Hello, I have a rather basic tube headphone amplifier with two stages, the second one being a cathode follower. Since I am not entirely happy with the sound signature I was thinking of inserting trimmers for all the tubes, both in the cathodes (for bias adjustment) and after the 1st stage plate load / 2nd stage cathode load (for plate voltage adjustment), in order to (fine)tune the operating points and therefore the sound. I will trace the changes on the diagrams to make sure the tubes stay linear and cool. For the job I looked around and saw that I could either get basic cermet resistor trimmers, such as this and this, or cermet axial potentiometers, or carbon axial pots, or plastic axial pots. I don't know what differences in performance there are between them.
For the bias trimmers I can get any rating as they only consume a few mW, so I'll probably get multiturn variants for finer control, but for the plate voltage ones I will need 1 W trimmers (1 watt is sufficient, barely). All of them will be placed in series with a fixed resistor for a minimum voltage (except for the CF load trimmer which might be placed in parallel). After I am done tweaking I may leave them in or replace them with fixed resistors (more likely).
My questions to anyone who has tried this are:
- Is there any (unwanted) side effect to using variable resistors instead of fixed ones in the signal path, such as possible parrasitic capacitances (?) affecting the sound or noise or other effects ? Perhaps more for the ones consuming close to 1 W ?
- Would any one of the types mentioned above perform better or worse than the others ?
I know the axial potentiometers have a much longer rotational life than basic trimmers, but I don't know what possible disadvantages they may pose (aside from size and price).
Any info and advice is welcome.
PS I did read Wavebourn's recommendations here which are good advice.
For the bias trimmers I can get any rating as they only consume a few mW, so I'll probably get multiturn variants for finer control, but for the plate voltage ones I will need 1 W trimmers (1 watt is sufficient, barely). All of them will be placed in series with a fixed resistor for a minimum voltage (except for the CF load trimmer which might be placed in parallel). After I am done tweaking I may leave them in or replace them with fixed resistors (more likely).
My questions to anyone who has tried this are:
- Is there any (unwanted) side effect to using variable resistors instead of fixed ones in the signal path, such as possible parrasitic capacitances (?) affecting the sound or noise or other effects ? Perhaps more for the ones consuming close to 1 W ?
- Would any one of the types mentioned above perform better or worse than the others ?
I know the axial potentiometers have a much longer rotational life than basic trimmers, but I don't know what possible disadvantages they may pose (aside from size and price).
Any info and advice is welcome.
PS I did read Wavebourn's recommendations here which are good advice.
Last edited:
Much better to learn about load lines etc. and design the bias. If you can hear the effect of a small change in bias then you are already in a significantly non-linear region.
While I agree with that idea in principle, it does not address the topic of this thread. I did not mention small or big changes. Let's say that I would like to experiment with different operating points and actually hear what the result of various calculations and plots sounds like.
So what type of trimmers or potentiometers are best for the job and what other (side) effects can I expect from using them ?
So what type of trimmers or potentiometers are best for the job and what other (side) effects can I expect from using them ?
First off, I would caution you to not leave any variable resistor in the amp for long; always replace with a fixed resistor. That said and with the understanding that one needs to start with load lines and proper engineering of the circuit before building, I do use high quality pots to fine tune a circuit. This is particularly useful when you have two mutually dependant active devices or there is a variable that is hard to account for. These 1k ohm ceramic wire wound pots have lasted me for 20 years and are a very useful tool. They are rated at 300V and 160mA or almost 50W. I generally use them in the cathode to vary bias, but sometimes in series with the plate supply to get the B+ to match the load line voltage I have selected. Notice they can be used in series or individually depending on the need. It appears as if the last use was in series adjusting a plate supply voltage, although they can be separated to provide two individual cathode resistors. Personally, I wouldn't use a little low wattage cermet trimmer potentiometer although I know they are popular in ST70 kits.
Attachments
I think one of the problems you face is that your controls are going to be interactive. Changing the cathode resistor changes the plate current and the plate voltage. Similarly, changing the plate resistor also changes both parameters.
Cheers
Ian
Cheers
Ian
Thanks for the replies. Yes they will be interactive but together they will cover more ground 🙂 Basically I'm adjusting the slopes of both the anode and cathode load lines with Ra and Rk (although the results will probably be quite similar for many combinations).
@Palustris thanks for the advice. I read trimmers generally tend to degrade over a long time but it was not clear what does that imply and how long is a long time 🙂 Anyway, although they are very durable, I thought about staying away from wirewound pots since they have an inductance which in theory will affect the sound and also some say they are noisy. What about the other types that I mentioned (cermet, carbon or plastic pots) ? Like I said I would only need them rated at 1W or at most 2W. Any idea which of them would be better and how they behave over time ?
Also can the theoretical dissadvantages of these pots (such as inductance, capacitance, noise) significantly affect the sound or are they negligible in practice ?
@Palustris thanks for the advice. I read trimmers generally tend to degrade over a long time but it was not clear what does that imply and how long is a long time 🙂 Anyway, although they are very durable, I thought about staying away from wirewound pots since they have an inductance which in theory will affect the sound and also some say they are noisy. What about the other types that I mentioned (cermet, carbon or plastic pots) ? Like I said I would only need them rated at 1W or at most 2W. Any idea which of them would be better and how they behave over time ?
Also can the theoretical dissadvantages of these pots (such as inductance, capacitance, noise) significantly affect the sound or are they negligible in practice ?
Trimmers can be dangerous as there is the risk of a shock at normal valve amplifier voltages especially if there is a fault. I use a plastic trimmer tool.
Exactly: you don't want to be casual with high voltages or cheap low wattage pots. That's why my pots are mounted to a small panel that screws onto the chassis, usually where the bottom panel would be screwed on. The panel is secure, the shafts facing me, and I can attach a meter to the terminals. Having several volt meters on clip leads saves a lot of probing into a live chassis.
For a capacitor bypassed cathode R in the first stage the wirewound's inductance won't matter, at least for test purposes. Once you've got your desired operating point locked in you can switch to an equivalent R value with a good resistor anyway.
Putting a few fixed resistors in series and clip leading across one or more of them for different ranges is what I do. A pot at the end of the chain can work for fine tuning as you mentioned, but remember that the power rating is for the whole pot winding not 5 watts dissipation across 10% of a 5 Watt pot. Manufacturer app notes mention the need to adjust power rating accordingly for the portion of total resistance that is in circuit when using the pot as a rheostat (partial load). For my own experiments I would further divide it by at least 2, usually 4.
Don't use conductive plastic for this. If you inadvertently heat them up (which we do! don't we?😀) the plastic can warp and the wiper will become intermittent or completely open.
Putting a few fixed resistors in series and clip leading across one or more of them for different ranges is what I do. A pot at the end of the chain can work for fine tuning as you mentioned, but remember that the power rating is for the whole pot winding not 5 watts dissipation across 10% of a 5 Watt pot. Manufacturer app notes mention the need to adjust power rating accordingly for the portion of total resistance that is in circuit when using the pot as a rheostat (partial load). For my own experiments I would further divide it by at least 2, usually 4.
Don't use conductive plastic for this. If you inadvertently heat them up (which we do! don't we?😀) the plastic can warp and the wiper will become intermittent or completely open.
These are some very good points. Since the power consumption on these trimmers looks like serious business 🙂, I guess I'll see to it that they operate at less than 1/4 of their rated power.
The idea of using a rotative switch to select resistors of progressive value (in series or parallel) did occur to me as the most sound option. It was however also the most work intensive solution so I put it aside for the time being 🙂
Now assuming I would like to keep some of these potentiometers in the signal path, either those in the cathode (more likely) or the plate. I am now only referring to cermet or carbon pots, not wirewounds. Regarding the side-effects to the sound due to their inductance, capacitance, etc., are these effects negligible for continuos use or do they affect sound significantly in your opinion ?
Thanks again for all the info and advice
The idea of using a rotative switch to select resistors of progressive value (in series or parallel) did occur to me as the most sound option. It was however also the most work intensive solution so I put it aside for the time being 🙂
Now assuming I would like to keep some of these potentiometers in the signal path, either those in the cathode (more likely) or the plate. I am now only referring to cermet or carbon pots, not wirewounds. Regarding the side-effects to the sound due to their inductance, capacitance, etc., are these effects negligible for continuos use or do they affect sound significantly in your opinion ?
Thanks again for all the info and advice
Everything makes a difference. The question is, can you hear it , and if you can does it matter to you. These things depend on your system and the kind of listener you are. There's no way around it, you have to try and listen.
So you use the engineering guidelines to make sure that some day when your system is on and you step out for a quick run down to the corner store , or into the next room for a nap that the system doesn't fail in a way that you'll regret. Then, after that, it's all try and see.
My limited experience with carbon pots and resistors told me they suck, I hear it like a blanket has been thrown over the speaker - but that's in circuits that I built the way I built them, they might be the perfect match for another. . . .
You've already written off the wirewounds, yet I would take them over any other in many cases , especially in a bypassed cathode circuit.
I use cermet trimmers in circuits where they are for adjusting bias on sand devices and current is limited. There is also a wiper current rating to be considered.
Back to your question about sound and continuous use. I've learned that something can sound really good, and then you turn it off and walk away. Yet a very subtle change can keep you happily sitting there, dropping the comparison and forgetting about other things you thought you needed to be doing. The difference between the two doesn't need to be dramatic yet the effect on listening behavior can be and is easily caused by the change in a part type. So it comes back to trying various things and learning how they affect you directly. There's no other way.
So my recommendation? Try the carbon. Don't write off the wirewound. . . . . and in the end, the fixed R (or equivalent sand part) will probably be the best. Heh heh.
So you use the engineering guidelines to make sure that some day when your system is on and you step out for a quick run down to the corner store , or into the next room for a nap that the system doesn't fail in a way that you'll regret. Then, after that, it's all try and see.
My limited experience with carbon pots and resistors told me they suck, I hear it like a blanket has been thrown over the speaker - but that's in circuits that I built the way I built them, they might be the perfect match for another. . . .
You've already written off the wirewounds, yet I would take them over any other in many cases , especially in a bypassed cathode circuit.
I use cermet trimmers in circuits where they are for adjusting bias on sand devices and current is limited. There is also a wiper current rating to be considered.
Back to your question about sound and continuous use. I've learned that something can sound really good, and then you turn it off and walk away. Yet a very subtle change can keep you happily sitting there, dropping the comparison and forgetting about other things you thought you needed to be doing. The difference between the two doesn't need to be dramatic yet the effect on listening behavior can be and is easily caused by the change in a part type. So it comes back to trying various things and learning how they affect you directly. There's no other way.
So my recommendation? Try the carbon. Don't write off the wirewound. . . . . and in the end, the fixed R (or equivalent sand part) will probably be the best. Heh heh.
- Status
- Not open for further replies.