Help with Preamp Schematic

[jsefer]

Hello First of, I want to point out, I am no expert, nor do I claim to be. I am just a seasonal DIYer and a recording engineer w moderate electronics knowledge enough to follow schematics, know components and mod by example.

I have recently acquired a pair of AKAI-M8 mono blocks and Ive decided to make HIFI preamps out of them. I have looked at many schematics, and looked at alot of other designs, including the NY Daves  MILA, RED47 and a few others on the net. With some help also from some great people that replied to my requests for advice I have come up with the design attached.

So Please excuse my ignorance, I just want to build a great preamp that I can use in for years to come. Im choosing all high quality parts, rebuilding the power section, and replacing all caps resistors w quality ones.

I am also sourcing a few VU meters and drivers from jlmaudio.com  which I will use a simple DP Switch to switch between channel outputs.

Im using a Telefunken Ef86 and a Tele ecc81. My goal is to build two of these sections w different characteristics, giving me 4 channels of quality preamps that I can use as a bus analogue amp or as a mic preamp for my studio.

My question is if some of the members on this board can look over my schematic and shoot me a few pointers or point me in the direction so that I may find what I need.

DETAILED IN RED ARE THE CHANGES THAT I AM LOOKING TO DO.

All help is appreciated.



Mario.

 

[volki]

Regarding the Hi Z jack inputs: If you actually want to use those for high impedance sources , IMO you should think about getting jack sockets that interrupt the connections to the Xformers when a jack is plugged in (meaning, you insert the jack btwn xformer and tube).
Otherwise your high Z sources will interact with the xformers' inductive (and stray capacitive) impedances. Many DI inputs are designed that way.

Also, you should use coupling cap's, otherwise grid leak current might get into the outputs of instruments you plug in.
In I/P 2, this would also prevent the 10K resistor (probably to tame the xformer's resonance?) to load down the DC resistance of the grid input (which obviously wants the 500K resistor).

Your O/P's (especially O/P1 coming from the plate of a pentode) will have a Z of at least a few KOhms - which might cause trouble when interfacing to the rest of your recording chain, especially if your next i/p is xformer-balanced.

In case you're planning on using both of the I/P's at the same time, you should decouple their plate supplies.

 

[DaveP]

Hi jsefer,

Mic pre's have to be carefully designed to avoid noise.

I wouldn't put a 100k pot on an EF86 grid, you need a low noise metal film resistor; there is already a 500k pot on the output.

If you are putting a 10k on the TX secondary before the 1st ECC81 then you don't need the 500k, it will only add noise.

The gain of those 2 ECC81 is going to be over 60dB plus the gain from the TX, but what you still haven't got is any driving power.  You have several options here are a few:-

1)  Use the 2nd half as a cathode follower and take the output from the cathode via your 100uF.  But some people don't like cathode followers.

2)  Parallel the ECC81's and run hot and take a plate output through a step-down TX.  The EMI REDD47 takes this approach but uses an ECC88.

3)  Just make a single mic pre out of both tubes, use the EF86 for voltage gain with the output going to paralleled ECC81 to give you some poke as described in (2)

There are plenty of other ideas out there including having variable feedback to control the volume, as in the REDD47.  Feedback will reduce distortion and noise and reduce the output impedance making it easier to drive low Z loads.

You will get lots of ideas from this forum, best thing is to try a few to see what you like, then rebuild it with your quality components. I would advise star grounding if its not already installed, and DC heating is often used on mic pre'e to reduce hum levels.
good luck
DaveP

 

[jsefer]

Regarding the Hi Z jack inputs: If you actually want to use those for high impedance sources , IMO you should think about getting jack sockets that interrupt the connections to the Xformers when a jack is plugged in (meaning, you insert the jack btwn xformer and tube).
Otherwise your high Z sources will interact with the xformers' inductive (and stray capacitive) impedances. Many DI inputs are designed that way.

Great I will definitely Do this, thanks.

Also, you should use coupling cap's, otherwise grid leak current might get into the outputs of instruments you plug in.
In I/P 2, this would also prevent the 10K resistor (probably to tame the xformer's resonance?) to load down the DC resistance of the grid input (which obviously wants the 500K resistor).

That 10k was recomended by Dave @ Cinemag, in regards to the coupling caps, isnt that what that C9 and C11 are?? Im confused now, you saying I should use another coupling caps??

Your O/P's (especially O/P1 coming from the plate of a pentode) will have a Z of at least a few KOhms - which might cause trouble when interfacing to the rest of your recording chain, especially if your next i/p is xformer-balanced.

In case you're planning on using both of the I/P's at the same time, you should decouple their plate supplies.

thanks, I will add a 50k resistor and a 20uf 400v cap between stages, would you say this should suffice??

Thanks for your time Volki

 

[jsefer]

Hi jsefer,

Mic pre's have to be carefully designed to avoid noise.

I wouldn't put a 100k pot on an EF86 grid, you need a low noise metal film resistor; there is already a 500k pot on the output.

So In your opinion, just having a 100k and a 500k output pot is sufficient? Great I will try that. I thought It would be good to have a gain Pot at the input side.

If you are putting a 10k on the TX secondary before the 1st ECC81 then you don't need the 500k, it will only add noise.

So removing it and just leaving the 10k should suffice? Great I will try this also

The gain of those 2 ECC81 is going to be over 60dB plus the gain from the TX, but what you still haven't got is any driving power.  You have several options here are a few:-

1)  Use the 2nd half as a cathode follower and take the output from the cathode via your 100uF.  But some people don't like cathode followers.

2)  Parallel the ECC81's and run hot and take a plate output through a step-down TX.  The EMI REDD47 takes this approach but uses an ECC88.

3)  Just make a single mic pre out of both tubes, use the EF86 for voltage gain with the output going to paralleled ECC81 to give you some poke as described in (2)

There are plenty of other ideas out there including having variable feedback to control the volume, as in the REDD47.  Feedback will reduce distortion and noise and reduce the output impedance making it easier to drive low Z loads.

You will get lots of ideas from this forum, best thing is to try a few to see what you like, then rebuild it with your quality components. I would advise star grounding if its not already installed, and DC heating is often used on mic pre'e to reduce hum levels.
good luck
DaveP

Great, thanks dave I will research this..

 

[volki]

Also, you should use coupling cap's, otherwise grid leak current might get into the outputs of instruments you plug in.
In I/P 2, this would also prevent the 10K resistor (probably to tame the xformer's resonance?) to load down the DC resistance of the grid input (which obviously wants the 500K resistor).

That 10k was recomended by Dave @ Cinemag, in regards to the coupling caps, isnt that what that C9 and C11 are?? Im confused now, you saying I should use another coupling caps??

The 10k is needed for even frequency response of the xformer, but the 500k is useless then since it's effectively in parallel, so your grid i/p resistance will be 10k instead of 500k. It may bias the tube differently, but just try and see (read: hear) if you like it.
I wasn't referring to o/p coupling caps, you want them at the i/p's as well: jack - coupling cap - grid (including grid resistor).

Your O/P's (especially O/P1 coming from the plate of a pentode) will have a Z of at least a few KOhms - which might cause trouble when interfacing to the rest of your recording chain, especially if your next i/p is xformer-balanced.

In case you're planning on using both of the I/P's at the same time, you should decouple their plate supplies.

thanks, I will add a 50k resistor and a 20uf 400v cap between stages, would you say this should suffice??

Depending on the plate current, the voltage at the plate will drop according to V=R*I, lowering your headroom.  You could also feed the EF86 directly off the PSU rather than in series with the triode, making the voltages more independent. As for capacitance, someone else will gladly fill you in regarding value suggestion, but that also depends on if you want to use those amp stages individually, or in combination, as DaveP suggested.

 

[jsefer]

Also, you should use coupling cap's, otherwise grid leak current might get into the outputs of instruments you plug in.
In I/P 2, this would also prevent the 10K resistor (probably to tame the xformer's resonance?) to load down the DC resistance of the grid input (which obviously wants the 500K resistor).

That 10k was recomended by Dave @ Cinemag, in regards to the coupling caps, isnt that what that C9 and C11 are?? Im confused now, you saying I should use another coupling caps??

The 10k is needed for even frequency response of the xformer, but the 500k is useless then since it's effectively in parallel, so your grid i/p resistance will be 10k instead of 500k. It may bias the tube differently, but just try and see (read: hear) if you like it.
I wasn't referring to o/p coupling caps, you want them at the i/p's as well: jack - coupling cap - grid (including grid resistor).

Got it I will work out some values and try this also.

Your O/P's (especially O/P1 coming from the plate of a pentode) will have a Z of at least a few KOhms - which might cause trouble when interfacing to the rest of your recording chain, especially if your next i/p is xformer-balanced.

In case you're planning on using both of the I/P's at the same time, you should decouple their plate supplies.

thanks, I will add a 50k resistor and a 20uf 400v cap between stages, would you say this should suffice??

Depending on the plate current, the voltage at the plate will drop according to V=R*I, lowering your headroom.  You could also feed the EF86 directly off the PSU rather than in series with the triode, making the voltages more independent. As for capacitance, someone else will gladly fill you in regarding value suggestion, but that also depends on if you want to use those amp stages individually, or in combination, as DaveP suggested.

Great note, I will rewire the stages, I want to use them independently, So this is a great point... Im reading now on NY Daves One bottle, I had gave it a look a while back, and now its reminding me of the benefits of adding a negative feedback system. Im trying to work out the values and the design.. Essentially I have 2 of these mono blocks and for what I want to use them Id like to build a few different designs, so that I have in the end 4 different mic preamps w different characteristics at my disposal...

So To recap, If i feed voltages directly to the tubes will i still need to decouple them?

 

[DaveP]

Hi jsefer,

If you intend to run four separate channels then you need to run 4 resistors to 4 caps from a common supply.  This was the principle behind monoblocks anyway, non interaction of stereo channels.

If two stages share a common cap it is when the signals are opposite phase so they self cancel which helps the decoupling.

The output Z of the EF86 is 250k//2.5M so its 227k, the ECC81 is 250k//25k which is 22.7k.  You will have to be careful what you connect to these as the wrong Z will affect the frequency response.

Aitkin explains this very well:-
http://www.aikenamps.com/

good luck
DaveP

 

[jsefer]

Hi jsefer,

If you intend to run four separate channels then you need to run 4 resistors to 4 caps from a common supply.  This was the principle behind monoblocks anyway, non interaction of stereo channels.

If two stages share a common cap it is when the signals are opposite phase so they self cancel which helps the decoupling.

The output Z of the EF86 is 250k//2.5M so its 227k, the ECC81 is 250k//25k which is 22.7k.  You will have to be careful what you connect to these as the wrong Z will affect the frequency response.

Aitkin explains this very well:-
http://www.aikenamps.com/

good luck
DaveP

Well actually its 2 channels per, I have two of these sections each containing its own power supply, shared by 2 valves.., Thanks I will look at this. I am actually thinking now, talking to you, that It would be wise to maybe build one  with two channels and one of them as a one channel, EF86 feeding the ecc81 as you suggested.

 

[jsefer]

If anyone else has any pointers on the design Id really appreciate it.