Hello,
I've been studying others' solutions to setting up different bias conditions for different output tubes, without having to physically replace resistors.
To begin with I was thinking to put a 560 Ohm and 250 Ohm resistor in series to act as the cathode bias, then paralleling the 250 Ohm via a switch in order to select either 810 Ohm (560 + 250) or 560 Ohm so as to create bias conditions for EL34 and KT88 respectively. However I'm not sure this is particularly elegant due to physical space constraints with the PCB.
I notice r_knize has a six position switch that puts one of six resistors in parallel with the one on the board in order to enable bias conditions for most tubes. Nice!
This all looked a bit fiddly though, and I was considering putting in a 1000 Ohm variable resistor or potentiometer (don't really know the difference!) in order to have complete control of the bias conditions rather than be limited to discrete values.
Am I completely bonkers to consider this?
I've been studying others' solutions to setting up different bias conditions for different output tubes, without having to physically replace resistors.
To begin with I was thinking to put a 560 Ohm and 250 Ohm resistor in series to act as the cathode bias, then paralleling the 250 Ohm via a switch in order to select either 810 Ohm (560 + 250) or 560 Ohm so as to create bias conditions for EL34 and KT88 respectively. However I'm not sure this is particularly elegant due to physical space constraints with the PCB.
I notice r_knize has a six position switch that puts one of six resistors in parallel with the one on the board in order to enable bias conditions for most tubes. Nice!
This all looked a bit fiddly though, and I was considering putting in a 1000 Ohm variable resistor or potentiometer (don't really know the difference!) in order to have complete control of the bias conditions rather than be limited to discrete values.
Am I completely bonkers to consider this?
I'm no expert, so please don't think this is a great idea either...unless someone else thinks so.
I chose to do this, but made a couple of choices for different reasons. First, I didn't want myself to blow stuff up by accidentally shorting the cathode to ground by turning the pot the wrong way. So I knew I was going to use a resistor in series with the pot. Next, I have a great western electric voltmeter I wanted to use to measure the bias, so it worked nicely to have a fixed resistance value to measure across to calculate the current. The voltmeter is, unfortunately, 0-100v, so I needed a sizeable resistance to generate enough of a voltage drop to get a readable number, so I picked 300R. This also seemed to be a good number just beyond the max dissipation for many tubes at my boards voltages. I then chose to use a 500R pot, but 1k would have also been good.
Remember to wire the pot in the correct direction so that if it fails open (most common) the full resistance of the pot is used.
I chose to do this, but made a couple of choices for different reasons. First, I didn't want myself to blow stuff up by accidentally shorting the cathode to ground by turning the pot the wrong way. So I knew I was going to use a resistor in series with the pot. Next, I have a great western electric voltmeter I wanted to use to measure the bias, so it worked nicely to have a fixed resistance value to measure across to calculate the current. The voltmeter is, unfortunately, 0-100v, so I needed a sizeable resistance to generate enough of a voltage drop to get a readable number, so I picked 300R. This also seemed to be a good number just beyond the max dissipation for many tubes at my boards voltages. I then chose to use a 500R pot, but 1k would have also been good.
Remember to wire the pot in the correct direction so that if it fails open (most common) the full resistance of the pot is used.
Thank you, I hadn't really considered the potential to blow bits up by accidentally setting zero R and shorting the cathode - rookie error! Your idea of putting the pot in series with a resistor that, on its own, won't cause smoke sounds like good fail-safe design.
Did you have trouble physically arranging this on the PCB, given that there are heating concerns due to the proximity of PCB and metal chassis? (Assuming we're talking SSE with the PCB under chassis)
Kind regards.
Did you have trouble physically arranging this on the PCB, given that there are heating concerns due to the proximity of PCB and metal chassis? (Assuming we're talking SSE with the PCB under chassis)
Kind regards.
The cathode resistor will dissipate a few watts, so you'll need to find a pot with a 5W dissipation rating or higher. Probably a wirewound. Then you need a way to figure out what the tube is actually doing. I suppose you can measure the voltage across the fixed resistor so you know the plate current.
Yeah thanks for the warning; I came to the same conclusion this morning regarding the power rating for the potentiometer - and 5W pots are probably going to be prohibitively expensive.
It occurred to me too, that if I'm going to go to the bother and expense of installing a variable bias pot, then I ought to measure the plate current directly, rather than inferring it from voltage measurements. I came across this 100mA analogue ammeter on eBay that would do the job. Just build it into the chassis and put the bias pot next to it - perfect bias for any tube every time!
New DC 100MA Analog AMP Current Panel Meter Ammeter on eBay (end time 24-Mar-11 16:26:12 GMT)
Perhaps it's bad though to have such a relatively long signal path - I don't know.
I just liked the idea of turning into a bit of an experiment station with old-school needles and such-like, but maybe I should get out more!
P.S. Looks like I'm committed now - just won the auction for 99p.
It occurred to me too, that if I'm going to go to the bother and expense of installing a variable bias pot, then I ought to measure the plate current directly, rather than inferring it from voltage measurements. I came across this 100mA analogue ammeter on eBay that would do the job. Just build it into the chassis and put the bias pot next to it - perfect bias for any tube every time!
New DC 100MA Analog AMP Current Panel Meter Ammeter on eBay (end time 24-Mar-11 16:26:12 GMT)
Perhaps it's bad though to have such a relatively long signal path - I don't know.
I just liked the idea of turning into a bit of an experiment station with old-school needles and such-like, but maybe I should get out more!
P.S. Looks like I'm committed now - just won the auction for 99p.
Yes, I think two ammeters is a bit excessive to begin with. I'm trying to run before I can walk and will probably fall flat on my face.
I hadn't thought of using just one voltmeter to measure the cathode resistance drop, with a DPDT to switch between the two cathodes. Using Ohm's Law is elegant but relies on the accurate measurement of the total cathode resistance, which will increase as the temperature rises, and really needs to be measured in situ I would have thought, via banana sockets also wired to the DPDT switch I guess.
Sorry, I'm thinking out loud; thank you for the prompts.
I hadn't thought of using just one voltmeter to measure the cathode resistance drop, with a DPDT to switch between the two cathodes. Using Ohm's Law is elegant but relies on the accurate measurement of the total cathode resistance, which will increase as the temperature rises, and really needs to be measured in situ I would have thought, via banana sockets also wired to the DPDT switch I guess.
Sorry, I'm thinking out loud; thank you for the prompts.
I did some computations on an EL34 with 450V B+ 5K load, with 750 ohm cathode bias resistor using a 560 ohm base resistor and 190 ohm resistor, the wattage across the 560 ohm resistor is 1.7W and the 190 ohm resistor is 0.56W totaling 2.24W across the 750 ohm cathode bias resistors (560+190). Computations were based on the Tube & Transformer page made by George on his website.
So if there's a 250ohm to 300ohm pot which can handle at max 1W, maybe it might work. I could be wrong as I myself am going about building my first simple SE too and sourcing the parts while researching along the way.
So if there's a 250ohm to 300ohm pot which can handle at max 1W, maybe it might work. I could be wrong as I myself am going about building my first simple SE too and sourcing the parts while researching along the way.
If i remember my 30+ years old lessons, the power dissipation rating of a potentiometer is for the complete track / wire length. So if you trim it down to its middle position, it can only dissipate half of its rated power. A better approach is to have one resistor and commute a few other R in parallel with a switch. It has been done by others.
There have been several threads about this in the Tubelab forum. The usual method for adjustable bias is to put the highest value 5 watt cathode resistor in the board, usually 820 or 750 ohm, then use a switch with a selection of resistors that get put in parallel with the 750 ohm resistor on the board to raise the current.
It is also possible to put a 750 ohm resistor in the board, then wire a pot and series resistor combo across that. Using a pot by itself allows for accidentally setting the resistance to zero or some other value that's low enough to force too much current through the tubes.
Since the switched resistors or pot are always used in parallel with the on board part, the current / power dissipation is shared between them. This allows for use of 3 watt parts if desired.
Simple, but not as easy to use method....
Put the 750 ohm resistor in the board. Make some resistors with alligator clips on the ends in the 2K to 3K range. Connect the clipped resistors in parallel with the 750 on the board as needed.
I used the 750 ohm by itself with an EL34 in triode mode. That gave me about 55 mA with the tubes I had. I clipped on some 2.2K resistors when I used 6550's in UL mode. This combo gave me 560 ohms for about 80 mA.
It is also possible to put a 750 ohm resistor in the board, then wire a pot and series resistor combo across that. Using a pot by itself allows for accidentally setting the resistance to zero or some other value that's low enough to force too much current through the tubes.
Since the switched resistors or pot are always used in parallel with the on board part, the current / power dissipation is shared between them. This allows for use of 3 watt parts if desired.
Simple, but not as easy to use method....
Put the 750 ohm resistor in the board. Make some resistors with alligator clips on the ends in the 2K to 3K range. Connect the clipped resistors in parallel with the 750 on the board as needed.
I used the 750 ohm by itself with an EL34 in triode mode. That gave me about 55 mA with the tubes I had. I clipped on some 2.2K resistors when I used 6550's in UL mode. This combo gave me 560 ohms for about 80 mA.