The Ultimate Cap Dump

[Faraday88]

What about this ssr?

Hi Why _ me ,

SSR in Cap dump ciruits have to be tried out ..they look promising as far as the ease of switching inputs are concerned, however, as John K pointed out, the Discharge peak currents are severe to handle and hence you may need to match several SSR in parallel for reliable operation. i need to tinker on your schematic to get an understanding of what you intend to do!!
Rgds,
Faraday88.

[John_Koorn]

Why-me, the cap size looks more appropriate but now I can't see a connection to the negative so I don't think it will charge. You don't need that series cap and diode because as long as the cap is higher in voltage than the battery it will dump down to the battery voltage.

I dont want want to rain on your parade but it seems that what you're trying to do could be achieved better with what Peter Lindemann demonstrated at the last conference. I suggest you should try and get access to a computer where you can watch the presentation. From everything I've seen Peter's system is by far the best.

John K.

[John_Koorn]

Forum members, I removed why-me from the forum because of his inappropriate last post (which I deleted) and because he chose to delete all of his posts on this thread. Faraday and I tried to help him out with his project and all we got back was selfish whining because he didn't get what he wanted. He also pulled the same trick on Energetic Forum.

That sort of behaviour ticks me off and I have no time for people like that.

John K.

[AlvaroHN]

Cap dump

Hello,

I am playing with very low voltage system at the moment, and the comparator circuit that I have is not good for this, because the energy in and out is lower than normal so the comparator circuit eats up the CAP charge.

I made the zener+scr cap dump, it works but for low capacitance on the cap, if the cap is 100uf or more, it gets latched (I was with solid state in that test, maybe the ssg with rotor might allow the scr to unlatch.

But anyway, I came up with this idea to make that scr dump bigger caps,



the idea is to have a regular cap, C1, and that charging a very small cap (C2) through a resistor, the SCR+zenner discharges the small cap into the gate of the mosfet which opens and discharges C1.

I have to try it but I was hoping that someone with electronics knowledge could take a look to see if everything is connected the right way.

best,

Alvaro

[Bradley Malone]

the only input i would have on the ciruit is you may have to make the gate to source resistor variable as C2 could discharge faster/slower than you want if the values are set so you need a variable discharge time on C2 to allow the discharge of C1 time to enter the battery. If you have an o-scope then you could adjust the resistors while watching C1 charge and discharge until you had the correct timing.

[AlvaroHN]

didn't work

But anyway, I came up with this idea to make that scr dump bigger caps,

[ATTACH=CONFIG]5739[/ATTACH]

the idea is to have a regular cap, C1, and that charging a very small cap (C2) through a resistor, the SCR+zenner discharges the small cap into the gate of the mosfet which opens and discharges C1.

Boooooo!!!! .... it didn't work!

the zener connected as the drawing didn't do anything, I moved it from negative to positive of C2 and it did dump the cap... but just once, it did get latched.

It seems that once the SCR turned ON it wont turn off until C2 is completely discharged, so the gate of the MOSFET is seeing a voltage curve starting 15V or so, but slowly going down to 0, so the MOSFET starts fully open and gradually start to raise its internal resistance, so basicly it didn't work, it latches the SCR and gets the MOSFET conducting all the time.

[min2oly]

Boooooo!!!! .... it didn't work!

the zener connected as the drawing didn't do anything, I moved it from negative to positive of C2 and it did dump the cap... but just once, it did get latched.

It seems that once the SCR turned ON it wont turn off until C2 is completely discharged, so the gate of the MOSFET is seeing a voltage curve starting 15V or so, but slowly going down to 0, so the MOSFET starts fully open and gradually start to raise its internal resistance, so basicly it didn't work, it latches the SCR and gets the MOSFET conducting all the time.

Hi Alvaro,
This is the problem with the SCR cap dump. If you're insistant on using it, I have found that adjusting the "input" from the SSG makes a difference. Big, spaced, single pulses from the SSG will give it time to un-latch. If your goal is to switch a "lower energy" why not use the breakdown voltage of a single transistor collector - emitter...
KR - Patrick

[AlvaroHN]

Hi Alvaro,
This is the problem with the SCR cap dump. If you're insistant on using it, I have found that adjusting the "input" from the SSG makes a difference. Big, spaced, single pulses from the SSG will give it time to un-latch. If your goal is to switch a "lower energy" why not use the breakdown voltage of a single transistor collector - emitter...
KR - Patrick

Hi Patrick,

I have used the breakdown transistor cap dump a long time ago, but I remember that it just dumped like 2 volts, from say 11V to 9V. I might be wrong as I didn't have an oscilloscope back then.

I was already thinking in other way, related to the breakdown of transistor, the esaki effect led flasher, it is a led flasher using the reverse break down of a transistor, maybe I can change the led for an opto, and the opto fully opening the mosfet.

Best,

Alvaro

[min2oly]

Hi Patrick,

I have used the breakdown transistor cap dump a long time ago, but I remember that it just dumped like 2 volts, from say 11V to 9V. I might be wrong as I didn't have an oscilloscope back then.

I was already thinking in other way, related to the breakdown of transistor, the esaki effect led flasher, it is a led flasher using the reverse break down of a transistor, maybe I can change the led for an opto, and the opto fully opening the mosfet.

Best,

Alvaro

Hi Alvaro,
yes, different transistors have different breakdown characteristics. Test what you have on hand using voltmeter, if too fast, swap out the CAP for a larger CAP that will not charge as fast... Also you can add transistor in series...

I like your opto idea! Let us know how it works out. Keeping things simple has always been my drive.
KR - Patrick

[AlvaroHN]

Ok so the transistor break down + opto + MOSFET worked!!!

First I used 4000uf 25V CAP for C1 but it wasn't dumping all, just 1 or 2 volts.

then I changed that cap for a photoflash cap 300v 220uf and now it works like a charm, the variable resistor before the positive of C2 change the dump voltage higher resistance higher dump voltage, and viceversa. Then I also added more caps and it worked great untill 800 uF worth of caps, I guess that to dump larger capacity with this circuit the only change would be to make C2 larger to have longer ON TIME on the OPTO.

I also added a trim pot between pin 2 of opto and negative rail, my intention was that it would change ON time, but it does not work quite as expected, it does change ON time a little but if the resistance is too much the opto won't switch.

Here is the circuit and some pics. in the pic with the scope it is dumping twice a second from 17v down to 8v.





best,

Alvaro

[min2oly]

Ok so the transistor break down + opto + MOSFET worked!!!

First I used 4000uf 25V CAP for C1 but it wasn't dumping all, just 1 or 2 volts.

then I changed that cap for a photoflash cap 300v 220uf and now it works like a charm, the variable resistor before the positive of C2 change the dump voltage higher resistance higher dump voltage, and viceversa. Then I also added more caps and it worked great untill 800 uF worth of caps, I guess that to dump larger capacity with this circuit the only change would be to make C2 larger to have longer ON TIME on the OPTO.

I also added a trim pot between pin 2 of opto and negative rail, my intention was that it would change ON time, but it does not work quite as expected, it does change ON time a little but if the resistance is too much the opto won't switch.

Here is the circuit and some pics. in the pic with the scope it is dumping twice a second from 17v down to 8v.

[ATTACH=CONFIG]5740[/ATTACH]

[ATTACH=CONFIG]5741[/ATTACH]

best,

Alvaro

Hi Alvaro,
very nice! I might have to round up and opto and give this a go.
Thanks - Patrick

[AlvaroHN]

Hi Alvaro,
very nice! I might have to round up and opto and give this a go.
Thanks - Patrick

Your welcome Patrick , I tested it a little more and its great. R2 changes dump voltage from min 8v and Max I don't know... I went until 20+volts and stopped there because the irfp260 are 20v Max Gate voltage.

R4 changes min dump voltage, the higher the resistance the shorter the ON time of the MOSFET.

The transistor is a 2n2222.

If you use a larger CAP for C1 you will need to use a larger one for C2. I am using 200~800uf for C1 and 10uf for C2.

[AlvaroHN]

I added more CAPs in my cap dump, now I am with 3000uF on C1, and had to change C2 to be able to dump C1 correctly.

This is the clean up schematic:



I suspect that if I add a diode from positive rail to R1 the ON time of the Opto and Mosfet could be greater, because I think that when C1 is dumping and the voltage get lower than C2 voltage C1 drains back C2 and the dump is interrupted. For example if I dump with a short circuit instead of CHARGE BAT, C1 never reaches 0 on the dump, it dumps from whatever voltage is set with R2 until aprox 4 volts.

I will try to add that diode and see what happens.

Other thing that I have doubts about is that in all mosfet cap dump I have seen, when there is an opto, the opto is triggering a NPN transistor and that npn is triggering the mosfet/s. In my circuit I am triggering the mosfet just with the opto... and it is working.

best,

Alvaro

[RS_]

Hi Alvaro,

you should add a 15 to 18 Volt Zener from the gate to ground on the Mosfet, so that you can pulse at over 20V and not kill the gate........

also, the opto by it's self, may not be able to turn on the Mosfet fully, and/or turn it on fast enough for that really sharp fast dump that you want to see......

[AlvaroHN]

Hi Alvaro,

you should add a 15 to 18 Volt Zener from the gate to ground on the Mosfet, so that you can pulse at over 20V and not kill the gate........

also, the opto by it's self, may not be able to turn on the Mosfet fully, and/or turn it on fast enough for that really sharp fast dump that you want to see......

Tnx for the advice RS_ !! do you think a 2n2222 would be ok between opto and mosfet?

Also do you think I should add a diode between positive rail and R1 ???

best,

Alvaro