Hi,
I made a veeery long time ago an AMB M3 headphones amp, and incorporated inside the modified linkwitz crossfeed board offered at that time by TangentSoft.
I opted at that time to switch the crossfeed on/off and mid/high settings with a rotary switch three positions.
Time passed, and it seems this switch has become a tad rusty, introducing some distortion (audio signal had to come and go between the crossfeed board and the rotary switch).
A better solution would be to have the rotary switch pilot good quality relays close to the crossfeed board.
Now I never did this... The M3 has a 24VDC power, that will be used to actuate the relays. My Idea is to have one relay for ON/OFF crossfeed, being NC to ON and NO to OFF, and a second relay for mid/high crossfeed settings, being NC to mid and NO to High, so my usual listening position will be the no electricity required one 🙂.
But I have no clue which DPDT relay (OMRON or else) to use, nor the kind of active/passive components I need around the relays to have them working with the 24VDC power present inside the amp.
Thanks already for all your help 🙂.
Greg
I made a veeery long time ago an AMB M3 headphones amp, and incorporated inside the modified linkwitz crossfeed board offered at that time by TangentSoft.
I opted at that time to switch the crossfeed on/off and mid/high settings with a rotary switch three positions.
Time passed, and it seems this switch has become a tad rusty, introducing some distortion (audio signal had to come and go between the crossfeed board and the rotary switch).
A better solution would be to have the rotary switch pilot good quality relays close to the crossfeed board.
Now I never did this... The M3 has a 24VDC power, that will be used to actuate the relays. My Idea is to have one relay for ON/OFF crossfeed, being NC to ON and NO to OFF, and a second relay for mid/high crossfeed settings, being NC to mid and NO to High, so my usual listening position will be the no electricity required one 🙂.
But I have no clue which DPDT relay (OMRON or else) to use, nor the kind of active/passive components I need around the relays to have them working with the 24VDC power present inside the amp.
Thanks already for all your help 🙂.
Greg
Hi,
All I need truely is
If anyone knows this, I would much appreciate you hepl on this 🙂.
All the best,
Greg
All I need truely is
- good DPDT gold contacts 24VDC relays suggestions
- how to feed the 24VDC the amp is using to them in order
- not to pollute the amp power
- use the relays properly (capacitors, diode, whatever they need).
If anyone knows this, I would much appreciate you hepl on this 🙂.
All the best,
Greg
For the relays, I find this that could match:
Omron G6A-274P-ST-US. It exist in 24VDC, and has Ag-Au alloy contact and are fully sealed.
Now, I suppose I cannot simply put the supply 24V through the relay and the front switch... I need some components around the relay to regulate the flow of current, right?
Omron G6A-274P-ST-US. It exist in 24VDC, and has Ag-Au alloy contact and are fully sealed.
Now, I suppose I cannot simply put the supply 24V through the relay and the front switch... I need some components around the relay to regulate the flow of current, right?
Often a so-called freewheeling diode is put across the coil, connected such that it is in reverse when the coil is energized. Its purpose is to provide a path for the current through the coil when you switch off the relay. Relay manufacturers often prefer a diode with a Zener in anti-series, although that mainly matters for relays with a heavy electric load.
Theoretically, when you try to abruptly switch off the current through an ideal inductor, you get an infinite voltage peak. In real life, you get a finite but quite high voltage peak. When the relay is driven from a transistor, the voltage peak may kill the transistor. The freewheeling diode with or without Zener clamps the voltage to a safe value. Mechanical switches are more forgiving, but excessive arcing doesn't do their lifetime any good, so it may be a good idea to include the clamping circuit anyway.
The advantage of putting a Zener in anti-series with the diode is that the magnetic energy in the coil gets dissipated faster than with just a diode, so the relay switches off faster, reducing arcing in its own contacts - which probably doesn't matter anyway for line level audio signals.
Theoretically, when you try to abruptly switch off the current through an ideal inductor, you get an infinite voltage peak. In real life, you get a finite but quite high voltage peak. When the relay is driven from a transistor, the voltage peak may kill the transistor. The freewheeling diode with or without Zener clamps the voltage to a safe value. Mechanical switches are more forgiving, but excessive arcing doesn't do their lifetime any good, so it may be a good idea to include the clamping circuit anyway.
The advantage of putting a Zener in anti-series with the diode is that the magnetic energy in the coil gets dissipated faster than with just a diode, so the relay switches off faster, reducing arcing in its own contacts - which probably doesn't matter anyway for line level audio signals.
Such as the third solution here?:
But do I need the Q3? or can my rotary simply open/close the circuit driving the relay?
Any advice as for the diode and zener to use for 24VDC supply?
But do I need the Q3? or can my rotary simply open/close the circuit driving the relay?
Any advice as for the diode and zener to use for 24VDC supply?
Yes, that's what I meant. Your switch should be able to drive the relay without any Q3. The coil current is only 8.3 mA (according to the Digi-Key website), so a 1N4148 and a 24 V, 400 mW Zener should easily be able to handle it.
Maybe an 8.2 V Zener is better. I just checked some rotary switch datasheets and some have lower switching voltage ratings than I expected.
The current then flows through the diode and the Zener until they and the coil's wiring resistance have converted the magnetic energy in the coil into heat (which shouldn't take longer than a few milliseconds). When the magnetic field has become weak enough, the relay switches off.
But beware: The "best" solution has the disadvantage that it transmits an electrical impulse to the contacts due to the high reverse voltage. You could hear that.
With an 8.2 V Zener and 24 V supply, you get a 33 V voltage jump instead of 24.7 V. Some 3 dB more, not decades more.
I'm not sure I quite catch... The electrical impulse is transmitted to what contacts? Those of the audio signal (NC/NO/COMMON) or those of the rotary switch I'd use to switch on/off the relays? Or maybe to the ground / 24VDC power supply, that from there could creep back into the audio?
You will always have some capacitive coupling from the coil to the relay contacts. Not a lot, but some. With an 8.2 V Zener, that will lead to a 3 dB bigger disturbance in the audio than with just a freewheeling diode.
Last edited:
So, As I understand the thing, when closing the circuit, the coil will actuate the relay, connecting common to NO.
When opening the circuit, the coil will send back some energy through the diode and the zener will let it pass until the voltage drops below 8.2V. This drop and release of the coil can emit an impulse that can be heard through the common and NC terminals. These, being in my audio signal path (crossfeed filter), this could be heard.
Is this correct?
Second solution could be better at dissipating this energy? Or a cap somewhere in the parallel circuit with the relay? I don't mind the actuation of the relay to be somehow delayed in comparison to the rotary switch. Let the relay engage or disengage when it's ok for it, minimising this impulse. How one could achieve this? To have the voltage dropping slowly from 24 to 0?
When opening the circuit, the coil will send back some energy through the diode and the zener will let it pass until the voltage drops below 8.2V. This drop and release of the coil can emit an impulse that can be heard through the common and NC terminals. These, being in my audio signal path (crossfeed filter), this could be heard.
Is this correct?
Second solution could be better at dissipating this energy? Or a cap somewhere in the parallel circuit with the relay? I don't mind the actuation of the relay to be somehow delayed in comparison to the rotary switch. Let the relay engage or disengage when it's ok for it, minimising this impulse. How one could achieve this? To have the voltage dropping slowly from 24 to 0?
I doubt very much that any of it will be heard. I frequently switch line level signals with relays driven with freewheeling diodes with Zener diodes in antiseries and never had any issue.
If you should want to try slow relay switch-off, you would have to put an RC series network across its coil. The resistor has to be chosen such that the peak current through your rotary switch at turn on remains within the contact ratings of the rotary switch. The capacitor could be an electrolytic capacitor of 10 uF or so. Normally this is bad for the reliability of the relay contacts, but you will probably get away with it as long as the electric load is far below the relay contact rating. You could get temporary short circuits if the next relay turns on before the present one turns off.
If you should want to try slow relay switch-off, you would have to put an RC series network across its coil. The resistor has to be chosen such that the peak current through your rotary switch at turn on remains within the contact ratings of the rotary switch. The capacitor could be an electrolytic capacitor of 10 uF or so. Normally this is bad for the reliability of the relay contacts, but you will probably get away with it as long as the electric load is far below the relay contact rating. You could get temporary short circuits if the next relay turns on before the present one turns off.
Last edited:
Ok, not the best idea then for the relay contact life, even if line level is not critical.
Switching between the crossfeed settings is seldom used. It is sometimes insightful to go back and have a listen without just to gain back the idea it was properly doing it's job, or sometimes when the track is a true binaural recording that does need no treatments. So switching to off mode (relay 1 activated) is calmly done, then back to mid position (both relay off), then sometimes switching to high mode (relay 2 on), just to compare. So both relays should never be on at the same time.
With the zener and diod in antiserie, in parallel with the coil, is there an advantage to put the rotary switch before or after all this (close to the +24VDC, or close to the gnd?
Is there also a need to 'clamp' the entire circuit with diods on top and at the bottom, close to the power rails? These are used to feed the M3 amp, so They have to be kept as clean and stable as possible.
Switching between the crossfeed settings is seldom used. It is sometimes insightful to go back and have a listen without just to gain back the idea it was properly doing it's job, or sometimes when the track is a true binaural recording that does need no treatments. So switching to off mode (relay 1 activated) is calmly done, then back to mid position (both relay off), then sometimes switching to high mode (relay 2 on), just to compare. So both relays should never be on at the same time.
With the zener and diod in antiserie, in parallel with the coil, is there an advantage to put the rotary switch before or after all this (close to the +24VDC, or close to the gnd?
Is there also a need to 'clamp' the entire circuit with diods on top and at the bottom, close to the power rails? These are used to feed the M3 amp, so They have to be kept as clean and stable as possible.