separate supply. The wires you see on the photo
It should be fine for a pic that use 5V.
Doubled had a brilliant idea how to supply the PWM. The circuits are earlier on this thread.
I can refer you to it if you cannot find it

[ldissing]

They mentioned on the interview that they used TTL. It was 1984.
I do not know if this will work on the SG chip as you need at least 8V to run it. JB said that his did not work. He used the words "goofed out".
That is why he supplied his with a separate supply. The wires you see on the photo
It should be fine for a pic that use 5V.
Doubled had a brilliant idea how to supply the PWM. The circuits are earlier on this thread.
I can refer you to it if you cannot find it

I have the schematics posted by DoubleD. Yes, those schematic had some good stuff in them too.

I guess the "ground" shown in the attachment by JB was not necessarily 0 volts. It could have been +7V. Are we in agreement with where the +12 and -12 are in this schematic? There are CMOS versions of all those chips that could run at 12+ V though, I have most of them on hand. DoubleD was using 12V too, I think.

Leroy

[ldissing]

They mentioned on the interview that they used TTL. It was 1984.
It should be fine for a pic that use 5V.

nvisser,

I just hooked a cap to a 9V battery + through 60 ohms and through 20 ohms and through 10 ohms to negative, and I get 9V on the cap with any of the resistors.

The schematic on JBs website is a dead give away, the cap is connected to bat+ (through a diode) to bat- on the other side. THE TESLA SWITCH.

Let me know if you still believe the cap is at 5V and maybe we can figure out why I am wrong....again!

Leroy

[vallentin]

nvisser,

I just hooked a cap to a 9V battery + through 60 ohms and through 20 ohms and through 10 ohms to negative, and I get 9V on the cap with any of the resistors.

The schematic on JBs website is a dead give away, the cap is connected to bat+ (through a diode) to bat- on the other side. THE TESLA SWITCH.

Let me know if you still believe the cap is at 5V and maybe we can figure out why I am wrong....again!

Leroy

Not really a dead end, Leroy.
Look at the file attachment (the part evidentiated in red). Using a Zenner diode and a cap, you can drive your TTL the way you want. After the initial start from outside source.
The SW switch does that.
It connects the 555 driving circuit to the battery, immediately after the aux. source was removed.
It looks like back in the 1984 they had the same dillema.
And my guess is that they figured it out then.

Quote from the same page THE TESLA SWITCH : "I ONLY HAD AN IDEA ABOUT A PORTABLE BATTERY CHARGER YOU NEVER HAD TO PLUG INTO THE WALL FOR POWER"

If you are concerned about the driving circuit power consumption, go SMD.
But you'll need "doctored" tools, and steady hands.
No alcohool alowed

[ldissing]

Not really a dead end, Leroy.
Look at the file attachment (the part evidentiated in red). Using a Zenner diode and a cap, you can drive your TTL the way you want. After the initial start from outside source.
The SW switch does that.
It connects the 555 driving circuit to the battery, immediately after the aux. source was removed.
It looks like back in the 1984 they had the same dillema.
And my guess is that they figured it out then.
If you are concerned about the driving circuit power consumption, go SMD.
But you'll need "doctored" tools, and steady hands.
No alcohool alowed

Thank you for the excellent drawing...marked in red....there are so many circuits out there. I wasn't talking about that one in particular (but more towards Bill and Rays drawing), but the one presented by JB that was drawn by somebody else and reposted by nvisser...it did not have a zener, so the cap charges to full battery voltage.

I have to be concerned about power consumption...we all do, but without the zener, it does charge to the battery voltage. I was only talking about the circuit as presented, it is not TTL. Put a 5V zener in (TTL V) and you are good to go at 5V, but the zener will dissapate the "excess" voltage, hence another loss in the circuit.

Leroy

[nvisser]

I have to be concerned about power consumption...we all do, but without the zener, it does charge to the battery voltage. I was only talking about the circuit as presented, it is not TTL. Put a 5V zener in (TTL V) and you are good to go at 5V, but the zener will dissapate the "excess" voltage, hence another loss in the circuit.

Leroy

Ask John K to mail you the Bill and Ray interview and listen for yourself.

[ldissing]

Ask John K to mail you the Bill and Ray interview and listen for yourself.

I've heard the interview, I think John already emailed it to me. Just be careful if you hook it up to a TTL, or worse, a PIC that is expecting 5V ...that is all I'm saying. I've done the tests for it, so you are on your own. No more discussion needed.

Leroy

[nvisser]

Hi Carroll
I was playing around tonight with the TS diagram to see if I could get rid of some diodes or transistors to see if I could get less voltage drop over the components. I came up with something like a double scalar charger where batteries replace the caps and then I remembered that I have seen that before. So I went back on the forum and found your circuit just like that. You did not report any more on results from that circuit.
See post 1825
As there are only 2 transistors and the load on a given time between the series batteries and each one of the parallel batteries the voltage drop over the 12v battery should be higher then 14v. If you loose say 2.3v over each transistor (I think that was what I measured when switched on for a few seconds)and then use a high wattage load with a low resistance that also will have a low voltage drop over it (say 1.4v), than the available voltage should be
24-2.3-2.3-1.4=18V . A lot don’t you think.
That give you a much better criteria than that JB asked for.
Now i think the switching pulse width is the important one and since I am still waiting for my pic I cannot really test it.
If the pulse width is to long you will drain your series batteries to fast specially as the load is low impedance. JB spoke about 1usec pulses but I think he meant 1msec as the rising edge of a opto is already 3usec. I don’t think with a 1msec pulse you will even see any glow on the bulb and the voltage drop over it is anyway to low. The low impedance bulb can easily let a lot of current through and that is why I think he says we must see how much potential we can switch before any current starts to flow.
Did any of you guys tried very short duty cycles of a few msec on a 500msec period(2 Hz)or 2 cycles/sec?
So its two things. 1) more than 14v available over charge battery
2) Fast switching before current flow
Then of course there is the Tom Bearden explanation that we need a very sharp leading edge on the pulse which is a bit difficult with a opto.
John Stone described today that the cap and diode on the base improves the leading edge.
The way I understand it is that the pulse has to reach more than 0.6v – the drop over the diode, before it will start to switch the base. By that time the pulse is past it slow rising edge slope and we get faster switching. The cap I’m not sure of, It can keep the transistor on a bit after the pulse collapsed. Not sure why it is necessary. But I can be completely wrong on this!!
Then again JB said the TS is a current pump!!
I just can’t win

[ldissing]

Did any of you guys tried very short duty cycles of a few msec on a 500msec period(2 Hz)or 2 cycles/sec?

Then again JB said the TS is a current pump!!

I believe that JB said the scalar charger is a current pump! Doesn't necessarily mean that the 4 battery TS is not, but that is what he said.

I've tried everything, including short "on" times of a few msec and a longer delay like anywhere from 128ms to 1 sec and I can do longer. Things are looking up with the 12V 20W bulbs I secured today. A little charging going on...so it seems. Time will tell.

Leroy

P.S. nvisser, I don't mean to rag on you. I'm still learning like the rest of us all.....Someday, JB will decide to come back...hopefully, and set us straight! I hope it is soon, running out of ideas.