4in1 tube mic preamp SE/PP TRI/PEN

Help Support GroupDIY:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
No. A triode is not a constant-current sink. When the upper triode receives a positive on its grid, current increases, as well as Gm, and Rp decreases. At teh same time, the lower triode sees its plate voltage increase, as do current and Gm, for a decrease in Rp; this tends to linearize operation. Not as much as in a true push-pull, though.
OK thanks for the technical clarification.

 
Last edited:

emrr

Well-known member
Joined
Apr 12, 2006
Messages
7,875
Location
NC, USA
Have you made measurements/analysis of distortion in the different positions?
I would think the fact that the second triode is still connected in "SE" mode actually tends to produce less waveform assymetry than a true SE.
Indeed this output stage is in the old books, meant to allow this sort of output connection with a transformer, I believe the last stage cancels it’s own second harmonic generation regardless of AC signal.

in many cases you lose obvious bass response in the transformer if DC is imbalanced, this effect swamps any perceived harmonic difference.
 

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
Last edited:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
I would like to incorporate a gain switch in this design... The gain switch works for the microphone input but not the DI. Is there away around this? I mean the DI can´t have elevated ground like the feedback network of IPT secondary....
 

Attachments

  • 217454C8-2E14-4041-BF6C-55726926AEDB.jpeg
    217454C8-2E14-4041-BF6C-55726926AEDB.jpeg
    230.2 KB · Views: 112
Last edited:

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,018
Location
Marcelland
There's a big mistake in this document.
"What will be the influence of cathode feedback in single ended amplifiers? This truly will lower the plate resistance of B2. "
This guy does not understand (or forgot) that the effects of NFB are dependant on the way it's applied. There are 4 cases: voltage-to-voltage, voltage-to-current, curren-to-current and current-to-voltage.
In the case of an un-bypassed cathode, current is sensed in the cathode resistor and applied in series with the signal. Since it tends to regulate the cathode current - which is also the output current, it actually increases the output impedance. That's the particularity of current-to-voltage NFB, It increases the input Z, and also the output Z.
Some textbooks call the different implementations of NFB parallels, series, and combinations of both; I think it's ambiguous.
 

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
I'm afraid no...
You could apply voltage-to-current NFB by creating a path from the output to the input, but that would decrease the input Z, which you don't want for a DI.
Yep I tough so too. The DI still work ok without FB. I wanted to apply FB on the mic input because its a little hot with condenser mics... (y) Low Z-in is not desirble...

Yeas the cathode resistor in the push-pull gives NFB which tend to decreas distortion but increases output impedance... And its works fine in a coloured pre, but for objective performance you bypass the cathode and maybe have some feedback between anode and grid... then you have low distortion and low output impendance...
 
Last edited:

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,018
Location
Marcelland
Yep I tough so too. The DI still work ok without FB. I wanted to apply FB on the mic input because its a little hot with condenser mics... (y) Low Z-in is not desirble...
Mmmh... Depends. Modern xfmr-less condenser mics are very tolerant to load impedance, but those with a xfmr output react with the load, generally in an unfavourable way.
All passive dynamic and ribbon mics benefit from a large load impedance.
 

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
Mmmh... Depends. Modern xfmr-less condenser mics are very tolerant to load impedance, but those with a xfmr output react with the load, generally in an unfavourable way.
All passive dynamic and ribbon mics benefit from a large load impedance.
Agreed... My schematic is probably the best compromise.... (y)
 

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
Gonna try a Pultec style OPT from carnhill VTB2425 15k:600r:24r.

The tertiary winding makes it possible to apply some feedback to cathode of PP stage. Im also making a new Connector board so its upright. When laying flat it pick up noise from the PT.
My plan is to switch between cathode resistor feedback and bypassed resistor with tertiary FB

The serial resistor without cap has great midrange and break up I suspect the tertiary feedback to give a cleaner tone and better bass response
 

Attachments

  • 2C23F118-1277-4412-ABB1-8FD2DCCE3293.jpeg
    2C23F118-1277-4412-ABB1-8FD2DCCE3293.jpeg
    79.1 KB · Views: 59
  • FE8814C8-6107-41D9-94CC-588151301F4B.jpeg
    FE8814C8-6107-41D9-94CC-588151301F4B.jpeg
    158 KB · Views: 50
Last edited:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
Its hard to find info on this transformer but in AML:s building guide for Pultec you can see the trafo hook up
 

Attachments

  • 2FA9BE7C-DA71-4DA4-AE65-7BD6DAA05D2B.jpeg
    2FA9BE7C-DA71-4DA4-AE65-7BD6DAA05D2B.jpeg
    169.4 KB · Views: 55

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
The vtb2425 with tertiary feedback i cleaner and gain is higher. But its also erradicates the sonic characters of SE and PP. SE no longer has that softer mellower flavour. The other problem with the 2425 its picking noise from the PT that the 2291 does not… But of you are in the game of building a clean PP mic pre you should give 2425 a try…
 

Attachments

  • 3CE6602A-CA1D-4318-A7F6-2437AB91072C.jpeg
    3CE6602A-CA1D-4318-A7F6-2437AB91072C.jpeg
    143.8 KB · Views: 52
Last edited:

Tubetec

Well-known member
Joined
Nov 18, 2015
Messages
3,119
If your feedback is increasing the gain its not negative after all , Remember the signal is phase inverted at the anode relative to the grid and cathode , try swapping the +/- on the feedback windings .


Heres an interesting output stage employing feedback , they cross couple the cathodes and grids and the tertiary feedback gets applied at the input stage cathode . Uses an equivalent of the ECC81 medium mu triode.

http://www.bbceng.info/ti/eqpt/LIM_5.pdf
 

Attachments

  • Limiter output amplifier.PNG
    Limiter output amplifier.PNG
    29.7 KB · Views: 54

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
I now👍Both modes got feedback, in one I have a unbypassed resistor 560r that gives less gain than bypassed resistor + tertiary feedback. 24 r is not a huge amount of feedback. So both give less gain than just bypassing the resistor with cap👍
 
Last edited:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
If your feedback is increasing the gain its not negative after all , Remember the signal is phase inverted at the anode relative to the grid and cathode , try swapping the +/- on the feedback windings .


Heres an interesting output stage employing feedback , they cross couple the cathodes and grids and the tertiary feedback gets applied at the input stage cathode . Uses an equivalent of the ECC81 medium mu triode.

http://www.bbceng.info/ti/eqpt/LIM_5.pdf
Its some PP stage with a twist very intresting topology havens sen that before 👍
 
Last edited:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
Going to try a partially bypassed cathode resistor. Its a compromise between feedback, gain and bass performance. I like the sound and the lower gain of unbypassed resistor, but it suffers a bit in the bass compartment compared to bypassed cap + tertiary feedback

If I done the math right 6n6p Rp= 2K Rk= 100r (unbypassed) mu=22 100*22=2200 + 2k= 4200. 2291 primary inductance is 100H but from C.T the tube sees 25H. 4200r at 25H puts the -3 dB at ca 30 Hz. This equation is for a singel triode stage does it change when you calculate a PP stage?
 

Attachments

  • 1F807A46-29B6-4A8E-87C1-0E0671D30792.jpeg
    1F807A46-29B6-4A8E-87C1-0E0671D30792.jpeg
    161.7 KB · Views: 46
Last edited:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
The partially bypassed cathode resistor worked out well, it´s a keeper... 560r with bypass cap + 100r. And I´ll stick with the VTB-2291 trafo in 4+4:1. After the summer maybe I´ll try the LL1689 a more dedicated PP trafo....

Lessons learned so far... The feedback around the input stage works great.... it takes down gain a notch, more detail och spacial info etc... but implementing feedback around the PP stage, aside from some smal amount of serial feedback, eradicates the sonic signatures of SE and PP and does not work well in this design... The PP stage breaks up very nice with minimum amount feedback. But if you are designing a clean PP stage I would definitly explore the tertiary feedback winding route....

Lessons still to be learned. When I get REW to work I will take some measurments and see what it has to say....
 
Last edited:

solkatten

Well-known member
Joined
Apr 3, 2021
Messages
200
Location
sweden
Here it comes THD measurements in REW... -15 db output PAD was engaged so I measured at around OdBu

Black: THD
Red: H2
Orange: H3
 

Attachments

  • PP PEN.jpg
    PP PEN.jpg
    111 KB · Views: 24
  • PP TRI.jpg
    PP TRI.jpg
    112.9 KB · Views: 19
  • SE PEN.jpg
    SE PEN.jpg
    113.2 KB · Views: 23
  • SE TRI.jpg
    SE TRI.jpg
    112.9 KB · Views: 35
Last edited:
Top