Do the resistor values in the pattern switch matter?

[trans4funks1]

Do the resistor values of the pattern switch matter?

I understand that you use an even number of resistors and an odd number of switch positions so that you may divide the voltage.

How do you figure out what the total resistance of all of them in series should be so that you can divide and figure what each should be?

Thanks.

 

[RuudNL]

I have used 8 x 470 K in my C12 pattern switch.

 

[trans4funks1]

Thanks for making your comment.

I have a bunch of schematics and can see many examples that have been used. I am hoping to better understand the idea behind the choice. I imagine there must be some specific basic electronics math that makes it easy to understand so I am wondering if someone can explain it.

 

[0dbfs]

I would expect a similar total value to what the divider network feeding the backplate has (including the hi-z resistor/caps). That would at least make sense to me since "it works" in that position and provides similar bias voltage to the capsule configuration.

I would like to better understand the "why" or "why-not".

Cheers,
jonathan

 

[dmp]

They are a simple voltage divider to set the polarization voltages at the different steps, with a high total value to keep a low current from flowing through them. If you sum them up, and use ohm's law, you can calculate how much current for some typical designs. There's downsides if it is too high or too low.

 

[trans4funks1]

I think I am starting to realize that the total value of the series of resistors in the pattern switch is used to maintain a voltage supply that is twice the voltage that is being supplied to the back plate via the B+.

I guess the idea is to work along the B+ supply to the back plate and determine the voltage at the back plate and then calculate a value that will only drop the B+ to the point which it equals twice that which the back plate is seeing. That value will be the total value of the series of resistors plus any other resistors that are in series between the pattern switch and the rear capsule.

Does that idea seem like I am on the right track?


Thanks!

 

[Matador]

Think in the extremes:  use nine, 1 ohm resistors.  The low resistance would clobber the B+ voltage (120V across 9 ohms is over 10 amps!).  But it would work if B+ could supply the current.

Now go the other way:  nine, 1G resistors.  Now there is almost no current, so no loading problems.  But the output of the switch need to charge the filter cap:  1G into 1uF has time constant of many days...nobody wants to wait weeks for the pattern to change on the mike!  But again, in principle it works.

So we pick something in the middle:  a few MOhm doesn't load down a passive B+ supply too much, and gives a reasonable 10-ish second response time to pattern changes.

 

[trans4funks1]

Thanks Matador.