Hi all,
I'm rather new to tube amps and working on upgrading a little "Darling Clementine" amp I bought recently (before embarking on a big ST-70 upgrade.)
I thought it might be an interesting and fairly inexpensive experiment to allow a switch to be used to change the cathode bypass caps on the output tubes. (Thinking 5 options: no cap, 0.1uF, 1uF, 10uF, 100uF).
I realize some folks might say it's a waste of time and a few bucks but I'm not concerned with that because I like the tinkering and I think it adds something sorta unique to this little project.
So a couple of questions:
1) is there downside to doing this?
2) Right now the best place to mount the switch and all the caps is under the large central power transformer (in orange in the picture). Can that induce noise into the audio? I'm hoping that a) it's hard to induce noise via a capacitor if it's even possible at all, and b) the fact that one end of the cap goes to ground rather than into the circuit might eliminate that risk if it would otherwise be a risk.
Any advice much appreciated!
Ross in Denver
I'm rather new to tube amps and working on upgrading a little "Darling Clementine" amp I bought recently (before embarking on a big ST-70 upgrade.)
I thought it might be an interesting and fairly inexpensive experiment to allow a switch to be used to change the cathode bypass caps on the output tubes. (Thinking 5 options: no cap, 0.1uF, 1uF, 10uF, 100uF).
I realize some folks might say it's a waste of time and a few bucks but I'm not concerned with that because I like the tinkering and I think it adds something sorta unique to this little project.
So a couple of questions:
1) is there downside to doing this?
2) Right now the best place to mount the switch and all the caps is under the large central power transformer (in orange in the picture). Can that induce noise into the audio? I'm hoping that a) it's hard to induce noise via a capacitor if it's even possible at all, and b) the fact that one end of the cap goes to ground rather than into the circuit might eliminate that risk if it would otherwise be a risk.
Any advice much appreciated!
Ross in Denver
Make sure your ''No Cap'' switch position does not just short the cathode to ground, it must be an open disconnect. Just disconnect the cap for that step. This will be a short term experiment so I'd just use some jumpers out to the bench. From full bypassing, down through the lower cap values, you should just hear the mids to bass fall away as the gain for those frequencies is attenuated by the degenerative FB that isn't bypassed. But, also, depending on any GNFB the amp uses, you may hear next to no changes between caps at all, as that overrides the changes in output frequencies. Have fun.
Last edited:
Thanks, 20to20. Yes, definitely open and not shorting cathode to ground.
My bigger question is whether locating the capacitors below the transformer is a likely source of noise in this configuration. I hope not...
My bigger question is whether locating the capacitors below the transformer is a likely source of noise in this configuration. I hope not...
You can use a copper tube for AC wiring or signal wiring past the transformer. Some brake lines work great and are easy to bend, plus you can solder them to a bracket and fasten to a tube socket.
Your ground path is far likely to cause you grief.
Your ground path is far likely to cause you grief.
Hey Anatech,
I'm working in a very tight space. Probably only a 3- or 4-inch wire from cathode terminal on PCB to switch and probably 2 inches from end of capacitors to ground. So I don't think I have a lot of flexibility in putting them in things, if that makes sense.
I guess part of the question is whether that stuff is really part of the signal path in a way that can inject noise?
I don't understand your second sentence.
I'm working in a very tight space. Probably only a 3- or 4-inch wire from cathode terminal on PCB to switch and probably 2 inches from end of capacitors to ground. So I don't think I have a lot of flexibility in putting them in things, if that makes sense.
I guess part of the question is whether that stuff is really part of the signal path in a way that can inject noise?
I don't understand your second sentence.
No problem. I use tubing in my designs as a precaution, so it is designed in. It doesn't take much room up, but would be a devil to install later.
Every circuit element can either radiate or receive noise under the right conditions. Things with large signals or current would be the radiators, the others are "victim" circuits that are impacted by radiation.
The one question I always asked new technicians was "when is a ground not a ground?". At very low signal levels or high frequencies, a wire isn't a perfect wire, same for the chassis. They have resistance and inductance. Other currents running through the same conductor will create voltage drops, small usually but those can impact your circuit. So on a new build or design, you have to pay close attention to what you call ground. Call it "common" for whatever circuit you are looking at. Your power supply has a common in it. So does the power supply return to each audio circuit and you may have a "dirty ground" where decoupling capacitors connect. Each "common" serves a distinctly different purpose, and mixing them can bring grief. They typically join together at one point, sometimes some commons may connect before the main reference common point.
Every circuit element can either radiate or receive noise under the right conditions. Things with large signals or current would be the radiators, the others are "victim" circuits that are impacted by radiation.
The one question I always asked new technicians was "when is a ground not a ground?". At very low signal levels or high frequencies, a wire isn't a perfect wire, same for the chassis. They have resistance and inductance. Other currents running through the same conductor will create voltage drops, small usually but those can impact your circuit. So on a new build or design, you have to pay close attention to what you call ground. Call it "common" for whatever circuit you are looking at. Your power supply has a common in it. So does the power supply return to each audio circuit and you may have a "dirty ground" where decoupling capacitors connect. Each "common" serves a distinctly different purpose, and mixing them can bring grief. They typically join together at one point, sometimes some commons may connect before the main reference common point.
Hey anatech, i think i basically understand your point. I'm learning this stuff as we go along. FWIW, the "ground" I'd be connecting to IS attached to the green wire of the power cord. Not a trace on a board.
OK, I'm a moron. I just figured out (thanks to someone telling me to check again) that there's a 100uF cap there already and the 0.1uF film cap is bypassing that one.
BUT I still find this topic interesting and I'm grateful for your input as I still consider the project.
BUT I still find this topic interesting and I'm grateful for your input as I still consider the project.
Hi Rossputin,
Doing that may generate some serious noise issues. Remember, all kinds of currents flow though the AC safety connection (the AC plug ground). It exists to protect you from shock and must be bonded securely to the chassis, a very good electrical connection.
Doing that may generate some serious noise issues. Remember, all kinds of currents flow though the AC safety connection (the AC plug ground). It exists to protect you from shock and must be bonded securely to the chassis, a very good electrical connection.
Last edited:
indeed...it's soldered to a thing that's well attached to the chassis and I'd be soldering just below where the green wire attaches to the terminal.
Thanks for your responses, by the way!
Thanks for your responses, by the way!
so let me restate my main question:
Do you experts think that simply having a capacitor (that goes from tube cathode to ground) mounted under a transformer can inject noise into the circuit because of being under the transformer?
Do you experts think that simply having a capacitor (that goes from tube cathode to ground) mounted under a transformer can inject noise into the circuit because of being under the transformer?
Can you show a schematic? Picture?
You normally don't want to connect your audio common directly to the AC safety ground.
You normally don't want to connect your audio common directly to the AC safety ground.
Everything is in the signal path, so yes it can inject noise, but if this is a PP stage and you are careful to keep things symmetrical it should cancel.
It would be more accurate to note that there isn’t a single signal path. These are often drawn as though they go through the valve horizontally, from grid to plate, but that’s not so. There is one coming from the previous stage going down the grid leak resistor to ground, and a new derived signal coming up the cathode circuit from ground, through the valve vertically, and out through the plate.
Hey ejp, I have a lot of studying to do before really understanding that but I get some of it. Looking forward to having more time to read the tube books, maybe even take a class somewhere if one exists about designing tube-based audio circuits.
Some things are common to all design, audio and other. You have to understand it all.
Morgan Jones is a great author for tube product. Bob Cordell and Douglas Self for audio even though they are more solid state, the concepts are critical for tube audio as well.
Morgan Jones is a great author for tube product. Bob Cordell and Douglas Self for audio even though they are more solid state, the concepts are critical for tube audio as well.
I have a Jones book on Building Valve Amplifiers! Just haven't had time to get to it yet. Hoping to really study this a lot when I retire but it'll be slow until then.
