Unfortunately I can hear up to 17.5k and “feel” up to 18.8k… even though i’m almost 40 and have spent many years at music events. Trying to set make a high sensitivity speaker to satisfy my needs was a very long and expensive endeavor LOL! I would describe my hearing as very extended but overall low sensitivity….
Back to the filament supply… i’ve switched things up to a regular LDO (LM1084it-5.0) instead of my “zener shunt” and swapped out the 266LA12 for a 266K18 transformer and TBH it’s not running much cooler. The Bridge is maybe a touch cooler at 52 instead of 60s, but the filament transformer (266K18) is running at 70 degrees C… i’ve tried dropping the primary voltage with a variac to just above point where the regulator is able to do its thing but not much change in bridge heat, although LDO is obviously dissipating less. I’m hoping a 266L16 which is 35VA instead of 27VA of the prior trafos, and closer in voltage to what the whole thing appears to need - thus less to dissipate. because of my space constraints, I might even consider a switching setup just to keep things cool.
Back to the filament supply… i’ve switched things up to a regular LDO (LM1084it-5.0) instead of my “zener shunt” and swapped out the 266LA12 for a 266K18 transformer and TBH it’s not running much cooler. The Bridge is maybe a touch cooler at 52 instead of 60s, but the filament transformer (266K18) is running at 70 degrees C… i’ve tried dropping the primary voltage with a variac to just above point where the regulator is able to do its thing but not much change in bridge heat, although LDO is obviously dissipating less. I’m hoping a 266L16 which is 35VA instead of 27VA of the prior trafos, and closer in voltage to what the whole thing appears to need - thus less to dissipate. because of my space constraints, I might even consider a switching setup just to keep things cool.
That's an interesting point that I have actually seen in real life. My new old Keithley bench meter required some repairs and the logic power supply was a 4 amp bridge into an LM317 regulator. The regulator gets pretty warm but the 4 amp bridge is scorching hot. The board is black underneath it. Thinking maybe something was wrong I replaced it with a 6 amp bridge which gets equally as burning hot. The strange part is the transformer never gets too warm.
I tried the experiment suggested here - dropping the first filter cap down from 10,000u to 3,300u (which is what I had on hand) the smaller cap still allows for a steady 5v…. however the downstream caps (C6, C13) are still quite large and seems a bit silly now… but none of the cap swapping made much of a difference, the biggest change was as observed in putting in a more appropriate transformer - switched Hammond 266K18 (27VA) for 266L16 (35VA).. which doesn’t struggle as much and has less for the LDO to burn off.
I think the other issue is:
- small surface area bridge (GBP406)
- bridge Vf not as low as it could be (1v as opposed the 0.9v that’s available)
- compromised thermal design (on the vertical part of a thin interior L-bracket instead of directly against an exterior wall of the chassis.
Again, I don't work much with SS components but I think that bridge is too small. I used an old-style square-bottom, epoxy-embedded 50A bridge for my Aikido experiments (four 6SN7s) and it got pretty warm. But I am not much of a technician so I can only offer anecdotal advice.
1. With a very high impedance signal source and the volume control at Max, I see worst a case of 50k Ohms driving the miller capacitance of the input stage grid.
2. I see the parallel impedances of: The input stage plate rp, plate load Rp, and 2nd stage Rg driving the 2nd stage grid's miller capacitance.
(Miller capacitance is much higher for a Triode wired pentode, than for the same pentode wired as a pentode).
3. The output transformer.
Suppose each of the three above are - 1dB at 20kHz.
The result is - 3dB at 20kHz (3 x - 1dB).
2. I see the parallel impedances of: The input stage plate rp, plate load Rp, and 2nd stage Rg driving the 2nd stage grid's miller capacitance.
(Miller capacitance is much higher for a Triode wired pentode, than for the same pentode wired as a pentode).
3. The output transformer.
Suppose each of the three above are - 1dB at 20kHz.
The result is - 3dB at 20kHz (3 x - 1dB).
Thanks for these insights! I will try pentode-wiring the pentode.
As for the input of the first stage, I could probably increase the value of of the grid leak resistor or omit it altogether, I just don’t like the idea of it floating while the pot is being turned…
Sounds like my theme here might be a whole bunch of unnecessary resistance all over the place… i’ll work from the lowest hanging fruit upwards.
As for the input of the first stage, I could probably increase the value of of the grid leak resistor or omit it altogether, I just don’t like the idea of it floating while the pot is being turned…
Sounds like my theme here might be a whole bunch of unnecessary resistance all over the place… i’ll work from the lowest hanging fruit upwards.
My typical source (for the time being) will likely be some common op-amp output (Carver C2) until I get around to my tube preamp project which will tentatively be based of the Luxman CL35
I just received the Picoscope and getting accustomed to it now, so far seems like a decent piece of kit - the internal relay noises are somewhat reassuring that it's got a "real" scope front end lol.
I just received the Picoscope and getting accustomed to it now, so far seems like a decent piece of kit - the internal relay noises are somewhat reassuring that it's got a "real" scope front end lol.