I agree, try Lar's fix - it will work quite a bit better than the current implementation.I'll describe the result.
By adding NFB, sound reproduction becoming a little bit "dull" and very sensitive to NFB's resistors materials.
By using Lars solution, the result totally different from original.
Bass becoming poor, boomy, loose, more like using very-very bad sub-woofer.
Treble also poor, very dull, seems already roll-off in 15kHz.
Midrange frequency very dominant, more like midrange booster.
All musical instrument loss it's timbre, for ex: differentiation between pianos brand and type becoming blur or sound same.
Sound stage becoming narrow.
All sound reproduction like "record sound" not real.
It seem odd-order harmonics becoming greater.
Last, it's degrading all my system performance.
Any other ideas?
By adding NFB, sound reproduction becoming a little bit "dull" and very sensitive to NFB's resistors materials.
By using Lars solution, the result totally different from original.
Bass becoming poor, boomy, loose, more like using very-very bad sub-woofer.
Treble also poor, very dull, seems already roll-off in 15kHz.
Midrange frequency very dominant, more like midrange booster.
All musical instrument loss it's timbre, for ex: differentiation between pianos brand and type becoming blur or sound same.
Sound stage becoming narrow.
All sound reproduction like "record sound" not real.
It seem odd-order harmonics becoming greater.
Last, it's degrading all my system performance.
Any other ideas?
That should work very well (though gain still too high for most modern applications). It's important to bias up the cathode follower's heater ~75V so as not to challenge the heater-to-cathode voltage rating- I'm sure you know that, but Edwind may not, and we'd hate to hear about sparks and flames.
I already lift up the heaters to nearly 80V (from GE 12AU7 datasheet, 100V is maximum rating for DC component and 200V for DC component + peak).
Ah I forget to wrote before, that all components that I used to build my 1st post and Lar's are same quality (all low grade components), and all tubes that I used all are NOS tube, matched pair Mullard (±1960) EF86 and Siemens ECC81 (± 1960). Using ECC82 became more terrible, very loud midrange, bad bass quality, treble more dull than using ECC81.
I'll draw the Lar's schematic including it's voltage values and post soon.
Please wait!
Ah I forget to wrote before, that all components that I used to build my 1st post and Lar's are same quality (all low grade components), and all tubes that I used all are NOS tube, matched pair Mullard (±1960) EF86 and Siemens ECC81 (± 1960). Using ECC82 became more terrible, very loud midrange, bad bass quality, treble more dull than using ECC81.
I'll draw the Lar's schematic including it's voltage values and post soon.
Please wait!
Edwind,
If you built like it like my topology you have done something wrong anyway. Schematic with working voltages please. Should not be any big differences with ECC81.
This is a preamp with the same topology we produced some decade ago. Extremely well recevied. reVintage.se
If you built like it like my topology you have done something wrong anyway. Schematic with working voltages please. Should not be any big differences with ECC81.
This is a preamp with the same topology we produced some decade ago. Extremely well recevied. reVintage.se
Attachments
Last edited:
You can use it like a line preamplifier also, i have build it and im very satisfied how it sound.
It's not totally the same, I'll draw complete one including it's voltages.
Please wait for it, I must redraw it from paper.
Looks like it should. The only things that are off limits is B+ that should be ca 50V higher and Ua of EF86 that might be a little higher, with 250 B+ maybe 120V. Good enough, will not harm performance much.
No matter what, it will still have a THD of well below 0,05% at 2V out, mainly 2nd order. Also a bandwith that ends at a few hundred kHz. So how can you hear harmonics and a muffled highend in that? Probably something else that is broken!
No matter what, it will still have a THD of well below 0,05% at 2V out, mainly 2nd order. Also a bandwith that ends at a few hundred kHz. So how can you hear harmonics and a muffled highend in that? Probably something else that is broken!
I wasn't being awkward, just aware that some of the people on here might be confused. We do get people asking how to wire a volume pot. I know you know what you are doing. My comment was meant as a minor correction, not a criticism.revintage said:
PS is it possible that Edwind is missing the noise and distortion of his original circuit, and misinterpreting this? The new circuit looks OK.
Last edited:
I don't know about THD and it's order values for both circuit since I don't have that equipment.
Maybe DF96 true if I lost some noises and harmonics differences between two circuits.
As I know from my experience, an in-phase devices sound better than an inverted-phase devices.
But in this case just the same in spite of I inverted a speaker's cable to correct the phase.
One question, is there any good simulation softwares?
I had bad experiences with it, the differences between simulation and real are too great.
Maybe DF96 true if I lost some noises and harmonics differences between two circuits.
As I know from my experience, an in-phase devices sound better than an inverted-phase devices.
But in this case just the same in spite of I inverted a speaker's cable to correct the phase.
One question, is there any good simulation softwares?
I had bad experiences with it, the differences between simulation and real are too great.
Most simulation software is good. The problems come from the device models:
1. They implement a reasonable approximation to a typical device.
2. Unless you intervene, all devices of a given type in your circuit will be identical.
As a result, simulations can give good results for gain and frequency response but only a vague guess for distortion. The only way round this is to accurately measure the characteristics of the actual devices you intend to use, then construct a separate mathematical model for each of them, and remember to use only these valves in the final build. Simpler to build the circuit and test it.
The audibility of absolute phase is something which people argue about. They also argue about whether music sources preserve absolute phase. My understanding is that some people can detect absolute phase on some music or test signals, but I have been told that policy among UK broadcasters was to make no attempt to preserve absolute phase (on the assumption that studios etc. did the same).
1. They implement a reasonable approximation to a typical device.
2. Unless you intervene, all devices of a given type in your circuit will be identical.
As a result, simulations can give good results for gain and frequency response but only a vague guess for distortion. The only way round this is to accurately measure the characteristics of the actual devices you intend to use, then construct a separate mathematical model for each of them, and remember to use only these valves in the final build. Simpler to build the circuit and test it.
The audibility of absolute phase is something which people argue about. They also argue about whether music sources preserve absolute phase. My understanding is that some people can detect absolute phase on some music or test signals, but I have been told that policy among UK broadcasters was to make no attempt to preserve absolute phase (on the assumption that studios etc. did the same).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.