^ ^ ^ ^ ^ Yeah, what he said! ^ ^ ^ ^ ^
And hey, if you want to stick to US-made tubes, 6J4P = 6AU6, which is a common little 7-pin pentode that looks good wired triode.
You could even wire it up with the plate and suppressor grid connected to ground, use the screen grid as the 'plate', use the control grid as normal. That would make a nice triode with mu of about 30 using up only about 5mA 'plate' current. Check this out:
6AU6 Triode Mode
6AU6A data sheet with triode curves:
http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6AU6A.pdf
--
And hey, if you want to stick to US-made tubes, 6J4P = 6AU6, which is a common little 7-pin pentode that looks good wired triode.
You could even wire it up with the plate and suppressor grid connected to ground, use the screen grid as the 'plate', use the control grid as normal. That would make a nice triode with mu of about 30 using up only about 5mA 'plate' current. Check this out:
6AU6 Triode Mode
6AU6A data sheet with triode curves:
http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6AU6A.pdf
--
Last edited:
Experimenting...
cc: famousmockingbird --
Last night I installed some nice enamel-coated 100R 5W wirewound resistors across the OPT secondaries (from 50 ohm tap to Common). The idea is to smooth out the impedance wobbles of my Sennheiser HD650 cans, and to bring the HD650's 310 to 500 ohm impedance down closer to the nominal 50 ohm impedance of the OPT secondary.
The strangest thing happened. The resulting sound from the HD650 is improved quite a bit. It was really damned good. Now there's enough bass. While still a tad bright sounding, the tonal balance is quite good. Apparently the 100R across the secondary does help optimize the OPT for that load, and I enjoyed the sound very much.
But when I connected my other headphones, 50 ohm Fostex T50RP, I noticed they now sounded dark with overactive midbass (in the 100Hz region maybe) and rolled off highs. They didn't sound like this without that 100R resistor.
Any idea what's happening here?
I'm thinking of mounting a DPDT switch near the output jack so I can switch in/out those 100R resistors.
Adding the 100R resistors reduced the gain a bit, but the amp's gain is still adequate. With the Fostex cans, which aren't very sensitive, turning the amp up all the way results in a sound that's still subjectively clean-sounding, but would make me well and truly deaf if I listened that loud.
--
cc: famousmockingbird --
Last night I installed some nice enamel-coated 100R 5W wirewound resistors across the OPT secondaries (from 50 ohm tap to Common). The idea is to smooth out the impedance wobbles of my Sennheiser HD650 cans, and to bring the HD650's 310 to 500 ohm impedance down closer to the nominal 50 ohm impedance of the OPT secondary.
The strangest thing happened. The resulting sound from the HD650 is improved quite a bit. It was really damned good. Now there's enough bass. While still a tad bright sounding, the tonal balance is quite good. Apparently the 100R across the secondary does help optimize the OPT for that load, and I enjoyed the sound very much.
But when I connected my other headphones, 50 ohm Fostex T50RP, I noticed they now sounded dark with overactive midbass (in the 100Hz region maybe) and rolled off highs. They didn't sound like this without that 100R resistor.
Any idea what's happening here?
I'm thinking of mounting a DPDT switch near the output jack so I can switch in/out those 100R resistors.
Adding the 100R resistors reduced the gain a bit, but the amp's gain is still adequate. With the Fostex cans, which aren't very sensitive, turning the amp up all the way results in a sound that's still subjectively clean-sounding, but would make me well and truly deaf if I listened that loud.
--
Well I would assume it's having the same effect on the Fostex headphones as it has on the HD650s, but the Fostex headphones are already at around the right impedance. If we do a simple parallel resistance calculation, you're at around 33.33 ohms.
The ideal solution would be to have an OPT with multiple taps on the secondary, so maybe 50, 300 and 600 ohm taps. Something along those lines.
The ideal solution would be to have an OPT with multiple taps on the secondary, so maybe 50, 300 and 600 ohm taps. Something along those lines.
Of course. I could just spend money and have transformers custom wound, even. But what fun is that? I'm hoping this will be a learning experience...
Right. And it would be a lot more expensive, and a bigger hassle. I'm currently looking at having a pair of 50W transformers custom wound... I talked to Electra-Print, but I'm not sure I want to blow $400 in transformers. I talked to Edcor, but I don't want to wait two months. I contacted MagneQuest and Heyboer, and have yet to get a response. Though to Heyboer's credit, I left a message on a national holiday... they might not be back from vacation yet. Though I did email them about two weeks ago and still haven't gotten a response. I sure hope they aren't closing shop.
And Transcendar isn't taking orders
And I'm told that James isn't making transformers anymore, plus that would be shipping from Taiwan.
So if you can use something off-the-shelf, that's always preferable. I'm seriously considering just buying something from Hammond and being done with it.
So to make a very long story short, just stick with your $15 Edcors if you like the sound
And Transcendar isn't taking orders
And I'm told that James isn't making transformers anymore, plus that would be shipping from Taiwan.
So if you can use something off-the-shelf, that's always preferable. I'm seriously considering just buying something from Hammond and being done with it.
So to make a very long story short, just stick with your $15 Edcors if you like the sound
https://www.innerfidelity.com/images/FostexT50RPMk3.pdf
I am guessing they are just dark headphones. Look at the db vs frequency, it's highest is 100Hz and the midband is -10db down. Frequency looks flat out to 10k so I don't know if there is some sort of compensation built in?
The 12HL7 has an impedance of 1.3k, so the reflected back impedance of a 33 ohm load would be ~5k which is adequate for loading a triode. Maybe the distortion has changed or something where the sound at low frequencies is now more distorted? The 100 ohm resistor value is a tradeoff, ideally with the 50 ohm headphones you don't want it there so maybe make it switched? That way there with the 50 ohm headphones you can take the resistor out of circuit but with the 300 ohm headphones switch it back in to smooth things out.
That all makes a lot of sense. The T50RP are 'dark' sounding headphones, but usually in a 'good' way (as in 'mellow sounding' as opposed to 'muffled'). Did you notice that its impedance curve is ruler-flat at about 52 ohms?
I think I will put in that switch. I have some suitable mini-toggles on hand. It could be that the 100R across the secondary is making these 50 ohm headphones unhappy, but might do nice things for some 32 ohm headphones that sound too harsh/zippy. A tone control of sorts. Definitely worth trying.
Thanks for your input on this.
I think I will put in that switch. I have some suitable mini-toggles on hand. It could be that the 100R across the secondary is making these 50 ohm headphones unhappy, but might do nice things for some 32 ohm headphones that sound too harsh/zippy. A tone control of sorts. Definitely worth trying.
Thanks for your input on this.
Oh, thanks. I don't think this amp can make 1W out into 50 ohms. I haven't measured it yet, but I calculate that with -2V grid bias, triode mu of about 30, and OPT stepdown ratio of about 12.5:1, that the most voltage swing I could possibly get at the output would be in the vicinity of 3V rms (2V pk input * 30x gain = 60V pk into the OPT primary, then 12.5:1 stepdown ratio = 60/12.5 = 4.8V pk = 3.4V rms). That equates to 230mW into 50 ohms, or 38mW into 300 ohms.
I agree that is plenty. On this amp, turning the volume all the way up when fed by a DAC with fixed level output results in a very loud but still subjectively clean sound from the headphones. That's louder than I'd ever want to listen, so I figure the gain and available headroom are at a spot where it's good enough for my purposes.
It's a hideous waste of power to make that 0.5W of stereo audio output from a circuit that dissipates 7W of heater power and 16W of plate current, but hey... SE Triode Amp!
I agree that is plenty. On this amp, turning the volume all the way up when fed by a DAC with fixed level output results in a very loud but still subjectively clean sound from the headphones. That's louder than I'd ever want to listen, so I figure the gain and available headroom are at a spot where it's good enough for my purposes.
It's a hideous waste of power to make that 0.5W of stereo audio output from a circuit that dissipates 7W of heater power and 16W of plate current, but hey... SE Triode Amp!
I wired in that switch with 100 ohm resistors. In doing so, I forgot which way is with the 100 ohms in, and which way is out of circuit. So I have to rely on my ears to tell me which is which. Well...
I think I can just barely tell the difference with the amp driving the 60-ohm Fostex cans (the dark sounding ones). The bass sounds a bit looser and more boomy one way. But it's very subtle.
Switching to the 300-ohm Sennheiser cans, the difference is a bit more apparent, but more subtle than I thought it would be. I can definitely hear one switch position is 'softer' sounding than the other. Just the slightest bit different.
I think that means your idea of putting those 100 ohm resistors across the secondary certainly does no harm, and may be doing some good (smoothing out impedance variations in the load).
Whaddya know... A learning experience!
Oh, and by the way: The amp sounds very good. I'm very pleased with it.
The only shortfall is that it doesn't have enough gain to get my not-very-sensitive Fostex cans to go really loud from a smartphone headphone jack or the 1V rms output fix-level outputs on the PC soundcard I use. From the soundcard, I can crank the volume control on the amp all the way up but all I get is merely 'loud' instead of deafening. But that's not the case with 32 ohm cans, which do get deafening from this amp. Interestingly, the 300-ohm Sennheiser goes louder and sounds brighter too, whichever way I have the 100 ohm switch set.
The amp does drive either set of headphones louder than I can stand when driven by my Raspberry Pi/Allo Boss/MoOde music player. I think it's just a matter of my phone and this soundcard having less than the 'standard' 2V rms max line level audio output.
I'm hoping that when I can add the voltage regulators a la Elvee that it will sound a bit tighter in the bass. It can't get any quieter, though. There is absolutely no noise from hum. If I crank the volume all the way up I can hear some tube noise; microphonic tinkles and the like.
Well now I'm curious what huge and efficient horn speaker you have. Did you build it, woodworking and all? Did you design it, including crossovers?
I dabbled with some vintage Klipsch speakers and found they weren't to my taste (harsh upper mids). However, I came across a post claiming that if you want to get a taste of what a competently designed, up-to-date oblate-spheroid, controlled-directivity speaker sounds like, check out the JBL LSR305. So I bought a pair (they're not expensive). Driven by my Impasse-oid balanced-out preamp, I think they're just what I was looking for. Now if I could only get a 95dB/1W efficient 10" 2-way to sound like that driven by 5W of class A DHT amp, I'd be happy as a clam.
But now I'm going waaaay off-topic...
--
I think I can just barely tell the difference with the amp driving the 60-ohm Fostex cans (the dark sounding ones). The bass sounds a bit looser and more boomy one way. But it's very subtle.
Switching to the 300-ohm Sennheiser cans, the difference is a bit more apparent, but more subtle than I thought it would be. I can definitely hear one switch position is 'softer' sounding than the other. Just the slightest bit different.
I think that means your idea of putting those 100 ohm resistors across the secondary certainly does no harm, and may be doing some good (smoothing out impedance variations in the load).
Whaddya know... A learning experience!
Oh, and by the way: The amp sounds very good. I'm very pleased with it.
The only shortfall is that it doesn't have enough gain to get my not-very-sensitive Fostex cans to go really loud from a smartphone headphone jack or the 1V rms output fix-level outputs on the PC soundcard I use. From the soundcard, I can crank the volume control on the amp all the way up but all I get is merely 'loud' instead of deafening. But that's not the case with 32 ohm cans, which do get deafening from this amp. Interestingly, the 300-ohm Sennheiser goes louder and sounds brighter too, whichever way I have the 100 ohm switch set.
The amp does drive either set of headphones louder than I can stand when driven by my Raspberry Pi/Allo Boss/MoOde music player. I think it's just a matter of my phone and this soundcard having less than the 'standard' 2V rms max line level audio output.
I'm hoping that when I can add the voltage regulators a la Elvee that it will sound a bit tighter in the bass. It can't get any quieter, though. There is absolutely no noise from hum. If I crank the volume all the way up I can hear some tube noise; microphonic tinkles and the like.
One reason why I love my huge and efficient horn speaker system, I can get away with a low powered single ended amp with only a handful of bottles. I think my amp is ~30watts power consumption.
Well now I'm curious what huge and efficient horn speaker you have. Did you build it, woodworking and all? Did you design it, including crossovers?
I dabbled with some vintage Klipsch speakers and found they weren't to my taste (harsh upper mids). However, I came across a post claiming that if you want to get a taste of what a competently designed, up-to-date oblate-spheroid, controlled-directivity speaker sounds like, check out the JBL LSR305. So I bought a pair (they're not expensive). Driven by my Impasse-oid balanced-out preamp, I think they're just what I was looking for. Now if I could only get a 95dB/1W efficient 10" 2-way to sound like that driven by 5W of class A DHT amp, I'd be happy as a clam.
But now I'm going waaaay off-topic...
--
Attachments
Last edited:
12HG7 or 12GN7 are pin-compatible with 50% more transconductance - worth a try if you come across some. I didn't find a detailed data sheet with a quick web search, though.
BTW, 1.3K plate resistance will transform to about 10 Ohms output impedance, so no surprise that 100 Ohm load has limited effect. A good article on the interaction of headphone and source impedance here: NwAvGuy: Headphone & Amp Impedance
BTW, 1.3K plate resistance will transform to about 10 Ohms output impedance, so no surprise that 100 Ohm load has limited effect. A good article on the interaction of headphone and source impedance here: NwAvGuy: Headphone & Amp Impedance
Last edited:
I was looking at the 12GN7 and 12HG7.
12HL7 - https://frank.pocnet.net/sheets/135/1/12HL7.pdf -- (gm = 21mA/V, pentode rp = 55k)
12GN7 - https://frank.pocnet.net/sheets/135/1/12GN7A.pdf -- (gm = 36mA/V, pentode rp = 50k)
12HG7 - https://frank.pocnet.net/sheets/202/1/12HG7.pdf -- (gm = 32mA/V, pentode rp = 60k)
I was shying away from 12GN7 as I thought that the higher the gm, the higher the possibility of problems with microphonic noises. Of the three, 12GN7 has the lowest rp, but 12HL7 is really close. 12HG7 has slightly higher rp than either of the others.
I should try a pair of 12GN7A to see if the slightly higher triode mu makes an audible difference.
--
Re: NwAvGuy's discussion of Zout and headphones:
Good point there! I read through the article. Good food for thought.
If the OPT's turns ratio is 12.6:1 (8k:50 ohms, or roughly 160:1 impedance ratio), then wouldn't it apply that same 160X transformation to the 12HL7's plate resistance? If so, then 1300 ohms/160 = Zout of 8.125 ohms. Maybe more like 10 ohms after transformer losses, lower than expected tube gm, etc.
So now I can understand why the Fostex cans with their roughly 55 ohm impedance sound slightly more woolly and loose when run from the 12HL7 amp than when run from my Objective2 amp with its Zout of roughly 0.5 ohm.
On the other hand, it's pretty easy to understand how the HD650 cans would sound fine through either amp. Their roughly 350 ohm impedance is a full 30x the 12HL7 amp's Zout. That might be beyond the law of diminishing returns as far as damping factor goes.
Interesting to think about.
12HL7 - https://frank.pocnet.net/sheets/135/1/12HL7.pdf -- (gm = 21mA/V, pentode rp = 55k)
12GN7 - https://frank.pocnet.net/sheets/135/1/12GN7A.pdf -- (gm = 36mA/V, pentode rp = 50k)
12HG7 - https://frank.pocnet.net/sheets/202/1/12HG7.pdf -- (gm = 32mA/V, pentode rp = 60k)
I was shying away from 12GN7 as I thought that the higher the gm, the higher the possibility of problems with microphonic noises. Of the three, 12GN7 has the lowest rp, but 12HL7 is really close. 12HG7 has slightly higher rp than either of the others.
I should try a pair of 12GN7A to see if the slightly higher triode mu makes an audible difference.
--
Re: NwAvGuy's discussion of Zout and headphones:
Good point there! I read through the article. Good food for thought.
If the OPT's turns ratio is 12.6:1 (8k:50 ohms, or roughly 160:1 impedance ratio), then wouldn't it apply that same 160X transformation to the 12HL7's plate resistance? If so, then 1300 ohms/160 = Zout of 8.125 ohms. Maybe more like 10 ohms after transformer losses, lower than expected tube gm, etc.
So now I can understand why the Fostex cans with their roughly 55 ohm impedance sound slightly more woolly and loose when run from the 12HL7 amp than when run from my Objective2 amp with its Zout of roughly 0.5 ohm.
On the other hand, it's pretty easy to understand how the HD650 cans would sound fine through either amp. Their roughly 350 ohm impedance is a full 30x the 12HL7 amp's Zout. That might be beyond the law of diminishing returns as far as damping factor goes.
Interesting to think about.
Nothing special or fancy, just some Klipsch La Scala's. I say huge because they are a lot bigger than the majority of speakers that I encounter in day to day life at friends houses etc.
105db/1watt
Stock crossovers including original capacitors. I have found that anything up to 1 ohm output impedance does great. The most 3 dimensional sound I get is with single ended amps, specifically triodes. I have a separate subwoofer for the low end thump but the La Scala's does everything above say 50Hz well.
I have some 12HL7's but no iron to try your setup. I have been looking for a cheap tube headphone amp to build but I have no money to finish my latest amp let alone a headphone amp. I have some 70v line transformers but they have hardly any inductance and no gap. So they would need to be AC coupled by a very low output impedance.
Now I'm way off topic
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.