Many vacuum tubes work best at a few hundred volts. Just the nature of the beast. Making electrons leap through the void is harder than getting them to traverse a P-N junction.
High GM tubes can work ok at lower voltages, which is why you see ecc88 and ecc82 in lower voltage designs.
the 12ae6, 12fm6, and 12fk6 are "space charge" tubes designed for a b+ no higher than 24vdc, so the original millet hybrid is using them within their original design spec. They were designed to be used in car radios without using a 'vibrator' type voltage multiplier.
But the ecc82 is absolutely a better audio tube, as is the ecc88.
One of my favorite headphone amps is my Cavalli "Bijou" - which is an all-tube design (with some silicon diode biasing) at near 300 volts. Uses ecc88 types in the gain stage and 6n6p in the output stage.
High GM tubes can work ok at lower voltages, which is why you see ecc88 and ecc82 in lower voltage designs.
the 12ae6, 12fm6, and 12fk6 are "space charge" tubes designed for a b+ no higher than 24vdc, so the original millet hybrid is using them within their original design spec. They were designed to be used in car radios without using a 'vibrator' type voltage multiplier.
But the ecc82 is absolutely a better audio tube, as is the ecc88.
One of my favorite headphone amps is my Cavalli "Bijou" - which is an all-tube design (with some silicon diode biasing) at near 300 volts. Uses ecc88 types in the gain stage and 6n6p in the output stage.
Some food for thought, I know tubes work best with a few hundred volts across them, but that seems complete overkill for a headphone amp.
Understand about "space charge tubes" and wish they were more available in the UK.
ecc82 and ecc 88 correct me if I am wrong, but is 300V not over kill for headphones? I can build 5-20 Mullard mono's now, but it is not what I need.
I am looking for a decent hybrid amp.
Can build these of the shelf, and I know they are rarer than hens teeth, but not fit for purpose.
Necessary? No. Overkill? Depends on your design goals. It could be argued that all fancy headphone amps are overkill.
Ever listen to an Objective2 headphone amp? Costs about $100 to diy including the pretty faceplate. Based on opamps that cost less than a dollar for both.
And crushes many more expensive and extravagant designs easily.
.:Sent by pneumatic tubes
Ever listen to an Objective2 headphone amp? Costs about $100 to diy including the pretty faceplate. Based on opamps that cost less than a dollar for both.
And crushes many more expensive and extravagant designs easily.
.:Sent by pneumatic tubes
I will agree with you that many hybrid headphone amps are overkill.
That is why the starving student amp seemed so elegant.
The only headphone amps I have listened to are from mixers that produce ear ringing levels or a Musical Fidelity X Cans V2, but for headphones (correct me if I am wrong) I don't believe you need +200V.
That is why the starving student amp seemed so elegant.
The only headphone amps I have listened to are from mixers that produce ear ringing levels or a Musical Fidelity X Cans V2, but for headphones (correct me if I am wrong) I don't believe you need +200V.
The output is only a few volts rms, but the plate voltage is to get the tube to a desired operating point.
.:Sent by pneumatic tubes
.:Sent by pneumatic tubes
And most tubes want more than 12 volts to get into their linear region. I'd just go solid state, and forget about trying to get a hybrid to sound good.
jeff
This is where I am stuck tbh. Go solid or hybrid. I have had good results from hybrid before, and from solid.
My situation has changed and headphones will dominate for some time, a hybrid design would be more desirable but a SS design would be just as good. As mentioned I was left lacking sound from dual LM380.
So am unsure what direction to go.
A elegant design like starving student in valve or SS would be good.
Want to keep it simple.
Cavalli's EHHA design uses a couple ECC86 and a +/-30v supply. With a beefy enough power supply and heatsinking on the output devices, it'll drive small speakers.
EHHA Rev A – Cavalli Audio
EHHA Rev A – Cavalli Audio
People are quick to dismiss using normal tubes at low voltages, but they can be surprisingly linear... for example, look at these 6SN7 curves from Steph Bench:
You will get similar linearity from an ECC88, which is remarkably similar to the ECC86, which is intended for low voltage use. I've tested a lot of tubes between 12 and 48V on the plate, and most triodes are usable there, but it is true that most will generate more distortion than they would at a higher plate voltage.
That said - designs like the Millett hybrid are not particularly low distortion designs. They are characteristically "tubey", which to me means that they have a significant even-order harmonic content. Whether this is good or bad depends on the listener.
Like many tube amps, you really need to consider this as a signal processor, not just a "wire with gain". Certainly a simple opamp based headphone amp will measure better, and to many listeners, sound better. It also depends on the headphones driven - some sound better with a little even order distortion, at least to me, and some do not. Similarly, some sound better with a high source impedance (some as high as 100 ohms), and some need very low source impedance to achieve decent damping.
There are too many variables to give you a clear answer, I'm afraid. You will need to try some things for yourself.
Pete
You will get similar linearity from an ECC88, which is remarkably similar to the ECC86, which is intended for low voltage use. I've tested a lot of tubes between 12 and 48V on the plate, and most triodes are usable there, but it is true that most will generate more distortion than they would at a higher plate voltage.
That said - designs like the Millett hybrid are not particularly low distortion designs. They are characteristically "tubey", which to me means that they have a significant even-order harmonic content. Whether this is good or bad depends on the listener.
Like many tube amps, you really need to consider this as a signal processor, not just a "wire with gain". Certainly a simple opamp based headphone amp will measure better, and to many listeners, sound better. It also depends on the headphones driven - some sound better with a little even order distortion, at least to me, and some do not. Similarly, some sound better with a high source impedance (some as high as 100 ohms), and some need very low source impedance to achieve decent damping.
There are too many variables to give you a clear answer, I'm afraid. You will need to try some things for yourself.
Pete
Yeah - I liked your Rev. MH hybrid design with the 12fm6 and buf634, liked it less with 12ae6a. I forget if i ever acquired the 12fk6. If i remember right, isn't the voltage amplification with the 12fk6 slightly negative?
I then liked the old hybrid more with steinchen's diamond buffer daughtercard. Built one with OnSemi mje parts and one with Toshiba parts. Liked the Toshiba buffer so much i soldered it in.
I may have been the first person to build the mosfet max configuration. It, also, is pretty good.
But like you say, they're about introducing even-order harmonic distortion, which is a completely legitimate psychoacoustic goal. Particularly the way most music is produced, recorded, and distributed these days.
I've never listened to the lm380. I've heard an EET mumbling something about the lm386 being perfectly good for headphones, if you give it a clean enough power supply, because the CMRR isn't very good.
Lately most of my music comes out of a Stax SRM-001mk2, which is based on a Toshiba TL062 variant. Which is hardly a high end part.
This is the tube audio forum, so I figure we should stick to tube related designs. But there are plenty of simple solid state designs that could offer plenty of fidelity and can drive headphones to earsplitting volume if you care to build them for that.
I then liked the old hybrid more with steinchen's diamond buffer daughtercard. Built one with OnSemi mje parts and one with Toshiba parts. Liked the Toshiba buffer so much i soldered it in.
I may have been the first person to build the mosfet max configuration. It, also, is pretty good.
But like you say, they're about introducing even-order harmonic distortion, which is a completely legitimate psychoacoustic goal. Particularly the way most music is produced, recorded, and distributed these days.
I've never listened to the lm380. I've heard an EET mumbling something about the lm386 being perfectly good for headphones, if you give it a clean enough power supply, because the CMRR isn't very good.
Lately most of my music comes out of a Stax SRM-001mk2, which is based on a Toshiba TL062 variant. Which is hardly a high end part.
This is the tube audio forum, so I figure we should stick to tube related designs. But there are plenty of simple solid state designs that could offer plenty of fidelity and can drive headphones to earsplitting volume if you care to build them for that.
Last edited:
Thank you for the replies. If I could get hold of the tubes I would go for the Millett design.
But we have a lot of ECC82s so going to do the 12v Class A design and wire it point to point and see if I like the sound first. Done things backwards here as the PCB is already designed lol.
Just a quick question regarding LM317T can LM388T be used in place? No LM317T in stock at work.
But we have a lot of ECC82s so going to do the 12v Class A design and wire it point to point and see if I like the sound first. Done things backwards here as the PCB is already designed lol.
Just a quick question regarding LM317T can LM388T be used in place? No LM317T in stock at work.
Stupid question after just reading the design notes a 10dB gain is to be had, can this be compared to some thing in the real world?
Answered my own question it has been a long day, 10dB is the gain I get now using the LM380.
Anyway of hiring the gain of the ECC82 from the design, like a 13dB gain.
Quick question about the ECC 82, built a little prototype at work today and everything worked, and I can't find a 12v Supply at home anywhere so I can test it at line level. Could I put 15v across the heaters?
Thanks for the quick replies, no resistors at home or would have dropped the heater voltage that way and no 6v PSU.
Will have to wait untill tomorrow.
Will have to wait untill tomorrow.
Well tried it at home, after sorting out the "hum" with the regulated PSU I quickly put together.
For a portable amp I think it works, but for driving line level I wasn't impressed, don't think there is enough voltage swing to drive the headphones good enough. The LM380 works better even though I am only getting around 1/4 watt.
Back to looking for another design or purchase an X-Cans V2. But it has been fun doing the research and debugging etc.
For a portable amp I think it works, but for driving line level I wasn't impressed, don't think there is enough voltage swing to drive the headphones good enough. The LM380 works better even though I am only getting around 1/4 watt.
Back to looking for another design or purchase an X-Cans V2. But it has been fun doing the research and debugging etc.
- Status
- Not open for further replies.