The sparkling electrolytic capacitor/rectifier
Tesladownunder, Sat Apr 15 2006, 04:11PMThis is an unusual experiment which is very simple and is detailed on my site based on info from here. Basically two aluminium electrodes in a sodium bicarbonate solution will glow and sparkle with sufficient applied voltage. One electrode is aluminium foil and the other is an aluminium bar.
Here's my take on what is happening. If you insert an aluminium electrode in a solution then it will form an oxide layer and when the metal is positive will block current flow. This, as I understand, is the basis for electrolytic capacitors and the reason they are polarized. Here we have two aluminium electrodes which is the equivalent of back to back electrolytics which are then non-polarized.
Initially there is a lot of current flow e.g. 10V 1A but this rapidly drops as the oxide layer builds up. After a few minutes there is little flow even at 100 VAC. There is a diffuse glow starting at about 60 VAC but the sparkling becomes more prominent with increasing voltage.
Exceed 100VAC, however and things start to happen fast. At 150 VAC the current rises to 5 A (750 W) and widespread sparking is seen particularly at the electrolyte/air/electrode junction. I interpret this as breakdown of the oxide layer with voltage. Note that the alkali solution boils rapidly with the heat. Eye protection is very important here.
Peter
Re: The sparkling electrolytic capacitor/rectifier
vasil, Sat Apr 15 2006, 05:05PM
Do you think that would work as a gap in a TC tank, at mains voltages?
vasil, Sat Apr 15 2006, 05:05PM
Do you think that would work as a gap in a TC tank, at mains voltages?
Re: The sparkling electrolytic capacitor/rectifier
Tesladownunder, Sat Apr 15 2006, 05:47PM
Doubt that it would work as a gap. Might at a pinch be able to be used as a tank cap in the same way as a non-polarised electrolytic. Might even have self healing properties as the oxide layer will reform.
Capacitance of this simple arrangement was a huge 0.33uF, leakage resistance was 1.0Mohm and there was 0.15V of potential across it. Possibly different aluminium alloy types making an inefficient battery. It should be able to be made low inductance and should hold 50 - 100V AC
So if you use a 20 cap MMC arrangement to get 20 kV at 0.016uF that should be OK. Could be an interesting project.
This photo shows the glow at 60 VAC. It is best seen if power is switched off and on.
Peter
Peter
Tesladownunder, Sat Apr 15 2006, 05:47PM
Doubt that it would work as a gap. Might at a pinch be able to be used as a tank cap in the same way as a non-polarised electrolytic. Might even have self healing properties as the oxide layer will reform.
Capacitance of this simple arrangement was a huge 0.33uF, leakage resistance was 1.0Mohm and there was 0.15V of potential across it. Possibly different aluminium alloy types making an inefficient battery. It should be able to be made low inductance and should hold 50 - 100V AC
So if you use a 20 cap MMC arrangement to get 20 kV at 0.016uF that should be OK. Could be an interesting project.
This photo shows the glow at 60 VAC. It is best seen if power is switched off and on.
Peter
Peter
Re: The sparkling electrolytic capacitor/rectifier
Marko, Sat Apr 15 2006, 07:45PM
cool, but I doubt it could be really used for anything like this.
Leakage resistance of 1M is low and ewlectrolyte resistance is relatively high.
Also because of large volume of entire capacitor there are lot of corona losses like with bottle caps (if they could be made to withstand several kilovolts).
For now you can just use one small 330nF 100V capacitor hat is far more efficient, smaller and dry, easily replacing such bottle-electrolytic cap.
Real electrolytics (polarised use bunch of foil rolled around the center with some porous insulating material between and dipped into electrolyte.
Electrolyte itself is the another electrode and body of the cap will not form the oxide, so the cap conducts when wrongly polarised.
Maybe you could try something similar (use bunch of rolled aluminium separated with papper, and let the another electrode be of some other conductive material.
Anybody for home-made high energy electrolyc caps?
good luck..
Marko, Sat Apr 15 2006, 07:45PM
cool, but I doubt it could be really used for anything like this.
Leakage resistance of 1M is low and ewlectrolyte resistance is relatively high.
Also because of large volume of entire capacitor there are lot of corona losses like with bottle caps (if they could be made to withstand several kilovolts).
For now you can just use one small 330nF 100V capacitor hat is far more efficient, smaller and dry, easily replacing such bottle-electrolytic cap.
Real electrolytics (polarised use bunch of foil rolled around the center with some porous insulating material between and dipped into electrolyte.
Electrolyte itself is the another electrode and body of the cap will not form the oxide, so the cap conducts when wrongly polarised.
Maybe you could try something similar (use bunch of rolled aluminium separated with papper, and let the another electrode be of some other conductive material.
Anybody for home-made high energy electrolyc caps?
good luck..
Re: The sparkling electrolytic capacitor/rectifier
Hazmatt_(The Underdog), Sat Apr 15 2006, 08:37PM
This sounds like a lot of fun actually. Maybe I'll play with it when I have my big supply finished, then I can anodize the electrodes properly for capacitor construction.
Now I have something to look forward to over the summer :p
Hazmatt_(The Underdog), Sat Apr 15 2006, 08:37PM
This sounds like a lot of fun actually. Maybe I'll play with it when I have my big supply finished, then I can anodize the electrodes properly for capacitor construction.
Now I have something to look forward to over the summer :p
Re: The sparkling electrolytic capacitor/rectifier
Tesladownunder, Sun Apr 16 2006, 01:50AM
Corona is trivial at 10kV in a high current primary circuit. What I have here is a non-polarised electrolytic as back to back electrolytics. Of course commercial caps will be better, but the values have to be chosen properly to allow doubling up and a sufficient voltage rating.
For example take 132 times 4uF 330V. Half of these connected in reverse and hence for each half cycle, not contributing to to capacitance or voltage. This 132 cap MMC would be 0.015uF, 21780VAC rated. There may be significant issues with internal resistance and inductance. Self healing is possible but might need a slowly increasing low frequency AC voltage to do this.
Peter
Tesladownunder, Sun Apr 16 2006, 01:50AM
Firkragg wrote ...
..Leakage resistance of 1M is low and ewlectrolyte resistance is relatively high.
At 100V leakage current is 0.1mA, not a big deal in a primary circuit...Leakage resistance of 1M is low and ewlectrolyte resistance is relatively high.
Firkragg wrote ...
Also because of large volume of entire capacitor there are lot of corona losses like with bottle caps (if they could be made to withstand several kilovolts).
For now you can just use one small 330nF 100V capacitor hat is far more efficient, smaller and dry, easily replacing such bottle-electrolytic cap.
Real electrolytics (polarised use bunch of foil rolled around the center with some porous insulating material between and dipped into electrolyte.
Electrolyte itself is the another electrode and body of the cap will not form the oxide, so the cap conducts when wrongly polarised.
Also because of large volume of entire capacitor there are lot of corona losses like with bottle caps (if they could be made to withstand several kilovolts).
For now you can just use one small 330nF 100V capacitor hat is far more efficient, smaller and dry, easily replacing such bottle-electrolytic cap.
Real electrolytics (polarised use bunch of foil rolled around the center with some porous insulating material between and dipped into electrolyte.
Electrolyte itself is the another electrode and body of the cap will not form the oxide, so the cap conducts when wrongly polarised.
Corona is trivial at 10kV in a high current primary circuit. What I have here is a non-polarised electrolytic as back to back electrolytics. Of course commercial caps will be better, but the values have to be chosen properly to allow doubling up and a sufficient voltage rating.
For example take 132 times 4uF 330V. Half of these connected in reverse and hence for each half cycle, not contributing to to capacitance or voltage. This 132 cap MMC would be 0.015uF, 21780VAC rated. There may be significant issues with internal resistance and inductance. Self healing is possible but might need a slowly increasing low frequency AC voltage to do this.
Firkragg wrote ...
Maybe you could try something similar (use bunch of rolled aluminium separated with papper, and let the another electrode be of some other conductive material...
The advantage of home construction is that both caps can be made in one with both electrodes being aluminium. It would only take 6 dozen beer cans then......Maybe you could try something similar (use bunch of rolled aluminium separated with papper, and let the another electrode be of some other conductive material...
Peter
Re: The sparkling electrolytic capacitor/rectifier
Marko, Sun Apr 16 2006, 12:29PM
It is OK for 100V but if we want to really use it in SGTC (without stacking hundreds of such bottles in series) it must work at least in lower kV range and with 1M leakage you get tens of watts of dissipation (resulting in boiling electrolyte).
Corona may take some uA away (like foil-beer bottle caps) but it can be omitted.
Only problem for now seems how to make such a cap withstand higher voltage without breakdown (and also lower leakage resistance if possible).
How is that actually done in commercial electrolytic capacitors?
Maybe another electrolyte needs to be used, rather than sodium bicarbonate?
If you really make such thing work in SGTC you would have regenerative dielectric wich is great thing, despite maybe-not-the-best pulse charachteristics
I meant this if you want to make polarised caps (wich is easier as you need to roll only one electrode) of some more serious capacitance for pulse power experiments
You can also separate two layers of aluminium foil separated with paper (just to ensure they are not touching and that there is electrolyte between) to make a non-polarised high energy cap (evil grin)
It may be an interesting experiment also.
cheers, Marko
Marko, Sun Apr 16 2006, 12:29PM
At 100V leakage current is 0.1mA, not a big deal in a primary circuit.
It is OK for 100V but if we want to really use it in SGTC (without stacking hundreds of such bottles in series) it must work at least in lower kV range and with 1M leakage you get tens of watts of dissipation (resulting in boiling electrolyte).
Corona may take some uA away (like foil-beer bottle caps) but it can be omitted.
Only problem for now seems how to make such a cap withstand higher voltage without breakdown (and also lower leakage resistance if possible).
How is that actually done in commercial electrolytic capacitors?
Maybe another electrolyte needs to be used, rather than sodium bicarbonate?
If you really make such thing work in SGTC you would have regenerative dielectric wich is great thing, despite maybe-not-the-best pulse charachteristics
The advantage of home construction is that both caps can be made in one with both electrodes being aluminium. It would only take 6 dozen beer cans then......
I meant this if you want to make polarised caps (wich is easier as you need to roll only one electrode) of some more serious capacitance for pulse power experiments
You can also separate two layers of aluminium foil separated with paper (just to ensure they are not touching and that there is electrolyte between) to make a non-polarised high energy cap (evil grin)
It may be an interesting experiment also.
cheers, Marko
Print this page