I think that might be best. One transformer that supplies the B+ and heater, and another that supplies the mosfets. I don't think having one custom-made (I wouldn't try to make it myself) would be time or cost efficient. AI says we would need a single toroidal transformer with:
https://www.hammfg.com/files/parts/pdf/369AX.pdf
- High-Voltage Winding: 275V AC @ 150mA.
- Heater Winding: 6.3V AC @ 5A.
- Low-Voltage Winding: 15V AC (center-tapped) @ 500mA for rectification into ±15V DC.
https://www.hammfg.com/files/parts/pdf/369AX.pdf
I would do my own take on a Nutube + MOSFET hybrid, and if I have a spare hour or 2 on the weekend, I may do just that and sketch some ideas.
Unless there's some compelling need for a HV tube, I would avoid any HV stuff, or transformer coupling, just for convenience and ease of development. I'd start with something like the suggested Nutube schematics, or some variation, and couple that to a MOSFET gate.
I'd even use the MOSFET as a gain stage and concoct my own mixed-mode NFB circuit, "doing it my way".
However, if I simply attach a sketch, you may be at a complete loss to understand my reasoning, unless it's accompanied by a tonne of documentation. And that's usually better done in a dedicated design / build thread, or off-site. Hence the great advice others have given, to start off with a proven project someone else has already developed.
It may be that there is, in fact, a shortage of hybrid projects, but have you checked?
The Pete Millet Nutube HPA was already mentioned. How's your knowledge of op-amps?
Unless there's some compelling need for a HV tube, I would avoid any HV stuff, or transformer coupling, just for convenience and ease of development. I'd start with something like the suggested Nutube schematics, or some variation, and couple that to a MOSFET gate.
I'd even use the MOSFET as a gain stage and concoct my own mixed-mode NFB circuit, "doing it my way".
However, if I simply attach a sketch, you may be at a complete loss to understand my reasoning, unless it's accompanied by a tonne of documentation. And that's usually better done in a dedicated design / build thread, or off-site. Hence the great advice others have given, to start off with a proven project someone else has already developed.
It may be that there is, in fact, a shortage of hybrid projects, but have you checked?
The Pete Millet Nutube HPA was already mentioned. How's your knowledge of op-amps?
Well, it had been my intention to avoid using op-amps or integrated circuits. I thought that since I am using AI to design it for me, it would be interesting to make as old-school an amp as would properly serve my 32Ohm headphones. I did look, but did not find any hybrid-transistor headphone amp kits or designs. I did not try looking for a speaker amp design or kit that I could adapt for headphones. Why? I'm a dummy often. I'm not always a dummy, though. But you'll catch me slipping often, and I'll try not to vex anyone's patience. I was, in retrospect, quite eager to involve AI in this. But if I had found a tube-transistor hybrid headphones amp kit that puts 1W into 32Ohms, I'd have bought and built that. And again, I have looked.
I also didn't want to use any designs that were push-pull or had output transformers. It's gotta leave those 2nd order harmonics intact from the tube.
I guess that's another way to look at this amp idea: a Bottlehead Crack with low output impedance, keeping it otl and single-ended.
I know AI is prone to error right now. I don't know which AI you used, but some are better than others.
I know AI is prone to error right now. I don't know which AI you used, but some are better than others.
Upon reflection, I believe I've learned this much so far:
1. AI isn't ready. It can't envision the whole amp and still have room to reason properly. The less it knows, the better it thinks. The more it knows, the worse it thinks. We all struggle against our limitations.
2. Me just having learned what these components are and the beginnings of what they do does not enable me to have the grand vision of the amp that it would take to oversee AI's thinking about each section (which it currently can't do convincingly for experts).
3. I can't export that thinking and vision to either AI or the good people here. Actually, either would be ill-advised. The best thing for me to do would be to get an education, as you all say, and revisit the idea in the 2030s sometime. That would actually benefit me the most, and be the most satisfying and potentially fruitful course of action in the long run, if AI regresses.
4. I have to fall back and re-group. But this has been a very eye-opening experience for me. I truly appreciate your participation so far.
I expect that I will be back at some point in the future when it seems that AI finally can do it. I give you all my word that when I come back talking about this idea, I will have been working with an AI advanced enough to resolve all of these issues, and provided me with a schematic, gerber file, and anything else you might want to see. And I will use everything you've taught me and everything I'll continue to learn about this amp and its components and design in evaluating what AI gives me in the future. For improving my ability to prompt AI, I am forever grateful. And I will return to this forum and re-read what you've said and the links you've suggested. In other words,
When I return!
1. AI isn't ready. It can't envision the whole amp and still have room to reason properly. The less it knows, the better it thinks. The more it knows, the worse it thinks. We all struggle against our limitations.
2. Me just having learned what these components are and the beginnings of what they do does not enable me to have the grand vision of the amp that it would take to oversee AI's thinking about each section (which it currently can't do convincingly for experts).
3. I can't export that thinking and vision to either AI or the good people here. Actually, either would be ill-advised. The best thing for me to do would be to get an education, as you all say, and revisit the idea in the 2030s sometime. That would actually benefit me the most, and be the most satisfying and potentially fruitful course of action in the long run, if AI regresses.
4. I have to fall back and re-group. But this has been a very eye-opening experience for me. I truly appreciate your participation so far.
I expect that I will be back at some point in the future when it seems that AI finally can do it. I give you all my word that when I come back talking about this idea, I will have been working with an AI advanced enough to resolve all of these issues, and provided me with a schematic, gerber file, and anything else you might want to see. And I will use everything you've taught me and everything I'll continue to learn about this amp and its components and design in evaluating what AI gives me in the future. For improving my ability to prompt AI, I am forever grateful. And I will return to this forum and re-read what you've said and the links you've suggested. In other words,
When I return!
Very few headphones needs more than +30dB.
Most can do with +20dB.
Here is the power needed for amplification of headphones.
+10dB 3.16mW
+20dB 10mW
+30dB 31.6mW
+40dB 100mW
+50dB 316mW
+60dB 1 Watt
Most can do with +20dB.
Here is the power needed for amplification of headphones.
+10dB 3.16mW
+20dB 10mW
+30dB 31.6mW
+40dB 100mW
+50dB 316mW
+60dB 1 Watt
No one said that you must get an engineering degree to do what you intend, but a bit of background in electronics would be very beneficial. Try finding some old audio magazine or electronic hobbyist mags and cook books to get a better grasp of electronics, the components and even the jargon used by these people. You putting a target date of 2030, is plain ridiculous.
Kill this thread and maybe start a new one with something along these lines: "I need a headphone amp, can someone help" Don't start mixing technologies from the word go (tubes and semi) that already poses a threat to any newcomer. Start something simple just to get the jest of it. Feel what it feels like to build something that actually works, then take it a step further.
I mean you no disrespect, I simply feel that pitching AI against these very helpful folk here was a wrong approach from the outset. Besides copy and pasting AI response makes someone immediately defensive, you could have at least ask or disclose something in your own words and understanding of the matter. AI is probably better put to use for online dating in my opinion.
Kill this thread and maybe start a new one with something along these lines: "I need a headphone amp, can someone help" Don't start mixing technologies from the word go (tubes and semi) that already poses a threat to any newcomer. Start something simple just to get the jest of it. Feel what it feels like to build something that actually works, then take it a step further.
I mean you no disrespect, I simply feel that pitching AI against these very helpful folk here was a wrong approach from the outset. Besides copy and pasting AI response makes someone immediately defensive, you could have at least ask or disclose something in your own words and understanding of the matter. AI is probably better put to use for online dating in my opinion.