does a neve 2254 really work this way

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

pucho812

Well-known member
Joined
Oct 4, 2004
Messages
14,831
Location
third stone from the sun
I have never seen a 2254 schematic or ever used a 2254. But one of my co-workers was insisting that a 2254 takes the line level signal knocks it down to mic level. does the compression stuff and then boosts it back up to line level at about 70db and that the make up up gain is really an an attenuator....  Never used one but the idea just sounds wrong to me....

what's really going on in there?
 
fwiw:
i'm looking at an uncle fester schematic not the original--don't know the 'trunny' [slang is bugging me (transformer)] ratios but after the input xfmr I see about a 10 or 20:1 signal knock down b-fore the fets that feeds another trans and some boosters sidechain etc...don't about the attenuator but it don't sound too wrong to me--shunt-fet agc don't allow for big signal at the active portion afaict.
 
shabtek said:
fwiw:
i'm looking at an uncle fester schematic not the original--don't know the 'trunny' [slang is bugging me (transformer)] ratios but after the input xfmr I see about a 10 or 20:1 signal knock down b-fore the fets that feeds another trans and some boosters sidechain etc...don't about the attenuator but it don't sound too wrong to me--shunt-fet agc don't allow for big signal at the active portion afaict.

I never seen a schematic.... so thanks for that...
 
> what's really going on in there?

I don't have a clue what you are talking about.

However most variable-gain schemes have a narrow dynamic range. Often, the noise is not so high, but the maximum level is quite low.

Diode-bridge is clean to 20mV, distorts bad past 60mV.

OTA-like gain-cells the same.

JFETs go wonkers somewhere above 50mV-200mV.

Opto-resistors are not so abrupt. 200mV is pretty darn clean. I could hear the change when I built with 3V limiting level, noticable IM.

The DBX et al VCAs would be nasty above 20mV and servo-away their input so the actual emitter level is a teeny fraction of a mV (ideally zero).

Tube limiters may reach several Volts on the grids, but distortion is rising. Start-of-limiting is usually 100mV-200mV.

Potentiometers can handle line level but have trouble with speed.

Any of these lower levels "might" look like "mike level". And some of these would naturally use something like a "mike preamp" after the gain-cell.
 
The level coming out from the diode bridge is -40dBm, see below signal flow diagram.
http://danalexanderaudio.com/neveinfo/neve2254aSchem.jpg
 
jensenmann said:
PRR said:
Potentiometers can handle line level but have trouble with speed.

What does that mean, PRR? Could you pleasse elaborate this?

hand_anatomy_intro01.jpg
 
Panasonic used to make affordable motorized pots. I'm not seeing the low-price ones at DigiKey today. They do have a $50 motorized slide-pot. You can also salvage motorized volume knobs from 1980s stereos.

If sound is too loud, apply a negative voltage to the motor to turn the knob and volume "down".

When sound is not excessive, but was, apply a positive voltage to the motor to restore the original gain.

If motor speed is proportional to voltage, then attack/release slew rates may be set with voltage stops.

An advantage: this limiter pot may also be your main gain pot, keeping the audio system simple.

Speed: IIRC the slew was OTOO 20dB per second. A smooth manual fade. This might be OK for a slow leveler. It is far too slow to limit transients. The motor voltage could be raised above specs and speed would increase. Motor would not overheat because the over-volt period is short. But 10X voltage is still only 200dB/Sec or 0.2dB/milliSecond or 2dB in 10mS. And 10X voltage might cause arcing in the brushes and rapid pitting and early failure.

There are fastor motors, and Mechano or robotics hardware to couple motor to pot. But electric motors rarely hit the millisecond speed we need for rapid audio limiting.

We could try voice-coils. Lower mass. However a long-travel voice coil's self-mass gives about a millsecond max response, slower with pot mass, and probably leverage to convert to wide rotary swing (or long linear swing) from short linear motion.

Air pistons and valves are a bit quicker.

All fast mechanical contraptions are prone to be clunky so may have to be isolated from the monitor room.
 
What about a flywheel and a clutch?

Or maybe compressed air? You'd need a compressor to power your compressor.


Justin
 

Latest posts

Back
Top