Transformerless Feedforward Vari-Mu

[Heikki]

Transformerless vari-mu based on the U73 compressor/limiter.

Design files

notU73.zip

(10.92 MiB) Downloaded 218 times

-

Schematic

73.pdf

(148.42 KiB) Downloaded 281 times

-

Layout

73layout.pdf

(2.37 MiB) Downloaded 250 times

-

Fully tried and tested design. I'll post measurements and more pictures later.
-

Gerber files here
viewtopic.php?t=75
-

Boards for sale
viewtopic.php?t=49
-

not73.JPG (3.72 MiB) Viewed 35630 times

[Heikki]

Specifications

Input impedance ≈ 12k ohms, balanced line-to-line *

S/N > 86 dB re +4 dBu, 22 kHz BW, unity gain

Frequency response = 20 - 20 kHz, ±0.5 dB

Gain = 0 dB at limit mode, 6 dB at compress mode

Output impedance = 50 ohms, balanced line-to-line
-

* INA2137 Internal resistors are ratio matched but have ±25% absolute value

[scott2000]

I can make out U5 on the DCDC board pic. What model Recom is used?

[Heikki]

Recom R12-150B

Regulators for the DC-DC converters are 12V and regulator for the heaters is 5V.

R1 = 1 ohms

R2 = 2.2 ohms

R3 = 4.7 ohms

-

[Script]

Would 6BA6 tubes work ? Read that they are not fully equivalent to 6K4P, something to do with internal resistance.

[Heikki]

6BA6 is essentially equivalent to 6K4P. Only somewhat significant difference is that in 6KP4 suppressor grid is internally connected to cathode. Difference in internal connections won’t matter since tube pins 2 & 7 are connected together on the board. Any vari-mu pentode with same pinout as 6BA6 or 6K4P will work, with other tubes R13 might need to be changed.

[Script]

R18 and R19 are attack and release, right?

Worth making them variable, despite feedward then possibly tending to sound weird at 'wrong' setting?

[Script]

For stereo would need two main boards. Can they be powered from one PSU board?

[Heikki]

Yes R18 is attack resistor and R19 release resistor. I connected 6 position rotary switch with different resistors and capacitors to J10 & J11 to change the release time. J10 & J11 marked TC on the board connect parallel to C9 & R19. R18 could be made variable too without any ill effects.

Gain and threshold are fixed to avoid situation where ratio goes beyond infinity. Gain and threshold can be made variable by wiring RV5 & RV1 to the front panel if you don't mind the possibility of beyond infinity ratio.

One PSU board can power up to 4 main boards.

[Script]

I read the description of the U73b, document on groupdiy.

In that text they talk about 'solid release times' of 0.3 - 0.6 - 1.2 seconds. These are easy to derive by putting resistors or caps on a rotary.

But they also talk about 2.5 - 6 -1.2s as being ' influenced by their respective program content'. What do they mean by this -- technically? Is it just longer release times realized thru simple RC or is there more to it in terms of circuitry?

I think they just mean to say that longer release times can see another attack kicking in that changes overall release behaviour, right? So, simple RC circuit.