General Audio Research - REDD47 build thread

Help Support GroupDIY:

delwood

New member
Joined
Nov 8, 2019
Messages
3
Basically tubes are relatively noisy high impedance input devices. To obtain decent noise performance from a tube mic you need to raise the input signal as much as is feasible before feeding it to the grid of the first stage (which is where most of the noise is added).

Really helpful for a learner like me to understand that - thank you.

A 1:5 input transformer is relatively straightforward to make whilst retaining an excellent frequency response and contributing very little noise. With care a 1:10 transformer can be make with little compromise in frequency response and noise. With extreme care, and a lot of money, a 1:20 transformer can be made.

Is it possible to utilise a hybrid design, with a 1:5 input transformer coupled to a solid-state gain stage to bring the gain up to be comparable to a 1:10? I appreciate not an idea for the purist... just curiosity more than any intention.
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,931
Location
Norfolk - UK
Is it possible to utilise a hybrid design, with a 1:5 input transformer coupled to a solid-state gain stage to bring the gain up to be comparable to a 1:10? I appreciate not an idea for the purist... just curiosity more than any intention.
There are quite a few transformer-less tube mic pre designs (commercial and DIY) that have a nice quiet FET shoved up the cathode of the first tube. If you want to use a transformer plus SS then probably the optimum transformer ratio is 1:2. If your SS stage had 20dB gain the the whole is equivalent to a 1:20 transformer.

Cheers

Ian
 

HyprDrivr

Active member
Joined
Nov 5, 2020
Messages
44
Update.

Since GAR just came out with a PSU board for his builds i decided to try it out. Looks very neat and a practical way to get all voltages on the boards and phantom power.
One thing i am a bit concerned about is the AC in to HT is recomended 200V. rectified that leaves something like 283V and the circuit calls for 280v B+
My thinking was using a 240v AC instead, but i have to figure out R1 and R2 values. I was thinking that R1 is used to drop the voltage and that would be pretty simple to calculate, but i am a bit confused about the purpose of R2 that goes goes from R1 to ground...

20211109_113029.jpg
20211109_094624.jpg
20211109_094649.jpg
 

dmp

Well-known member
Joined
Oct 28, 2009
Messages
3,083
Location
Madison, WI
The documentation on the GAR site says this:
for redd 47 & v72 input 200vac R1= 1k R2= 470k total output =280vdc

No schematic, which is frustrating. But I would guess the 470k is the bleeder resistor also. It needs to be able to handle the power dissipation. Whether the B+ is 280V or 240V, it is fine. P=V^2/R=0.16 W

1k is also fine, it sets the filtering with the caps.

The final B+ will depend on the load current. With zero load, the B+ will be at the peak 283V. Under load the B+ will drop as there is voltage drop across R1 and L1
 

HyprDrivr

Active member
Joined
Nov 5, 2020
Messages
44
The documentation on the GAR site says this:


No schematic, which is frustrating. But I would guess the 470k is the bleeder resistor also. It needs to be able to handle the power dissipation. Whether the B+ is 280V or 240V, it is fine. P=V^2/R=0.16 W

1k is also fine, it sets the filtering with the caps.

The final B+ will depend on the load current. With zero load, the B+ will be at the peak 283V. Under load the B+ will drop as there is voltage drop across R1 and L1
If you look at the first image I posted you can see the traces and the circuit is fairly simple.

I would guess from the schematics avalible that current draw is something like 20mA, but i am going to use two preamps so that would have to be compensated for.
So just to get a starting point i am going to assume 40mA draw. 240VAC rectified is something like 340VDC so R1 would have to be something like 1k5 3watts.

Let me know if i am missing something or if there is something wrong with my math.
 
Last edited:

dmp

Well-known member
Joined
Oct 28, 2009
Messages
3,083
Location
Madison, WI
Just build it and see. You can always change R1 after you take a measurement.
If you are trying to get more voltage drop you can skip the inductor and just use another resistor in its place
The point of an inductor is to get the filtering without DC voltage drop.
 

HyprDrivr

Active member
Joined
Nov 5, 2020
Messages
44
Just build it and see. You can always change R1 after you take a measurement.
If you are trying to get more voltage drop you can skip the inductor and just use another resistor in its place
The point of an inductor is to get the filtering without DC voltage drop.
Yes, thats the idea.
Ah, ok! I was wondering about the inductor. Thanks for the info.
 

beatnik

Well-known member
Joined
Oct 18, 2009
Messages
916
Location
Dorset, UK
Looks to me like the DC filament supply section isn't properly designed. The 7806 has 1A output current rating and the EF86 and ECC88 filaments draw 0.5A each channel, so for a stereo pair the 7806 will be at its maximum rating, plus there is not much space for a proper heathsink, it will have to be rather small with no mechanical attachment to the circuit board and it also looks like it will end up being very close to the capacitors, potentially shortening their lifespan.
Perhaps it's best to leave the DC filament unregulated and adjust the voltage with resistors.
 
Last edited:

HyprDrivr

Active member
Joined
Nov 5, 2020
Messages
44
I did not think of that..

I dont think isnt properly designed, only that i am attempting to make it do something it wasnt designed for...

I got this board so i could build a nice clean looking unit, but at this point i am thinking that i might just build a supply from scratch since i am looking at modifying all three parts of it.
 

HyprDrivr

Active member
Joined
Nov 5, 2020
Messages
44
It seems i have a 7806CV on there that should be able to handle 1,5amps if i am reading the data sheets right.

I have this heat sink on it, the only one i have that fits. If it gets extremely hot i could always move it off board.

20211118_122129.jpg
 

dmp

Well-known member
Joined
Oct 28, 2009
Messages
3,083
Location
Madison, WI
I would recommend trying AC heaters with tightly twisted, proper wire routing. I have built a couple tube preamps with AC heaters.
My V76 build used AC heaters and hit the very low Telefunken noise spec.
 

moamps

Well-known member
Joined
Sep 24, 2005
Messages
1,905
Location
Croatia, HR
I have few objections to the design, but I will only suggest for now that capacitors in the filament part of the power supply do not need to be rated for 63V. It is better to use 4700 or 10000uF/16V or 25V depending on output DC voltage (6.3 or 12.6V).
 

moamps

Well-known member
Joined
Sep 24, 2005
Messages
1,905
Location
Croatia, HR
That is rather expensive for what looks to me like an unscreened input transformer.

The picture of the transformer shows a screen, something in the style of Lundahl transformers. In addition, the core of the transformer is made of mu metal, so to me that price does not seem so excessive. (I am not affiliated with the manufacturer in any way).
1637746888886.png
 

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,931
Location
Norfolk - UK
The picture of the transformer shows a screen, something in the style of Lundahl transformers. In addition, the core of the transformer is made of mu metal, so to me that price does not seem so excessive. (I am not affiliated with the manufacturer in any way).
View attachment 86628
Cinemag and Jensen equivalents with full mu metal screening cans are the same price as this.

Cheers

Ian
 
Top