I see now on my pcb that HGD is connected to the B+ with a 75K + 240k voltage divider, which should bias the HTR to +76V. Pretty close to what you suggested.
I just got back from vacations, but I will do some more testing in the coming week. I got rid of the hiss with a simple l-pad on the output. And I can use almost the whole volume control range. Not my favorite solution, but it is what it is.
There's a 50k shunt attenuator on the input. Can I add a feedback resistor without adding an extra series resistor on the input?
I just got back from vacations, but I will do some more testing in the coming week. I got rid of the hiss with a simple l-pad on the output. And I can use almost the whole volume control range. Not my favorite solution, but it is what it is.
There's a 50k shunt attenuator on the input. Can I add a feedback resistor without adding an extra series resistor on the input?
By an L-pad do you mean a voltage divider, or something different than that?
Is the 50k shunt attenuator something very different from a rotary potentiometer? For adding the NFB loop, negative feedback should still work with an attenuator at the input, like this:
The junction of R23 (the series resistor) and the 620k resistor from the output (the feedback resistor) forms a 'virtual ground'. So a volume control that's outside of the feedback loop won't affect V1B.
The input impedance of the circuit is R23 (pretty much), so the 50k volume control's 12.5k ohm output impedance is working into roughly 150k ohms. That should be OK.
If the 50k attenuator is something very different from what's shown above, then all this may not work.
I'm not saying the above is anything like a fully optimized circuit. It's just an example adding on to what I believe may already be built.
For discussion's sake.
I tried this with 100k+432k input and feedback resistors. It works fine. Output level is reduced, but hiss is still somewhat audible in listening position. Could I reduce the value of the feedback resistor even further, or would that put too much of a load on the tube?
The input volume control is a common relay attenuator ladder. Nothing special. And the attenuator I had put on the output was a regular 10k voltage divider. Have taken it out again for now.
The input volume control is a common relay attenuator ladder. Nothing special. And the attenuator I had put on the output was a regular 10k voltage divider. Have taken it out again for now.
I see that you also changed the resistor on the output to gnd to 220k instead of the original 1M. Is there a specific reason for that?
That was to make the initial charge-up of C16 go a little faster, to keep any high voltage spikes down at power on, etc. There are better ways to do that. This is not too important if you're running this preamp into a tube power amp, or if your amp has DC blocking caps on its inputs. This would only be important if you're running your preamp into a solid state power amp or powered speakers with DC coupled inputs (no DC blocking caps on its inputs). The 220k load resistor allows C16 to charge up more quickly, hopefully reducing high voltage transient output at power on. If you're worried about those transients, there are better ways to deal with the issue.
I wonder if the hiss is coming from the power supply...
Another reason could be a cold solder joint, or maybe a noisy resistor. Did you use any carbon composition resistors in this build?
Another reason could be a cold solder joint, or maybe a noisy resistor. Did you use any carbon composition resistors in this build?
As a cathode follower per design offers low output impedance, 470 kΩ isn't a noticeable load to it. You could even decrease it to, say, 100 kΩ if you want.
Best regards!
I've never tried this kind of feedback loop with the shunt resistor lower in value than the series resistor.
In theory, a 150k shunt resistor with a 150k series resistor should result in unity gain. But perhaps because the 12AU7 has such limited open loop gain, the shunt resistor value will need to be actually -smaller- than the series resistor value to get the circuit to unity gain (gain of 1x).
But then, isn't a cathode follower a better unity gain buffer?
In theory, a 150k shunt resistor with a 150k series resistor should result in unity gain. But perhaps because the 12AU7 has such limited open loop gain, the shunt resistor value will need to be actually -smaller- than the series resistor value to get the circuit to unity gain (gain of 1x).
But then, isn't a cathode follower a better unity gain buffer?
Could be from the power supply. I don't think there are any cold solder joints. Most resistors are vishay dale. The brown ones.
I guess check the zener string. How much current is being drawn through it (to ground)? I think that should be at least 5mA. Morgan Jones wrote that they get noticeably quieter with 10mA or more drawn through them.
Hmm, I'm not able to measure any current through the zeners. Maybe either my meter or one of the zeners is broken? 😵
If Q1 is blown, the question is now 'what made it blow?'
What value capacitors do you have for C2 and C3? 4.7uF as specified in the schematic?
It looks like R1 sets the current through the zener string. The original value is 330R, which should set the zener string current to only about 1.5mA. Maybe you want a little more current drawn through the zener string, perhaps up to 5mA?
Yes, they are 4.7uF. I don't know what's happened. Maybe it's from all the fiddling around I've bern doing.
Ok, so I probably want to reduce the R1 value then? But first I need to replace the Q1 🙃
Right now I'm running in passive, and it definitely doesn't sound as good as with the tube pre.
Ok, so I probably want to reduce the R1 value then? But first I need to replace the Q1 🙃
Right now I'm running in passive, and it definitely doesn't sound as good as with the tube pre.
How do you know the 2N5401 is blown? Does it measure open-circuit between its pins?
I guess the first thing to do would be to replace Q1 to see if it brings the circuit back to working. If the 2N5401 is blown and the replacement blows too. at least those transistors are very inexpensive (about $0.25 each).
If the replacement transistor doesn't blow then you might experiment with increasing the current through the zener string. You could put a second resistor in parallel with the R1 330R that's there now to reduce the value of R1 to your preference. R1*R2/R1+R2.
I guess the first thing to do would be to replace Q1 to see if it brings the circuit back to working. If the 2N5401 is blown and the replacement blows too. at least those transistors are very inexpensive (about $0.25 each).
If the replacement transistor doesn't blow then you might experiment with increasing the current through the zener string. You could put a second resistor in parallel with the R1 330R that's there now to reduce the value of R1 to your preference. R1*R2/R1+R2.
Thanks. The transistor is visibly chipped. Like half of the transistor blown away. And when I hooked up the amp there was an audible hum. Everything else looks fine, but I'm a little worried that something else has blown up too. Unfortunately I don't have more 2N5401's laying around, so I will have to wait a few days.
Will try with a parallel resistor at R1 once I get it up and running again.
Will try with a parallel resistor at R1 once I get it up and running again.
I'm starting to think a cathode follower with unity gain would be better for me. But I honestly would prefer a kit of some kind. Don't want to go through the whole process of routing pcb's again.
Something like the Aikido 6SN7 buffer would be perfect. Aikido cathode follower
Something like the Aikido 6SN7 buffer would be perfect. Aikido cathode follower
Wow. Does it look like this happened because of a slip of the soldering iron, or did it blow up?
Have you checked the zener string to make sure it's not shorted out? If the zeners burnt up and failed short circuit, Q1's collector would have been shorted to ground, and POP.
I just noticed something. In the posted schematic, Q1 is connected with its collector to the junction of the top of the zener string and R3.
Shouldn't Q1 be connected as an emitter follower/constant current source? If so, shouldn't Q1 be connected like below?
Or am I misunderstanding the circuit?
You are right. Looks like it could be a mistake in the diagram. Inverted Q1. Maybe that's why I didn't measure any current over the zeners