Anyone tried using super capacitor (ex seen on ebay : 2.5v 3.3F or 5.5v 4.7f) stacked for a simple dc heater supply?
For indirectly heated tube.
Transformer --- > bridge ---> adjustment resistor ---> Super capacitor
Or for DHT
Transformer --- > bridge ---> adjustment resistor ---> Super capacitor ---> commun mode choke
For indirectly heated tube.
Transformer --- > bridge ---> adjustment resistor ---> Super capacitor
Or for DHT
Transformer --- > bridge ---> adjustment resistor ---> Super capacitor ---> commun mode choke
There are two kinds of "super caps" you'll find in places like ebay.
The little ones - double-layer electrolytic caps designed for memory backup - are small and have very high ESR, on the order of 1 - 10 kohms. I can't think of much use for these in a tube amp.
At the opposite end of the spectrum are the big honkin' caps they sell for car audio. These have very low ESR.
The car stereo types would be interesting to try. I think you'd want to feed them with a CCS though... as was mentioned, using one as a first cap in a cap input filter would result in enormous peak current spikes, and lots of noise.
On the other hand, would they be any better than a lead-acid battery? I dunno...
Pete
The little ones - double-layer electrolytic caps designed for memory backup - are small and have very high ESR, on the order of 1 - 10 kohms. I can't think of much use for these in a tube amp.
At the opposite end of the spectrum are the big honkin' caps they sell for car audio. These have very low ESR.
The car stereo types would be interesting to try. I think you'd want to feed them with a CCS though... as was mentioned, using one as a first cap in a cap input filter would result in enormous peak current spikes, and lots of noise.
On the other hand, would they be any better than a lead-acid battery? I dunno...
Pete
Those Super Caps are primary used for Memory backup which consume very very small current in magitude of less than 1uA.
I really don't think they are capable for use for filtering in the filament power supply.
There are some very large size super caps used in Car Audio. Those are good but their size is big as a beer bottle.
Johnny
I really don't think they are capable for use for filtering in the filament power supply.
There are some very large size super caps used in Car Audio. Those are good but their size is big as a beer bottle.
Johnny
I tough theses caps where used in electric welder, RC car for faster acceleration ... I've seen informations about ultra-capacitor for the futur of hybrid car also.
My welborne terraplanes 300b are dead silent using CLC--->comon mode choke for the 5 volts dc. The most silent amplifiers I ever had home.
I was trying to improve my directly coupled el84 SE(t). I have 1-2mv at the output using AC filament. The problem may be on the b+ but I measured it pretty clean.
I have no space left of chokes and it will be used on front horn so noise is still an issue at 1mv.
I've tried the usual way of reducing hum in ac config.
My welborne terraplanes 300b are dead silent using CLC--->comon mode choke for the 5 volts dc. The most silent amplifiers I ever had home.
I was trying to improve my directly coupled el84 SE(t). I have 1-2mv at the output using AC filament. The problem may be on the b+ but I measured it pretty clean.
I have no space left of chokes and it will be used on front horn so noise is still an issue at 1mv.
I've tried the usual way of reducing hum in ac config.
I wonder if in 2021 anyone has tried this with parts like this one:
https://www.mouser.de/datasheet/2/257/CE_2017_Datasheet_2_7V325F_3001962_EN_1-1274163.pdf
...1.7milliOhm ESR does not sound that bad to me...and the guys in the digital-reclocking threads are very positive...which does not mean that it sound good...but lets imagine a LCRC PSU, with a limiting R when charging and a normal filter R when on...?
Anyone tried ? I imagine it as a kind of battery supply for filament purposes...no ?
https://www.mouser.de/datasheet/2/257/CE_2017_Datasheet_2_7V325F_3001962_EN_1-1274163.pdf
...1.7milliOhm ESR does not sound that bad to me...and the guys in the digital-reclocking threads are very positive...which does not mean that it sound good...but lets imagine a LCRC PSU, with a limiting R when charging and a normal filter R when on...?
Anyone tried ? I imagine it as a kind of battery supply for filament purposes...no ?
I have only used those for applications that have nothing to do with audio. They are meant for regenerative breaking in electric vehicles and other short-term energy storage applications.
A couple of things to keep in mind:
-They can last for years at room temperature, but their lifetime goes down very quickly with increasing temperature.
-They have a large leakage current that slowly (over several hours) drops when you keep them under voltage. It is mainly due to dielectric absorption, so when you then discharge and charge them again, you get most of the leakage back for hours.
-You really have to get used to the capacitance that is in a different order of magnitude than what you are used to. For example, when you have three in series, the time constant that you get when charging them from a 1 ohm source is almost two minutes. I measured the leakage current of a 300 F cell as a function of time, switched my meter from the 10 A to the 200 mA range and had to wait for a quarter of an hour for the current to settle again, due to the five minute time constant of the 1 ohm shunt resistor of my meter and the 300 F capacitor.
-You will need parallel resistors to ensure a good voltage distribution. That doesn't help immediately after charging, but after settling.
A couple of things to keep in mind:
-They can last for years at room temperature, but their lifetime goes down very quickly with increasing temperature.
-They have a large leakage current that slowly (over several hours) drops when you keep them under voltage. It is mainly due to dielectric absorption, so when you then discharge and charge them again, you get most of the leakage back for hours.
-You really have to get used to the capacitance that is in a different order of magnitude than what you are used to. For example, when you have three in series, the time constant that you get when charging them from a 1 ohm source is almost two minutes. I measured the leakage current of a 300 F cell as a function of time, switched my meter from the 10 A to the 200 mA range and had to wait for a quarter of an hour for the current to settle again, due to the five minute time constant of the 1 ohm shunt resistor of my meter and the 300 F capacitor.
-You will need parallel resistors to ensure a good voltage distribution. That doesn't help immediately after charging, but after settling.
Last edited:
Best plan of action it to buy one if these, preferably second hand and compare to standard linear supplies. If not dramatically better let it go, hopefully at a modest loss. If dramatically better - great news as it doesn't appear like something too hard to reverse engineer.
Farad Super 3 Power Supply
Farad Super 3 Power Supply
You need a charging mechanismn, which is either a NTC or a resistor or a mkre complicated mosfet or chip solution...as this is just an experiment to see how they sound in this applicatiin, I would make it very simple... either a resistor or a variac (which I use currently anyhow to play with different chokes and R in the filament).. Start up may take a little longer, but in the end it willl be charged i guess.
hello Frank, Are you interested in the caps for the low ESR?
Working with your 801As would require 3 in series, and to get 7.5V on the filament, the feeding (charging) circuit would need very careful design to avoid any one cell getting more than 2.7V (remembering to account for any filtering between them and the filament).
Running without constant charging would not really work, since the runtime for 325F is only ca. 3 minutes for 801A (1,25A), even if you allow the voltage to drop to -10%.
The paragraph in the data sheet saying:
"Electrolyte vapor and gases generated during normal operation may escape the package"
does not inspire confidence for use in domestic situations!
Working with your 801As would require 3 in series, and to get 7.5V on the filament, the feeding (charging) circuit would need very careful design to avoid any one cell getting more than 2.7V (remembering to account for any filtering between them and the filament).
Running without constant charging would not really work, since the runtime for 325F is only ca. 3 minutes for 801A (1,25A), even if you allow the voltage to drop to -10%.
The paragraph in the data sheet saying:
"Electrolyte vapor and gases generated during normal operation may escape the package"
does not inspire confidence for use in domestic situations!
Hi Rod,
Well...I am maybe a bit simple in this regard, but I was inspired by IAn Canada's extensive usage of these caps in his FIFO and DAC-solutions. I could even use his boards in theory...and maybe I misunderstood their purpose, but his boards seem to be a kind of battery buffer, constantly charged, with an inital charging time of 10 minutes.
My very simply experiment solution would be three in series with three small resistors in parallel and than use a variac when powering on and watch the voltage until we got around 7.3V (which is enough for the 801A).
So, unless I misunderstood Ian's usage completely, there are already hundreds of DIYAUDIO-colleagues using these devices (for digital, which does not mean that they sound good for analog).
Well...I am maybe a bit simple in this regard, but I was inspired by IAn Canada's extensive usage of these caps in his FIFO and DAC-solutions. I could even use his boards in theory...and maybe I misunderstood their purpose, but his boards seem to be a kind of battery buffer, constantly charged, with an inital charging time of 10 minutes.
My very simply experiment solution would be three in series with three small resistors in parallel and than use a variac when powering on and watch the voltage until we got around 7.3V (which is enough for the 801A).
So, unless I misunderstood Ian's usage completely, there are already hundreds of DIYAUDIO-colleagues using these devices (for digital, which does not mean that they sound good for analog).
Hello Frank,
Ah, OK its for the low impedance then I guess.
Using unregulated feed for the caps is a very high risk. As Marcel says, the cell-voltage balancing is important. I would imagine that allowing the voltage of ANY CELL to get above 2.7V could result in unpleasant and likely dangerous venting of the electrolyte. Please don't do this! Use a proper control & charging controller!
Ah, OK its for the low impedance then I guess.
Using unregulated feed for the caps is a very high risk. As Marcel says, the cell-voltage balancing is important. I would imagine that allowing the voltage of ANY CELL to get above 2.7V could result in unpleasant and likely dangerous venting of the electrolyte. Please don't do this! Use a proper control & charging controller!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.