Hi
I haven't selected any output transformers as I'm working out what I can get down here and what I will have to import from the US.
I wasn't aware of all the variations available and can fully understand why George doesn't supply a kit.
I'm also waiting for a tube book to arrive from the UK what will give me more of an understanding on what I'm getting myself in for as at this stage my knowledge in regards to tube amps is slim.
There's such a price variation in tubes it's like going to the bottle shop and trying to select a good bottle of wine without spending the earth.
One question and it will probably sound a little strange but does the type of tube apart from the circuit also govern the gain of the amp as at this stage my speakers are inefficient at 89db/1 watt at 6 ohms being the Polk Audio 10b.
Thanks for your advice!
Cheers
I haven't selected any output transformers as I'm working out what I can get down here and what I will have to import from the US.
I wasn't aware of all the variations available and can fully understand why George doesn't supply a kit.
I'm also waiting for a tube book to arrive from the UK what will give me more of an understanding on what I'm getting myself in for as at this stage my knowledge in regards to tube amps is slim.
There's such a price variation in tubes it's like going to the bottle shop and trying to select a good bottle of wine without spending the earth.
One question and it will probably sound a little strange but does the type of tube apart from the circuit also govern the gain of the amp as at this stage my speakers are inefficient at 89db/1 watt at 6 ohms being the Polk Audio 10b.
Thanks for your advice!
Cheers
If you meant to write:
" .... but does the type of tube apart from the circuit also govern the power output of the amp as at this stage my speakers are inefficient at 89db/1 watt at 6 ohms ... "
Then, the answer is yes.
A 45 will do a watt, maybe a watt and a half.
A 2A3 will do about double a 45, or between two and three watts.
I don't have any 300B's, but I've read they can do about eight watts. I'd imagine six or seven watts is a more dependable number.
How loud do you like to listen? With 89db speakers, you may wish to consider an SSE.
Apologies if I have mangled your question.
Win W5JAG
" .... but does the type of tube apart from the circuit also govern the power output of the amp as at this stage my speakers are inefficient at 89db/1 watt at 6 ohms ... "
Then, the answer is yes.
A 45 will do a watt, maybe a watt and a half.
A 2A3 will do about double a 45, or between two and three watts.
I don't have any 300B's, but I've read they can do about eight watts. I'd imagine six or seven watts is a more dependable number.
How loud do you like to listen? With 89db speakers, you may wish to consider an SSE.
Apologies if I have mangled your question.
Win W5JAG
That's kind of difficult to answer other than I prefer quality audio now that I'm a little older.
With all the hype these days with regards to power per channel It was a little difficult to accept three watts per channel but I'm sure after what I have seen and read it's totally worth it. Apart from better bass I wasn't aware that the SSE will give a larger output so maybe I need to review my selection.
Cheers
With all the hype these days with regards to power per channel It was a little difficult to accept three watts per channel but I'm sure after what I have seen and read it's totally worth it. Apart from better bass I wasn't aware that the SSE will give a larger output so maybe I need to review my selection.
Cheers
The SSE with KT88 output tubes will give around 14 watts in ultralinear mode, this should be enough to drive your 89db efficient speakers
to loud levels ( that is if they are indeed 89db efficient, some commercial speaker manufacturers are not very reliable with their numbers).
It all depends on the kind of music you listen to, how loud you like to listen and the size of your listening.
Electro Harmonix KT88 tubes are an excellent value for the money based on my personal listening.
14 Watts seems more than enough for me. I will eventually look at a new set of speakers so the Harmonix will be a good starting point.
As with anything the build cost is sort of escalating so until I do a bit more research in regards to transformer parts I'm holding back a little.
Many Thanks and full credit to all the Amp builders I've seen so far!
As with anything the build cost is sort of escalating so until I do a bit more research in regards to transformer parts I'm holding back a little.
Many Thanks and full credit to all the Amp builders I've seen so far!
Hear, hear!
Thanks, jdg123. I've been meaning to post my own update, too. Since modifying the 300B voltage regulators, I found the PT and choke at R4 getting extremely hot. I replaced it with two lower resistance, higher current chokes (Hammont 159T [2.5H, 300mA, 43 ohm]) and 200uf oil caps, as follows:
IMG_4161 by jeffdrouin, on Flickr
IMG_4153 by jeffdrouin, on Flickr
IMG_4170 by jeffdrouin, on FlickrSo now I have not one but two reactors on top to enhance the Cold War look!
It results in a little more volume, but more accurately improved loud-soft dynamics and transients. It sounds incredibly clean, detailed, and effortless. Thanks to George for such a great design!
I might want to tweak my new operating points. I have 19 more volts of B+ (374, up from 355), but the rest could possibly use some fiddling.
B+ = 374V
B- = 144.3 V
5842 plate = 148V (down from 163V)
300B = -68V, 98mA (used to be -60V, 92mA)
I could get more plate voltage out of the 5842s if I replaced them; one of them is dropping more and more below the other, so I wonder if that means it's failing. I can keep them equal at 148, but the other one can go much higher.
Hear, hear!
Thanks, jdg123. I've been meaning to post my own update, too. Since modifying the 300B voltage regulators, I found the PT and choke at R4 getting extremely hot. I replaced it with two lower resistance, higher current chokes (Hammont 159T [2.5H, 300mA, 43 ohm]) and 200uf oil caps, as follows:
So now I have not one but two reactors on top to enhance the Cold War look!
It results in a little more volume, but more accurately improved loud-soft dynamics and transients. It sounds incredibly clean, detailed, and effortless. Thanks to George for such a great design!
I might want to tweak my new operating points. I have 19 more volts of B+ (374, up from 355), but the rest could possibly use some fiddling.
B+ = 374V
B- = 144.3 V
5842 plate = 148V (down from 163V)
300B = -68V, 98mA (used to be -60V, 92mA)
I could get more plate voltage out of the 5842s if I replaced them; one of them is dropping more and more below the other, so I wonder if that means it's failing. I can keep them equal at 148, but the other one can go much higher.
George mentioned that its hard to match the 5842s. You're in the range but if you want you could drop the coin for an experimental batch...
You've already spent well with the PSU improvements 🙂
The free tweek would be to drop the bias 20mA at a time until you've gone too far and come back 5mA at a time, thus selecting an operating point by ear. A lot of folks settle in between 60 and 80 percent of maximum dissipation;-)
Thanks. Yes, I'll play with it next week. I liked it best with the bias current around 90mA (the OPTs are rated to 100mA).
I made a mistake in the values I wrote above. At 98mA the grid voltage is at -63.5. At 90mA it was -68.5V, which I was afraid would be too high, given that the optimal grid voltage in the 300B datasheets is listed as -60V.
As long as a high grid voltage is not going to damage the 300Bs, I'll play around with lower current bias just to hear how it sounds.
I forgot to mention that after two hours of playing, the PT gets very warm but not burning hot as before; I can keep my hand on it. The chokes and OPTs are cool to the touch.
I made a mistake in the values I wrote above. At 98mA the grid voltage is at -63.5. At 90mA it was -68.5V, which I was afraid would be too high, given that the optimal grid voltage in the 300B datasheets is listed as -60V.
As long as a high grid voltage is not going to damage the 300Bs, I'll play around with lower current bias just to hear how it sounds.
I forgot to mention that after two hours of playing, the PT gets very warm but not burning hot as before; I can keep my hand on it. The chokes and OPTs are cool to the touch.
Your numbers indicate you're running between 85 and 92% of maximum dissipation. Kinda hot but if you like it, that's the whole point. The 300B is a tough tube. For me it's tougher to replace. It could be influencing my liking around 75 percent of max...or it's the speakers or its the room, who knows. Anyway, you should be fine.
Cheers!
Cheers!
Please help... stumped
I built a working 300B Tubelab SE with ~405V B+ @ ~85mA. 5842's were biased to about 170V. Everything sounded good and I listened to it for hours.
I decided to look at square wave output so I hooked it up to the generator and everything looked about as I expected it look and sine looked perfect. The next day I went to re-bias the amp to make sure everything was still in spec and I noticed the output tube current was high so I bumped it down but when I did, the oscope displayed a funky shaped wave:
I immediately thought that the DC offset on the generator was off so I tried to adjust it only it didn't change... I quickly realized that you have to pull the knob out to adjust. Well, I pulled it out and instantly it put a 8-10V negative DC into the input of the amp and fried what I assumed were the CCS's.
I replaced both the CCS's and the MOSFET's, inspected the board for visual failures of other components, and tested nearly every passive component in the board. Everything is up to spec. Except now, every time I power it on, the bias current of the output tube goes right past 200mA within a few seconds (no matter where the trimmers are set) before it goes outside the meter's range and I have to turn the amp off. I have no idea what's going on since I haven't changed anything and all components seem to be perfectly fine.
The one thing that concerns me is there's a near -300V bias voltage... but I figured that's only until the B+ comes up (I'm using the Hammond 273BX - 700V). But the B+ can't get to where it's supposed to be though because the plate current goes so high.
Could the 300B's be bad?
I appreciate your help,
Kevin
I built a working 300B Tubelab SE with ~405V B+ @ ~85mA. 5842's were biased to about 170V. Everything sounded good and I listened to it for hours.
I decided to look at square wave output so I hooked it up to the generator and everything looked about as I expected it look and sine looked perfect. The next day I went to re-bias the amp to make sure everything was still in spec and I noticed the output tube current was high so I bumped it down but when I did, the oscope displayed a funky shaped wave:
I immediately thought that the DC offset on the generator was off so I tried to adjust it only it didn't change... I quickly realized that you have to pull the knob out to adjust. Well, I pulled it out and instantly it put a 8-10V negative DC into the input of the amp and fried what I assumed were the CCS's.
I replaced both the CCS's and the MOSFET's, inspected the board for visual failures of other components, and tested nearly every passive component in the board. Everything is up to spec. Except now, every time I power it on, the bias current of the output tube goes right past 200mA within a few seconds (no matter where the trimmers are set) before it goes outside the meter's range and I have to turn the amp off. I have no idea what's going on since I haven't changed anything and all components seem to be perfectly fine.
The one thing that concerns me is there's a near -300V bias voltage... but I figured that's only until the B+ comes up (I'm using the Hammond 273BX - 700V). But the B+ can't get to where it's supposed to be though because the plate current goes so high.
Could the 300B's be bad?
I appreciate your help,
Kevin
Last edited:
But that funky o scope shot was prior to the damage?
If I looked at your picture correctly, it shows 0.75 volts peak to peak which would only be 9 milliwatts into an 8 ohm load, if that waveform is at the speaker terminals.
What part of the amp was being measured?
Win W5JAG
If I looked at your picture correctly, it shows 0.75 volts peak to peak which would only be 9 milliwatts into an 8 ohm load, if that waveform is at the speaker terminals.
What part of the amp was being measured?
Win W5JAG
Thanks for responding, yes prior to damage.
The probe I was using is a x10 at the speaker outs across 8-ohm resistor. It's old so there could be some noise but it should be 5V/div.
After I find some resistors to replace those, I'll hook it back up and look at the wave again. Could a DC offset (a little one) cause that top peak to invert like that?
I don't know; it's an interesting question. My cheap generator does not have an offset adjustment that I know of, so I can't try that to see what it does. I don't know if my ancient HP does or not - it's too big for my bench, so I never use it.
Maybe one of the more knowledgeable people on here could answer that question.
I don't use a probe when I am measuring mine - I just use a good quality clip lead.
Win W5JAG
Maybe one of the more knowledgeable people on here could answer that question.
I don't use a probe when I am measuring mine - I just use a good quality clip lead.
Win W5JAG
It is strange but I have been having some of the same problems as you with my SE amp. My latest try resulted in burned up grid resistors too. It did not happen until I dropped the needle of my turntable and it saw signal. Someone said I may not be measuring the bias correctly across the plate resistors. I have set them for 62 ma a number of times. I have actually put this build on hold until I give it some more thought. I have checked and rechecked most everything on the board for mistakes and I have found no problem even though I replaced some parts just for the heck of it. This is not my first tube build, more like my 6th or 7th but all the ones before were point to point builds. Very frustrating that I cannot get this one to work. It has made me wonder about my output tubes as well. I am building another 300B without CCS or transistors so I should see shortly if my tubes are still OK of which I think they are.
After I replaced the grid resistors on the output tubes everything is back to working fine. This time I used 1/2 watt resistors. George recommended 1/4 watt (which is probably fine, I'm pretty sure it was negative DC that fried them) but with a significantly higher current and B+, I went ahead and upgraded.
Remember that there's no cap on the input of this amp so any DC can have a catastrophic effect on the semiconductors.
The best way is to have the 10ohm 1% (R18 and R29) installed and measure the voltage across it. On a meter set to 2V, centi-volts should be equal to milli-amps. In other words, .700 on the meter should equal 70mA.
The current on my tubes has drifted substantially over the hours since I first powered it up. I think George even recommends to keep the meters hooked up for the first X number of hours.
I think the tubes are pretty resilient but the solid state components are probably the first point of failure.
Possible LFO from tonearm resonance?
There is no grid stop resistor on the 5842 and that's a high gm tube. Also the cathode bypass cap is pretty large, I would think even a 100uF is enough. Since that stage has a lot of gain I would want to set the low frequency gain a bit lower, especially if using a turntable.
Cathode bypass cap formula is:
Rk | | ( ( Ra + ri ) / ( mu + 1 ) ) = R
1 / ( 2 * pi * R * f ) = C
Rk is cathode resistor
Ra is plate load
ri is internal resistance
f is lowest frequency
Since the CCS is such a high impedance we can pretty much assume Rk will dominate.
For an R of 300, and a low frequency of 5Hz, we get 100uF
There's not one on the schematic, but there is on the boards: R31, R32. Here he recommends increasing the value to 4.7K using carbon comps:
Tubelab SE update | Tubelab