Redd.47 Output Level Сontrol

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I used 1k pot with two Cinemag 30k:600 output models, 5:1 Edcor and some home made 5:1 to try them out. 1k pot works well in practice although T pad is a bit better. Especially when very good transformers are used and one wants the rest to be as good as possible. "The problem" with Redd47 is very limited gain range, not lack of gain with 46dB max, it really needs an output pad.
Did anyone notice how sound is changing when switching gain positions? At higher input levels 34dB position sound kind of falls apart a bit, i like 40dB best, 46dB has more of sometimes annoying high frequency distortion. I never tried to extend the range of gain switch for this reason, maybe it is something to try out and hear how lower than 34dB of gain sounds. IIRC, this is what Chandler did down to 16dB while retaining stability.
If it's not difficult, please draw a diagram of your attenuator with a 1K resistor?
 
Last edited:
And a constant impedance atternauator will not work 'properly' unless fed from the correct impedance and loaded with the correct impedance. They worked fine in the old days when you could pretty much guarantee that Zout = Zin = 600 ohms. Add to that that most of todays three pot bridged T attanuators are nothing like constant impedance and you can see why the current paradigm of low Z out and high Z in predominates because it is much kmore predictable and much less susceptible to componenet variables.

Cheers

Ian
Isn't sound with some of this pads changing because output stage feels different loads at different positions of the pad? We used to recommend loading outputs without pad with 600r resistors in order to get it sound as if it was driving 600r input, when it is driving modern bridging inputs.
You probably remember old threads by NYD and some others with small series resistors before 1k pot, it has minimum impedance of around 600r with ~4dB of insertion loss. This should probably be better than 1k pot only if out and in impedance are known (easier with tubes).
Uneeda-audio is great site for this things because they take in account parameters mentioned by you, beside giving k factors, etc.
 
If it's not difficult, please draw a diagram of your attenuator with a 1K resistor?
Did you translate my English post to Russian when quoting? Feels like provocation, if you did delete it.
 
Last edited:
Did you translate my English post to Russian when quoting? Feels like provocation, if you did delete it.
Well, yes, a mistake .. - I corrected ... :mad:

I tried to do according to the scheme, which is in my first image.
But the more I reduce the level, the more distortion at high frequencies.
 
Last edited:
Well, yes, a mistake .. - I corrected ... :mad:

I tried to do according to the scheme, which is in my first image.
But the more I reduce the level, the more distortion at high frequencies.
I'm sorry if it wasn't provocation, current situation is incredible...
First post has simplest form of this attenuator as used by API and many members here:
https://groupdiy.com/threads/access-312-ez1290-1k-pot-output-attenuator.80724/I also used 1k pot with series resistors and different wiring, another version on switch too. We better see what Ian thinks about loading of the output stage because it is important. This is a little complicated matter, i must admit something seemed wrong on your drawing but i didn't catch mistake at first.
What do you need it for, what is input impedance of units driven by this?
This site is great for understanding various pads, it gives formulae for calculating and some examples:
http://www.uneeda-audio.com/pads/
 
It is a complex question. Any distortion is almost certainly dominated by the output stage tube itself. The transformer does not care what the load is; its distortion depends primarily on the applied primary voltage. So the question becomes how does the output stage react to a varying load reflected to the transformer primary? Now the REDD 47 was designed in the days of constant impedance loads so the original design expected a load of 200 ohms all the time. So to be true to that design intent you should use a constant impedance bridged T attenuator with a 200 ohm load on its output (or a 600 ohm version with 600 ohm load if the transformer ratio has been changed for 600 ohms).

A bridged T attenuator needs only two variable resistors but both need to be able to vary from zero ohms to infinity which is hard to do with regular pots. In addition, the law required is nothing like any of those commonly available so the three pot versions are something of a compromise. The only way to find out how much of a compromise is to load a 600 ohm one with 600 ohms and measure its input resistance as you rotate it. It should read 600 ohms all the time. If it reads within 10% of that value across the range that that should be good enough.

If you want a true bridged T attenuator then a double pole switch is the way to go. There are plenty of on line calculators that will work out the resistor pair values for each position for you.

Edit: Going back to the OPs original question: A 1K pot wired as in his schematic should work. If loaded with a regular 10K line input then there is no reason the distortion should vary with the position of the pot. It seems more likely the pot has been wired incorrectly and is shorting the transformer secondary output which will make the output stage distort more. IIRC that is exactly the scheme used in the so called REDD DI box.

Cheers

Ian
 
I record vocals (hobby), but I can't afford to buy a Chandler Redd.47, so I'm trying to do something similar myself, based on Sowter 9970 and 9980 transformers.

My knowledge of this is negligible, so I just do what I see on the diagram.
Everything works well, only the preamp has a very large output and I need to make it adjustable without changing the sound much (I have a regular E-MU 0404 usb line-in audio interface).

The image (Chandler Redd.47) shows that the Output circuit uses a potentiometer, resistor and capacitor.
But it is not possible to consider a more detailed scheme.

Can someone tell me how this is implemented in Chandler Redd.47?
 

Attachments

  • Chandler REDD 47 POT 3.jpg
    Chandler REDD 47 POT 3.jpg
    69.3 KB
Last edited:
I don't see a capacitor in that image. Is the rotary switch the output level control? How is it labelled on the front panel? With that info it should be easy to come up with a schematic for it.

Cheers

Ian
 
On the image I wrote the name of this potentiometer, I also drew what I could see in this photo.
Maybe there will be some ideas how this attenuator can be implemented?
 

Attachments

  • Chandler REDD 47 POT 3 POST.jpg
    Chandler REDD 47 POT 3 POST.jpg
    119.8 KB
Now I can see the capacitor, thanks.

I suspect the capacitor is wired directly across the transformer secondary. There is no need for a series capacitor and any way the one in the picture looks way too small for that job. It is common to wire a series resistor capacitor combination across the secondary of a transformer to smooth the frequency response - values depend on the transformer. The pot is almost certainly wired exactly as in your first post.

Cheers

Ian
 
I tried to make different options like in image 1.
As I already wrote, the lower I make the signal level, the more distortion.

Next, I did as in image 2.
Everything works well, only this is a fixed attenuator option, and I need to make an adjustable attenuator.

I decided to make an unbalanced output as in image 3.
Everything works and is regulated well.
But it still confuses me that this is already an unbalanced output ..
Although this may not be so important, given the line level of the signal and the short cable from the preamp to the audio interface.
But maybe I'm wrong and a balanced output is definitely needed?
 

Attachments

  • REDD.47 POST 01.jpg
    REDD.47 POST 01.jpg
    28.1 KB
  • REDD.47 POST 02.jpg
    REDD.47 POST 02.jpg
    15.4 KB
  • REDD.47 OUT UNBALANCED.jpg
    REDD.47 OUT UNBALANCED.jpg
    16.7 KB
How is the picture?
If so, then I tried it, but the distortion became even greater as the output level was reduced..

The preamplifier is included in the line input of the regular E-MU 0404 USB audio interface.
 

Attachments

  • REDD.47 POST 022.jpg
    REDD.47 POST 022.jpg
    13.8 KB
This question with this preamp comes up all the time, and I always wonder what actual max levels are encountered, given the rating is for +14dBm “before severe audible distortion becomes apparent”. It should be in hard clipping pretty soon after that.
 
emrr makes a very good point. I cannot see why you would need to attenuate the output by more than 20dB if you are feeding it into a line input. The line input of the EMU will take an input level of up to +14.2dBu which is pretty close to the maximum output of the REDD47. What happens when you feed the REDD 47 straight into the EMU line input?

Cheers

Ian
 
This question with this preamp comes up all the time, and I always wonder what actual max levels are encountered, given the rating is for +14dBm “before severe audible distortion becomes apparent”. It should be in hard clipping pretty soon after that.
Very interesting, I did not know that such problems occur with this preamp..
If such a problem happens often, then perhaps this is due to the implementation of the output attenuator?
I hope there is no such problem in the version of REDD.47 from Chandler?
 
Last edited:

emrr makes a very good point. I cannot see why you would need to attenuate the output by more than 20dB if you are feeding it into a line input. The line input of the EMU will take an input level of up to +14.2dBu which is pretty close to the maximum output of the REDD47. What happens when you feed the REDD 47 straight into the EMU line input?

Cheers

Ian
When connected to the E-MU line input, the signal level is too high.
For example, on the VOLTAGE GAIN is 34dB - on the E-MU indicator it is already close to -3dB.
If I switch to 40 and even more so 46dB, then the E-MU is already moving into the clip.
But at the same time, the preamplifier does not distort the signal in the way that my "adjustable" output attenuator does (hard clipping).

If I do according to scheme 3 - not a balanced output - everything works and is regulated remarkably.

I'm already ready to make the regulated output unbalanced, but there are some doubts.
Perhaps in the future I will use a compressor (1176) after the preamplifier.
I don't know how it will work in unbalanced mode..
Perhaps this will degrade the sound with such a connection?

Therefore, I still want to do it in a balanced version.
In Chandler REDD.47 they did it somehow ..
 
Last edited:
When connected to the E-MU line input, the signal level is too high.
For example, on the VOLTAGE GAIN is 34dB - on the E-MU indicator it is already close to -3dB.
If I switch to 40 and even more so 46dB, then the E-MU is already moving into the clip.
But at the same time, the preamplifier does not distort the signal in the way that my "adjustable" output attenuator does (hard clipping).

OK, Aassuming you have turned the EMU gain down as low as it will go, this implies the output from the RED 47 is already at about +11dBu. You do not say what source you are using but it looks like even the lowest gain setting on the 47 is too high. So possibly what you really need is an input attenuator
If I do according to scheme 3 - not a balanced output - everything works and is regulated remarkably.

I'm already ready to make the regulated output unbalanced, but there are some doubts.
Perhaps in the future I will use a compressor (1176) after the preamplifier.
I don't know how it will work in unbalanced mode..
Perhaps this will degrade the sound with such a connection?

Therefore, I still want to do it in a balanced version.
In Chandler REDD.47 they did it somehow ..
This is all very confusing. Wiring one side of the transformer to ground to make the output unbalanced should not make any difference. The implication is that there may be something wrong with the build of your REDD 47. Did you build it yourself? DO you have a 20dB input pad switch like the Chandler does?

Cheers

Ian
 
Back
Top