Scenario: a mostly happy outcome, yielding the hoped-for "warm" even-harmonic coloration with my first foray into tube circuit construction, owing to a dirt-simple introductory experience found here:
The “Universal” Preamp | Cascade Tubes
With my battered old ears, "noise floor" is not so critical as for you young-uns, but if I could just cut this particular instance down by 50%-ish, particularly the line hum component, I could upgrade the project results from "mostly" to "completely" satisfactory.
Because I was trying to keep the parts tally restricted to salvaged junk, I did not follow Matt's above-linked instructions to the exact letter, and rightly expect a good tongue-lashing.
Instead of the traditional twin-diode tube arrangement, I took the lazy coward's way out, capped off the center tap and screwed a nice potted bridge rectifier to the cannibalized, cut-down old Leslie amp chassis, followed by a series-wired double-pi filter with nigh onto half a farad worth of random junkbox caps and two "proper" gapped B+ chokes (at least I think they're gapped, since they only have 2 leads apiece) of unknown inductance values and fairly beefy construction. One measures around 400Ω, the other about 180Ω, yielding 256V of B+ to the plate feed resistors.
I've had decent luck making little switch boxes and suchlike without using an isolated signal ground, and only being rigorous about shielding, so that's what I did with this thing. Was this a yuge mistake, since it involves a circuit that has a bit of gain?
Along with the just-barely-tolerable hum level, (practically inaudible at my general listening levels) there's a less pronounced, dirty sort of "pinkish" noise when the output pot is at midrange and the well-behaved Denon receiver it drives has the volume cranked up over 0 dB, that seems attributable to the natural character of a 12AU7 in such a simple circuit. Or could this be "Johnson-Nyquist" detritus from the blasphemous solid-state rectifier? Was the use of a switch-mode DC filament supply and referencing common (pin 9) to chassis ground not such a bright idea after all?
I used an "X-wired" DPDT on-off-on toggle power switch, and have noticed that there's an audible change in the timbre of the hum when hot & neutral are swapped. Lifting the earth ground has no noticeable effect. With this Brobdingnagian old beast of a power transformer, the supply draw seems improbably low at under 1/4 amp, following the momentary whole-house lamp-dimming of inrush current when initially fired up. But this was measured with a beat-up old $10 thrift store amp-clamp meter, so plenty of skepticism is warranted.
My desecrated version of Matt's lovely circuit is herein attached, should anyone have some sage words or well-deserved ridicule to dispense. 🙂
- - -
UPDATE - 11/02/21 - Huge progress! The desired outcome appears to have been attained. With all due thanks for advice received thus far, please forgive the awkward graphic, as I haven't yet figure out how to make a single-file HTML page out of it.
https://ydgqsa.dm.files.1drv.com/y4...ob6GCIOQYKhgqw/Cheapo 12AU7 preamp.jpg?psid=1
The “Universal” Preamp | Cascade Tubes
With my battered old ears, "noise floor" is not so critical as for you young-uns, but if I could just cut this particular instance down by 50%-ish, particularly the line hum component, I could upgrade the project results from "mostly" to "completely" satisfactory.
Because I was trying to keep the parts tally restricted to salvaged junk, I did not follow Matt's above-linked instructions to the exact letter, and rightly expect a good tongue-lashing.
Instead of the traditional twin-diode tube arrangement, I took the lazy coward's way out, capped off the center tap and screwed a nice potted bridge rectifier to the cannibalized, cut-down old Leslie amp chassis, followed by a series-wired double-pi filter with nigh onto half a farad worth of random junkbox caps and two "proper" gapped B+ chokes (at least I think they're gapped, since they only have 2 leads apiece) of unknown inductance values and fairly beefy construction. One measures around 400Ω, the other about 180Ω, yielding 256V of B+ to the plate feed resistors.
I've had decent luck making little switch boxes and suchlike without using an isolated signal ground, and only being rigorous about shielding, so that's what I did with this thing. Was this a yuge mistake, since it involves a circuit that has a bit of gain?
Along with the just-barely-tolerable hum level, (practically inaudible at my general listening levels) there's a less pronounced, dirty sort of "pinkish" noise when the output pot is at midrange and the well-behaved Denon receiver it drives has the volume cranked up over 0 dB, that seems attributable to the natural character of a 12AU7 in such a simple circuit. Or could this be "Johnson-Nyquist" detritus from the blasphemous solid-state rectifier? Was the use of a switch-mode DC filament supply and referencing common (pin 9) to chassis ground not such a bright idea after all?
I used an "X-wired" DPDT on-off-on toggle power switch, and have noticed that there's an audible change in the timbre of the hum when hot & neutral are swapped. Lifting the earth ground has no noticeable effect. With this Brobdingnagian old beast of a power transformer, the supply draw seems improbably low at under 1/4 amp, following the momentary whole-house lamp-dimming of inrush current when initially fired up. But this was measured with a beat-up old $10 thrift store amp-clamp meter, so plenty of skepticism is warranted.
My desecrated version of Matt's lovely circuit is herein attached, should anyone have some sage words or well-deserved ridicule to dispense. 🙂
- - -
UPDATE - 11/02/21 - Huge progress! The desired outcome appears to have been attained. With all due thanks for advice received thus far, please forgive the awkward graphic, as I haven't yet figure out how to make a single-file HTML page out of it.
https://ydgqsa.dm.files.1drv.com/y4...ob6GCIOQYKhgqw/Cheapo 12AU7 preamp.jpg?psid=1
Attachments
Last edited:
Depending on layout you might be getting induced hum from the power transformer into the tube and perhaps the choke(s).
A 12AU7 is likely the better choice. A 12AX7 will likely have too much gain and will have a too-high output impedance too.
You might want to float the filament relative to the B+ (see attached Broskie PS).
Steve
A 12AU7 is likely the better choice. A 12AX7 will likely have too much gain and will have a too-high output impedance too.
You might want to float the filament relative to the B+ (see attached Broskie PS).
Steve
Attachments
Filament isolation is nowhere nearly so daunting as taking it all apart to do the signal / chassis ground separation thing. Thanks for the feedback, (amplifier pun intended) and double that if it yields improvement.
BTW, I already did a fair amount of fiddling around with shielding, placing bits of copperclad board and ferrous sheet metal between the transformer, filter chokes, tube, etc., and observed no noticeable difference.
I have the same experience. The only material that has some shielding effect is permalloy sheet.
Sometimes rotating a choke or transformer 90 degrees will reduce - or maybe increase - hum. If you are VERY careful this can be done live if the leads on a choke are sufficiently long. It's worked for me.
See the attached link to Mr. Carlson's Lab on YouTube. Start at about the 1 hr. and 25 minute mark.
Two Six Pac's On The Bench - YouTube
See the attached link to Mr. Carlson's Lab on YouTube. Start at about the 1 hr. and 25 minute mark.
Two Six Pac's On The Bench - YouTube
Now I'm redoing the whole thing from scratch, using an isolated ground and humdinger pot this time around, there's going to be lots more shielding between the inductive bits and the preamp circuit, and the main filter choke is not only getting rotated 45°, but attached to the other side of the chassis.
As regards using caution when fiddling with live B+, thanks for the reminder. Sometimes I forget all those years of groping around with a well-worn VoltOhmyst probe, inches away from horizontal output plate caps and second anode leads, and manage to occasionally light myself up even though I should know better by now.
As regards using caution when fiddling with live B+, thanks for the reminder. Sometimes I forget all those years of groping around with a well-worn VoltOhmyst probe, inches away from horizontal output plate caps and second anode leads, and manage to occasionally light myself up even though I should know better by now.