UPDATED; Audio Historians - REDD/RCA/Langevin/RFT Lorentz/WSW

Help Support GroupDIY:

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
There was a thread on Gearslutz (as it was called then) a few years ago where a lucky guy found a REDD.43 and a REDD.59, intact with all cables and racks of REDD.47 amps.
The thread is probably still up there so it'll be searchable.
Here's a screen shot I grabbed showing what he found but there'll be other shots still up there no doubt:
 

Attachments

  • Screen Shot from Gearslutz.png
    Screen Shot from Gearslutz.png
    770 KB · Views: 39

Matt Syson

Well-known member
Joined
Dec 17, 2005
Messages
82
Location
France
Hi
I suppose for my 2 centimes worth I could point out that switched gain controls have the potenial (although probably not realised) of being more 'equivalent' in sound at different gain settings as different 'networks' can be introduced at each step so both resistance and capacitance which could compensate for changing feedback response in a way similar to frequency compensated gain switches in gear like oscilloscopes or wideband metering gear although there is a difference i one being 'series' and the other shunt. This sort of 'compensation' cannot be applied when a pot is used for the gain setting variable. Not seen any single shaft pots with variable caps (think Medium wave tuning caps) added to the rear to keep the high frequency gain response constant available 'off the shelf'.
i suppose some of the earlier Rupert Neve designed (and others) mic amplifier units had three 'optimised' gain stages here for low gain requirements one stage would be bypassed.
Thinking aloud and the talk of quadrant faders on another thread, I suppose frequency compensated 'stud contact' faders SHOULD be the thing. 600 Ohm types would be near constant HF response but certainly 10K if feeding into screened cable to the following amplifier stage will see varying HF response (arguments about 'Passive preamps' come to mind.
 

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
Hey Matt,
yep all good points there, worth more than 2 centimes for sure.

On the Neve early amps, the gain of the one stage of the three that "did" alter fb for different gain only did so over a somewhat limited 15dB swing, from the 40dB to 55dB area. There were certainly others of that time that did things in a similar way too.

With EMI, the valve REDD.47 amp would be an example of what you mentioned of having a differing compensation scheme at each gain setting, with the 40 dB and 34dB positions each introducing a C in parallel with the respective feedback series R to compensate for phase shift from the output transformer.

It's also a very good point regarding higher Z faders (or pots) feeding a length of screened cable and then an amplifier with an input capacitance which then changes the high end response within the audio band, depending on setting. Your -6dB point along that fader's travel being the place where your top end is most reduced of course.

The 200 ohm or 600 ohm Painton, Elcom, Daven etc. studded Bridgted "T" or ladder quadrant faders were, as you say, never an issue when run unbalanced within a desk environment.
When we get everything sorted with the guys who came up with quadrant fader retrofit parts idea, I'll probably go with using 1K audio taper P&G's as used in API desks, which will do me quite nicely :)

5K audio as for Neve's, or the standard 10K will be fine too as long as you use 'em as and where intended.
 

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
..would make sense if they referred to 1mW into 200 Ohms, not 600 Ohms,,

Regarding .447V ?
Yes sir, they did.
I just learned last night that Olympic Studios used the same reference voltage.
Was that common throughout Europe in the 50's and '60's Jakob?
 
Last edited:

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
13,386
Location
Norfolk - UK
We discussed some of this "compensated switched gain" stuff in another thread. The basic problem is that if you have a fixed open loop gain and vary the closed loop gain simply by altering the NFB, then the amount of feedback (open loop gain dB - closed loop gain dB) varies over a large range and this is what makes compensation difficult.

What you need is a way of keeping the amount of feedback more or less constant despite changing the closed loop gain. One way to do this is to vary the open loop gain at the same time as you vary the closed loop gain, thus keeping the amount of feedback nearly constant so a single compensation is all that is needed. This sounds complicated but in practice it is not too hard to do with discrete designs. The V76 did it and I think one of the earlier V40 series did too. The Neve three transistor amp does it as does the classic Helios three transistor amp. I do it in my EZTubeMixer design and the well known TransAmp does it too.
Cheers

Ian
 

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
Yes the early designed NWDR unit was the V41 and you're right, it uses the same method as the later V76. The V41 was developed just post WWII I believe (1947/48) and was very clever for its time.

I'd say 20dB is about the limit you'd be able to easily compensate a valve amplifier and expect good results all the time using standard schemes. Yet I think both the V41 & 76 have 40dB gain swing (I'm excluding the lower level settings which are pads) which is a huge difference for sure.

This certainly keeps the amp (V76) stable with enough phase margin at each setting.
However, what does change on the V76 is the nature of the overall throughput of the amplifier's THD/IMD, since part of the scheme involves the degree to which the first stage cathode is bypassed.
It's in this sense that I'm still fond of the lowly 40dB V72 with its fixed negative and positive feedback down to DC. Everything in harmonious balance so to speak.
Of course, that's just my own personal preference on those 2 (72 vs 76) amps. Worth no more than that 🙂
 
Last edited:

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
Bumping up the RCA content, here's a pic of the desk from Studio A at RCA Italiana. V76 pre amps visible in the "penthouse"
 

Attachments

  • RCA Italiana.png
    RCA Italiana.png
    485.3 KB · Views: 16

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
Close up of the right-hand wing. Looks like a row of Neuman TEV germanium transistor EQ's, 3 Pultec EQP's and four Fairchild solid state comp modules top right:
 

Attachments

  • Italiana right side.png
    Italiana right side.png
    672.7 KB · Views: 21
Last edited:

Winston OBoogie

Well-known member
Joined
Jun 3, 2004
Messages
2,427
Location
UK.
Stuck in 'My Own Private Idaho', or solitary confinement...

So, not sure I've seen a frequency response graph online anywhere for the REDD desk "Bass Lift".

To re-cap: this doo-dat thing-umy was used to compensate for the bass loss which occurs with figure-eight condenser mics when used in mid-field/far-field positions, such as with a Blumlein pair on orchestra, a string quartet, or an Eleanor Rigby rigmarole.

Not a circuit diagram, but this is a sim of the frequency and phase characteristics. It also includes the relevant (to the low end) characteristics of the REDD A92 input transformer as utilized in the REDD.47 amps,

This thing is, naturally, passive, at mic level (with approx 10dB insertion loss), and 3dB per octave slope.
Sorry, ignore the top right mini screen info, just trying to get the pointer out of the way
 

Attachments

  • REDD 'Bass Lift' with REDD A92 input transformer.png
    REDD 'Bass Lift' with REDD A92 input transformer.png
    81.6 KB · Views: 12
Last edited:

Latest posts

Top