I've brought this up before and feel silly, but I'm stumped. I have a 6.3v panel mount imitation Fender jewel light from Angela and am not sure if it's AC or DC.
It takes a #47 bulb with a bayonet base. Automotive type bulbs are usually DC, I think, because they run off the car battery. I emailed Angela to ask and they wrote back saying not to work on tube amps because I could kill myself.
A little late for that.
There's no indication anywhere I can find -- on the bulb, on the housing, online -- that this thing is either AC or DC. Maybe it doesn't matter?
I'm just trying to figure out whether to wire it to T-7 and T-8 or F-4V and F-6V.
It takes a #47 bulb with a bayonet base. Automotive type bulbs are usually DC, I think, because they run off the car battery. I emailed Angela to ask and they wrote back saying not to work on tube amps because I could kill myself.
A little late for that.
There's no indication anywhere I can find -- on the bulb, on the housing, online -- that this thing is either AC or DC. Maybe it doesn't matter?
I'm just trying to figure out whether to wire it to T-7 and T-8 or F-4V and F-6V.
A #47 bulb, or any other incandescent bulb doesn't care whether it eats AC, DC or some combination of the two.
You can connect it to T1-7 and T1-8 for 6.3 volt AC
You can connect it to T1-7 and T1-8 for 6.3 volt AC
Thanks, George! And thanks for designing such a superb circuit. I've got about 12 hours listening on mine and it continues to open up and improve. I'll look into adding a motor run cap to C5 soon to see if it improves transients.
5842 heater
Hey George,
May I ask was it your intention in this design to run the 5842 at a lower voltage (maybe to help tame the tube) or do you suspect I have done something wrong to get my strange heater numbers: 5.5VDC with 0.3VAC?
My transformer is putting in it's required 6.3VAC...
Hey George,
May I ask was it your intention in this design to run the 5842 at a lower voltage (maybe to help tame the tube) or do you suspect I have done something wrong to get my strange heater numbers: 5.5VDC with 0.3VAC?
My transformer is putting in it's required 6.3VAC...
I've got about 15+ hours of listening to the TSE and I definitely notice some changes to the sound as it burns in. It's generally more coherent, and the tone and timbre of the instruments is starting to improve.
Now I'm thinking about adding a motor run cap in parallel at C5, to see if it improves the transients as suggested on the tubelab site. The existing cap is per the BOM spec: 150uF 450V electrolytic capacitor.
Two questions:
1. How would I calculate the value to choose in the parallel motor run cap? On the Tubes & Applications page, George says:
"non polarized polypropylene capacitor with a low ESR characteristic. There are capacitors made for this purpose such as the Solen “fast cap”. This improves the transient response and reduces hum. I added a large 100 uF, 370 volt AC polypropylene motor RUN cap to my 300B."
Should I just go with a 100uF/370VAC or might there be some more optimal value, given my measurements. The C5 that's in place now is seeing 371 V at the positive lead.
2. Is it worth going with a boutique cap for the motor run at C5? There's this Obbligato one for $50 at DIY HiFi Supply, or this $15 Temco model at eBay. I've seen them as low as $10 elsewhere.
Now I'm thinking about adding a motor run cap in parallel at C5, to see if it improves the transients as suggested on the tubelab site. The existing cap is per the BOM spec: 150uF 450V electrolytic capacitor.
Two questions:
1. How would I calculate the value to choose in the parallel motor run cap? On the Tubes & Applications page, George says:
"non polarized polypropylene capacitor with a low ESR characteristic. There are capacitors made for this purpose such as the Solen “fast cap”. This improves the transient response and reduces hum. I added a large 100 uF, 370 volt AC polypropylene motor RUN cap to my 300B."
Should I just go with a 100uF/370VAC or might there be some more optimal value, given my measurements. The C5 that's in place now is seeing 371 V at the positive lead.
2. Is it worth going with a boutique cap for the motor run at C5? There's this Obbligato one for $50 at DIY HiFi Supply, or this $15 Temco model at eBay. I've seen them as low as $10 elsewhere.
FWIW, There are 10 schools of thought online depending on who you listen to! 😀
Jung suggested 1/10th as a bypass. Others say 1/2 the value. Some say BOTH! I don't think it all that critical to be honest. I think all of them will sound good as the bottom line is you are lowering the ESR of the last cap that the B+ draws directly from which will increase it's reaction to the current draw. Multiple values will allow the "bank" of caps to cover more frequencies (See Cparallel below):
Essentially a cap has a resonance where the impedance is lowest. By combining multiple caps of differing values, you lower the impedance across the curve. Regardless of whether it is 1/10th or 1/2, it adds another impedance dip which is a good thing. The same value cap will not accomplish that.
More than likely, the frequencies are beyond the audio spectrum which then it becomes a moot point. Without modeling the circuit and making simulations of various cap values, regardless of what you do, adding a low ESR cap in parallel is a good thing.
FWIW, I plan on using an Obbligato 47uF PSU Film in parallel with mine. I may also later add a 4.7uF Obbligato Copper as well.
Experiment and have fun! But as a rule, divide the voltage present by 0.8 (80% derating) and ALWAYS choose a cap that voltage or higher. Give yourself a 20% safety margin. Especially in a tube amp as the heat in the chassis will decrease the voltage ratings. Granted films are more tolerant than "wet" electrolyte types like electrolytic and film oils, but better safe than smoke! 😉
Jung suggested 1/10th as a bypass. Others say 1/2 the value. Some say BOTH! I don't think it all that critical to be honest. I think all of them will sound good as the bottom line is you are lowering the ESR of the last cap that the B+ draws directly from which will increase it's reaction to the current draw. Multiple values will allow the "bank" of caps to cover more frequencies (See Cparallel below):
Essentially a cap has a resonance where the impedance is lowest. By combining multiple caps of differing values, you lower the impedance across the curve. Regardless of whether it is 1/10th or 1/2, it adds another impedance dip which is a good thing. The same value cap will not accomplish that.
More than likely, the frequencies are beyond the audio spectrum which then it becomes a moot point. Without modeling the circuit and making simulations of various cap values, regardless of what you do, adding a low ESR cap in parallel is a good thing.
FWIW, I plan on using an Obbligato 47uF PSU Film in parallel with mine. I may also later add a 4.7uF Obbligato Copper as well.
Experiment and have fun! But as a rule, divide the voltage present by 0.8 (80% derating) and ALWAYS choose a cap that voltage or higher. Give yourself a 20% safety margin. Especially in a tube amp as the heat in the chassis will decrease the voltage ratings. Granted films are more tolerant than "wet" electrolyte types like electrolytic and film oils, but better safe than smoke! 😉
Last edited:
Excellent post Craigtone, I've often had it mentioned but never quantified.
As a student of human perception i can say the jury is still out on frequencies "beyond human hearing". Like ABX testing the concept of the vaulted "TEST!" Does not line up with the expereince. Then again, If you dont personally hear said difference why bother. Buy the bargin wine 🙂
As a student of human perception i can say the jury is still out on frequencies "beyond human hearing". Like ABX testing the concept of the vaulted "TEST!" Does not line up with the expereince. Then again, If you dont personally hear said difference why bother. Buy the bargin wine 🙂
For my own personal experience with my TSE 300B, I got all the numbers lined up. I've checked everything five, six times. It's sounding agressive.
I'm using ASC poly in the PSU and my friend who gave them to me said I'll need to give then up to 500 hours for them to relax. I'll report back in the spring.
If you dont believe in cap break in or any kind of break in period, ignor all my posting past, present and future.
I'm using ASC poly in the PSU and my friend who gave them to me said I'll need to give then up to 500 hours for them to relax. I'll report back in the spring.
If you dont believe in cap break in or any kind of break in period, ignor all my posting past, present and future.
I am using Angela Instruments motor run cams and I believe they are ASC. I use two 50uF parallel with a 470uF snap in electrolytic. Not agressive at all..I am also using a 20uF Axion 1200 volt poly cap as the first cap. Using a Electra print choke.
So, the voltage regulator failed while we were testing the TSE on a scope today. It's always gotten so hot.
The engineer I've been working with suggests adding a second regulator and associated circuitry so that there's one per channel, off-board and properly heat-sinked with TO-220 units. That should reduce the heat and also eliminate some cross-talk.
I should be OK using one PQ5EV5 per channel, but are there any recommendations for a more robust regulator?
The engineer I've been working with suggests adding a second regulator and associated circuitry so that there's one per channel, off-board and properly heat-sinked with TO-220 units. That should reduce the heat and also eliminate some cross-talk.
I should be OK using one PQ5EV5 per channel, but are there any recommendations for a more robust regulator?
Last edited:
Personally I would avoid the Sharp regulator. It is an end of life part and is getting harder to find. Measure your DC input to the regulator and choose an appropriate LDO (low dropout regulator) that will accept the voltage at C1 and still output 5V. Also make sure to give yourself the standard 20% safety margin which the Sharp does not, so a MINIMUM of 3.125A per tube.
You may also want to check out the Tent Labs DHT Filament Supplies. They are pricey but the reviews are very good and they were designed specifically for DHT amplifiers.
Suitable transformer for 300B?
Found this one from Hammond:
260K Hammond Manufacturing | Mouser
650V CT @ 230ma, 5V @ 3A, 6.3V @ 5A.
Open frame, so not very pretty. 🙂
There is also the more expensive 373EX, with 345ma rating.
Found this one from Hammond:
260K Hammond Manufacturing | Mouser
650V CT @ 230ma, 5V @ 3A, 6.3V @ 5A.
Open frame, so not very pretty. 🙂
There is also the more expensive 373EX, with 345ma rating.
Last edited:
Check out the Weber WPT30. This is what I used and works perfect. No hum or mechanical buzz and barely gets warm. Will save you a few bucks too!
Craig,
Yes, I've looked up that one too - but high voltage is rated for 600V CT, which is a bit lower than what George recommends (~650V CT). Did you try the WPT30 with 300B tubes? Understand that you are enjoying the magic of 45 tubes now. 🙂
With JJ 300B tubes, I am aiming for 380-390 volts for B+, 80-85ma bias (yes, a bit hot for the 300B).
I have a Hammond 373BX also, but that is rated at 700V CT. Evanc had advised PSUD2 to test various values of the first cap in the CLC filter. But at 700V CT i am guessing that the first cap value will become rather small, and have read on some other thread that a very low first cap value is not good.
Yes, I've looked up that one too - but high voltage is rated for 600V CT, which is a bit lower than what George recommends (~650V CT). Did you try the WPT30 with 300B tubes? Understand that you are enjoying the magic of 45 tubes now. 🙂
With JJ 300B tubes, I am aiming for 380-390 volts for B+, 80-85ma bias (yes, a bit hot for the 300B).
I have a Hammond 373BX also, but that is rated at 700V CT. Evanc had advised PSUD2 to test various values of the first cap in the CLC filter. But at 700V CT i am guessing that the first cap value will become rather small, and have read on some other thread that a very low first cap value is not good.
You don't have to run 300B's at 400V. You can run them fine at lower voltages just adjust the bias and plate load accordingly to get into the linear portion of the tube curve. The prob with the higher voltage PT is you can't run the 45. With this transformer I can run any of the 3 tubes. Just be aware that you may need to recalculate the bias resistors to make sure you don't bias the FET's too hard.
Two and a half months later...
Reporting back after extensive burn in and fixing some issues I can say clearly now how the TSE and my 45 amp compare.
The TSE has come up a few pegs in my estimation. It is worth the patience required to adjust and tune this amp to your liking. In my case I had to bring the bias way up from my target number to get it to sing. I'm at 330VDC for B+ and 80mA on the 300B tubes. Now it's got legs!
It is punchy and has oodles of detail. It is close to the seduction of the 45. I think I could get it even better if I invested in first tier output tubes and output transformer.
I have not followed the intended BOM perfectly. I found and installed a pair of the original Toshiba transistors for the power drive just because I could, but I don't think it made a difference in the sound. I swapped in a pair of Nichicon Muse BP 470uF/16V caps for cathode bypass because I found I liked them more in testing than the original 4700uF Panasonics (60uF PIO were even better but won't fit in my chassis). And I have built a completely isolated and independent B+ and bias supply (which makes the range of B+/bias possibilities better and provides more stability). But I feel I have captured what George wanted to get out of this design. Like I said, it's punchy and clean; it feels like/sounds like a larger more powerful amp than it is while still being identifiably a class A, DHT, SET 🙂 I can't see how anyone just building one to spec would be anything but mightily pleased but my own trip was made better by being able to find the just right operational point. That did require a second transformer but really, they're cheap compared to tubes.
I send out a mighty Cheers! to all who assist here in the DIYaudio forums! It is not a perfectly clear path we travel (to all but a few) but very rewarding with a little help now and then. 😉
Reporting back after extensive burn in and fixing some issues I can say clearly now how the TSE and my 45 amp compare.
The TSE has come up a few pegs in my estimation. It is worth the patience required to adjust and tune this amp to your liking. In my case I had to bring the bias way up from my target number to get it to sing. I'm at 330VDC for B+ and 80mA on the 300B tubes. Now it's got legs!
It is punchy and has oodles of detail. It is close to the seduction of the 45. I think I could get it even better if I invested in first tier output tubes and output transformer.
I have not followed the intended BOM perfectly. I found and installed a pair of the original Toshiba transistors for the power drive just because I could, but I don't think it made a difference in the sound. I swapped in a pair of Nichicon Muse BP 470uF/16V caps for cathode bypass because I found I liked them more in testing than the original 4700uF Panasonics (60uF PIO were even better but won't fit in my chassis). And I have built a completely isolated and independent B+ and bias supply (which makes the range of B+/bias possibilities better and provides more stability). But I feel I have captured what George wanted to get out of this design. Like I said, it's punchy and clean; it feels like/sounds like a larger more powerful amp than it is while still being identifiably a class A, DHT, SET 🙂 I can't see how anyone just building one to spec would be anything but mightily pleased but my own trip was made better by being able to find the just right operational point. That did require a second transformer but really, they're cheap compared to tubes.
I send out a mighty Cheers! to all who assist here in the DIYaudio forums! It is not a perfectly clear path we travel (to all but a few) but very rewarding with a little help now and then. 😉
Personally, I would save for first tier speakers and room acoustics. If I have money left, then output transformer. If I have more money left, then output tubes maybe.
HI
Can some one please explain to a vacuum tube beginner the selection criteria for C4 and C5 and what value of choke I should use on a 240vac supply 50 Hz.
At this stage I will probably look at using the 300b but as I live in Australia I do not have the resources to purchase reasonably priced components for general testing.
Thanks
Pasquale
Can some one please explain to a vacuum tube beginner the selection criteria for C4 and C5 and what value of choke I should use on a 240vac supply 50 Hz.
At this stage I will probably look at using the 300b but as I live in Australia I do not have the resources to purchase reasonably priced components for general testing.
Thanks
Pasquale
Hi Pasquale,
For a TSE with 300B tubes running under 400v B+, you can follow George's recommendation on the TSE Parts with capacitors rated at 450v list for C4 and C5:
Parts List | Tubelab
This is also assuming that your power transformer will be within the recommended values too.
If you plan to run the 300B tubes higher than 400v (not recommended), then 500v rated capacitors of the same values are needed.
For the choke, George recommends a 5-10 Henry, 150-200 mA, 150 Ohm Choke. An example:
C-14X Triad Magnetics | Mouser
Do you have any particular power transformer and output transformers in mind?
For a TSE with 300B tubes running under 400v B+, you can follow George's recommendation on the TSE Parts with capacitors rated at 450v list for C4 and C5:
Parts List | Tubelab
This is also assuming that your power transformer will be within the recommended values too.
If you plan to run the 300B tubes higher than 400v (not recommended), then 500v rated capacitors of the same values are needed.
For the choke, George recommends a 5-10 Henry, 150-200 mA, 150 Ohm Choke. An example:
C-14X Triad Magnetics | Mouser
Do you have any particular power transformer and output transformers in mind?