I want to examine the possibility of creating a variable resistor to retain the parameters of a normal, good quality fixed value resistor. The main quality is little or no change with temperature, current, voltage and frequency.
A potentiometer, I found, is too noisy and unstable, does not like DC through it and so on.
A simple transistor can act as a variable resistor between collector and emitter, unfortunately I do not know how to control it so its "resistance" stays the same regardless of Vce and Ic.
Any ideas? I skipped-read the thread about the LED controlled resistor thingy, and maybe this is a solution?
A potentiometer, I found, is too noisy and unstable, does not like DC through it and so on.
A simple transistor can act as a variable resistor between collector and emitter, unfortunately I do not know how to control it so its "resistance" stays the same regardless of Vce and Ic.
Any ideas? I skipped-read the thread about the LED controlled resistor thingy, and maybe this is a solution?
Ah I see, a huge switch with dozens of resistors. Just like what I have on my ageing bench power supply.
Forgot to say that is certainly a possibility and in that case I'd use just 4 settings, I would not need more. A rotary switch with 4 positions, and 4 resistors. Maybe 6 positions and 6 resistors.
But if I wanted something that will do the job on the PCB, not at the front panel so the signal does not have to travel over extra wires.
But if I wanted something that will do the job on the PCB, not at the front panel so the signal does not have to travel over extra wires.
I think the potentiometer problems you experience are from cheap pots. Try a wire wound in the 500 ohm size.
Like this except PCB? 43C1500 Honeywell Potentiometers
or this? Digi-Key - SP140-1.0K-ND (Manufacturer - 140-0-0-102)
Like this except PCB? 43C1500 Honeywell Potentiometers
or this? Digi-Key - SP140-1.0K-ND (Manufacturer - 140-0-0-102)
These pots are very expensive, I can buy two Eminence Stonehenge 12" speakers for that price, and the speakers are the most crucial and expensive part.
My solution will be to use 4 or 6 resistors, placed very close to the transistor (1/4 "), so that the voltage across them stays close to the transistor. There will be 4 or 6 short leads/tracks to a rotary switch (1.5" track + 2-3" leads), and those will be connected to earth. The whole switch will be earthed. Hopefully this will limit noise and distortion.
My solution will be to use 4 or 6 resistors, placed very close to the transistor (1/4 "), so that the voltage across them stays close to the transistor. There will be 4 or 6 short leads/tracks to a rotary switch (1.5" track + 2-3" leads), and those will be connected to earth. The whole switch will be earthed. Hopefully this will limit noise and distortion.
the 12 position rotary switch is quite cheap.
It can be bought in 4 versions:
12w 1p
6w 2p
4w 3p
3w 4p
It can be bought in 4 versions:
12w 1p
6w 2p
4w 3p
3w 4p
Yes I will use a rotary switch, I already have a bunch in my toolbox. The comment about "pots are very expensive" refered to the previous post and some very very expensive wire-wound pots
What are you trying to do? Is this something that will be adjusted often like a volume control, or an internal function like bias that's just adjusted every now and then?
CH
CH
Gain control, but not using op-amps, and not using feedback across vaious stages. I did it using a dual-gang pot, one modifies the bias the other the local feedback (Re). It works, in that it allows the gain to be set, but turning the pot results in moderate disturbances of the signal.
My new idea as I said above is to have 4 or 6 resistors very close to the transistor so the PCB layout and the signal stays tight, and they get earthed via a rotary switch. I could select 4 or 6 gain levels, that will be more than adequate. Bias does not need to change, as long as we do not make radical changes to the feedback resistor and with a 24.5 volt supply there is plenty of room.
My new idea as I said above is to have 4 or 6 resistors very close to the transistor so the PCB layout and the signal stays tight, and they get earthed via a rotary switch. I could select 4 or 6 gain levels, that will be more than adequate. Bias does not need to change, as long as we do not make radical changes to the feedback resistor and with a 24.5 volt supply there is plenty of room.
I have had more problems with cheap switches than quality pots. The need to make the POT PCB mount makes it expensive. If a panel mount or panel hole could be placed in the PCB then very short wires and the pot cost is much less. The CTS pots I use are like $7. If you do go with a switch sealed and gold contact like a grayhill telecom switch is good.
Isn't it better, in principle anyway, to try to find something with the same mode of operation as a 270 deg. pot? It's then just a matter of sourcing.
Apart from the fakes issue, can't a solution be found with an ALPS or similar product? They're not all that expensive and some grades are said to be superb.
Oh, something wiser heads have recommended: Conductive plastic element pots, as used in studio and pro. equipment, I believe. Does anyone know about these and how they might be tracked down or even if they still exist?
Apart from the fakes issue, can't a solution be found with an ALPS or similar product? They're not all that expensive and some grades are said to be superb.
Oh, something wiser heads have recommended: Conductive plastic element pots, as used in studio and pro. equipment, I believe. Does anyone know about these and how they might be tracked down or even if they still exist?
Part of the issue is how fine of adjustment is needed. If 6 steps is more than enough a switch can make sense. More than 10 a pot starts looking a lot better.
Pots come in conductive plastic, carbon film, cermet, and wire wound. cermet are temperature stable but difficult to adjust and are noisy. CP is used a lot for audio and is much quieter the carbon film and much better with the DC. Carbon film is the cheap one. Wire wound are used where quality and even better performance than the CP is needed, even lower noise and better DC. CP is used at RF with like 50ohm impedance matched applications. I use almost all wire wound for any audio and instrumentation. Carbon film is good for high voltage.
Pots come in conductive plastic, carbon film, cermet, and wire wound. cermet are temperature stable but difficult to adjust and are noisy. CP is used a lot for audio and is much quieter the carbon film and much better with the DC. Carbon film is the cheap one. Wire wound are used where quality and even better performance than the CP is needed, even lower noise and better DC. CP is used at RF with like 50ohm impedance matched applications. I use almost all wire wound for any audio and instrumentation. Carbon film is good for high voltage.
Thanks, I take also the point guys, of switching. Though quality switches, unless you have access to high spec. overstocks, is nowa pursuit of the rare and expensive. You need good make-before break styles and common moulded, ball-contact types are well....not good.....for nothin'.
I have no problem with on-line buying better gear in the USA for example. I just glaze over at the interminable "stock-out" responses when you really only want a small number of items.
I have no problem with on-line buying better gear in the USA for example. I just glaze over at the interminable "stock-out" responses when you really only want a small number of items.
Perhaps what you trying to do doesn't lend itself to any easy method of control unless the output is muted (to avoid the noise). If you are simultaneously altering bias (DC conditions do you mean ?) and feedback factors (is that a DC coupled network too) then I'm not surprised it causes "disturbance"
Perhaps if you posted a circuit of some sort so we could see what you are trying to do.
http://embrace.grayhill.com/embrace/IMAGES/PDF/I-56.pdf
Good switch and PC mount... may be an oxymoron. This is a very nice one.
Good switch and PC mount... may be an oxymoron. This is a very nice one.
Tektronix hasn't had a problem with pots in the last 60 years. They even use super-stable ten turn pots for precision timing. A volume control, that doesn't even care if it's +/-3dB, is a noticably less involved application.
Ironically, Tek uses switches now, but that's because the whole thing is digital. They use encoders. "Volume" control is digital (usually relays switching precision dividers, but a 1-2-5 series probably wouldn't be too handy for audio).
Tim
- Status
- Not open for further replies.