Opinions - One or two gain knobs?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

BluegrassDan

Well-known member
GDIY Supporter
Joined
Mar 17, 2009
Messages
507
Location
Elizabethton, TN
Howdy,

Thanks to the helpful folks here, my NFB issues are resolved. Now I need help deciding how to configure the gain in a most user-friendly way.

We've got two tube gain stages with a basic rotary switch attenuator between them. The first tube stage seems to work very well with up to 18dB of NFB as a gain control. Thus allowing condenser mics to be used on the input without "having" to pad it down. Of course, there is still a typical -20dB switchable pad if needed.

I now have some options to consider with the engineer in mind:
Option 1: Two rotary switches (similar to attached photos). "Input" adjusts the NFB to reduce the first gain stage. "Gain" or "Level" or "Atten" is the between stage.
Option 2: One rotary "gain" switch, combination mid-stage attenuator and NFB adjust. For example, max gain would have no NFB for full 60dB gain. The lowest gain would have the most NFB. This could be similar to the Gyraf G9 where part of the switch covers attenuation while the other positions adjust NFB.

Biggest drawbacks:
Option 1. Users may overdrive the front end without realizing it. This could be helped by making the -6dB NFB actually be centered at "0" (see picture).
Option 2. It takes the power to shape harmonics and saturation out of the user's hand.

Curious to know what everyone thinks.
 

Attachments

  • 2-ch-w-EQ,-Front-Panel-Design-(Draft-12-merged-for-layout)2.jpg
    2-ch-w-EQ,-Front-Panel-Design-(Draft-12-merged-for-layout)2.jpg
    264.8 KB · Views: 54
  • 2-ch-w-EQ,-Front-Panel-Design-(Draft-12-merged-for-layout).jpg
    2-ch-w-EQ,-Front-Panel-Design-(Draft-12-merged-for-layout).jpg
    79 KB · Views: 54
I guess the question must be do you want a clean preamp or coloured preamp? It seems like you are designing a tone shaping colouring pre then the 2 pot options would be nice.... Thats just my opinion (y)
 
Your first stage can probably swing to +30dBu before it begins to clip. With 6dB of amp gain and 20dB in the transformer, this means the input could reach +4dBu before the first stage clips. With your 20dB pad at the front you could probably feed it line level. What I am saying is that lightly loaded tube first stages like yours tend to produce very little distortion. It is the output stages that create colour and even then only when fully loaded.

Cheers

Ian
 
And yet another way to make it more of colour pre would be to have a output PAD (-15db?) and one gain knobb... so line stage have to work harder... or you have switchable load resistor in parrallel with OPT seconadary like 330r-1K range... for increased THD.... (y)

The output PAD is a user freindly clean or coloured mode

Nice looking front...
 
Last edited:
And yet another way to make it more of colour pre would be to have a output PAD (-15db?) and one gain knobb... so line stage have to work harder... or you have switchable load resistor in parrallel with OPT seconadary like 330r-1K range... for increased THD.... (y)

The output PAD is a user freindly clean or coloured mode

Nice looking front...
Since distortion is mostly proportional to output level it makes sense to control the amount of colour with a pot on the secondary of the output transformer. So I would advise using the combined switch and if somone wants extra colour, offer them the additional output control.

Xheers

Ian
 
After hearing from Jakob (Gyraf), he shares my concerns about gain staging issues from users with the two switch options.

So, here is another idea. Thoughts?
 

Attachments

  • Gain-Rotary-Option-1.jpg
    Gain-Rotary-Option-1.jpg
    105.9 KB · Views: 51
Looks a bit like a job for a dual gang pot if you want a single knob control as used by Midas (IIRC) and several others with the obvious problem that they would need to be custom assembled and depending on your actual circuit design might be a log and a reverse log section as well as different resistances.Of course as mentioned earlier it depends if you want to be able to offer 'colour' and 'clean' on the one control, depending on setting. The 'goal' of having a mic preamp gain control that could take signals from say -72dBu (gain) but could also handle up to +24dBu (line level) input took a fair degree of effort.
 
Seems to me this is really a question for your target market. If you are planning to mass produce it then you should undertake some real market research. If it is a kind of boutique product then you can probably afford to make several versions available.

Cheers

Ian
 
Seems to me this is really a question for your target market. If you are planning to mass produce it then you should undertake some real market research. If it is a kind of boutique product then you can probably afford to make several versions available.

Cheers

Ian

I certainly ain't quitting my day job!

Ultimately, this whole process has been about discovering what sounds good to ME and functions well in use. The first gain stage clipping doesn't really sound that appealing to my ears, so I'm inclined to make the front end relatively clean.

Could always put a switchable -15dB output pad on the front, like this. Maybe even make the first position of the rotary switch a "line in" setting.
 

Attachments

  • 2-ch-w-EQ,-Front-Panel-Design-(Draft-09-merged-for-layout).jpg
    2-ch-w-EQ,-Front-Panel-Design-(Draft-09-merged-for-layout).jpg
    213.8 KB · Views: 32
Could always put a switchable -15dB output pad on the front, like this. Maybe even make the first position of the rotary switch a "line in" setting.
One of the very earliest Neve channel amps, the 1063 did exactly this. It used a single input transformer (the classic 10468 but in a plug in octal can) and via a 3 wafer switch altered the gain from 80dB to line level.

Cheers

ian
 
Seems like a separate -30dB (or so) PAD would be a better way to allow for line level. Wouldn't there be some risk of oversaturating the 1:10 input transformer?
Yes there is. With a 1:10 input transformer I use an input pad made of a couple of 3K3 resistors and a 150 ohm resistor which gives about 33dB attenuation and an input impedance of about 6K6. The Neve transformer ratio was only 1:2 so they could get away with a secondary only attenuator (and a very good transformer design).

Cheers

Ian
 
Back
Top