rotary selector switch

[moamps]

Hi,

The best way to implement the buss routing would be to analize the pan circuitry (KM-PS in E510 sheet) which is permanently connected to the desired buss. I believe that duplicating that circuit (with or whitout the PAN function) and using 2-pole/3-position selector (break-before-make or make-before-break) instead of MUTE switch will do the job.

Regards,
Milan

P.S. I've got some modules from that console but without the shematics so I'm wondering if there's any way I can get them (schem.) from you?

 

[salomonander]

Hi,

The best way to implement the buss routing would be to analize the pan circuitry (KM-PS in E510 sheet) which is permanently connected to the desired buss. I believe that duplicating that circuit (with or whitout the PAN function) and using 2-pole/3-position selector (break-before-make or make-before-break) instead of MUTE switch will do the job.

Regards,
Milan

P.S. I've got some modules from that console but without the shematics so I'm wondering if there's any way I can get them (schem.) from you?

here is the pan module. but nothing happening there except resistors. they could have used a pot instead i guess.

sure, i can supply you shematics. just tell me what you need. most is in this thread anyways by now.
i think there is hardly a point in racking any of these modules (unless you get the proper transformers etc) - except the mic preamp that is. i have seen lots of people racking them. that said. i would happily buy the modules from you (especially channel modules) for i still have free slots in my desk.

 

[salomonander]

regarding the switching i should maybe just do some testing.
i remember using the console with only 5 modules inserted at some point. the levels might have been slightly off (im not sure it was in the beginning) but there certainly were no obvious problems (no humm, buzz etc).
and when leaving the channels out ofthe frame i essentially have the bus-feed resistors floating i guess (nothing connected to them).

would this be a legit test:
- insert one module to the frame and measure frequency response and level
- insert a few more modules and see if anything changes in the first channels behaviour

if there is no change i should be good with a simple 2x3 rotary right?
as i have said - i think i ran it with 5 modules once and had no issues.

 

[ruffrecords]

If it is passive mixing it will run fine with any number of modules. The problem is, the bus loss is different depending on how many channels are fitted. So if you fit one module and send it a test signal and set the output for 0VU, then add a couple more modules and run the test again, you will probably find the level is no longer at 0VU but somewhat lower. If that happens you definitely have passive mixing.

Cheers

Ian

 

[salomonander]

If it is passive mixing it will run fine with any number of modules. The problem is, the bus loss is different depending on how many channels are fitted. So if you fit one module and send it a test signal and set the output for 0VU, then add a couple more modules and run the test again, you will probably find the level is no longer at 0VU but somewhat lower. If that happens you definitely have passive mixing.

Cheers

Ian

Hey Ian

i just did some testing. and yes, you were spot on. the desk certainly uses passive summing. the more modules i insert, the more level drop is get. thank you so much for guiding me through this!
if you have an ingenious way of using a 3x3 rotary for three groups and still allow for grounding the bus feed resistors - let me know. otherwise i will use a 6 pole rotary switch so i can ground any unused bus feed resistor. maybe i try to solve this with relays...we'll see.

either way i guess i have some other things to do now. since i have to extend the bus wiring in order to allow all channels to be routed to any group and L/R, i will have to re-calibrate all group amps right?
or would i have to change the bus feed resistor values as well?
i guess not since auxes use 3.6k as well and these busses reach all channels now.

so my question:
- can i simply extend the bus wiring (using the same resistor values), re-calibrate the group busses, make sure all unused  bus feed resistors are grounded when unused and get on with my life ?

in case all this is fine:
- shorting vs non-shorting rotary doesn't matter in this application (what about clicks)?
- i should not switch the ground right? only audio signal?

thank you so much

 

[moamps]

Hi,

According to the attenuator schematic, it is clear that the buss resistors always see the source impedance, which varies between 0 ohm and 1 Kohm (mute channel) depending on the position of the pan switch. Consequently, the buss is not sensitive to such impedance variations so you don't need to connect unselected buss inputs to the ground.

If you want the channel to be fed into only one buss pair at a time, buss routing can be simply implemented as shown in the picture below.

However, if you wish to be able to feed the channel into more than one buss pair at the same time, then I'd recommend using (building) push buttons or toggle switches.

Thank you for publishing all those schematics. It would be great if you could complete the set with the schematics for the filter modules (KM-PA07 and KM-AT) as well.

Please contact me via PM about the modules you're interested in and I'll look around if I have any left over.

Regards,
Milan

 

[salomonander]

Hi,

According to the attenuator schematic, it is clear that the buss resistors always see the source impedance, which varies between 0 ohm and 1 Kohm (mute channel) depending on the position of the pan switch. Consequently, the buss is not sensitive to such impedance variations so you don't need to connect unselected buss inputs to the ground.

If you want the channel to be fed into only one buss pair at a time, buss routing can be simply implemented as shown in the picture below.

However, if you wish to be able to feed the channel into more than one buss pair at the same time, then I'd recommend using (building) push buttons or toggle switches.

Thank you for publishing all those schematics. It would be great if you could complete the set with the schematics for the filter modules (KM-PA07 and KM-AT) as well.

Please contact me via PM about the modules you're interested in and I'll look around if I have any left over.

Regards,
Milan

hey Milan

are you sure? because i just tried it and i do get volume drops if not all channels are in the frame. so basically my groups gain would always vary depending on how many channels are routed to them. so when any bus feed resistor is ungrounded the volumes change. when i mute any of the inserted channels nothing happens since the mute switch pulls the bus feed resistor to ground (i guess )

thanks

ps: ill upload the shematics in a second

 

[salomonander]

Here are the shematics for the filter section. all these were given to me my H&E studiotechnik (thanks again Andy):

 

[salomonander]

treble and bass

 

[salomonander]

power regulator

 

[salomonander]

aux modules

 

[salomonander]

overload display

 

[salomonander]

and finally the inputs. this should cover an entire channel. if anyone needs shematics for busses shoot me a pm

 

[moamps]

The insertion of each additional module may cause a slight decrease in buss volume. However, once a module has been inserted, it does not make much difference whether the non-routed buss pan switches are connected to the ground (via the buss selection switch) or not.
The picture below shows a buss selection switch design that allows the non-routed buss outputs to be connected to the ground, using a 3-pole switch.

Regards,
Milan

 

[ruffrecords]

so my question:
- can i simply extend the bus wiring (using the same resistor values), re-calibrate the group busses, make sure all unused  bus feed resistors are grounded when unused and get on with my life ?

Basically yes.

in case all this is fine:
- shorting vs non-shorting rotary doesn't matter in this application (what about clicks)?
- i should not switch the ground right? only audio signal?

thank you so much

Shorting switches are only really necessary where you are switching capacitors that may be charged and hence cause clicks. This is not the case here so no-shorting should be OK.

Yes, you should switch the signal only.


And in answer to moamps point about the source resistance via the pan pot he is not quite correct. Each half of the pan pot is as he says just under 1K but, since the pot is fed from a very low source impedance (the output of an amplifier with negative feedback) the worst case apparent source impedance of the pan pot is one quarter of this value or approximately 250 ohms (at the half way point with about 500 ohms in parallel with 500 ohms). You can now see that the bus resistors are more than 10 times this value which is necessary for passive mixing channels not to audibly interact with each other. Therefore the bus resistors NOT routed to the pan pot MUST be connected to 0V to maintain the bus impedance.

I am not 100% clear about how you intend the routing to work with your three position switch. Is the intention to switch the output of the pan pot to either groups1/2 OR groups 3/4 OR groups L/R. If this is the case then you will need a 6 pole switch which, as moaps suggests, might be easier to implement with mechanically latched push button switches.

Cheers

Ian

 

[moamps]

And in answer to moamps point about the source resistance via the pan pot he is not quite correct. Each half of the pan pot is as he says just under 1K but, since the pot is fed from a very low source impedance (the output of an amplifier with negative feedback) the worst case apparent source impedance of the pan pot is one quarter of this value or approximately 250 ohms (at the half way point with about 500 ohms in parallel with 500 ohms). You can now see that the bus resistors are more than 10 times this value which is necessary for passive mixing channels not to audibly interact with each other. Therefore the bus resistors NOT routed to the pan pot MUST be connected to 0V to maintain the bus impedance.

Actually, what I said was “it does not make much difference” because even in the worst case scenario the difference in impedance would be too small to have any significant effect on summing. The interaction between channels is negligible here because, statistically speaking, buss is always loaded with (3.6K+Rpan/2)/N . Besides, the non-selected channels are anyway disconnected from the buss.

Regards,
Milan

 

[salomonander]

so my question:
- can i simply extend the bus wiring (using the same resistor values), re-calibrate the group busses, make sure all unused  bus feed resistors are grounded when unused and get on with my life ?

Basically yes.

in case all this is fine:
- shorting vs non-shorting rotary doesn't matter in this application (what about clicks)?
- i should not switch the ground right? only audio signal?

thank you so much

Shorting switches are only really necessary where you are switching capacitors that may be charged and hence cause clicks. This is not the case here so no-shorting should be OK.

Yes, you should switch the signal only.


And in answer to moamps point about the source resistance via the pan pot he is not quite correct. Each half of the pan pot is as he says just under 1K but, since the pot is fed from a very low source impedance (the output of an amplifier with negative feedback) the worst case apparent source impedance of the pan pot is one quarter of this value or approximately 250 ohms (at the half way point with about 500 ohms in parallel with 500 ohms). You can now see that the bus resistors are more than 10 times this value which is necessary for passive mixing channels not to audibly interact with each other. Therefore the bus resistors NOT routed to the pan pot MUST be connected to 0V to maintain the bus impedance.

I am not 100% clear about how you intend the routing to work with your three position switch. Is the intention to switch the output of the pan pot to either groups1/2 OR groups 3/4 OR groups L/R. If this is the case then you will need a 6 pole switch which, as moaps suggests, might be easier to implement with mechanically latched push button switches.

Cheers

Ian

hey Ian

great! yes, i want the group selector switches to be able to route the channels to groups 1/2, 3/4 or L/R (only one at a time obviously). so i do need a 6 pole 3 position switch. OR:
i take a totally different approach and use a single pole 3 postion switch that triggers relays. i think this might be the best solution since i planned on fitting a mute relay in every channel anyways.
now if i'd fit three relays (one dpdt relay for each bus output) that by default ground the bus feed resistors, i guess i could open one at a time with a logic input (or 0v) coming from the 1pole rotary switch. if i now have the mute button cut the rotary's supply line id have mute and bus selection covered in one. what do you say?

pushbuttons are no option. i do have some space at the back of the channel modules but hardly any room towards the frontplates. i can literally fit one rotary switch. thats it im afraid. there was some room - but i used it up for eq bypass, insert bypass and rotary switches for the shelving eq's frequency points.

 

[ruffrecords]

hey Ian

great! yes, i want the group selector switches to be able to route the channels to groups 1/2, 3/4 or L/R (only one at a time obviously). so i do need a 6 pole 3 position switch. OR:
i take a totally different approach and use a single pole 3 postion switch that triggers relays. i think this might be the best solution since i planned on fitting a mute relay in every channel anyways.
now if i'd fit three relays (one dpdt relay for each bus output) that by default ground the bus feed resistors, i guess i could open one at a time with a logic input (or 0v) coming from the 1pole rotary switch. if i now have the mute button cut the rotary's supply line id have mute and bus selection covered in one. what do you say?

Yes, I think that i8s a good solution that will work well.

Cheers

Ian

 

[ruffrecords]

And in answer to moamps point about the source resistance via the pan pot he is not quite correct. Each half of the pan pot is as he says just under 1K but, since the pot is fed from a very low source impedance (the output of an amplifier with negative feedback) the worst case apparent source impedance of the pan pot is one quarter of this value or approximately 250 ohms (at the half way point with about 500 ohms in parallel with 500 ohms). You can now see that the bus resistors are more than 10 times this value which is necessary for passive mixing channels not to audibly interact with each other. Therefore the bus resistors NOT routed to the pan pot MUST be connected to 0V to maintain the bus impedance.

Actually, what I said was “it does not make much difference” because even in the worst case scenario the difference in impedance would be too small to have any significant effect on summing. The interaction between channels is negligible here because, statistically speaking, buss is always loaded with (3.6K+Rpan/2)/N . Besides, the non-selected channels are anyway disconnected from the buss.

Regards,
Milan

What you actually said was:

According to the attenuator schematic, it is clear that the buss resistors always see the source impedance, which varies between 0 ohm and 1 Kohm (mute channel) depending on the position of the pan switch. Consequently, the buss is not sensitive to such impedance variations so you don't need to connect unselected buss inputs to the ground.

It is the phrase "Consequently the buss is not sensitive to such impedance variations' that I misunderstood. I think we were effectively saying the same thing.

I do not understand your statement above:

Besides, the non-selected channels are anyway disconnected from the buss.

By this do you mean channels that are not plugged in?

Cheers

Ian

 

[moamps]

Hi Ian,

My point in this discussion was to underline the peculiarity of this particular buss routing solution. The select/deselect (mute) switch is not directly connected to the buss summing resistor but to the input of the pan attenuator, which has a relatively low impedance.

In this particular case, this means that the switch does not necessarily need to connect the unselected channel to the ground (contrary to most passive summing mixers).

One could even argue that the grounding of the input of the pan attenuator via the channel ground may be counterproductive because it may induce some ground loop issues. If you take a closer look at the schematic, you’ll see that the summing ground is separated from the channel ground (pictures in replies nos. 12 & 13).

It follows that if the input of the pan attenuator is not connected to the ground when the channel is inactive, the channel remains unaffected because it is disconnected from the buss. Also, the unwanted change in the buss load is, in my opinion, insignificant (3.6k +500 ohm vs. 3.6k+250 ohm).

The one thing that may be subject to speculation is arguably higher crosstalk between the busses when the input of the pan attenuator is not connected to the ground.

Given these design features, I presented 2 possible solutions for routing, using a 1-pole switch for the non-grounded version and a 3-pole switch for the grounded version. I hope things are clearer now.

I'd like to ask moderators to copy these useful schematics that salomonander posted here to the Library, if that’s not too much trouble. TIA.

Regards,
Milan