Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: Lectroid on January 22, 2024, 06:58:26 am
-
My new 18W amp is up and running well. It's got all the volume I can use without going to bigger tubes. The problem is that it brings on blocking distortion when I really overdrive it. Shooter suggested using a scope to visualize it. Would I set the scope probes between the 6BQ5 grid and cathode? If not, where?
Q: What do builders usually do about blocking distortion?
A: Just avoid dialing in peak power, or try to change the circuit to minimize it?
Aiken gives a list of possible responses to limit blocking distortion:
- Reduce value of coupling caps
- Reduce value of grid bias feed resistor (grid leak?)
- Reduce value of grid stopper resistor
- Reduce or remove any negative feedback
Any suggestions or preference on which of these has the best effect?
Thanks!
-
Don't play so hard.
-
Q: What do builders usually do about blocking distortion?
A: Just avoid dialing in peak power, or try to change the circuit to minimize it?
Aiken gives a list of possible responses to limit blocking distortion:
If you know you will always want to drive the output tubes hard, use Paul Ruby's zener-diode clamp (https://www.18watt.com/storage/18-watter_buzz_info_311.pdf).
- Blocking happens because the positive-swing in the grid exceeds the bias, charges the coupling cap, and shifts bias.
- The example amp bias-shifted negatively more then 2x bias, so both tubes were cut off (severe crossover distortion).
- The added zener diode has a voltage slightly larger than the bias (you select your zener based on your amp's bias).
- The zener still allows the tube-grid to conduct & clip the positive-going side.
- The zener-plus-1N4007 clips the negative-going side of the input to prevent it from bias-shifting far into grid-blocking.
-
Another option might be to adjust the LTP operating point, to reduce the maximum signal level it can put out.
It’s already got a large cathode resistor, but the cathode voltage isn’t very high, so it could be increased, the bias resistor could be reduced to increase current draw, anode resistor value increase, an additional HT node added for the LTP - a lower supply voltage will reduce the maximum output voltage
-
If you know you will always want to drive the output tubes hard, use Paul Ruby's zener-diode clamp (https://www.18watt.com/storage/18-watter_buzz_info_311.pdf).
- Blocking happens because the positive-swing in the grid exceeds the bias, charges the coupling cap, and shifts bias.
- The example amp bias-shifted negatively more then 2x bias, so both tubes were cut off (severe crossover distortion).
- The added zener diode has a voltage slightly larger than the bias (you select your zener based on your amp's bias).
- The zener still allows the tube-grid to conduct & clip the positive-going side.
- The zener-plus-1N4007 clips the negative-going side of the input to prevent it from bias-shifting far into grid-blocking.
https://el34world.com/Forum/index.php?topic=31264.msg344586#msg344586
It works, as I mentioned in your other thread, simple, try it.
EDIT: Linked fixed --- HBP
-
Another option might be to ....
Yet another technique is a significant series resistor to the grid. Roughly half the grid to ground gridleak resistor. This was good enough for Williamson. (But he was classical occasional-clip, not heavy-metal constant-clip).
The resistor will probably be so big it should be ON the socket, minimum lead between resistor body and grid pin.
"Half" assumes R4(say) is leaking all the time and R1 only charging half the time. That's infinite overdrive, can't happen, so R1 should be less than half. With 470k at R4, R1 could be 150k or even less (but 5k is not about blocking reduction. more RFI).
If you do go over 200k you may have fall-off in the top overtones. This may even be good? If you want the sizzle back, bypass R1 with maybe 30pFd.
Yes, you might spend the rest of your life hand-picking resistors and caps to cover all the songs and style and venues in your repertoire.