Stepped Relay Ladder priority encoded temporary erroneous 3dB

Howdy,

Playing with a concept for a stepped relay ladder of resistors, that has native "priority encoding" in the topology for ladder/resistor values built into the relay ladder encoding. (e.g., the relay to the right in the ladder will defeat the relay/resistor value adjacent to the left)

Typically, only one relay at a time is actuating as a rotary switch is rotated (some 74HC logic in a remote location encodes/decodes coil voltage sequences for the ladder as the ladder is "wrapped" to take advantage of DPDT unused contacts)

There may be a point whence one relay is in mid-throw (not connected on any of the contacts), which defeats the priority functionality, to where two of the stepped ladder resistors may be in series...

Thusly, when switching from 5dB to 6dB there may be a short period where the total ladder resistance effects a 3dB output (could be worse I suspect).

The question is, a 3dB change in output level for about, say, 4~8 ms throw noticeable?  Will there be a noticeable "thud" or "click"?

... Area under the curve and human hearing...

This is for a mastering gain/attenuator (active) before hitting converters, so not a lot of gain after this proposed device..  Set it and forget it whilst bouncing... it's just the initial setup during playback that may be the concern/annoyance...  I would a*sume no time varying dynamics processors post this device but one never knows...

Thoughts?  Cheers!

Twenty Log said:

Howdy,

Playing with a concept for a stepped relay ladder of resistors, that has native "priority encoding" in the topology for ladder/resistor values built into the relay ladder encoding. (e.g., the relay to the right in the ladder will defeat the relay/resistor value adjacent to the left)

Typically, only one relay at a time is actuating as a rotary switch is rotated (some 74HC logic in a remote location encodes/decodes coil voltage sequences for the ladder as the ladder is "wrapped" to take advantage of DPDT unused contacts)

There may be a point whence one relay is in mid-throw (not connected on any of the contacts), which defeats the priority functionality, to where two of the stepped ladder resistors may be in series...

Thusly, when switching from 5dB to 6dB there may be a short period where the total ladder resistance effects a 3dB output (could be worse I suspect).

The question is, a 3dB change in output level for about, say, 4~8 ms throw noticeable?  Will there be a noticeable "thud" or "click"?

... Area under the curve and human hearing...

This is for a mastering gain/attenuator (active) before hitting converters, so not a lot of gain after this proposed device..  Set it and forget it whilst bouncing... it's just the initial setup during playback that may be the concern/annoyance...  I would a*sume no time varying dynamics processors post this device but one never knows...

Thoughts?  Cheers!

Click to expand...

A schemo would help here. I don't understand your concern about audibility since you mention that it's set and forget.
The 3dB hump may be audible, or not, depending on the nature of the program. On a pure sinewave it would be definitely audible, on standard musical program, it should not.

Rig up ONE relay, with a 3dB jump, and *try it* on many various programs.

Thanks for taking a look...

Enclosed is pic of relay ladder priority encoding (simplified). 

As can be seen by throwing relays, for example in the transitions from 5 to 6 dB and back again from 6dB to 5dB:

There are temporary periods of time that the resistor values of BOTH 5 and 6 dB positions (10k and 12k58) will be in series adding up to a total resistance that is pretty close to the 3dB resistance value (24k3). 

Relay switching time might be 4ms or so (assuming 8ms just in case different relays are used)...

Relays are thrown one at a time, but are quick to engage, and delayed to release (Hex Schottky inverters with slow down capacitors on the coil control voltage driver chips with unequal pullup and pulldown resistors -- on purpose)  Such that, for example, the 5dB could engage quickly while the 6dB relay disengages some short time later (and vice-versa).

This network can be used for both gain or attenuation topologies without changing resistor values shown (same ladder works for both in hierarchical upstream opamp topology not shown)... gain changes might not be bad to temporarily decline down to 3dB of gain whence in the 5~6dB range, but attenuation changes declining temporarily will mean more "louderer" for a short amount of time when jumping from 6dB of attenuation to 3dB of attenuation (imagine if we were talking 12dB desired with jumps downward?)...

My Best Thanks...  Thanks again for taking a look...

For the life of me, I don't understand your circuit. where's the input? Where's the output? What does this bus joining all relays for? What feeds the first (wrap around) relay?

Ah... Sorry... So the input and output is the "WHOLE_DB_A" and "WHOLE_DB_B" net labels.

The ultimate goal is to use the unused DPDT contacts as dB setting for the upper 7dB+ range using the "wrap around" relay but can be ignored for now.

The Whole dB "A" and "B" ports are destined to be the resistor values (for the "mystery" resistors R8 and R9, which are one in the same really) in the circuit described here: http://groupdiy.com/index.php?action=dlattach;topic=48402.0;attach=12935

The bus creates the "priority encoding" such that the relay to the right of any relay takes precedence to the relay to the left if there is more than one relay engaged at one time as relays transition (except of course when a relay is completely out of circuit during transition as is the original conundrum)...  But there will be times when two relays are engaged for a brief moment.

The nice thing is that the relay ladder only need to be instantiated once, and switched into either R8 or R9 positions (with another relay) to effect a gain or an attenuation as indicated by the resistor value selected, as such, the value of the relay ladder will provide the same value of gain or same value of attenuation in dB depending on its location as R8 or R9 in the schematic [in the above PDF].

Cheers!