Although I'm in the process of building a 46 (or 47) driver with a SUT and a hybrid mu-follower, here is my latest implementation of the 300B:
http://www.bartola.co.uk/valves/300b-se-amplifier/
I'm a big fan of DHT drivers, however the D3a has a sterling performance in the role.
Cheers
Ale
http://www.bartola.co.uk/valves/300b-se-amplifier/
I'm a big fan of DHT drivers, however the D3a has a sterling performance in the role.
Cheers
Ale
Hi Ale - I'm sure it's good. You are very familiar with the technology by now. And it will be a schematic others can be helped by. My NE5534 > 46 solution is a lot simpler and more modest, but I'm liking the sound. You know how fanatical I am about the timbre and tone of acoustic instruments and about overall clarity.
Hi Andy
I know and I always follow your guidance
The 46 is a superb DHT. Much cheaper than the 45. Also the 47 plays same way. As you may remember I used it as driver with the crazy idea of implementing filament bias on it. I could cook fried eggs any time! The sound was amazing though.
Cheers
Ale
I know and I always follow your guidance
The 46 is a superb DHT. Much cheaper than the 45. Also the 47 plays same way. As you may remember I used it as driver with the crazy idea of implementing filament bias on it. I could cook fried eggs any time! The sound was amazing though.
Cheers
Ale
Yes - I got the PCBs from Jim's Audio in Hong-Kong. They're brilliant - heavy, well made and places for all the components you need. I tried several single op-amps and liked the NE5534 best for tone and timbre. I'm using x5.5 gain without any compensation network.
Stereo opamp pre-amplifier PCB featuring good rail decoupling ! | eBay
Well, this reminds me of the talk I had with a DIY friend.
"I do not like the hybrid or sand(transistor) in the circuit."
"But your CDP has a lot of sand in it already."
"......................"
Other DIY friends always talk strongly for '3 stages amp for the full swing' when I prefer 2 stages. Recently I put a nice transistor preamp between my DAC and 10Y tube pre, and I think I am done with the idea of 3 stages. I just regard my DAC has a high output.
"I do not like the hybrid or sand(transistor) in the circuit."
"But your CDP has a lot of sand in it already."
"......................"
Other DIY friends always talk strongly for '3 stages amp for the full swing' when I prefer 2 stages. Recently I put a nice transistor preamp between my DAC and 10Y tube pre, and I think I am done with the idea of 3 stages. I just regard my DAC has a high output.
Thanks for your post, Chul. This is exactly what I mean. For those of us who only use a DAC it's just a question of boosting the signal up to a point where it drives a 2 stage DHT amp like 10Y into 300b. I've found that op-amps are good at this low level gain and they don't need a coupling cap which is pretty important in itself. A 2v output for a DAC is simply a convention and can be revised as needed to more than that.
Yes, and sometimes this idea is also good in terms of S/N ratio with DHT pre and sensitive speakers.
What brand of NE5534? Haven't heard much of a difference. They say original Signetics are best, but I just use the current ones. TI NE5534AP
NE5534AP | NE5534AP Texas Instruments, Op Amp, 10MHz, 8-Pin PDIP | RS Components
NE5534AP | NE5534AP Texas Instruments, Op Amp, 10MHz, 8-Pin PDIP | RS Components
Hi Andy my system beginning from the source is, DAC with 2.0V output + Jfet pre (gain of X 2.5) + 10Y pre with Gyrator + Single transistor follower amp + heavily modified Altec Valencia.
As you see I need a preamp, although many times I listen to other power amps (45, 2A3 SE with 12AT7 or 12AX7 as the driver).
Hi Andy,
I think you may be operating the 46 rather at low voltage. Definitely the 46 likes +25mA and 30mA being optimal. I know the restriction of filament bias is pushing you to bias it around 6V but this gives poor headroom to drive the 300B at higher volume levels.
The 46 likes 200V/30mA ideally. That would require a bias of about 27V which is not practical with filament bias. I’ve been there, done that. A mu-follower is a good option to avoid bypassing the cathode resistor.
I’m using a hybrid mu-follower with an unbypassed cathode resistor. SUT is 1:8 at the input with an LL7903. I used this setup before and loved its performance.
I can get 0.2% THD at 200Vpp. Harmonic profile is nearly perfect and sounds extremely detailed.
I found a few 46 I had from my used stock which were slightly irritating when playing at louder volumes. To my surprise, I put them under the inspection of the spectrogram (FFT) lens and found that they exhibited lots of higher rich harmonics when pushed above 100Vpp. Same effect can be seen at lower bias voltages.
I moved on to the 47, sounds as good and is less restrictive with the bias levels and gain is higher close to 7. Have a few globe 247 black plate NOS ones which are precious to me
Just sharing my experience.
Ale
I think you may be operating the 46 rather at low voltage. Definitely the 46 likes +25mA and 30mA being optimal. I know the restriction of filament bias is pushing you to bias it around 6V but this gives poor headroom to drive the 300B at higher volume levels.
The 46 likes 200V/30mA ideally. That would require a bias of about 27V which is not practical with filament bias. I’ve been there, done that. A mu-follower is a good option to avoid bypassing the cathode resistor.
I’m using a hybrid mu-follower with an unbypassed cathode resistor. SUT is 1:8 at the input with an LL7903. I used this setup before and loved its performance.
I can get 0.2% THD at 200Vpp. Harmonic profile is nearly perfect and sounds extremely detailed.
I found a few 46 I had from my used stock which were slightly irritating when playing at louder volumes. To my surprise, I put them under the inspection of the spectrogram (FFT) lens and found that they exhibited lots of higher rich harmonics when pushed above 100Vpp. Same effect can be seen at lower bias voltages.
I moved on to the 47, sounds as good and is less restrictive with the bias levels and gain is higher close to 7. Have a few globe 247 black plate NOS ones which are precious to me
Just sharing my experience.
Ale
Hi Ale - thanks for posting that. Very interesting. I'm using the 300b at a lower voltage and I don't listen loud, so I'm probably getting away with the 46 at it's present operating point. It does have 30mA through it. I'm not using filament bias - my NP plate choke has 180H inductance and this seems to be coping OK. I don't have your measuring skills or gear so I'm only guessing here but it sounds OK. I'll look more closely at your suggestions, and I'm sure this will be a help to others using a DHT driver for a 300b. Remind me how you are driving the 300b in your latest amp?
Hi Andy
For some reason I thought you were using the 46 in filament bias at low bias point, about 6V. My bad.
46 will ask for grid current when operating at low bias point in particular if you are using the Op Amp stage in between. It may cope but will increase distortion unnecessarily.
I’ve been playing the 300B for some time and love my last design.
Up until a week ago I had the D3a hybrid mu-follower driver stage with SiC diode cathode bias, coupled into a Source Follower which is DC coupled to the 300B grid. The design is here:
300B SE Amplifier – Bartola(R) Valves
Then moved to a 46 driver in hybrid mu-follower setup, unbypassed cathode resistor and a SUT of 1:8 into the 46 grid. This gives a gain of 45 and can swing 200Vpp at 0.2%. Hard to get better than that.
here is what I was playing before I moved to the 47:
300B SE Amp: 46 Driver (Part II) – Bartola(R) Valves
The 47 is exactly the same as previous 46 but with a different cathode resistor as curves are different. As said before with the 47 should get closer to 55-56 of gain with this setup.
I think I will leave with this for a quite a while. 2 DHT stages is what I wanted and this does the job really well.
Cheers
Ale
For some reason I thought you were using the 46 in filament bias at low bias point, about 6V. My bad.
46 will ask for grid current when operating at low bias point in particular if you are using the Op Amp stage in between. It may cope but will increase distortion unnecessarily.
I’ve been playing the 300B for some time and love my last design.
Up until a week ago I had the D3a hybrid mu-follower driver stage with SiC diode cathode bias, coupled into a Source Follower which is DC coupled to the 300B grid. The design is here:
300B SE Amplifier – Bartola(R) Valves
Then moved to a 46 driver in hybrid mu-follower setup, unbypassed cathode resistor and a SUT of 1:8 into the 46 grid. This gives a gain of 45 and can swing 200Vpp at 0.2%. Hard to get better than that.
here is what I was playing before I moved to the 47:
300B SE Amp: 46 Driver (Part II) – Bartola(R) Valves
The 47 is exactly the same as previous 46 but with a different cathode resistor as curves are different. As said before with the 47 should get closer to 55-56 of gain with this setup.
I think I will leave with this for a quite a while. 2 DHT stages is what I wanted and this does the job really well.
Cheers
Ale
This is approximately my 46 driver. The output has 100K to ground into the 300b. As I was saying, inductance of the amorphous NP plate choke is 180H - it's pretty big. I chose 6v bias to keep the unbypassed cathode resistor lower in value. I don't know if you feel that more than 6v would be an advantage. With 10v bias the cathode resistor goes up to 333 ohms.
Attachments
Last edited:
Hi Andy,
If you raise the HT to about 205V, cathode resistor of 470R, then anode voltage will go somewhere closer to 160V, bias around 13.8V to 14V with a 30mA anode current. This will give you about 160Vpp maximum output with an input that can swing nearly 13V before grid current becomes high enough.
I suspect you will get a stage which will have less odd harmonics (H3) versus even (H2 mainly). I'd bet you a beer it may even sound better. That will give your DAC more headroom.
Only tradeoff is the LF cut-off corner from 3.5Hz to about 5Hz which I think is good enough and worth trying. Also gain is just below 5 around 4.8-4.9.
You only need to change cathode resistor and adjust slightly the powers supply if that is possible?
Cheers
Ale
If you raise the HT to about 205V, cathode resistor of 470R, then anode voltage will go somewhere closer to 160V, bias around 13.8V to 14V with a 30mA anode current. This will give you about 160Vpp maximum output with an input that can swing nearly 13V before grid current becomes high enough.
I suspect you will get a stage which will have less odd harmonics (H3) versus even (H2 mainly). I'd bet you a beer it may even sound better. That will give your DAC more headroom.
Only tradeoff is the LF cut-off corner from 3.5Hz to about 5Hz which I think is good enough and worth trying. Also gain is just below 5 around 4.8-4.9.
You only need to change cathode resistor and adjust slightly the powers supply if that is possible?
Cheers
Ale