Phase Switches and Pads

[adamasd]

Is there a good reason that phase switches tend to be on the input transformer instead of the output transformer? Doing it after the amplification would make it less prone to picking up noise would it not? At least keep any noise picked up from being amplified. I knda feel that there is an issue with the low impedance of the output that causes the problem here, but I don't really know.

Thanks,
adam

 

[emrr]

If I use them at all (rarely), I always put them on the output to keep the low level input signal switching at minimum. No good reason I can think of for having them at input, other than convenience in the various circuit board input stage kits.

 

[adamasd]

Good to know. Personally I have no intention of using a phase switch, but curiosity made me wonder why it was always stuck on the front. If I decide I need one I can put it on the output without fear and be happy.

Thanks

 

[Samuel Groner]

Depending on output stage topology there are reasons why you don't want it at the output as running the output unbalanced may not work well (or at all) with flipped polarity.

Samuel

 

[JohnRoberts]

In consoles it is routinely done at the input, usually in combination with pad circuitry if used.

The reason for correcting polarity up front is so signals routed from insert points will be consistent with outputs or other internal routing.

In a standalone preamp it's pretty much academic whether you manage polarity up front or at the tail end. I see no compelling performance advantage to either other than that output side switching is less susceptible to noise pickup.

JR

 

[emrr]

[quote author="JohnRoberts"]

In a standalone preamp it's pretty much academic whether you manage polarity up front or at the tail end. I see no compelling performance advantage to either other than that output side switching is less susceptible to noise pickup.

JR[/quote]

My point and context exactly.

 

[adamasd]

The reason for correcting polarity up front is so signals routed from insert points will be consistent with outputs or other internal routing.

I kinda suspected it would have something to do with a habit carried over from somewhere else in the field, but most of these things are.

Depending on output stage topology there are reasons why you don't want it at the output as running the output unbalanced may not work well (or at all) with flipped polarity.

What topologies can you switch on the output transformer and cause problems? I can not think of any, but I have been focusing on just a couple topologies for my learning.

thanks for clearing this all up for me

 

[JohnRoberts]

The only scenario that I can think of where polarity swap at output might be problematic is the case of a say a single ended output followed by a single ended input. If the one hot output line gets routed over to the input that isn't being looked at.

I am still inclined to manage polarity at the input but concede it out of habit. I'm not sure what stand alone mic preamps are good for (blasphemy alert). I think its blatant prejudice that any time there are more than two preamps in a box they are somehow inferior.

JR

 

[adamasd]

I'm not sure what stand alone mic preamps are good for (blasphemy alert). I think its blatant prejudice that any time there are more than two preamps in a box they are somehow inferior.

For me its just a habit carried over from all the time I used to spend playing with synths. A modular design with a fair amount of variety is just more natural for me. It is also easier to make changes in the future or move the thing. And you know, I just do not need or want a full console.

I suppose putting the pad on the primary of the input is also habbit versus putting it on the secondary after the step up. Although I can see some advantages of having it on the primary if you want try and keep the sound of the transformer minimal.

 

[emrr]

[quote author="JohnRoberts"]
I'm not sure what stand alone mic preamps are good for (blasphemy alert).
JR[/quote]

For me it's about color choices and the luxury of playing with different inputs. I own a mixer that has preamps I don't like; I use it for mixing. I do remote work and carry exactly what I need for inputs, and monitor with a 16 input 1RU line mixer. It's also about the right amount of gain for the job, and the right mix of mics with preamps that combine in ways I like. So, in that context, I can't think of any reason to switch phase at a low level. Especially given that most all of my gear is transformer coupled on both sides or has output stages that drive both sides of the line. In the case of a console where polarity is almost never in need of flipping or a preamp with an unbalanced out, I usually argue in favor of inline polarity barrels patched as needed as opposed to a switch in the signal path that virtually never gets used. These installed switches work fine for long periods of time, but do oxidize, get dirty and intermittent at some point. I'd rather these effects happen with big signals rather than small ones that are about to see big gain.

 

[JohnRoberts]

[quote author="adamasd"]

I suppose putting the pad on the primary of the input is also habbit versus putting it on the secondary after the step up. Although I can see some advantages of having it on the primary if you want try and keep the sound of the transformer minimal.[/quote]

Not habit, pad should absolutely be in the front end.

My apologies about the stand alone pre comment, purely rhetorical.

JR

 

[adamasd]

Not habit, pad should absolutely be in the front end.

What is the difference of putting the pad on the secondary of the input transformer instead of the primary? Other then one way not driving the transformer as much.

Or did you misunderstand me and think I was still talking about before the output transformer/after the active gain?

 

[JohnRoberts]

[quote author="adamasd"]

Not habit, pad should absolutely be in the front end.

What is the difference of putting the pad on the secondary of the input transformer instead of the primary? Other then one way not driving the transformer as much.

Or did you misunderstand me and think I was still talking about before the output transformer/after the active gain?[/quote]

Transformers can saturate from too much level, a pad after the transformer would not offer much benefit.

JR

 

[adamasd]

So it depend on what your input transformer can take and personal preference as to if you like the sound of the transformer or want as clean of a sound as possible.

It looks like Hamptone likes to put them on the secondary, at least in his published designs. I think I will have to experiment with this one abit.

Thanks,
adam

 

[JohnRoberts]

I'm from the distortion is bad camp...

JR

 

[emrr]

High frequency losses add up faster with pads on high-Z tube type secondaries of transformers than they do on low-Z primaries.

 

[adamasd]

I'm from the distortion is bad camp...

Ah, I am from the distortion has its uses camp.

High frequency losses add up faster with pads on high-Z tube type secondaries of transformers than they do on low-Z primaries.

I have discovered that, I am currently building a pre around a 1731 and I split the load resistor to get a -12db pad. There is a slight bit of high frequency loss when the pad is on, but not enough to worry about. It may not be as flat padded, but it still sounds good. I did not notice any high frequency loss when I tried the pad on the primary, but to be fair I was not looking for it either. My next project is going to be to design a tube pre, so I have jotted that down with my ideas for it.

thanks

 

[Samuel Groner]

The only scenario that I can think of where polarity swap at output might be problematic is the case of a say a single ended output followed by a single ended input. If the one hot output line gets routed over to the input that isn't being looked at.

That's not the only one: if you have a single ended driver with a transformer and you use a single-ended receiver the driver sees--depending on transformer type--a pretty hefty capacitive load which usually results in much higher distortion.

Samuel

 

[adamasd]

That's not the only one: if you have a single ended driver with a transformer and you use a single-ended receiver the driver sees--depending on transformer type--a pretty hefty capacitive load which usually results in much higher distortion.

I must be missing something, if you reversed the primary of the output transformer, or the secondary for that matter, how would that change anything other then the phase? Maybe I just do not know what this type of transformer you are hinting about?

 

[JohnRoberts]

[quote author="Samuel Groner"]
That's not the only one: if you have a single ended driver with a transformer and you use a single-ended receiver the driver sees--depending on transformer type--a pretty hefty capacitive load which usually results in much higher distortion.

Samuel[/quote]

I'm not sure I follow. There may be subtle inter winding capacitance differences with one end of the output winding shorted or the other, but should not be a distortion issue with any adequate driver IMO.

A center tapped output with one leg shorted would complain in either polarity setting.


[quote author="adamasd"]

Ah, I am from the distortion has its uses camp.
[/quote]

Having designed lots of gear, my design philosophy is to make individual paths or blocks arbitrarily flat and linear. If you desire some euphonious non-linearity, put that in a separate black box with properly labeled bypass switch.

It is easy to fall into the trap of designing an individual circuit block to complement some other flawed block in the same audio path. Microphones are notoriously varied wrt response and sound character so designing a pre by ear is unlikely to land in the middle of the range by chance. Perhaps OK for DIY personal use, but when addressing the wider market, I prefer the errors to reside in some other block.

This philosophy is also influenced by the intent of the circuitry. I only designed one class of equipment (audio delay) that was arguably an effect. The studio delay/flanger was unquestionably a effect. In those circuit blocks while there are objective measurements, the subjective result trumps the numbers.

JR