Hi Guys

Attached a schematic from a JB patent that I want to build, has any one here built this and what was the out come?

Also can anyone please help me with the pin numbers on the PWM - SG3524N chip for the corresponding connections to the chip as shown, (before I start making a PC-board.)

I think I have some of the pins right, if someone can please confirm it for me.

Also must some of the other pins not used be tied to ground or positive rail.

Thanks

Virus

Attached Images

	bedini_solid1.jpg (30.8 KB, 28 views)
	PWM - SG3524N - Pins.jpg (50.0 KB, 30 views)

Hi Virus,

There is quite a bit of discussion about this chip in the "Use for the Tesla Switch" thread. There is also a schematic posted by John B. Here is a link to that post.

Use for the Tesla Switch

Hope this helps,

Carroll

PS: We did learn that making the timing cap too large will cause the chip to act up and not change states. Try to go no larger than 1 mf. If you want to go for longer times just increase the timing resistance. That will allow for longer times and the chip will still work ok.

Hi Thanks Carroll

Think I will start on it now, anything else I need to be aware of, isn't there a thread that discussed this build, I look but could not find any?

I have a battery bank with 8 -150 amp 12 v Telko batteries, which radiant charging system would be better to charge them.

Is this the same/similar circuit used in the R-Charger?

Thanks for the help.

Virus

Hi Virus, Depending on exactly how you want you're circuit configured the
SG3535 chip can be better than the SG3524, the 3235 is only rated for 35 volts as
compared to the SG3524 at 40 volts but, the SG3524 has just single transistor
outputs, whereas the SG3525 has two transistors on each output arranged as a
basic mosfet driver. the SG3525 can drive mosfets much better, is very similar
but in my opinion a far superior chip. The features are very similar and both are
very good chips, I've used both and prefer the 3235 because of it's outputs.

I use one SG3525 to drive two mosfets on a variable frequency inverter for quite a while
now, and I just drive the mosfet gates directly with the 3525 outputs no
resistors just a connection. and it works just fine making nice square pulses up
to fairly high frequency when the gate drive from the chips becomes a bit
lacking in power.

Anyway I built the two phase cap dumping circuit, you'll have a lot of fun with
the high side drivers. The thing I didn't like about the circuit is it seems to be
only 50% duty for each stage, which means the solar panels can sit idle while
the cap is dumped. With my build the high side mosfets did get an inadequate
gate drive in some circumstances as well, causing them to heat up and dissipate energy.

If the circuit is tuned so that with at full panel power the cap is filled just
before dumping then in low light the dump is less than full voltage so less
effective. When the cap is discharged the panels sit idle. I found a way
around that but it was still not ideal. I didn't like the way the panels duty is
reduced to only 50% effectively. So the possible output of the panels is more
or less halved in my opinion..

Here was my small test circuit in operation.
Two Phase Pulser 2.wmv - YouTube

I think it needs some kind of way to be adjustable frequency if that is not
already in the schematic. I think microprocessor controlled is the way to go.

Cheers

P.S attached is the way I use 3424 as far as pins goes. But that circuit is for
HF timing values.

Thanks Farmhand

Will keep the SG3535 chip in mind you mentioned, have already received the SG3534 stock will try them first.

How did you get around the 50 % use of the solar panel as you mentioned, in the video.

I have an idea but not sure if it will work, by building 2 charge/dump circuits, connect them vice versa with the optos, iaw while the one charges the other side dumps and vice versa, that can bring the circuit closer to 100 % use of the solar panel, I see-in your vid that the charge time of the cap takes a long time and the dump to the battery is immediately.

Can one connect an LDR or something to the chip that will sense the light intensity and adjust the pulse rate accordingly?

Thanks

Virus

Hi Virus, What I did was use a big capacitor on the solar supply so that when
the cap was dumping the solar panels charged the capacitor rather than sit idle,
then when the device reconnected the panels/cap to the main dumping/charging
cap the panel/supply cap charged the dumping cap very quickly, I made the
panel/supply cap much higher value than the dumping capacitor. Not ideal but
it did improve low light/low frequency charging. With solar panels we will
always need to use less than 100% duty in order to keep the panel voltage up
for better power transfer.

I don't want to put you off at all, it is a good circuit to build and I used it only
with low frequency.

Cheers