Valve Headphone Amp

[simon]

Rather than hijack Steve's SEP thread with talk of headphones I thought I'd start a new one.

A long time ago I built an Aikido headphone amp and it was fine, worked well. I'd quite like to build another using a nice DHT for gain. The challenges as I imagine them are:

1 - Getting a DHT quiet enough for use with headphones. This should be okay with the current regs we have available to us these days.

2 - Getting enough gain from a DHT. I'm not sure how much I'd need, but maybe a 26 or 10Y with a notional gain of 8 might be enough?

3 - Driving the headphones, in this case 300 ohm Sennheiser 650s. An OPT would be the traditional way I suppose, if a suitable candidate existed. Lundahl look to offer some but I don't how suitable they would be. Cheaper would be some kind of follower/buffer, perhaps the second half of an Aikido?

Anything else I've missed? Thoughts on how 2 and 3 might be achieved? Or would I just be asking for a lot of trouble.

[simon]

I should probably add that I was thinking SE not PP...

[simon]

Looking through my parts box I've a pair of 15k 1:1 (I think) ITs made for 10Ys. Perhaps not ideal but I guess nothing would break if I tried them with a simple 10Y common cathode and connect the IT secondary to the headphones. Or is this a bad idea?

[Nick]

Will be applying the 300R load to the 10Y's. Not ideal.

How about a MoFo shape to drive the headphones. Could be smaller and lower current. Use the 10Y to provide gain and the MoFo as a buffer to driver the phones.

[simon]

Hmmm, yes of course, the Senn's impedance will dominate.

I'd had a passing thought about a MoFo buffer . Feels like a bit of overkill, but no more than other solutions I guess.

Would be easy enough to run a 10Y IT in to the MoFo I've already built to try.

[Nick]

Feels like a bit of overkill

Well you can scale everything down to match the 300R and power requirement.

[simon]

Something like this perhaps? Looks simple enough, perhaps too simple?

https://headwizememorial.wordpress.com/ ... ne-driver/

[ed]

A couple of things came to mind after your last post.

Do you want a headphone amp or do you want to experiment more with valves or do you insist on designing/ building a headphone amplifier. There doesn’t have to be, but there is a distinction.

Your post reminded me of the simple latfet hpa that I proposed a while back because there is a resemblance:

https://www.audio-talk.co.uk/phpBB3/vie ... php?t=8003

hpa cct.jpg (61.96 KiB) Viewed 3113 times

I was rather hoping that I would get some feedback on the sound from Ant as he has the spare boards. I’m not confident these days that I can see through the emotional investment of my own creations, but this HPA does sound good to me…and it’s cheap enough to breadboard.

[Ant]

Havent built it yet ed, things have gone rather sideways again lately. As usual, i lurch from one minor crisis to another. I have barely listened to any music in a few months, bought a streamer and havent used it, daughter is using the tidal subscription so at least thats not going to waste
Bought some records a week or so ago and i havent even taken them out of the bag.

Im sure it will sound good, i can pass the boards over to simon if he wants to build it?

[simon]

Some good questions ed.

I'm thinking valve front end just because I like valves really. I've some 10Ys and they glow nicely. I'm shallow. And I've a breadboarded preamp that I robbed a few bits from that should be easy to get working again. It's all power supply really, there's not much design to a common cathode IT coupled triode.

The buffer is more interesting - I'd quite happily use a suggested MoFo type buffer, 3 legged things make me break out in a cold sweat.

A different SS front end might sound better than a valve. My experience of the MoFo with opamp gain stage is it's remarkably good. It's not as good as my 45 amp though, but the parts probably cost 10x as much. You pays yer money. I have the bits for the 10Y stage from years of collecting so why not use them.

[simon]

Looking a bit closer at ed's circuit above I have a few questions - not questioning or wanting to re-engineer, just to understand better and overcome my aversion of 3 legged things!

The buffer part is from R2, the gate stopper, to the right, yes?

C2 (and C3) and R5 are required because the PS isn't bipolar so the output isnt at ground?

The LM317 in the Source is to set the current (bias) through the latfet?

Why use ECX10N20? Because it "sounds good"? Is suited to the task? It looks like it's "audio grade" but now obsolete - is it likely to be superior to say IRF510?

[simon]

Hmmm, scratch some of that, there's no gain stage before the ECX10N20 so the latfet must be a combined gain stage and buffer?

[Nick]

Just a buffer I think as it's a source follower.

[ed]

Why use ECX10N20? Because it "sounds good"? Is suited to the task? It looks like it's "audio grade" but now obsolete - is it likely to be superior to say IRF510?

not obsolete....

https://www.profusionplc.com/parts/ecx10n20
yes, very much superior to irf510 IMO

current builds I'm using the original Hitachis as I have a stash in the bit box...to all intents and purposes they are identical.

edit..apologies my only comparison has been with 610s and not 510s

[ed]

Why use ECX10N20? Because it "sounds good"? Is suited to the task? It looks like it's "audio grade" but now obsolete - is it likely to be superior to say IRF510?

not obsolete....

https://www.profusionplc.com/parts/ecx10n20
yes, very much superior to irf510 IMO

current builds I'm using the original Hitachis as I have a stash in the bit box...to all intents and purposes they are identical.

edit..apologies my only comparison has been with 610s and not 510s