Pulse Rectification question / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Pulse Rectification question

Move Thread

LAN_403

Thu Nov 09 2006, 12:31PM

Banned on April 7, 2007
Registered Member #277 Joined: Fri Mar 03 2006, 10:15AM
Location: Florida
Posts: 157

Nice to be back! Here is the short version. I have a hv spark jumping a 1-2 cm spark gap every few seconds, sending the spark energy into an inductor/cap combo causing it to ring the inductor for a few cycles each spark. Next, I'm using a standard full wave bridge rectifier (encapsulated, not discrete components) to rectify the 'ringing' of the inductor output which then charges a low voltage cap. This approach works okay, but I suspect I'm losing efficiency in the full wave bridge due to standard diode voltage drop. QUESTION: Can anyone recommend what type diodes I could use to build a full wave bridge rectifier that would have the least amount of power loss? ie: I'm looking for the most efficient full wave rectifier possible. (I've tried Schottky diodes in a full wave bridge configuration, didn't seem to work in this pulse type rectification scenario) Thanks in advance for your reply! CM

Sulaiman

Thu Nov 09 2006, 06:57PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

A spark of 1-2 cm means several kilovolts
the forward voltage drop of any common diode is negligible compared to kV
things that come to mind;
1_Reverse recovery time ... speed. Use fast/hyperfast/ultrafast etc. diodes
2_Reverse capacitance .. a schottky is close infinitely fast,
but has relatively high capacitance when not conducting
which absorbs the spark current
3_Maximum reverse voltage .. most important in this case
roughly, the maximum voltage that you can get on the capacitor is the reverse breakdown voltage of the diode(s)
Common schottky diodes have a low ( below 100v) breakdown voltage

So you need fast low-capacitance high-voltage diodes
currently I use BA159, HVRT300, BY8422 etc. for this kind of thing

...

Thu Nov 09 2006, 07:00PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Well... I would say that most of your energy is being lost in the spark gap, but... How much voltage are you getting out of the inductor? A normal diode only has 1-2v across it...

Also, what frequency does the lc circuit resonate at? Normal diodes are not good for anything high speed (I wouldn't use one for anything >1khz). You might try switching over to fast diodes...

Fri Nov 10 2006, 12:39AM

Banned on April 7, 2007
Registered Member #277 Joined: Fri Mar 03 2006, 10:15AM
Location: Florida
Posts: 157

Sulaiman:

I've checked the normal suspects like DigiKey, Mouser, Parts Miner, etc and also Googled for the diodes you mentioned, but not much luck other than companies with shipping originating from Asian. I am in the USA, would you be kind enough to tell me a USA distributor that I could purchase these diodes from? CM

Sulaiman

Fri Nov 10 2006, 09:25AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Off hand, no. BUT if you look through this forum for high voltage diodes I'm sure there are many useful references.

The actual setup - I'm guessing some electrical energy source charging up a capacitor until a sparkgap 'fires' and connects the capacitor to the inductor, hence the oscillations.
To efficiently collect the energy in a low voltage capacitor will require voltage transformation,
so I would use a secondary on the oscillating (Tank) inductor
and use low-voltage fast diodes - (the schottky diodes)
A VERY large transformation ratio is required (Thousands of volts : volts)
e.g. If the resonating inductor has 100 turns, the secondary could be a single turn - at a distance!
Alternatively you could use something like a xenon flash tube trigger transformer - backwards.
I can't imagine direct rectification of the hv resonance to the capacitor could ever be efficient
EDIT : Thinking about it ... IF you can get a flash unit from a disposable camera;
The trigger transformer is well designed - just good enough to work, but is useful
The diode that charges the flash capacitor is fairly fast and goood for over 300V
The xenon flashtube can be used as a fairly reliable sparkgap (around 2kV)
P.S. The idea of 'catching' sparks (miniature lightning bolts) probably goes through the mind of many many people - turns out to be a lot of effort for little return usually - hence no commercial lightning-powered-batteries.

Steve Conner

Fri Nov 10 2006, 04:38PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The diodes Sulaiman mentioned are European parts made by Philips or whoever, so they might be hard to come by in the US. I seem to remember Steve Ward mentioned some diode called a SF1500 or something. The UF4007 (800 PIV I think) or UF5408 (1000 PIV) spring to mind as fast diodes available in the states.

You seem very concerned about the efficiency of something that's inherently a very inefficient process. The huge impedance mismatch between "A voltage high enough to jump a 2cm spark gap" and "A voltage low enough for your low voltage capacitor" is probably the root cause of the problem, rather than bad diodes. Maybe discharging the spark into the HV end of an ignition coil and rectifying the output of the LV winding would work better.

I hope this isn't for any free energy type of application...

*edit* CM assures me that it's not any of this perpetual motion BS: it's a technique for harvesting naturally occurring ions in the air, which is known to work and I've seen cited elsewhere.

I don't see it solving the world's energy problems though: I once read that the total power generated by atmospheric electricity over the whole surface of the planet, including lightning, is only a few hundred megawatts.

*edit again* it is in fact very difficult to convert high voltage low current DC efficiently to a lower voltage, no matter what technique you use. Essentially, you need a switched-mode power supply, and in this case, the spark is your switch.

I think the method I suggested, where you discharge the spark into the HV winding of an ignition coil, and rectify the output of the LV winding, would be most likely to work. Or taking it one step further, you could try a Tesla coil run backwards with a rectifier on its primary. If you separate the topload from the secondary coil by a spark gap, and charge the topload until the spark gap breaks down, this will trigger an oscillation, and if you tune it right there will be an instant where all the energy will end up in the primary capacitor, transformed to a lower voltage. You could then open a transistor switch at just the right time to trap it there.

Tue Nov 14 2006, 11:02AM

Banned on April 7, 2007
Registered Member #277 Joined: Fri Mar 03 2006, 10:15AM
Location: Florida
Posts: 157

Conserve Ten:

Thanks. I recently received some of the BA159's for testings. Btw, you have Private Message from me. CM

Marko

Tue Nov 14 2006, 01:14PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I would simply try an AC flyback ran backwards, instad of TC; i think core would help a lot increasing efficiency.

Higher coupling would allow much faster energy transfer and less losses than with TC (wich would ring lots of cycles to transfet the energy adn waste it).

Ignition coil has lossy core and would probably be least efficient.

Regarding atmospheric charge collectors; just any large, ungrounded object put high enough in atmosphere will build up charge from triboelectric effect from air (just like clouds do), maybe little from influence of clouds and etc.

It actually converts kinetic energy of moving air into electricity, although with uber poor efficiency (simple wind turbine would do same thing thousands times better).

My proposal was to put a ballon wrapped in metal foil on some height and wait for it to build some charge.

If you manage to get something from it you could probably make a led flasher or power some small stuff for few seconds. (don't do it in stormy weather

)

I don't know if someone was actually silly/brave enough to do it but could be an interesting project.

It would probably work but not be of much use O_o

Nik

Tue Nov 14 2006, 10:43PM

Registered Member #53 Joined: Thu Feb 09 2006, 04:31AM
Location: Ontario, Canada
Posts: 638

I ran a 30awg copper wire up a kite string (on a clear day) jsut to see if i coudl get a charge. My volt meter read ~300v but this would drop quickly after taking the reading.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob