SG Cap Discharge Difficulties Musings

[Blargus]

Well the world keeps marching on and getting crazier since I have been messing with the SG over the years. I have succeeded in making a working replica of the one in the SG handbook and I get about 75% system COP and about 110% battery COP. I have made Peter Lindeman's cap discharge circuit and the timing part seems to work but when I hook it up to the capacitor and switching transistors it doesn't trigger the capacitor while the machine is running.

I'm wary of using it because I'm not that good at electronics and I'm a little lost for troubleshooting and also I don't want to mess around with a big capacitor and chance the discharge mechanism not working and the capacitor building voltage and exploding.

So I plan to continue with the spikey output to see what I can do perhaps with larger batteries or three battery method or something...but I wanted to ask if there's a foolproof cap discharge method maybe:

1. The momentary roller wheel switch seems I can find only 25 amps which seems too low and are $50 maybe.
2. A solid state relay in 100 amp range and use timing circuit to activate it? Again more than $50.
3. I have a 100 amp auto relay can I use the timing circuit from SG book Peter L's discharge circuit to activate it? Probably need to use transistor and take current from run battery? But wasted current to activate it...

Seems none is foolproof, was seeing if I could make a zener diode combined with a latching relay to interrupt trigger signal if capacitor builds too high 35 v or so still messing with it.

Anyway just musing the way things are going we're not post apocalyptic yet but sometimes I wonder where its all going. Thanks if anyone has any advice if I get good results I'll try to post.

[Aaron]

The pulley wheel with mechanical copper contact switch works perfect. Cheap and almost impossible to get it wrong. There's no components - just connections. If I can find a pic of my old bicycle wheel setup, I'll post it. I know I did in the past but Google has scrubbed most history so can't even find anything there anymore. If I can post it, you'll see how ridiculously low tech it was - as close to a scotch tape and bubble gum setup as you can get and it worked perfect.

[Blargus]

Thanks Aaron, right now leaning towards a small 300 mA switch to trigger the 100A relay but your description sounds right up my street low tech and cheap...if you can't find the photo can I ask what kind of switch, like the Delrin wheel on page 29 of SG book 2?

I wonder even opening up my 100A relay I could make a cam to close the contacts and save the relay coil activation energy...

I can envision etching a groove in my bicycle wheel axle area for a small belt and have room for a reduced wheel to drive something...will post if I get good result.

[Aaron]

The pulley method is in John's diagram that I posted in the cap dump thread with Mario.

[Blargus]

Thanks Aaron, found the photos from Mario's thread, hm silver contacts huh oh well. But I see the roller and contacts probably a better idea than my state of the art concept I am innovating took hours at the machine shop I can send you the CAD files if you sign an NDA :DSC01670.JPG

[Aaron]

Silver contacts are optimum but not necessary. I used simply copper strips.

[Blargus]

Ok, trying that route my progress so far with a cart wheel and some bearings, shafts and large o ring belts. So far got the cart wheel to I hope spin reliably at about 60 rpm without losing the belt. Tight O rings arranged on the bicycle wheel hub seems to work ok as a poor man's groove for another o ring belt. Will post if it works eventually:

[Blargus]

Here's my cart wheel upside-down hacked into relay deal for the record. Ran for a few hours so far no problems. I can adjust the discharge a bit by moving the relay up and down relative to the cam wheel seem to be able to mitigate the time the capacitor stays at battery voltage. Happens to trigger at about 75 rpm which lets it build to about 18 V before discharge. You have to size the pulley wheels based on your stable SG rpm to get discharges per second (ratio of input wheel to output wheel to reduce or increase rpm), or you could double it with a double cam or use a smaller cam to push the relay etc.

Had a separate question if anyone's interested related to Peter Lindemann's cap discharge timing circuit which I got to work just not the FETs part. I can use it to trigger a relay to discharge a capacitor (after my foray into this pulley deal), can I power his whole 555 circuit on the capacitor with a 7812 voltage regulator or will there be some strangeness with the dumping of the power source every second or so that might negatively affect the timing circuit? I was thinking as on output of a solar panel to charge a battery.

edit: some detail I made the pulley wheels from a cart wheel and two plastic discs from hardware store I cinched in a drill using a bolt and used a soldering gun to get a wide enough even enough groove--the belts are large o-rings from hardware store, the shafts are also from hardware store hobby brass pipe and a length of rod dunno where it came from bearings fit the shafts and wheels pretty well where it doesn't I wrapped electrical tape on the shaft to increase diameter. Used wood pieces glued together to hold the bearings to the frame with a single bolt and an adjustable area so I could move the shafts up and down on the frame a bit.

[Blargus]

Here's a capacitor safety circuit made out of a dpdt electronic 12V latching relay and a 30V zener diode. If the capacitor gets to 34V the zener breaks down and conducts current from the capacitor through the relay which is connected so that the trigger winding from the SG coil after the base resistor has a path when its reset and the path gets interrupted when it's tripped by the diode breaking down at 34V on the capacitor.

So it's like a breaker and makes the SG more foolproof, the machine won't run unless the relay is reset if its triggered by the capacitor reaching 34V. If the capacitor reaches 34V or so it will slowly wind down with no power to the transistors and the capacitor won't get any higher than the voltage it trips the relay.