Roughening the surfaces of a pair of parallel plates does only causes a small variation in the minimum distance the pairs of charges can approach — it is insignificant compared to the large distance between the plates. (unless in case of a megathick graphite coating LOL) The dielectric follows a smooth average line at the boundary.
A higher surface area does work in ultracapacitors as well as in electrolytic capacitors, but not in aircapacitors.
Last edited:
I couldn't agree more. If I contributed or provoked the harshness I sincerely apologize. As you look back on this thread you see that at one point I censored/removed a reply. I was about about to say the same as you are saying now. A bit later I said goodnight and goodbye because I was fed up with kind remarks that we're posted. I came here to learn and share not to be insulted and laughed at. I must say that the culture of the Dutch is more blunt than other cultures, but definitely not like this (I hope). Off-topic? No. Very much to the point.
C=eps*eps0*S/d, where eps is permittivity, eps0 vacuum(air) permittivity, S-area, d-distance between plates. Charge is related to the voltage at a given capacitance U=Q/C.
Having hair like structure will not help, unlike lytic capacitors where contact is provided by electrolyte and porous electrode is of a definite advantage.
Last edited:
I'm curious about charge density (but ignorant of the physics).
My cells (manufacturers originally to be inside Dayton-Wright speakers), made in the early '70's, charge very slowly as indicated by SPL. They take perhaps 18 hrs to be within maybe 2dB of their final SPL. The increase is uniform across their band, roughly 40-15kHz. There is a substantial resistance, maybe 100 MOhms, not sure, in series with the HW power supply.
What does that mean?
B.
My cells (manufacturers originally to be inside Dayton-Wright speakers), made in the early '70's, charge very slowly as indicated by SPL. They take perhaps 18 hrs to be within maybe 2dB of their final SPL. The increase is uniform across their band, roughly 40-15kHz. There is a substantial resistance, maybe 100 MOhms, not sure, in series with the HW power supply.
What does that mean?
B.
There is nothing peculiar in regard to sulfur hexafluoride besides density and breakdown voltage, so it's hard to tell. It well might be related to the coating resistance. If it's too high then even distribution will take too much time. It well might be that higher bias voltage means higher leakage thus less current available for charging diaphragm itself.
Thanks for info. I mounted the DW cells in open air frames, 6 cells to a panel. So the welding gas isn't around.
Yes, takes a while to charge the cells. But is the performance changed? Does the charge need to be re-charged after loud music passages? Can resistance be too large?
Would it be correct to think that if it took 18 hrs to charge, there couldn't be much leakage? Does the large series resistance matter much when talking of charging cells?
And do the amp and matching transformer care about the diaphragm charge?
B.
Hi Ben
Interesting panels you have but waiting 18hrs plus to hearing full SPL's is a little inconvenient perhaps or do you just keep them charged all the time like giant dust traps. Speaking of which without the gas I imagine youve got no dust covers and wonder if a membrane is leaking charge?
Diagnosis needs a good history:
Is this a change or have the panels always taken >18hrs?
What coating do you use? If its the original whats the consensus on what the coating is and how long it lasts?
Whats your high voltage source and specs - kV and mA, mono or stereo chargers?
Assuming your running stereo pairs are the SPLs similar ie playing a mono track is dead centre?
Have you tried the old neon bulb indicator trick to gauge charging rate?
If its a sudden change a panel or resistor may have shortened out with excessive leakage affecting the entire system, if you only have a single HV source for all panels.
If its been a slow change over years but is better on rainy days it might be time to re-coat.
Im no expert but others here certainly are so lets see what they think of:
Given your high panel count and the possibility just one leaky panel could degrade charging all the rest, there might be merit in trying the neon bulb indicator trick so each panel can be easily monitored. Neon lamps are still available and cheap or dirt cheap so maybe try paralleling a few for better current pass and faster panel charging. The neon bulbs are about $2 for twenty so you can wire each panel up and identify any gone bad. No flashing means no charge eg stray connection, resistor on open, coating discontinuity etc. Rapid flashing means a short or leakage.
You might want test and if need be upgrade your resistors at some stage. There might be merit in using multiple resistors eg two in series x two parallel for lowering voltage and current loads, purchase costs and to provide redundancy if one fails either way.
If you re-coat your panels you could do a really cool trial of different coatings given for have 2x8 panels. How about using four different coating, one coating then would be used on four panels. Then do SPL measurements and start a new thread.
Stay safe around all those kV's.
Interesting panels you have but waiting 18hrs plus to hearing full SPL's is a little inconvenient perhaps or do you just keep them charged all the time like giant dust traps. Speaking of which without the gas I imagine youve got no dust covers and wonder if a membrane is leaking charge?
Diagnosis needs a good history:
Is this a change or have the panels always taken >18hrs?
What coating do you use? If its the original whats the consensus on what the coating is and how long it lasts?
Whats your high voltage source and specs - kV and mA, mono or stereo chargers?
Assuming your running stereo pairs are the SPLs similar ie playing a mono track is dead centre?
Have you tried the old neon bulb indicator trick to gauge charging rate?
If its a sudden change a panel or resistor may have shortened out with excessive leakage affecting the entire system, if you only have a single HV source for all panels.
If its been a slow change over years but is better on rainy days it might be time to re-coat.
Im no expert but others here certainly are so lets see what they think of:
Given your high panel count and the possibility just one leaky panel could degrade charging all the rest, there might be merit in trying the neon bulb indicator trick so each panel can be easily monitored. Neon lamps are still available and cheap or dirt cheap so maybe try paralleling a few for better current pass and faster panel charging. The neon bulbs are about $2 for twenty so you can wire each panel up and identify any gone bad. No flashing means no charge eg stray connection, resistor on open, coating discontinuity etc. Rapid flashing means a short or leakage.
You might want test and if need be upgrade your resistors at some stage. There might be merit in using multiple resistors eg two in series x two parallel for lowering voltage and current loads, purchase costs and to provide redundancy if one fails either way.
If you re-coat your panels you could do a really cool trial of different coatings given for have 2x8 panels. How about using four different coating, one coating then would be used on four panels. Then do SPL measurements and start a new thread.
Stay safe around all those kV's.
Thanks for many thoughts and tyu too.
These are Dayton-Wright panels from mid-1970's, not DIY. Not able to readily disassemble to try new coatings. Used DD amp for a while.
Long delay no problem for me since they remain on except for vacations. And have been on for years albeit with a dust cover and vacuuming and heat treatment every year or so. Seem to work fine after 40 years (with a service gap of 25 years).
I have a bunch more in a box and have thought about a wall-of-sound speaker.
Single variable EMCO power supply, 5-8 kV. No noticeable change with humidity.
Info from neon testing would be interesting but not easy to do without a great number of re-charging cycles for my 12 cells and each has two diaphragm connections (actually 4 contact spots) and there are several dozen 22 meg resistors in the paths, some sealed into the cells.
I appreciate the efforts to help me explore and possibly fix these speakers that have always had exemplary SQ, as far as I can measure.
But what I would actually appreciate is learning what are the benefits and harms of having cells that seem to charge very slowly, perhaps due to very large series resistance or coating?
Thanks.
B.
I thougt the same... .....
By the way there is a youtube video by robert smith murray: "how to make conductive ink."
He plays around with the same graphite as you did.
Quite a rough and big way of making the fluid on that YouTube video.
I on the other hand use precision syringes like below to get the correct mix that is exact to 0,1 ml at all ingredients. The powder is weighed on my $ 10,00 China precision scale. For mixing I use a paint shaker, kitchen tools and just hands. Luckily no heating involved in my process. Each small batch is just for 4 ESL-63 panels. Tried larger batches, but couldn't get it to the right resistivity in those cases.
Quite a rough and big way of making the fluid on that YouTube video.
I on the other hand use precision syringes like below to get the correct mix that is exact to 0,1 ml at all ingredients. The powder is weighed on my $ 10,00 China precision scale. For mixing I use a paint shaker, kitchen tools and just hands. Luckily no heating involved in my process. Each small batch is just for 4 ESL-63 panels. Tried larger batches, but couldn't get it to the right resistivity in those cases.
Well I can do a much better job regarding precision as well but that is not the point. Robert shows some basics regarding conductive paints. It shows the key ingredients and tells you what to expect. These are guidelines. This might be useful for anyone who plays around with graphite stuff.
