This proposed circuit is little more than the driver from my Fi Primer Loftin-White & an output stage from an old AudioXpress. The driver has lots of gain & sounds good. The output stage seems like it could use a higher Z OPT & maybe a different operating point to better match the 6L6GC?, as I have a shirtload lying around left over from my guitar playing years. I'm quite pleased with the L-W, but can't seem to stop building amps.
Attachments
Hi 5u4,
nothing to really critizise, only a few remarks to consider:
6SL7 likes considerable higher plate voltages best, about the 195V total you halved for each system in the symmetrical totem pole. If you like it as is, okay, but I would prefer paralleled sections instead (also that would mean one heater winding less - max. heater-cathode voltage for 6SL7 is only 90V, IIRC).
Also, bandwidth will be more limited using the current trickler totem pole you showed than with paralleled 6SL7 sections. Don´t get confused by the "on paper" low output impedance of the totem pole - what counts instead is the peak current driving capability into the Miller-C of the trioded 6L6 (about 70pF, I guess). Your totem will be able to swing only a few hundred nano-Amps (at best) into that Miller-C without slewing in AF range.
You can happily reduce the coupling cap to 0,1uF or something.
A more appropiate plate load for trioded 6L6 or 807 would be around Ra=5k, but don´t expect more than, say, 1,5-2 Watts Po, either way (plate load).
The same plate load (about 5k) could give up to about some 8 watts or something when using tetrode mode instead. So, maybe supplying a switch to chose might be of interest. (For tetrode mode, your totem pole would be preferable, allowing for more gNFB due to higher gain - and it certainly will drive much lower Miller-C of 6L6 in tetrode mode to much higher freqs).
Given the gain reserves, 3-6 dBs of gNFB are possible even when using paralleled 6SL7 sections (which will produce less gain than your totem pole setup) and 6L6 in triode mode. And don´t miss the opportunity to try that. Might improve bandwidth and damping factor noticably.
But, as this all are only remarks, just start with the schematic as shown, it certainly will work as is.
Tom
nothing to really critizise, only a few remarks to consider:
6SL7 likes considerable higher plate voltages best, about the 195V total you halved for each system in the symmetrical totem pole. If you like it as is, okay, but I would prefer paralleled sections instead (also that would mean one heater winding less - max. heater-cathode voltage for 6SL7 is only 90V, IIRC).
Also, bandwidth will be more limited using the current trickler totem pole you showed than with paralleled 6SL7 sections. Don´t get confused by the "on paper" low output impedance of the totem pole - what counts instead is the peak current driving capability into the Miller-C of the trioded 6L6 (about 70pF, I guess). Your totem will be able to swing only a few hundred nano-Amps (at best) into that Miller-C without slewing in AF range.
You can happily reduce the coupling cap to 0,1uF or something.
A more appropiate plate load for trioded 6L6 or 807 would be around Ra=5k, but don´t expect more than, say, 1,5-2 Watts Po, either way (plate load).
The same plate load (about 5k) could give up to about some 8 watts or something when using tetrode mode instead. So, maybe supplying a switch to chose might be of interest. (For tetrode mode, your totem pole would be preferable, allowing for more gNFB due to higher gain - and it certainly will drive much lower Miller-C of 6L6 in tetrode mode to much higher freqs).
Given the gain reserves, 3-6 dBs of gNFB are possible even when using paralleled 6SL7 sections (which will produce less gain than your totem pole setup) and 6L6 in triode mode. And don´t miss the opportunity to try that. Might improve bandwidth and damping factor noticably.
But, as this all are only remarks, just start with the schematic as shown, it certainly will work as is.
Tom
Hi 5u4,
I feel a bit reluctant about 6L6 or 807 SE UL mode, since only very limited and scarce data is available (I know only the Sofia Audiomatica plot, see attachment below, which is almost useless for SE -designing due to the excessively high g2 voltage and very limited g1 and plate voltage range).
My best bet would be going along the STC 807 spec sheet class A SE tetrode mode recommendations with Zener- or VR-tube regulated screen feed, a 6SJ7 in front of that, and using about 12-15dB of gNFB.
Tom
I feel a bit reluctant about 6L6 or 807 SE UL mode, since only very limited and scarce data is available (I know only the Sofia Audiomatica plot, see attachment below, which is almost useless for SE -designing due to the excessively high g2 voltage and very limited g1 and plate voltage range).
My best bet would be going along the STC 807 spec sheet class A SE tetrode mode recommendations with Zener- or VR-tube regulated screen feed, a 6SJ7 in front of that, and using about 12-15dB of gNFB.
Tom
Attachments
I am currently using a 6L6GC in SE UL. I am using a 5K load with the $18 Edcor transformers. I use cathode feedback to improve the damping and lower distortion. No global feedback is used. I am running 445 volts B+ and cathode bias which puts about 400 volts across the tube. The cathode resistor is 560 ohms. This allows the use of several different output tubes without circuit changes. The driver is just a simple common cathode stage with a CCS load using 1/2 of a 12AT7 at 8 mA.
With a 6L6GC (Shuguang) I get 12 watts at the onset of clipping (5% distortion). Distortion is below 3% at 10 watts, and under 1% at 5 watts. Frequency response (3db) is 33Hz to 45KHz at 10 watts. I can use triode mode by flipping a switch, but I haven't made any power measurements yet. I would guess about 5 watts at clip.
Similar power and distortion readings are obtained with an EL-34, but the amp sounds different, more like a DHT SET especially in triode mode. More power (14 watts at clip) can be obtained by popping in a KT-88. This is the tube to use for loud rock music.
With a 6L6GC (Shuguang) I get 12 watts at the onset of clipping (5% distortion). Distortion is below 3% at 10 watts, and under 1% at 5 watts. Frequency response (3db) is 33Hz to 45KHz at 10 watts. I can use triode mode by flipping a switch, but I haven't made any power measurements yet. I would guess about 5 watts at clip.
Similar power and distortion readings are obtained with an EL-34, but the amp sounds different, more like a DHT SET especially in triode mode. More power (14 watts at clip) can be obtained by popping in a KT-88. This is the tube to use for loud rock music.
My cent...
I prefer running the 6SL7 with more current, and use 1K for both cathode resistors and around 200V.
I also dabbled with 807, starting with Gordon Rankin's design (6072 SRPP driver) and really liked it. But it uses 0C3 and 0D3 for the screen grid regulation.
Then I tried UL with the Hammond 5K w/ UL tap (the model scaped my mind right now), 350Vp and it sounds good to me too.
I like the tone of the 807 over the EL34 in most cases, but that's just me.
Experience confirms what many has said about the 807 G2- it should be well regulated. This means more parts but I think it is well worth the effort.
I prefer running the 6SL7 with more current, and use 1K for both cathode resistors and around 200V.
I also dabbled with 807, starting with Gordon Rankin's design (6072 SRPP driver) and really liked it. But it uses 0C3 and 0D3 for the screen grid regulation.
Then I tried UL with the Hammond 5K w/ UL tap (the model scaped my mind right now), 350Vp and it sounds good to me too.
I like the tone of the 807 over the EL34 in most cases, but that's just me.
Experience confirms what many has said about the 807 G2- it should be well regulated. This means more parts but I think it is well worth the effort.
I first ran the 12AT7 with an 18K resistor load. Measured voltage gain was about 30. With cathode feedback, this was not quite enough gain for all applications. I replaced the resistor with an IXYS CCS chip (no other changes) and the voltage gain went to about 55. Measured distortion for the entire amp went down slightly. This is enough gain to drive the amp to clipping in triode mode.
I have not tried the $75 Edcors yet, but I will when I have time. I will be working crazy hours for the next few weeks, leaving little time for tube fun.
I have found the absolute cheapest power transformer for this amp is the 6K7VG from Allied electronics ($43). I have been using their transformers for years and the only one that ever failed was the one that got wet during a hurricane. www.alliedelec.com
I have not tried the $75 Edcors yet, but I will when I have time. I will be working crazy hours for the next few weeks, leaving little time for tube fun.
I have found the absolute cheapest power transformer for this amp is the 6K7VG from Allied electronics ($43). I have been using their transformers for years and the only one that ever failed was the one that got wet during a hurricane. www.alliedelec.com
I have ran dozens of tubes through this amp, including some relatively high priced GE 6550's and RCA black plate 6L6's. I keep coming back to the $8 Shuguang 6L6GC's. My second choice is the old GE 6L6GB's that have already lived a hard life in a Bandmaster. These are very similar to 807's. I have a NOS pair of these somewhere, but I haven't found them yet. I also have lots of 807's, but I can't just plug them in.
I think that the EL-34's sound the most like DHT's, but I have a two 300B amps already.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.