Input, transformers, and hold options, meter options, bypass board, etc.
---------------------------------------------------------------------------

Chandler TG description.
TG* built with input level control Elma 21 pos. switch, arranged as approx. 
10k log pot, 2 db steps. Input transformer is custom Altran,
1:2 or 1:4 ratio. It can hold pretty high levels.
Neve mic. input or substitutes will not work here properly.
Hold control not used in Ch*r unit, it's wiper connected to ground.
This gives a plenty of gain from input amplifier/VCA, however,
doesn't improoves s/n ratio and it works in very distorting mode,
not recommended by EMI manual.
Output transformer is 50:600 Ohm, approx. 1:3 ratio, wound om M6 steel.
Nothing special in sound or specs. 
The rest is close to EMI schematic.

Original units were transformerless, and used HOLD control (10k lin pot).

Fester's clones can be built either with input pot or hold control.
Actually, HOLD control gives better s/n ratio, less unwanted distortion,
and closer to original units sound. 
10k lin potentiometer used in EMI units, gives control range more than 40db. 
To get more precision in range of +/-15 db gain, 
5k log pot connected in series with 5k resistor can be good option. 
To get precision steps, possible to use switch,
however, you'll need to connect 10k pot and measure its value at every step,
because control voltage depends on zener diodes used. 
1n751a is very close substitute to HS2051.
Optional 100-200k /HOLD trimer can be installed on front pannel
board to adjust both channels for precision "0" at the same position
of HOLD control. It connected in parallel to HOLD and "shifts" its "center".

It is not easy to get matched HOLD control even on two units,
seems this was the Ch's reason to change HOLD control to input pot.
To get precision matched pair of TG limiters, all HS2051's 
should be from same mfg batch and matched very close by Vz.
Matching procedure: 10V source, DPDT switch, 680 Ohm and 10k resistors, DC voltmeter. 
Measure Vz with 680R and 10k, write thes values on paper, next zener.
Matching should be better than 1% for both values for zener pairs.

Usually, 20-30 zeners from same mfg bach is enough to find 8 close-matched
for stereo pair of limiters.

Input/output transformers options, bypass board.
Transformer's bypass option reserved as well, input and output matching stages added.
When transformers bypassed, input of IC or discrete-opamp based balanced line receiver 
used to match Fester's clone to outside environment.
Its gain set by resistors Rin1/2/3/4 and should be same as input trafnsformer's gain.
When transformers are bypassed, output gain is matched to be same as with transformer
by output amp, located on bypass board. Its gain set by resistors Rop1/2.
OPA604 are recommended opamps, they feed from +/-24 v regulator
located on bypass pcb. Other opamps can be used with reducing supply voltage
to +/-18 V, by change of R91/92 on bypass PCB. 
Some variants of bypass board assembly shown in BYP_VERSIONS_Q.pdf.
Mind, this drawing done from older version bypass pcb, 
diodes across voltage regulators are shown wrong way here 
and power supply bypass 47uF capacitors are missing. 
These errors were fixed on new pcb. The rest is OK. 

Simplest way to build TG with iron and transformers bypass option
is to use 1:1 input and output transformers. In this case, use
1:1 input balanced line receiver and ommit output amp on bypass board.
Input trafo is 10k:10k Sowter etc, output is 600:600.

Another very nice sounding option is use of Neve 31267-like 10k:2k4
on input and Neve LO1173-like (70:600) as output transformer.
In this case, input balanced line receiver shoild have a gain of -6db, 
and output amp (on bypass board) should have +7..8 dB depending on
output transformer's load to match gain with and without transformers.

VU-meter.
Original units used 2.4 ma FS meter with scale marked -10...0...+10
and red regions. See decription.

To get nice looking meter, use Sifam supplied by Canford, 
http://www.canford.co.uk/products/productdetailpage.aspx?productid=58-379
with removed diode bridge.
Print scale on transparency and put paper from back.
To adjust the metar, use 500R-1k 10turn pot in parallel.  
Don't expect for high meter's precision. 

Before initial adjustment, read EMI circuit description taken from original console's manual.
Check everything again before first turn-on.
 
Testproc:
1). +/-20v
JP1...6, TP+20v, TP-20v
2).Unity gain
Initial gain is set by AOT1,2
3).Distortion w/o compression
minimum distortion adjustment AOT 7,9
4).Distortion at 20dB GR
distortion under hard compression adjusted by AOT8/10, usually 10R is fine
5).Com threshold
Com threshold adjustment: AOT3,4
6).Lim threshold
Lim threshold set by AOT5,6.
7).Check: reqresp, max gain, min  gain, S/N, timeconstants.

If hold control done like in Chandler's unit, vu-meter should be calibrated different 
than EMI scale.

Good luck with biulding this amazing sounding compressor!