-
Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.
Wiring the Tubelab SE without a volume pot
- Thread starter boywonder
- Start date
Yes.
Any value under 500K is OK for those resistors. They are there to provide a ground path for the grid under all conditions (no volume pot or bad spot in the volume pot). If your preamp prefers a different load resistance, use it. The value of this resistor was scientifically determined by the fact that I have a large supply of 121K low noise metal film resistors in my stash.
But it's good to have adjustable input level for the "gain structure" option which is important for low noise system.
It applies more for pro audio setting but it can also be used for home audio to an extent. It's basically a way to adjust level of various components to achieve low noise floor and good dynamic range without clipping. link
With a case of using preamp (AVR or processor) and amp, listener can adjust the output level of preamp and input level of amp to achieve low noise floor. It's not feasible for every home audio but for those with adjustable options, it's good to have a variable level control for amp.
I'm doing final debugging
OK, I have one minor issue..........Everything is checking out fine on the amp, and it's making beautiful sine waves (the sq waves don't look too bad either..) but I am running out of pot adjustment for the 5842 driver tube anode voltage adjustment pot.
The driver anode voltage for both channels maxes out around 165V. Do I need to change the value of the CCS current set resistor (the 332R) to bump the anode voltage up?
Also, this appears to be the first amp I've had on the bench that cannot be driven to clipping with my function generator, so I'm assuming that the input sensitivity is rather low.......I haven't done any number crunching yet other than to determine that with the func generator pegged, I'm getting 4W output.
OK, I have one minor issue..........Everything is checking out fine on the amp, and it's making beautiful sine waves (the sq waves don't look too bad either..) but I am running out of pot adjustment for the 5842 driver tube anode voltage adjustment pot.
The driver anode voltage for both channels maxes out around 165V. Do I need to change the value of the CCS current set resistor (the 332R) to bump the anode voltage up?
Also, this appears to be the first amp I've had on the bench that cannot be driven to clipping with my function generator, so I'm assuming that the input sensitivity is rather low.......I haven't done any number crunching yet other than to determine that with the func generator pegged, I'm getting 4W output.
Which value did you use for R14 and R25. Tubelab website lists 36 KOhm for B+ voltage of 350 to 400V.
Yes the 5842's vary a lot. That is the reason for the pot in the first place. 165 volts is probably OK, but it is easy to change. Some 5842's are just "off the end of the pot" so simply increase the resistor in parallel with the pot (R10 and R21). I have used 680 ohms in some boards.
It is a popular tweak to use an LED for bias in circuits like this. I have been asked why I don't recommend an LED here. As many have found out 5842's vary a lot and so do LED's, so you need a big box full of LED's and 5842's to find a set that match and are near 175 volts. I have LED's in my 307A TSE. I went through about 20 5842's and big box full of LED's. Does sound good though.
It is a popular tweak to use an LED for bias in circuits like this. I have been asked why I don't recommend an LED here. As many have found out 5842's vary a lot and so do LED's, so you need a big box full of LED's and 5842's to find a set that match and are near 175 volts. I have LED's in my 307A TSE. I went through about 20 5842's and big box full of LED's. Does sound good though.
Well, I bumped up R10 & R21 a little, and that fixed the problem of running out of pot adjustment on the 5842's.
The amp is presently functioning, although my B- is around -300V. I smoked R5 once already, so I replaced it and swapped out the MOSFETS, replacing the 2SK3563 with the 900V rated parts (2SK3564 IIRC).
I replaced R5 with a 3K R in an attempt to reduce the B- voltage, but then I ran out of pot adjustment on the 300B's......so I put a 270R back in for R5.
My transformer is a Hammond 273BX 350-0-350 and runs a bit high. I've added another RC stage in the PS; instead of just a choke in place of R4 I put the choke, then another RC stage. My B+ is about 400V with a 5AR4 when running about 65ma through the 300B's. But the excessive B+ is making the B- very -ve.
I've been calculating the various expected voltages down the bias circuit to ground, but I don't really know how much current is flowing through the MOSFETs to determine whats good and whats bad.
I just placed an order with Edcor today for a XPWR131, so the project goes on hold...
so....a few questions.....
What are the symptoms of blown MOSFETs?
Whats a reasonable B- voltage with the Edcor 330-0-330 transformer? -300V seems like a long way from -150V....I'm assuming the -150 is with a transformer suitable for 45 outputs.
The amp is presently functioning, although my B- is around -300V. I smoked R5 once already, so I replaced it and swapped out the MOSFETS, replacing the 2SK3563 with the 900V rated parts (2SK3564 IIRC).
I replaced R5 with a 3K R in an attempt to reduce the B- voltage, but then I ran out of pot adjustment on the 300B's......so I put a 270R back in for R5.
My transformer is a Hammond 273BX 350-0-350 and runs a bit high. I've added another RC stage in the PS; instead of just a choke in place of R4 I put the choke, then another RC stage. My B+ is about 400V with a 5AR4 when running about 65ma through the 300B's. But the excessive B+ is making the B- very -ve.
I've been calculating the various expected voltages down the bias circuit to ground, but I don't really know how much current is flowing through the MOSFETs to determine whats good and whats bad.
I just placed an order with Edcor today for a XPWR131, so the project goes on hold...
so....a few questions.....
What are the symptoms of blown MOSFETs?
Whats a reasonable B- voltage with the Edcor 330-0-330 transformer? -300V seems like a long way from -150V....I'm assuming the -150 is with a transformer suitable for 45 outputs.
Last edited:
Oh, I just realized that the -150V number was coming off the schematic. Yeah, that would be for 45s...the original application of this design. What are you using for R14/R25? The values on the schematic are too low for the 300B setup because of the higher B+ and lower B-. The FETs have to deal with too much dissipation.
This quote is from tubelab website. "R14 and R25 should be 36 K on boards that operate at 350 to 400 volts" Link
I'm using 36K.
I've been contemplating a small separate transformer to power the bias circuit, say a 120:240 tiny thing as an experiment......One problem is that my power transformer is on a separate chassis and my umbilical does not have enough wires to bring the mains voltage into the amp chassis.
I would then play with the various bias circuit R values to get the bias pots back in their happy place. Perhaps another RC stage for the bias as well. I suppose that adding more RC sections even off of the power transformer secondary would be interesting; burn some voltage and get more filtering.
I've been contemplating a small separate transformer to power the bias circuit, say a 120:240 tiny thing as an experiment......One problem is that my power transformer is on a separate chassis and my umbilical does not have enough wires to bring the mains voltage into the amp chassis.
I would then play with the various bias circuit R values to get the bias pots back in their happy place. Perhaps another RC stage for the bias as well. I suppose that adding more RC sections even off of the power transformer secondary would be interesting; burn some voltage and get more filtering.
Last edited: