CMOS switches / distortion, noise

[Rochey]

I looked at the Toshiba transistor they specc'd. It showed a typical VCE(sat) of 0.04V.
Which is quite a bit lower than the PN2222A's 0.3V!

/R

 

[JohnRoberts]

Vce sat at what collector current, and at what base current?

Not to mention the collector current and base drive are both constantly changing with signal voltage swing.

You can improve the distortion somewhat by driving the base with a current source instead of base resistor, but for this level of performance why bother?

JR

 

[Rochey]

John, I'm going to make you look at that schem again...

the base is driven by a constant voltage through a fixed resistor.

However, your right regarding the collector current. However, I thoguht that VCEsat was set by the base current?

/R

 

[jdbakker]

the base is driven by a constant voltage through a fixed resistor.

Look again, and consider what happens to the voltage across the base resistor (and thus the base current) when the opamps' outputs swing up and down.

However, your right regarding the collector current. However, I thoguht that VCEsat was set by the base current?

Base and collector current, amongst other things.

JD 'not JR' B.
[somewhat concerned about BE-junction reverse breakdown, but I'm sure Uli thought of that]

 

[Rochey]

Doh! Emmiter is all over the place!

 

[JohnRoberts]

John, I'm going to make you look at that schem again...

nah

the base is driven by a constant voltage through a fixed resistor.

nah

However, your right regarding the collector current. However, I thoguht that VCEsat was set by the base current?

/R

A little more complicated... just like life...

JR

PS: JDB... I think there are transistors made with higher base emitter reverse zener breakdown voltage, but why worry about a little extra distortion?

 

[clintrubber]

Looks like it's performing a gain switch? (switching a resistor into circuit).

Yep, it's the +4/-10 switch on the tape-returns, one switch for each 8 channels.

So if possible w.r.t. levels, preference for the -10 position, because it avoids the BJT in-closed-position.


At the end of the day no big problems here though, many people nowadays use desks like these mainly for monitoring.

Anybody happens to know how other brands implement this function ? Not unlikely that others do it cheap'n'nasty as well.

 

[clintrubber]

Ewww  you tricked me into looking at a Behringer schematic...  :-[

Please accept my apologies    8) ... wasn't my intention to derail this thread from improving stuff to
how not to...

Bye

 

[abbey road d enfer]

I've used a lot of DG308/411 with success, the latter being much better than the former. They are at least an order of magnitude better than the venerable 40xx.

They're also an order of magnitude more expensive, so that about evens out.

No, because there's nothing to even out. You want top performance, you pay the price. That's like saying a carbon mic is suitable for today's music production purpose, because it's cheaper than a Neumann.

 

[JohnRoberts]

I've used a lot of DG308/411 with success, the latter being much better than the former. They are at least an order of magnitude better than the venerable 40xx.

They're also an order of magnitude more expensive, so that about evens out.

No, because there's nothing to even out. You want top performance, you pay the price. That's like saying a carbon mic is suitable for today's music production purpose, because it's cheaper than a Neumann.

My read of his comments is that the cost and complexity of wrapping a feedback loop around the switches to get unmeasureable distortion, balances out with the performance difference.  He is in effect saying you get what you pay for, or more correctly pay for what you get.  Not sure what this has to do with carbon mics, unless you have some complex expensive way to make them not suck.

People i know still designing analog consoles use the new better switches. I used my complicated design approach to make them not suck back in the late '70s. Today PCB real estate and individual component placements make the complexity trade off even more costly.

JR