Hi all 🙂
I've got some holiday time coming up, which means it's time to sling some solder. One of the projects on my list is a headphone amplifier selector.
"Huh? Why would you want that?" some might ask. Purely out of curiosity, of course 😀. I've successfully built a number of headphone amps, ranging in complexity, and I'd like to be able to volume-level them and then A/B between them and listen for differences. Sure, I can easily plug and unplug my headphones, but human auditory memory is brief, and I'd rather not put undue wear on connectors.
Since we're not dealing with huge amounts of power, this shouldn't be too tricky to do, but I'd be grateful for any input.
Is a quality toggle switch good enough? Are mechanical relays a better option? Or should I look into solid-state / photoMOS relays?
How worried should I be about what this might do to the amplifiers? How much of a a delay should there be between disconnecting one amp and connecting the other? Obviously, if somehow both became connected at the same time that would be a Bad Thing.
Am I missing something important?
Thanks!
I've got some holiday time coming up, which means it's time to sling some solder. One of the projects on my list is a headphone amplifier selector.
"Huh? Why would you want that?" some might ask. Purely out of curiosity, of course 😀. I've successfully built a number of headphone amps, ranging in complexity, and I'd like to be able to volume-level them and then A/B between them and listen for differences. Sure, I can easily plug and unplug my headphones, but human auditory memory is brief, and I'd rather not put undue wear on connectors.
Since we're not dealing with huge amounts of power, this shouldn't be too tricky to do, but I'd be grateful for any input.
Is a quality toggle switch good enough? Are mechanical relays a better option? Or should I look into solid-state / photoMOS relays?
How worried should I be about what this might do to the amplifiers? How much of a a delay should there be between disconnecting one amp and connecting the other? Obviously, if somehow both became connected at the same time that would be a Bad Thing.
Am I missing something important?
Thanks!
Afraid I have to disagree on that, I would use a ssr every time. We routinely see ssr's used for high power speaker duty although admittedly these are discrete in nature (2 FET's and a photovoltaic coupler in their simplest form).
There is a ssr available in 8 pin DIP outline using the above technique but the part number evades me. Something like that would be ideal imo.
As to the problem of having two amps connected at once... just use a simple logic arrangement to drive the relays (of whatever flavour you choose). A common CMOS flip flop would be fine. A few uS overlap isn't going to upset anything, so no need for 'dead time' between the switching.
There is a ssr available in 8 pin DIP outline using the above technique but the part number evades me. Something like that would be ideal imo.
As to the problem of having two amps connected at once... just use a simple logic arrangement to drive the relays (of whatever flavour you choose). A common CMOS flip flop would be fine. A few uS overlap isn't going to upset anything, so no need for 'dead time' between the switching.
I would be scared about; it will add nonlinearities, or extra DCR in the path.
Or even crosstalk when switched off
no?
Or even crosstalk when switched off
no?
SSR's are linear like a resistor, just look at the graphs....
They have a specified DCR, just like a switch or relay, but they can be very low, and they don't get worn out or make noise....
The only thing you need to pay attention to is their off capacitance, which only matters when they're off, assuming they're in series....
Yes, the DC resistance can be very low indeed as soekris mentions. The FET's I use are actually quite modest (IRF2907) but have an on resistance of just 3.5 mΩ.
The capacitance (causing audio break through) could in theory be something to think about, although the impedances involved in backfeeding into amplifier output stages probably render it a non issue.
Its ssr every time for me.
The capacitance (causing audio break through) could in theory be something to think about, although the impedances involved in backfeeding into amplifier output stages probably render it a non issue.
Its ssr every time for me.
Thank you for the input 😀.
Living where I do, the availability of parts is somewhat limited. I did some digging on the RS website, but the vast majority of their DIP package SSRs have a current limit of 150mA or lower, which may be insufficient for some amplifiers.
I did find this one: http://docs-europe.electrocomponents.com/webdocs/133a/0900766b8133abda.pdf . It handles up to 1A. It has a turn on resistance of around 250mΩ though - will this be a problem?
Alternatively, I'm also happy to build a discrete SSR. Mooly, what sort of setup do you use?
Living where I do, the availability of parts is somewhat limited. I did some digging on the RS website, but the vast majority of their DIP package SSRs have a current limit of 150mA or lower, which may be insufficient for some amplifiers.
I did find this one: http://docs-europe.electrocomponents.com/webdocs/133a/0900766b8133abda.pdf . It handles up to 1A. It has a turn on resistance of around 250mΩ though - will this be a problem?
Alternatively, I'm also happy to build a discrete SSR. Mooly, what sort of setup do you use?
The discrete arrangement is pretty much identical to the device in your link. If you look carefully at the diagram of the FET in the IC package you will see it has two 'drains' which mirror the discrete connections.
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2618977
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2659578
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2660474
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2661555
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2618977
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2659578
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2660474
http://www.diyaudio.com/forums/solid-state/191449-output-relays.html#post2661555
Okay, so I read through that thread, and a couple others as well, and the optocoupler-MOSFET solution seems the simplest and best suited. Unfortunately, the photovoltaic driver mentioned is not available here, nor is the Vishay equivalent (VOM1271). But I think I've found a good alternative: Avago HCNR200.
It has the added benefit of being able to drive 2 channels from a single input. Would this be suitable?
As for the MOSFET, I imagine the part isn't all that important as long as its drain-source resistance is as low as possible, right?
Here is a quick schematic, let me know if I've done anything dumb. I've left out the additional protection diodes for clarity.
It has the added benefit of being able to drive 2 channels from a single input. Would this be suitable?
As for the MOSFET, I imagine the part isn't all that important as long as its drain-source resistance is as low as possible, right?
Here is a quick schematic, let me know if I've done anything dumb. I've left out the additional protection diodes for clarity.
Attachments
Last edited:
Your circuits fine but the opto coupler isn't.
What you need is a 'photovoltaic coupler', not an opto isolator. The opto isolator has been around for decades and is simply a light sensitive transistor illuminated by an infra red LED. It doesn't generate voltage. The photovoltaic coupler is like a tiny solar cell illuminated by an LED and as such it generates a true floating voltage of around 7 volts per cell.
The only one I could quickly find was this from RS. Look at the internal diagram of the device.
The FET's should ideally have as low Rds as possible and also a voltage rating equal to at least 50% (plus a safety margin) of the amplifiers total supply voltage. In a direct coupled amp with the load connected to ground the FET's can only see one rail or the other as a worst case condition and not both added together.
What you need is a 'photovoltaic coupler', not an opto isolator. The opto isolator has been around for decades and is simply a light sensitive transistor illuminated by an infra red LED. It doesn't generate voltage. The photovoltaic coupler is like a tiny solar cell illuminated by an LED and as such it generates a true floating voltage of around 7 volts per cell.
The only one I could quickly find was this from RS. Look at the internal diagram of the device.
The FET's should ideally have as low Rds as possible and also a voltage rating equal to at least 50% (plus a safety margin) of the amplifiers total supply voltage. In a direct coupled amp with the load connected to ground the FET's can only see one rail or the other as a worst case condition and not both added together.
- Status
- Not open for further replies.