Amp Channel switching (title edited)

[H.Manback]

There is something happening on my breadboard that I do not quite understand...

I am in the process of designing and prototyping a channel switching circuit for my p1 extreme. Now I wanted to do this with optocouplers and some digital switching circuit, and everything seems to be going as planned, the digital switching part is working, but now I don't quite understand how my optocoupler can be working correctly

I am kind of beginning to question myself because I think this shouldn't be possible. I am using a 4n25 optocoupler, which is a led and phototransistor type optocoupler. I tried to test the circuit by making a sort of mute box on a breadboard. The signal comes in and I just connect the tip and ground to the optocoupler transistors collector and emitter. Then I put a current through the LED (about 25mA) and I connected the output to the amplifier. I have a prototyping arrangment which has a bypass switch so I can bypass the circuit and feed the signal straight in to the amplifier. I hope that explanation is clear.

Now I don't hear a difference between the signal through the optocoupler and bypassed (except for a little bit of noise because the whole thing is on a breadboard). I was having doubts about this arrangement because an AC signal is flowing through the transistor, and it isn't biased. I connected the thing to my pc's microphone in and checked it with a pc oscilloscope, nothing strange, where I would expect something strange.

My question is, am I wrong in thinking current can only flow one way through the collector - emitter path? I think I am confused because of the 'odd' use of a transistor in this case. . Writing this down makes me think I am getting it mixed up with the base - emitter junction which is a one way path...

Also, one more thing, is a 4n25 a suitable choice for switching in this case? I want to switch the input and output of the preamp, so I want low noise there.

[R.G.]

Bipolars can actually work as AC switches for very small signals, turning on and off by base current.

[The Tone God]

I tried to test the circuit by making a sort of mute box on a breadboard. The signal comes in and I just connect the tip and ground to the optocoupler transistors collector and emitter.

I may be misunderstanding your wiring but have you tried the emmiter to ground and you signal on the collector ? Basically using the trasistior as a "kill switch".

Also, one more thing, is a 4n25 a suitable choice for switching in this case? I want to switch the input and output of the preamp, so I want low noise there.

4N25s seem to be one of those old standby parts that old techs/hams seem always turn too despite progress being made in this field. There are better optos out there but for the sake of testing it should still be ok but you might want to look at others for your final.

Andrew

[H.Manback]

@R.G.:  When you say very small signals, I don't suppose the signal going to the power amplifier falls in that category? If so I guess I should be looking for optocouplers using LDR or a solid state relay instead?

@The Tone God: That's how it is wired yes. Emitter to ground and collector to signal, although changing the two shouldn't make a difference I guess.

That 4N25 was what my usual electronics shop gave me when I asked for a plain optocoupler, so you hit the nail on the head there , they're tube techs from back in the days.

How should I go about finding a better one, what qualities should I be looking for?


Edit: I've been looking at SSRs (solid state relay) that I was considering earlier, and I think that might be a good idea. Only trouble is how I am going to get them, but if I can't get them any other way I guess Mouser is my best bet. Here's the datasheet on what I am looking at NEC PS7142-1A

[Paul Perry (Frostwave)]

One problem with the 4N25, it isn't designed for switching audio signals. And (although I don't know for sure) looking at the data for that solid state relay, I don't see any references for use with small audio signals. Sure, it would be great for robot motors, but for us?? maybe not!

[H.Manback]

I see your point, some other SSR I saw did mention that it is suitable for audio frequencies, and it can switch signals from nano volts to high voltages. Problem was, I couldn't find that one in Mouser's catalog... Here's that datasheet: Siemens SSRs. Something like the LH1517 in that sheet looks like it's a good one. Very low on-resistance, but it does have a switch offset of 0.1 V. The LH1516 might be better since it has 0.1 uV of switch offset..

[The Tone God]

@The Tone God: That's how it is wired yes. Emitter to ground and collector to signal, although changing the two shouldn't make a difference I guess.

Ok. 25ma seems alot for an opto so run a diode to make sure you didn't kill it. If it passes see if you can switch an LED on off with the opto's output. Confirm that the part is working.

How should I go about finding a better one, what qualities should I be looking for?

For an application like this I would look at the specs and find something with a high CTR (Current Transfer Ratio) among other things. Then I look at price. The 4N25 is kinda stinky in the area from what I remember.

There still will be a point where if the signal is big enough that the opto will not be able to ground the signal out completely that is if the opto ever could ground the signal out under normal condition.

Oh and I wouldn't use a SSR. The specs seem nice but they are not intended for audio. I would sooner use a low on-resistance analog solid state switch from say Maxim or the like. You may still need to deal with proper signal conditioning situations with those.

Andrew

[H.Manback]

Ah ok, well I don't believe the optocoupler is fried just yet... , it did switch ok, and turned the audio off. I checked the datasheet before I determined the LED current, and since it said at absolute maximum ratings that the maximum LED current was 60mA, I figured 25 would be safe. Also I checked the sheet again now, and it says the collector emitter breakdown voltage is 70V, and the emitter base is 7V... Since I don't use the base...do I get away with that? Here's the sheet by the way Vishay 4N25.

I found a online shop here in the Netherlands that has the Siemens SSR, so maybe I'll try it anyway, if it doesn't work well enough, too bad, have to take a chance some time I guess.

Oh and by the way, how do you get those Maxim products anyway? I always see interesting stuff there, but I have no clue how to get them!

[The Tone God]

Ah ok, well I don't believe the optocoupler is fried just yet... , it did switch ok, and turned the audio off. I checked the datasheet before I determined the LED current, and since it said at absolute maximum ratings that the maximum LED current was 60mA, I figured 25 would be safe. Also I checked the sheet again now, and it says the collector emitter breakdown voltage is 70V, and the emitter base is 7V... Since I don't use the base...do I get away with that? Here's the sheet by the way Vishay 4N25.

Cool. Yes you can just leave the base alone. The CTR is pretty lousy in that part which is what I roughly remembered. If you are using this to switch the signal before a power amp section then the signal might be too much for it.

I found a online shop here in the Netherlands that has the Siemens SSR, so maybe I'll try it anyway, if it doesn't work well enough, too bad, have to take a chance some time I guess.

Good luck. Atleast you will learn something new. Let us know how it turns out. Its been awhile since I tried that.

Oh and by the way, how do you get those Maxim products anyway? I always see interesting stuff there, but I have no clue how to get them!

They would probably have a list of distributers on their site. I know you can get their stuff at Digikey which is very good. They have things like 1-10 ohm solid state analog switches which you could use take into care their signal requirements.

Andrew

[H.Manback]

Well.. seems like I still have some trouble ahead

The Siemens SSR I found earlier, the LH1516, has a bad output capacitance of around 150pF at 1V... and it actually varies with the voltage . One SSR did have a low capacitance of about 5 pF, the LH1541, but I can't find seem to get a hold of that one.

I looked at analog switches now, but they don't really look like a good solution... Maximum signal is the supply voltage, which is typically at best 15V (or +/- 7.5V), and you need some headroom apparently because they distort signals close to the supply voltage.

What I also saw were the things Fender and Mesa Boogie used, I think they're called vactrols. Basically a LED and LDR in one package. Seems to me that is far from ideal since the off resistance is not that great (couple of kohms) and I read something about those things oxidating and failing as well . I'm guessing that's not a good solution..

I'm beginning to run out of ideas here, so I am considering using mechanical relays if I can't find anything else. I always had the idea that relays 'pop' when switching, is that true?

[d95err]

What I also saw were the things Fender and Mesa Boogie used, I think they're called vactrols. Basically a LED and LDR in one package. Seems to me that is far from ideal since the off resistance is not that great (couple of kohms) and I read something about those things oxidating and failing as well . I'm guessing that's not a good solution..

I'm beginning to run out of ideas here, so I am considering using mechanical relays if I can't find anything else. I always had the idea that relays 'pop' when switching, is that true?

Optocouplers with resistance output (i.e. an LED/LDR) is used in a lot of commercial amps, probably because the soft on/off means there will be no popping when switching. I don't know about the "couple of kohms" off resistance you mention. From the datasheets I've seen most have the same type of off resistance as an LDR, i.e. 10M or more. The on resistance varies a lot though. From 50 ohms to about 50k ohms. The Vactrol brand are often used in tube amps because some of them have
a fairly high effect rating (many LDRs can take just 0.1W or less). Depending on where in the amp you want to put them, this could be important.

I'm about to try LED/LDRs for channel switching in an amp I'm building. I'll let you know how it turns out (it will take a while though before I get that far, I've got the board toghether and I'm working on drilling the chassis...)

[H.Manback]

Oh yeah, I typed a couple of kohms but what I meant was a couple of 100 kohms , but I guess I must have been looking at the wrong ones if you say you saw 10M and more. I'll look in to those optocouplers some more then.