I don't remember where, but I recently read about Fluorinert caps, and the few posts I could find on the subject here seem to agree that it's pretty good. Because of its applications as a dielectric coolant, I have a couple of gallons in a barrel here, so is there any chance of getting some good sound out of it?
I remember reading a bit about fluorinated insulation and dielectric, like the sulphur hexafluoride used to insulate the national grid's breakers etc. A gas with some extremely weird properties!
All of the fluorinated dielectric liquids I looked at seemed to be excellent in terms of absorption etc, but only just slightly better than air in terms of their storage capacity.
Then you have the problem of it being liquid. You either need to hand roll your own caps and then infuse them with the liquid or build it with separated plated, like an air capacitor - which will give you an extremely low capacitance I would expect.
I think it's Elna who make capacitors with silk fibres in them, maybe you could hand roll some from silk for an experiment?
I made a big capacitor on my kitchen floor when I got bored using a roll of kitchen foil and a roll of grease proof oven paper. The oven paper was, incredibly, a few centrimetres wider than the foil, meaning it could all just go on top of one another like a sandwich. I used olive oil since it was the only thing immediately within my reach. By whetting the oven paper with the oil, the foil could be stuck to it quite nicely to stop it moving around and to help get a close contact. A gentle squeegee movement will help get air pockets out.
I then experimented measuring the capacitance and standing on the capacitor with it between two plates of glass. Unfortunately, there were one or two grains of sand or dirt on the glass and one must have pushed the foils through, because the plates shorted.
They hand rolled a capacitor in one of the polar observatory sites when their LIDAR died, using cling film and foil.
My advice for hand rolling...
1. Keep everything as clean as you possibly can to avoid things getting on the layers that could puncture them. I mean super clean... you don't want to waste all that time & money to have it explode on connection
2. If you're using paper, consider putting it in a sealed container with some drying agent to get the water out. Then soak it in the liquid you're using, then put the capacitor together quickly. That might help keep you from ending up with loads of atmospheric water absorbed into the dielectric. Might when to simultaneously dry the liquid before use as well.
3. Soldering to kitchen foil is a bitch. Which means either buying real tin foil (expensive and tricky to find) or trying to solder to kitchen foil using lots of flux (I've heard this works but that the connections don't last due to the bimetallic reaction / possibly flux contamination. It may work fine if you use something like lead free solder and wash the flux off properly.... I'd do a bit of experimenting first).
4. You have two form options. You can either have flat plates or roll it. Flat plates are easy to clamp between glass (backed with wood might be an idea) but hard to find small enclosures for. Rolled versions might give you more trouble when it comes to keeping them rolled, but they'd fit inside piping from a DIY store great and you could seal them pretty nicely too. It sounds a little crazy but a condom might actually be an easy solution to keeping the foils tightly rolled. If you use something like zip ties it'll concentrate dielectric stress at those points.
4.a If you get hungry for uF and start squeezing the layers tighter, you'll probably short it out via some tiny speck of grit or dirt you missed
5. If you hand roll, consider doing it onto a piece of dowel or something to stop the material flopping everywhere when you try to enclose it.
6. A hoover generates a remarkably good vaccuum.... I've seen people boil water at room temperature using regular tap venturi vacuum pumps
All of the fluorinated dielectric liquids I looked at seemed to be excellent in terms of absorption etc, but only just slightly better than air in terms of their storage capacity.
Then you have the problem of it being liquid. You either need to hand roll your own caps and then infuse them with the liquid or build it with separated plated, like an air capacitor - which will give you an extremely low capacitance I would expect.
I think it's Elna who make capacitors with silk fibres in them, maybe you could hand roll some from silk for an experiment?
I made a big capacitor on my kitchen floor when I got bored using a roll of kitchen foil and a roll of grease proof oven paper. The oven paper was, incredibly, a few centrimetres wider than the foil, meaning it could all just go on top of one another like a sandwich. I used olive oil since it was the only thing immediately within my reach. By whetting the oven paper with the oil, the foil could be stuck to it quite nicely to stop it moving around and to help get a close contact. A gentle squeegee movement will help get air pockets out.
I then experimented measuring the capacitance and standing on the capacitor with it between two plates of glass. Unfortunately, there were one or two grains of sand or dirt on the glass and one must have pushed the foils through, because the plates shorted.
They hand rolled a capacitor in one of the polar observatory sites when their LIDAR died, using cling film and foil.
My advice for hand rolling...
1. Keep everything as clean as you possibly can to avoid things getting on the layers that could puncture them. I mean super clean... you don't want to waste all that time & money to have it explode on connection
2. If you're using paper, consider putting it in a sealed container with some drying agent to get the water out. Then soak it in the liquid you're using, then put the capacitor together quickly. That might help keep you from ending up with loads of atmospheric water absorbed into the dielectric. Might when to simultaneously dry the liquid before use as well.
3. Soldering to kitchen foil is a bitch. Which means either buying real tin foil (expensive and tricky to find) or trying to solder to kitchen foil using lots of flux (I've heard this works but that the connections don't last due to the bimetallic reaction / possibly flux contamination. It may work fine if you use something like lead free solder and wash the flux off properly.... I'd do a bit of experimenting first).
4. You have two form options. You can either have flat plates or roll it. Flat plates are easy to clamp between glass (backed with wood might be an idea) but hard to find small enclosures for. Rolled versions might give you more trouble when it comes to keeping them rolled, but they'd fit inside piping from a DIY store great and you could seal them pretty nicely too. It sounds a little crazy but a condom might actually be an easy solution to keeping the foils tightly rolled. If you use something like zip ties it'll concentrate dielectric stress at those points.
4.a If you get hungry for uF and start squeezing the layers tighter, you'll probably short it out via some tiny speck of grit or dirt you missed
5. If you hand roll, consider doing it onto a piece of dowel or something to stop the material flopping everywhere when you try to enclose it.
6. A hoover generates a remarkably good vaccuum....
I've seen people boil water at room temperature using regular tap venturi vacuum pumpsEdit: I just had a go at soldering to some regular kitchen foil using some kind of water based plumbing flux with leaded and lead free solders.
I wasn't able to get either to stick, but may need to use a more aggressive flux.
I believe the problem with soldering to this foil is that the aluminium oxidises and creates the same kinds of problems it does for arc welding. So you'll need something to chemically reduce the oxide layer I suspect. That or, if you have some nitrogen handy, try abraiding the surface and soldering with nitrogen flowing over it (disposable welding cylinders perhaps).
I wasn't able to get either to stick, but may need to use a more aggressive flux.
I believe the problem with soldering to this foil is that the aluminium oxidises and creates the same kinds of problems it does for arc welding. So you'll need something to chemically reduce the oxide layer I suspect. That or, if you have some nitrogen handy, try abraiding the surface and soldering with nitrogen flowing over it (disposable welding cylinders perhaps).
Lars, I'm currently developing a Fluorinert cap using porous teflon as a carrier for the Fluorinert. I had some prototypes built using a paper carrier, and they handily beat the best dry teflons I have (Penta Caps, V-Cap, Sonicap). This latter result surprised me a bit because paper has high DA/DF. Wow, do they sound good.
Different Fluorinert types have different dielectric characteristics. For my part, I target lowest dielectric constant because DA level generally correlates with K.
Different Fluorinert types have different dielectric characteristics. For my part, I target lowest dielectric constant because DA level generally correlates with K.
I'm hoping it will, John. I'll be comparing it to a "dry" porous teflon cap which, for its part, will have a lower dielectric K (~1.4) and, with that, presumably lower DA. Liquids, for their part, seem to do interesting things as dielectrics, so I wouldn't be surprised if the Fluor/pptfe cap outdoes the dry version. I should know within a month or so.
I imagine the fluid probably helps create a more constant dielectric surface as well. Any pockets of liquid will represent errors, but errors that are closer than having pockets of air between the layers. Personally, if I was going to that much trouble in the first place, I'd vaccuum degas them as well. A fridge compressor would probably be fine - the 'in' port produces a pretty strong vaccuum. Need something strong to degas them in - a big high density (probably reinforced) plastic drum would be good but expensive, an old propane tank from the junk yard cut open would also just well and be a lot cheaper. If you roll them into copper plumbing pipe, they could have a degas port on the pipe to void the need for a separate chamber.
Where abouts are you getting the PTFE film from and in what kinds of dimensions?
I spent a lot of time thinking about the 'perfect capacitor'. The first would likely be some kind of huge vaccuum gap thing made with silver. Then maybe replace the vaccuum / air with sulphur hexafluoride to get the plates closer without arcing. Then it's to the good old fluorides films and liquids for usable uF!
Keep me up to do date with how you get along with this, sounds like just the kind of thing I'd do, fun! Love to see some pictures of your experiments.
Where abouts are you getting the PTFE film from and in what kinds of dimensions?
I spent a lot of time thinking about the 'perfect capacitor'. The first would likely be some kind of huge vaccuum gap thing made with silver. Then maybe replace the vaccuum / air with sulphur hexafluoride to get the plates closer without arcing. Then it's to the good old fluorides films and liquids for usable uF!
Keep me up to do date with how you get along with this, sounds like just the kind of thing I'd do, fun! Love to see some pictures of your experiments.
John, no question a liquid offers an overall better plate to dielectric contact (space charge contact). Liquid also moves physicallly, which leads me to speculate that DA recovery voltages, or some part of which, cancel by physical movement of the liquid dielectric, possibly thus imparting liquids an advantage over solid dielectrics. All of this exists in that broadest of categories called *nobody knows.* My ear tells me Fluorinert is quite a performer.
As to vacuum, Fluorinert does not take well to vacuum as it evaporates very quickly (tried that). Impregnation is via low-tech immersion.
My porous ptfe is 1.4" wide thereabouts. I bought it from one of the few manufacturers of the product after one year of off and on persuasion (minimum order quantity, after all).
I agree that perfect capacitor = air (vacuum, actually). Trouble is keeping the plates separated, and at a distance that does not vary with electrostatic charge.
As to vacuum, Fluorinert does not take well to vacuum as it evaporates very quickly (tried that). Impregnation is via low-tech immersion.
My porous ptfe is 1.4" wide thereabouts. I bought it from one of the few manufacturers of the product after one year of off and on persuasion (minimum order quantity, after all).
I agree that perfect capacitor = air (vacuum, actually). Trouble is keeping the plates separated, and at a distance that does not vary with electrostatic charge.
Originally posted by serengetiplains
Liquid also moves physicallly, which leads me to speculate that DA recovery voltages, or some part of which, cancel by physical movement of the liquid dielectric, possibly thus imparting liquids an advantage over solid dielectrics.
Interesting idea. The same movement may also cause movement in the plates. Whichever has the greater effect I'm not sure, but one would assume the molecules simply rotate on their axis (SF6 is symmetrical I think, not sure about the fluorinerts), so the plate movement would be minimal - if I'm thinking about it in the right way.
As to vacuum, Fluorinert does not take well to vacuum as it evaporates very quickly (tried that). Impregnation is via low-tech immersion.
Hmmm, that's annoying. Maybe you could do a quick investigation into the boiling pressure at room temperature for the fluid and then run your infusion at a slightly lower vacuum to minimise boil off. Might be tricky differentiating boiling of the fluorinert from degassing. If you've only checked this yourself, it might be worth a datasheet check to make sure it's not dissolved gas coming out under vaccuum.
If it is the fluorinert boiling, then maybe you could take some kind of pseudo method... dry the film & fluorinert with desiccant, roll the cap, vaccuum it, turn off the vaccuum and then backfill it with the fluorinert. The pressure will rise as the fluorinert enters the voids and help stop it boiling until it equalises with room pressure.
Boiling the fluorinert would drive off dissolved gas... but you seriously want to check the datasheets on temperatures since it may also decompose. And you always want to be super careful when there's potential for fluorine moving around.
Course... another idea would be to cool the fluorinert and everything else. That way, boiling would be minimised. Only drawback would be if the viscosity rises significantly in the cooling stage.
There's quite a wide range of fluorinerts to choose from. And I know for a fact there's at least one other manufacturer who makes a comparable range of fluoride fluids - I think here in the UK. A lot of degassing and infusion problems might be solved by swapping round the specific fluoride fluid used - e.g. one with a low viscosity at cool temperatures might allow it to be to degassed and infused under a deep vaccuum.
I think if you're using porous tape vaccuum infusion might make quite a difference, since the fluid will have to have an extremely low viscosity to make it into the microscopic pores at room pressure - if it ever does. You can dunk a sponge under the water in a bath and it won't absorb 100% of it's void volume, and the holes in a sponge are huge. Open cell and closed cell pores would make a big difference to fluid infusion as well, which is why paper may work well - because it doesn't have pores in the same sense. You'd want to use open cell porous tape - if it exists (haven't checked).
If it's of any interest, I'm in university doing electronic engineering at the moment and have a physics lab directly beneath mine. These capacitor ideas are something I'm interested in too so I'd be willing to help out with the experimentation. We have vaccuum pumps and pricey test equipment a'plenty if you need any help.
My porous ptfe is 1.4" wide thereabouts. I bought it from one of the few manufacturers of the product after one year of off and on persuasion (minimum order quantity, after all).
I was going to say, these things must be costing a fair amount. What are you using for the plates in your prototypes, kitchen foil or real tin? What are you doing in terms of rolling? By hand, jig, machine?
I agree that perfect capacitor = air (vacuum, actually). Trouble is keeping the plates separated, and at a distance that does not vary with electrostatic charge.
Yes, the only way I could think of solving that is with small cross-section, thick plates. But then obviously cost and size become gigantic.
Liquid also moves physicallly, which leads me to speculate that DA recovery voltages, or some part of which, cancel by physical movement of the liquid dielectric, possibly thus imparting liquids an advantage over solid dielectrics.
Interesting idea. The same movement may also cause movement in the plates. Whichever has the greater effect I'm not sure, but one would assume the molecules simply rotate on their axis (SF6 is symmetrical I think, not sure about the fluorinerts), so the plate movement would be minimal - if I'm thinking about it in the right way.
As to vacuum, Fluorinert does not take well to vacuum as it evaporates very quickly (tried that). Impregnation is via low-tech immersion.
Hmmm, that's annoying. Maybe you could do a quick investigation into the boiling pressure at room temperature for the fluid and then run your infusion at a slightly lower vacuum to minimise boil off. Might be tricky differentiating boiling of the fluorinert from degassing. If you've only checked this yourself, it might be worth a datasheet check to make sure it's not dissolved gas coming out under vaccuum.
If it is the fluorinert boiling, then maybe you could take some kind of pseudo method... dry the film & fluorinert with desiccant, roll the cap, vaccuum it, turn off the vaccuum and then backfill it with the fluorinert. The pressure will rise as the fluorinert enters the voids and help stop it boiling until it equalises with room pressure.
Boiling the fluorinert would drive off dissolved gas... but you seriously want to check the datasheets on temperatures since it may also decompose. And you always want to be super careful when there's potential for fluorine moving around.
Course... another idea would be to cool the fluorinert and everything else. That way, boiling would be minimised. Only drawback would be if the viscosity rises significantly in the cooling stage.
There's quite a wide range of fluorinerts to choose from. And I know for a fact there's at least one other manufacturer who makes a comparable range of fluoride fluids - I think here in the UK. A lot of degassing and infusion problems might be solved by swapping round the specific fluoride fluid used - e.g. one with a low viscosity at cool temperatures might allow it to be to degassed and infused under a deep vaccuum.
I think if you're using porous tape vaccuum infusion might make quite a difference, since the fluid will have to have an extremely low viscosity to make it into the microscopic pores at room pressure - if it ever does. You can dunk a sponge under the water in a bath and it won't absorb 100% of it's void volume, and the holes in a sponge are huge. Open cell and closed cell pores would make a big difference to fluid infusion as well, which is why paper may work well - because it doesn't have pores in the same sense. You'd want to use open cell porous tape - if it exists (haven't checked).
If it's of any interest, I'm in university doing electronic engineering at the moment and have a physics lab directly beneath mine. These capacitor ideas are something I'm interested in too so I'd be willing to help out with the experimentation. We have vaccuum pumps and pricey test equipment a'plenty if you need any help.
My porous ptfe is 1.4" wide thereabouts. I bought it from one of the few manufacturers of the product after one year of off and on persuasion (minimum order quantity, after all).
I was going to say, these things must be costing a fair amount. What are you using for the plates in your prototypes, kitchen foil or real tin? What are you doing in terms of rolling? By hand, jig, machine?
I agree that perfect capacitor = air (vacuum, actually). Trouble is keeping the plates separated, and at a distance that does not vary with electrostatic charge.
Yes, the only way I could think of solving that is with small cross-section, thick plates. But then obviously cost and size become gigantic.
Extremely handy link on fluoro based insulation
I've just saved a load of .pdf's on things like the electrohydrodynamics of dielectrics too. I'll look through them later and let you know if I find anything interesting.
I've just saved a load of .pdf's on things like the electrohydrodynamics of dielectrics too. I'll look through them later and let you know if I find anything interesting.
Seems these guys offer a porous teflon with a structure similar to cellulose paper - so the dielectric would flow nicely.
teflon tape
If I had to choose a fluid for this I think I'd use 3M Fluorinert FC-87 since it has a really low viscosity (about half that of water). Boiling point is also the lowest of the series, but so is it's minimum pouring point. So I'd stick it (dried) & the dry cap in the freezer and then vaccuum infuse.
One thing to offset the fluid's pocket filling behaviour is that the solid layer it's self looses some degree of homogeneity and dimensional stability when using porous tape - e.g. the more flexible the dielectric becomes (making it's structure porous like cellulose paper) the more the electrostatic forces between the plates can compress it.
It might be worth trying the same thing with a solid tape and blasting it with Fluorinert under vaccuum. That way you fill air pockets (and even if it's tightly rolled, I'm guessing there'll be a suprising amount of bubbling under vaccuum), you keep the structural strength of the solid tape and you use less fluid per cap (important given the cost).
Talking about macro properties of course, so it just depends which has the slightly larger effect. If the fluid produces a significantly improved discharge characteristic (due to it moving like you suggest), then the porous tape would still be the best idea - despite providing less plate support. Maybe ask the tape manufacturers if they produce the tape in different rigidities.
teflon tape
If I had to choose a fluid for this I think I'd use 3M Fluorinert FC-87 since it has a really low viscosity (about half that of water). Boiling point is also the lowest of the series, but so is it's minimum pouring point. So I'd stick it (dried) & the dry cap in the freezer and then vaccuum infuse.
One thing to offset the fluid's pocket filling behaviour is that the solid layer it's self looses some degree of homogeneity and dimensional stability when using porous tape - e.g. the more flexible the dielectric becomes (making it's structure porous like cellulose paper) the more the electrostatic forces between the plates can compress it.
It might be worth trying the same thing with a solid tape and blasting it with Fluorinert under vaccuum. That way you fill air pockets (and even if it's tightly rolled, I'm guessing there'll be a suprising amount of bubbling under vaccuum), you keep the structural strength of the solid tape and you use less fluid per cap (important given the cost).
Talking about macro properties of course, so it just depends which has the slightly larger effect. If the fluid produces a significantly improved discharge characteristic (due to it moving like you suggest), then the porous tape would still be the best idea - despite providing less plate support. Maybe ask the tape manufacturers if they produce the tape in different rigidities.
Nice source for porous ptfe, John. I get mine from DeWal. I will be using FC-72, which is the lowest dielectric constant version. I used FC-84 in my early prototype. Interestingly, the porous ptfe absorbs the Fluorinert like water in a paper towel. Electrochemicosomething attraction methinks.
Don't equal capacitor electrostatic forces exist on both sides of a plate, and thus cancel?
Don't equal capacitor electrostatic forces exist on both sides of a plate, and thus cancel?
Values and voltage ratings are unknown at present as the project is in the experimental phase. I'm currently using two Fluorinert in paper caps at 75VDC, which is the maximum I've put across them. With 0.001" teflon + Fluorinert, I'd think max V would be several hundred volts. Actual measurements will tell.
As to values, much depends on how physically large the units are. Values probably will be in the range of 0.1uF to 0.47uF.
As to values, much depends on how physically large the units are. Values probably will be in the range of 0.1uF to 0.47uF.
serengetiplains said:
Somewhere a poster discussed testing caps by connecting them in series with 6 ohms, driving the resultant network with an audio amp and listening to the cap. The claim was most caps were quite audible. Beyond silver mica I don't recall which weren't. A cheap and easy test.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.