Just a thought,
Has any one ever used switchable tubes in a SE or PP power amp.
Talking HIFI here.
For example, 3 X EL34 or EL84 SE and being able to switch off 1 or 2 of the EL34's etc to reduce power when not needed.
My question is the primary impedance of the O/P Tx and the change in load conditions / feed back. If this was cathode feedback and the EL34's were connected to the O/P Tx feedback winding as the common Gnd connection.
I guess the tubes would have to be fixed bias.
Regards
M. Gregg
Has any one ever used switchable tubes in a SE or PP power amp.
Talking HIFI here.
For example, 3 X EL34 or EL84 SE and being able to switch off 1 or 2 of the EL34's etc to reduce power when not needed.
My question is the primary impedance of the O/P Tx and the change in load conditions / feed back. If this was cathode feedback and the EL34's were connected to the O/P Tx feedback winding as the common Gnd connection.
I guess the tubes would have to be fixed bias.
Regards
M. Gregg
Last edited:
The biggest problem would be the DC current in the OPT, which will cancel out within the primary winding when in PP mode, but not so in SE mode, causing core saturation.
On a more abstract level, it doesn't sound like a particularly good idea to me. An amplifier, like any artifact of engineering, is a compromise that intends to achieve the best possible performance on a set of predetermined parameters. Starting out with two sets of potentially conflicting parameters sounds like a sure-fire way to get a crappy compromise for either set of parameters, or a good compromise for one set and a really bad one for the other.
If you need less power, then just install a volume pot.
On a more abstract level, it doesn't sound like a particularly good idea to me. An amplifier, like any artifact of engineering, is a compromise that intends to achieve the best possible performance on a set of predetermined parameters. Starting out with two sets of potentially conflicting parameters sounds like a sure-fire way to get a crappy compromise for either set of parameters, or a good compromise for one set and a really bad one for the other.
If you need less power, then just install a volume pot.
If you're doing parallel single ended then the primary of the OPT would have to change -- unless you use one transformer per power valve and then series the secondary windings (4R + 4R for an 8 ohm tap when in PSE mode). Then one whole section of the circuit could shut off for night-time listening, etc.
Thoughts are to shut the tubes down with Bias.
I know the other way is to triode connect to cut power.
I know that the the primary impedance is normally set for the tube, however how close has it got to be EL34 2.5K - 5k.
The other issue is cathode poisoning. (heater power off).
Its just a talking point at the moment.
Thoughts are 3XEL84 per side. Just for a fun test.
Regards
M. Gregg
I know the other way is to triode connect to cut power.
I know that the the primary impedance is normally set for the tube, however how close has it got to be EL34 2.5K - 5k.
The other issue is cathode poisoning. (heater power off).
Its just a talking point at the moment.
Thoughts are 3XEL84 per side. Just for a fun test.
Regards
M. Gregg
Oh right, now I see I read your post wrong all the time. I thought you were trying to develop a concept that would switch between PP and SE.
What would the purpose be of switching off power tubes? Increase THD? Create impedance mismatch problems? I can see why this could be desirable in a guitar amp, but for hifi, I don't really see the application.
What would the purpose be of switching off power tubes? Increase THD? Create impedance mismatch problems? I can see why this could be desirable in a guitar amp, but for hifi, I don't really see the application.
The easiest way to switch off an output valve is disconnect the heater. Unless you have silly HT voltages, no harm in it sitting there cold. Cathode poisoning occurs when the valve is conducting, but with temperature-limited emission i.e. cathode warm but not hot enough for normal space-charge-limited emission.
Thank's,
Do you see any issues with heater power up or down while another parallel tube is conducting. i.e. no output just idle current?
So in theory you could have a number of tubes sitting idle and select lets say an el34 or a 6550 and power up the heater to a tube.(As long as the bias was correct and OP Tx matched). So an "A" or "B" select. Or in a multi channel amp you could select just the channels required.
Regards
M. Gregg
Last edited:
Would this also be the case with DHT and pentode. I can't think of any OP Tx matches at the moment. However low power DHT switchable to pentode for power? I know the driver stage may be an issue.LOL
Regards
M. Gregg
The gain of the active devices multiplies the combined Cag of all the devices present. About the worst combination would be the gain of a pentode multiplying the capacitance of a big triode. You would need a good driver. Then the output capacitances, which simply add, could bring the OPT HF resonance down. I would not push this idea too far.
I agree its just ,
The only thought would be to switch one off and the other on (true triode / pentode switch), not in parallel. I'll leave that idea alone.
What are your thoughts on 3x EL84 parallel?
Regards
M. Gregg
Last edited:
Wavebourn,
Perhaps a tapped primary with non switched connection to Triode and Pentode?
Regards
M. Gregg
When you start turning off output tubes, you should change the tap on the OPT primary (or secondary) as well. If you want the fewer tubes to see the same output impedance anyway.
I'm not sure why you worry about efficiency. It's a TUBE amp. It's by definition inefficient. Class A, tons of heater power, etc. My 300B amp consumes over 100 W for 2x10 W output. Even worse... It consumes over 100 W for 0 W output as well (class A)... If you want energy efficiency, go class D.
But as a thought experiment, your idea is kinda interesting. One could probably also dream up a circuit that would switch tubes in/out depending on input signal level. Sorta like the newer engines from 'Merican auto makers that turn four out of eight cylinders off when the power is not needed. The challenge in your case is to do the switching cleanly.
~Tom
I'm not sure why you worry about efficiency. It's a TUBE amp. It's by definition inefficient. Class A, tons of heater power, etc. My 300B amp consumes over 100 W for 2x10 W output. Even worse... It consumes over 100 W for 0 W output as well (class A)... If you want energy efficiency, go class D.
But as a thought experiment, your idea is kinda interesting. One could probably also dream up a circuit that would switch tubes in/out depending on input signal level. Sorta like the newer engines from 'Merican auto makers that turn four out of eight cylinders off when the power is not needed. The challenge in your case is to do the switching cleanly.
~Tom
For guitar amp it would be interesting to experiment with a pot on control panel that changes local feedback ratio around output pentode. Remote cut-off pentode would be interesting as well, given that today tube amps tend to be low-output power, for studio usage. For example, a half watt output from 6BA6.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.