I read the Lynn Olsen website and he seems to present an argument in favor of DHTs and tubes in general based less IMD, at the time the articles were writted I believe it was difficult to measure IMD. But when I run RMAA or Audiotester I get pretty high IMD #'s on tube amps I've put together, these are line level/headphone types. Typically 0.1 to .5 %IMD on the RMAA. Is this what others typically see? And where does it leave Olsen's decade old postulations ?
Last edited:
There is nothing magic about DHTs. The issue is that they are old large designs, and so more likely to maintain cylindrical symmetry and have low mu. This can lead to lower distortion in the right circuit, because the grid and anode are the same shape. An indirectly heated valve from the same era would probably have a similar effect.
What is "low IMD"? To an SET fan, 5% is low. To a sand man 0.1% is high.
What is "low IMD"? To an SET fan, 5% is low. To a sand man 0.1% is high.
There is nothing magic about DHTs. The issue is that they are old large designs, and so more likely to maintain cylindrical symmetry and have low mu. This can lead to lower distortion in the right circuit, because the grid and anode are the same shape. An indirectly heated valve from the same era would probably have a similar effect.>>
That's not what I hear with DHTs. I'm talking specifically about smaller input tubes like 26, 10Y and such. All my listening experience over the last three years tells me that the sound is qualitatively different - directly heated filaments seem to be different. I don't think it's anything to do with the era of manufacture - new production DHTs can be excellent, and small DHTs like 1G4GT and 3a5 have the characteristic DHT sound. I don't think it's due to mu either - indirectly heated tubes with the same low mu (12b4, 6BX7 etc) don't sound anything like a 26 or a 10Y.
Andy
That's not what I hear with DHTs. I'm talking specifically about smaller input tubes like 26, 10Y and such. All my listening experience over the last three years tells me that the sound is qualitatively different - directly heated filaments seem to be different. I don't think it's anything to do with the era of manufacture - new production DHTs can be excellent, and small DHTs like 1G4GT and 3a5 have the characteristic DHT sound. I don't think it's due to mu either - indirectly heated tubes with the same low mu (12b4, 6BX7 etc) don't sound anything like a 26 or a 10Y.
Andy
Basically this :
And this:
And different variations in between.
Which are simply single-feed variations of the Espressivo from Olson's website:
Espressivo Headphone Amplifier
Basically this a triode in the purest form you are going to hear it (good headphones required.)
Any way when I read Olson and him being one of the first to say: hey look 1920's DHT's are the most linear active devices ever made for audio. And then he goes foward describing the effects of feedback adding upper harmonics at an exponential level, and doesn't he said that IMD is the worst form of distortion when actual music is being played (due to the thousands of tones)? Maybe I jumped to conclusions I shouldn't have.
Huge fan of Olson's insights, but there seems little info for the tube hobbiest now that we are able to hook up a $200 soundcard and take better measurements than Olson dreamed back when he wrote those brilliant articles.
BTW I am not a fan of RMAA, and think there are better programs, but lets face it we all have it and I think it would be a good thing if more folks shared some measurements in the tube forum.
I was surprised to see the WE417 at 0.5% IMD%Noise swinging only 3 volts. This is with the unmeasureable the noise floore below -120db (SSHV PS). I would have thought lower, and the DHT's aren't doing better.
All I wanted to know was ballpark RMAA figures folks are getting with DHT's, and maybe this IMD # from RMAA has very little correlation to SQ?
Last edited:
These are from a CCS loaded parafeed 6C45 using an 8K:32 OPT. It is running at ~125mW (2Vrms into 32R) so the tube is swinging something like 35Vrms (adding a bit for the copper losses, etc).
That's about 0.14% THD and 0.12% IMD. Not great, but maybe not totally awful for a tube w/o feedback. As you know, 125mW is pretty loud, and if you drop the level down to where you might actually listen, everything but the 2nd drops into the noise, and the 2nd is awfully close. Soundwise, it is really a great amp.
That's about 0.14% THD and 0.12% IMD. Not great, but maybe not totally awful for a tube w/o feedback. As you know, 125mW is pretty loud, and if you drop the level down to where you might actually listen, everything but the 2nd drops into the noise, and the 2nd is awfully close. Soundwise, it is really a great amp.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
That is incredible really. My single feed WE417 looks like this :
Really want to try parafeed 50%permalloy but not having much like finding them, Electraprint seems a bit ellusive in this request. I think the single feed and the low signal sensitivity of this M19 SE core are hurting me. I am working on the 60hz bump, but its -100dB down, not audible and seems to correlate with distance of OPT to power transformer more than anything. I think the upper harmonics are the ADC.
Really want to try parafeed 50%permalloy but not having much like finding them, Electraprint seems a bit ellusive in this request. I think the single feed and the low signal sensitivity of this M19 SE core are hurting me. I am working on the 60hz bump, but its -100dB down, not audible and seems to correlate with distance of OPT to power transformer more than anything. I think the upper harmonics are the ADC.
Last edited:
PS is external. Inside there is a linear reg before the CCS.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.