6A3sUMMER
You are correct in your observation regarding R4.
If this circuit won't work well ( ignoring the voltage observations) for a random mix of tubes, then mullard certenly made a mistake.
But there is a possibility that it does work and the spread in voltages are in fact benign and could be ignored. What do you think ?
You are correct in your observation regarding R4.
If this circuit won't work well ( ignoring the voltage observations) for a random mix of tubes, then mullard certenly made a mistake.
But there is a possibility that it does work and the spread in voltages are in fact benign and could be ignored. What do you think ?
petertub,
Yes, probably it works fairly well with a large spread of tube characteristics.
But I always strive to use proven circuits, or to design my own circuits that account for a fairly large spread from tube to tube.
If you (someone) can not predict the voltages and currents to within reasonable spreads, how do you know if the circuit is working properly? Then you wonder if it is wired correctly, or if there is a bad part (out of tolerance).
The spread of tubes makes me think about a family of tube curves. Curve tracers are a modern thing. In the old days (say 1930s), I can almost see them busy with multiple tubes, lots of mirrored scale meters, variable power supplies, and doing lots of data collection. Then they collated all the data into points of curves, and using a french curve to make the best fit drawings of the averages.
Yes, probably it works fairly well with a large spread of tube characteristics.
But I always strive to use proven circuits, or to design my own circuits that account for a fairly large spread from tube to tube.
If you (someone) can not predict the voltages and currents to within reasonable spreads, how do you know if the circuit is working properly? Then you wonder if it is wired correctly, or if there is a bad part (out of tolerance).
The spread of tubes makes me think about a family of tube curves. Curve tracers are a modern thing. In the old days (say 1930s), I can almost see them busy with multiple tubes, lots of mirrored scale meters, variable power supplies, and doing lots of data collection. Then they collated all the data into points of curves, and using a french curve to make the best fit drawings of the averages.
Hello,
Since there is a global NFB loop, the gain is not affected by this difference. The sound seems to be OK through my B&W DM2s, although the higher voltages in the tubes seem to pull down the power supply by a few volts, indicating higher power consumption. In the ideal case, I would take the amp to a tube store and select 2 tubes with the right voltages. I have some projects involving EF86's in the planning, I presume that sooner or later, I will order more and be able to select the ones with the right parameters.
Time to make a write up on the amp and document my experiences.
Since there is a global NFB loop, the gain is not affected by this difference. The sound seems to be OK through my B&W DM2s, although the higher voltages in the tubes seem to pull down the power supply by a few volts, indicating higher power consumption. In the ideal case, I would take the amp to a tube store and select 2 tubes with the right voltages. I have some projects involving EF86's in the planning, I presume that sooner or later, I will order more and be able to select the ones with the right parameters.
Time to make a write up on the amp and document my experiences.
Hopefully the other projects you have are designed to reduce the amount of voltage and current variations, versus the spread of EF86 tube characteristics.
Global feedback keeps the gain constant as long as there is enough gain of the amp without negative feedback, versus the gain with the negative feedback. There has to be a reasonable difference of that gain ratio.
Design an amp that has 23 dB of gain without negative feedback, and then apply negative feedback that is set for 23 dB gain, and you may not get 23 dB gain.
Global feedback keeps the gain constant as long as there is enough gain of the amp without negative feedback, versus the gain with the negative feedback. There has to be a reasonable difference of that gain ratio.
Design an amp that has 23 dB of gain without negative feedback, and then apply negative feedback that is set for 23 dB gain, and you may not get 23 dB gain.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.