Firstly, i wouldnt build this for myself. My personal headphone amps are parafeed / single end.
I am building a parts bin amp for my bro (senn 650s).
So, looking to crowdsource some improvements to this:
My ideas:
The input stage looks badly optimised (The triode curves for the 6n1p arent pretty at low currents). I was thinking ccs load, 8 to 10ma, at the same plate voltage (probably SIC diode bias at 1.6v?)
Possibly put some nfb from output to input? Will help lower the unecessary amount of gain and lower imp?
Bias the heaters at 75v ish. The article leaves them floating floating...
Constraints:
6n1p input and output. I have a load of them, they will cost me nothing.
OTL - he wont spend on txs
350v ish B+, i have an old TX.
£20 limit on new parts to enhance!
thoughts welcome!
I am building a parts bin amp for my bro (senn 650s).
So, looking to crowdsource some improvements to this:
My ideas:
The input stage looks badly optimised (The triode curves for the 6n1p arent pretty at low currents). I was thinking ccs load, 8 to 10ma, at the same plate voltage (probably SIC diode bias at 1.6v?)
Possibly put some nfb from output to input? Will help lower the unecessary amount of gain and lower imp?
Bias the heaters at 75v ish. The article leaves them floating floating...
Constraints:
6n1p input and output. I have a load of them, they will cost me nothing.
OTL - he wont spend on txs
350v ish B+, i have an old TX.
£20 limit on new parts to enhance!
thoughts welcome!
> curves for the 6n1p arent pretty at low currents). I was thinking ccs load, 8 to 10ma, at the same plate voltage (probably SIC diode bias at 1.6v?)
The "power" stage runs 10mA. You want to put the same amount of power in its grid driver?
I bet if you blow-up the "low current" part of the graphs (assuming they are honest), they look a lot like any other part of the terrain.
This "low current" operation seems to happen with 1V on cathode. Why do you figure 1.6V for higher current?
The BULK of the nonlinearity will be in the heavily loaded yet wimpy "power" stage.
The "power" stage runs 10mA. You want to put the same amount of power in its grid driver?
I bet if you blow-up the "low current" part of the graphs (assuming they are honest), they look a lot like any other part of the terrain.
This "low current" operation seems to happen with 1V on cathode. Why do you figure 1.6V for higher current?
The BULK of the nonlinearity will be in the heavily loaded yet wimpy "power" stage.
That's the headwise amp - I've seen Broskie has some updates over at tube cad but can't find the exact page. I'll see if I can find the page. OTL SSRP/PSRR searches always seem to pop up white cathode followers etc. For £20 on new parts I'd look at the power supply noise rejection. Possible minor alterations on the design to allow cancellation etc.
Have spent the best part of several months banging my head against the wall with OTL 32ohm headphone amps, it would be worth having a scan over tubecad - he has a mass of OTL including non-defacto non-6AS7 push pull designs.
A parallel output section would help the beef up the control over the bottom end?
Last edited:
I guess this is a kit.
Despite that, I'm not sure the 6N1P is a good candidate for the output stage.
Maybe another of the same family...say 6N3 or 6N5....unsure....I'm sure total gain and power would suffer, though the valves wouldn't be as heavily loaded.
Despite that, I'm not sure the 6N1P is a good candidate for the output stage.
Maybe another of the same family...say 6N3 or 6N5....unsure....I'm sure total gain and power would suffer, though the valves wouldn't be as heavily loaded.
Thanks for feedback so far. It's not a kit, I will P2P it. I can build 95% from parts bin and he won't pay me for it... So I am just being cheap.
My personal amps are a d3a parafeed with sshv V2 shunt, and my office amp is a 6e6p Dr parafeed. They drive anything and sound glorious, but the txs are best part of £200, I like my bro... But no that much.
My personal amps are a d3a parafeed with sshv V2 shunt, and my office amp is a 6e6p Dr parafeed. They drive anything and sound glorious, but the txs are best part of £200, I like my bro... But no that much.
6n6p would be beefier in the output and i have a couple in the bin, but the parts bin power tx wouldnt cope with the excessive heater current. Total power for heaters needs to be under 2a. Not sure I even need the input stage... Maybe just an srpp with 6n6ps... But then will need to drop b+ a bit, they look happiest around 120 -150 b+
Last edited:
If you are stuck with 6N1P, I would consider using 2 6N1P and parallel them. ie: Use 5 tubes instead of 3 tubes.
Hi again.
Does 6N5P meet heater requirements?
Checking it seems 6N1P is 600mA, as is the 6N5P. 6N3P is only 350mA.
I have a few 6N5P and (I think) some 6N3P too, if you wwnt to try those and have trouble finding them (though it wasnt too hard last time I looked they were still abundant)
Does 6N5P meet heater requirements?
Checking it seems 6N1P is 600mA, as is the 6N5P. 6N3P is only 350mA.
I have a few 6N5P and (I think) some 6N3P too, if you wwnt to try those and have trouble finding them (though it wasnt too hard last time I looked they were still abundant)
Last edited:
Build a solid state op-amp headphone amp, add a couple of dead/dying valves with heater glow but not connected - jobs a goodun. 😀
Seriously, despite Nick's comment, the simplest thing to do would probably be a MOSFET follower to take the load off the valves.
But valves are interesting things....(I've built a simple NE5532 headphone gain stage with, LH033 output stage, BD139/140 discrete output stage. Kinda made in one afternoon projects. Fun but not much to do.)
My favourite is my battery DHP Headphone amp (not using batteries)- so not altogether optimal.
I have KodaBMX's 6N3, 6N5 or 6N6 headphone amp PCB, which I started, then messed up (rubbish sockets I'd bought), so I will be starting over with that design soon.
But valves are interesting things....(I've built a simple NE5532 headphone gain stage with, LH033 output stage, BD139/140 discrete output stage. Kinda made in one afternoon projects. Fun but not much to do.)
My favourite is my battery DHP Headphone amp (not using batteries)- so not altogether optimal.
I have KodaBMX's 6N3, 6N5 or 6N6 headphone amp PCB, which I started, then messed up (rubbish sockets I'd bought), so I will be starting over with that design soon.
Last edited:
The 6n1p are weak... but the 650 are easy to drive. You should be fine current wise.
Otoh, a CCS loaded 6n1p will give you way too much gain and the output impedance will be highish. Consider adding feedback, as was done for the Morgan Jones. That helps on both front.
If your transformer is a bit weak, an old laptop supply will happily provide for 3 heaters in serie and can usually be found for (next to) nothing
Otoh, a CCS loaded 6n1p will give you way too much gain and the output impedance will be highish. Consider adding feedback, as was done for the Morgan Jones. That helps on both front.
If your transformer is a bit weak, an old laptop supply will happily provide for 3 heaters in serie and can usually be found for (next to) nothing
HD650 are quite easy to drive, being 300R.
About 100dB/mW, 105dB/V
50mW would probably be ample. My DHP amp gets about 50mW, but the output impedance is about 80R...
I have heard they (HD650) are sensitive to output impedance, or damping factor.
About 100dB/mW, 105dB/V
50mW would probably be ample. My DHP amp gets about 50mW, but the output impedance is about 80R...
I have heard they (HD650) are sensitive to output impedance, or damping factor.
Last edited:
Here is a cheap parallel feed option. A pair of Hammond 156C (150 Henries, 8mA, 3,700 ohms) about $36 for a pair and $16 a pair of
Radio Shack 70 Volt 10 Watt Line Transformers as outputs from the auction site. The Radio Shack (32-1031B) have plenty of taps to match your headphones and are good sounding outputs. That is $52 for the iron and it sounds better to my ears than the 6N1P OTL. Just a suggestion to get your buddy on valves and iron.
Edcor used to do nice series feed outputs like the XSM 10K/600 ohm for not much money but they do not appear to be catalog offerings. I’m sure they could wind them and that would drop your iron cost a little but it wouldn’t be parallel feed.
Matt
Radio Shack 70 Volt 10 Watt Line Transformers as outputs from the auction site. The Radio Shack (32-1031B) have plenty of taps to match your headphones and are good sounding outputs. That is $52 for the iron and it sounds better to my ears than the 6N1P OTL. Just a suggestion to get your buddy on valves and iron.
Edcor used to do nice series feed outputs like the XSM 10K/600 ohm for not much money but they do not appear to be catalog offerings. I’m sure they could wind them and that would drop your iron cost a little but it wouldn’t be parallel feed.
Matt
Edcor has a dedicated parafeed headphones transformer: EDCOR - PCW10K-7K/300-32
Very cheap in the US, twice the price in Europe though.
Very cheap in the US, twice the price in Europe though.
Sorry all if my sense of humour gets in the way 🙂
A MOSFET cathode follower would be a simple way of 'paralleling' the output without the heater burden. Only issue is with a push pull is the crossover and change of impedance for an OTL. Just needs a little planning.
Staying with tubes
Broskie's Aikido HPAs do very well with high impedance 'phones, IMHO. I've heard the 4-tube versions work well on Senn. 600s and 650s. See attachment for a 3-tube version from the Tubecad site a few years back.
Note: Only a 1 or 2uF output cap would be needed for high imp 'phones. You'd also want a bleeder resistor from the 'phones side of the output cap to ground.
Steve
Broskie's Aikido HPAs do very well with high impedance 'phones, IMHO. I've heard the 4-tube versions work well on Senn. 600s and 650s. See attachment for a 3-tube version from the Tubecad site a few years back.
Note: Only a 1 or 2uF output cap would be needed for high imp 'phones. You'd also want a bleeder resistor from the 'phones side of the output cap to ground.
Steve
Attachments
knocked it up this afternoon.
First tweak, it sounded loose. Needed a touch of NFB. I put 3 levels on a switch.
Second tweak, i have never needed to use DC heaters in my life... but this needed them.
Third tweak, DC heaters elevated to 70v.
Fourth tweak, i needed more capacitance in the PS to be hum free.
Fifth tweak, tried the CCS, it sounded better, but then i have my parafeeds for better. It sounds comfier..softer without (tubey as the kids would call it). Removed. this will be in my office, so warm and comfy is appreciated.
Sixth tweak - LED (yellow) for 2v in the SRPP lower cathode. just noticable in the bass.
Not bad for a parts bin amp. my only lie is the bridge rectifier, ran out. £2 ebay.
First tweak, it sounded loose. Needed a touch of NFB. I put 3 levels on a switch.
Second tweak, i have never needed to use DC heaters in my life... but this needed them.
Third tweak, DC heaters elevated to 70v.
Fourth tweak, i needed more capacitance in the PS to be hum free.
Fifth tweak, tried the CCS, it sounded better, but then i have my parafeeds for better. It sounds comfier..softer without (tubey as the kids would call it). Removed. this will be in my office, so warm and comfy is appreciated.
Sixth tweak - LED (yellow) for 2v in the SRPP lower cathode. just noticable in the bass.
Not bad for a parts bin amp. my only lie is the bridge rectifier, ran out. £2 ebay.
Last edited:
