circuit fade

[Mr RB]

To Mike; I think that circuit has issues because all the power is fed as DC (only half a phase) and the sound card output is always capacitor coupled, and power needs to be drawn from it as AC.

To Camerart; Instead of using the bridge rectifier to draw AC power from the soundcard, you could eliminate that rectifier voltage drop and put the soundcard output through a series resistor into the LED, then put another LED in parallel (reversed!) with the first LED. That will draw balanced AC power but now there is no voltage drop from the bridge rectifier. For the final app use a LED in parallel reverse with the LED inside the optocoupler.

Also your voltages out of the sound card still seem low, you should check all volume controls in Windows (double click the volume icon and turn everything right up) and also it's possible your morse generating software has a volume control or does not output max volume for some reason?

 

[camerart]

Attached is how the two circuits should look with the current-limiting resistor. This is to solve the sound-card overload issue.

What is the turns ratio of the transformer you are using. Line-Out of a sound card is usually about 1Vpp. If the transformer is 1:1, the 1V is not enough to forward-bias both the diode and the transistor base or the LED. You would need a transformer that has a turns ratio higher than 5:1 (hooked up backwards, so that it steps up the line-out voltage). An old output transformer from a "transistorized radio" 500Ω to 8Ω hooked up backwards would an ideal step-up. That would give you about 10Vpp to drive the rectifier and follow-on circuit...

Thanks for the diagrams MikeMl,

I have a 20K to 1K There are no markings about which way round it is. I used a resistance meter to get side 1/ = 1.2K side 2/ = 55R and 55R or 115R without tap. I am using one side 55R towards computer and 115K towards circuit.

 

[camerart]

To Mike; I think that circuit has issues because all the power is fed as DC (only half a phase) and the sound card output is always capacitor coupled, and power needs to be drawn from it as AC.

I hadn't realised that, so if the other half a phase isn't drawn, you seem to be saying it can damage the sound card. Very interesting!!

Also your voltages out of the sound card still seem low, you should check all volume controls in Windows (double click the volume icon and turn everything right up) and also it's possible your morse generating software has a volume control or does not output max volume for some reason?

Windows, isn't good for finding all of the settings, but I'm pretty sure I have them all turned up.

To Camerart; Instead of using the bridge rectifier to draw AC power from the soundcard, you could eliminate that rectifier voltage drop and put the soundcard output through a series resistor into the LED, then put another LED in parallel (reversed!) with the first LED. That will draw balanced AC power but now there is no voltage drop from the bridge rectifier. For the final app use a LED in parallel reverse with the LED inside the optocoupler.

What an elegant solution!!!

This is now working, on one computer, and the Morse type program, but no other programs. It seems not enough volume.

It doesn't work on my portable computer, and I need it to.

As there aren't many components, I can only change the transformer or resistor. I started with a variable resistor, and the led lit between 500R and 3K. I settled on 800R.
[/QUOTE]

 

[camerart]

I've just had a thought!! The interface is mono, and I have fitted a mono plug on it. Does it matter that I have plugged mono plug into a stereo output.

 

[MikeMl]

Yes, you are shorting one channel (the one that would normally be connected to Ring) to ground (Sleeve). You MUST use a stereo plug even if you are using only mono. Do not bridge L to R in an attempt to get more level.

 

[4pyros]

I've just had a thought!! The interface is mono, and I have fitted a mono plug on it. Does it matter that I have plugged mono plug into a stereo output.

Yes that's bad for your card, just use one side of a stereo cable.

 

[Mr RB]

...
This is now working, on one computer, and the Morse type program, but no other programs. It seems not enough volume.

It doesn't work on my portable computer, and I need it to.
...

OK congrats on getting something working! Laptops can have really wimpy sound output, basically they expect any speakers plugged into a laptop to be "powered" speakers (ie the speakers have their own amp).

So, it sounds like you will need something like a transistor to give you some gain, it can be triggered off a fraction of a volt and very little current where the LED needs >1.3v and >5mA or so.

Have you decided to keep the tranformer? I thought the whole point of the optocoupler was to get rid of the transformer as the opto provides the isolation?

Maybe to reduce all the back-forth interaction you should specifiy exactly what isolation you need ie none/opto/transformer and what type of output you need; relay contacts/transistor output etc and what it drives. Then we should be able to suggest a circuit that will work well first go even from a wimpy level sound output like a laptop headphones plug or an audio line-level signal.

 

[camerart]

OK congrats on getting something working! Laptops can have really wimpy sound output, basically they expect any speakers plugged into a laptop to be "powered" speakers (ie the speakers have their own amp).

I was able to get it working on the portable, after using Linux and other boosts, but this isn't satisfactory really.

So, it sounds like you will need something like a transistor to give you some gain, it can be triggered off a fraction of a volt and very little current where the LED needs >1.3v and >5mA or so.

Have you decided to keep the tranformer? I thought the whole point of the optocoupler was to get rid of the transformer as the opto provides the isolation?

No, the point of the optocoupler is to get rid of the relay that was there originally. Do you think this will work without the transformer?

Maybe to reduce all the back-forth interaction you should specifiy exactly what isolation you need ie none/opto/transformer and what type of output you need; relay contacts/transistor output etc and what it drives. Then we should be able to suggest a circuit that will work well first go even from a wimpy level sound output like a laptop headphones plug or an audio line-level signal.

See attachment: The Keyer is just a contact, from the radio 12V to earth @10mA No particular isolation, only computer damage protection.

I would like to know that if the 2 X channels are connected to the transformer primary as shown on the drawing, would this double the power?

MikeMl Pyros, I will use a stereo plug and not connect the 2 X channels together. [Checked online, tried 2 x channels into 1 using 10R resistors. On the portable: The output led lit, but not very brightly.]

 

[Mr RB]

I don't think you need the transformer, but keeping the optocoupler as your isolation between the PC and radio might be a good idea. However if you do need gain to make it reliable that means the PC side of the opto will need a power source to operate the transistor like a 9v battery.

You can't connect two stereo outputs to the transformer like your diagram. However connecting them with two 10 or 15 ohm resistors was probably safe enough, and is the way to use the two outputs to get more power (assuming they have the same signal and are in phase).

Maybe before committing to needing a powered gain stage you could test with both outputs through the transformer, and reduce that 680 ohm resistor? If it barely lights the LED you don't need so much resistance there.

Alternatively you could use both outputs to drive the LED directly, either through the two 10 ohm resistors alone or maybe adding another series resistor if the LED is too bright. (Remember in both cases you still need that reverse parallel LED to make the load AC).

 

[camerart]

Thanks Mr RB,

After doing lots of tests, I'm pretty sure the earphone output can't supply enough current for the LED/Opto coupler, either with or without a transformer. I think both computers have a fool proof earphone output circuit, which manages the power output, and protects against short circuit.

The transformer is to lift the voltage from around 1V to enough for a LED.

I replaced the 680R resistor with a variable 1K resistor. While running, I turned the VR from 1K to 0R with no difference in light output from the leds.

So, it seems I do need battery power for it to work.

(As a matter of interest, I have some 2n7000 fet transistors)

 

[Mr RB]

Ok, so if it's decided you must have gain then basically you are after a "VOX or "voice operated switch". This is a simple circuit that turns on the output when there is a sound signal present on the input.

Basically they use a transistor, probably a normal NPN, biased with its base at about 0.5v so it is JUST off. Then the audio is coupled to its base through a capacitor, so even with a small signal like 0.1 or 0.2v audio will be enough to turn the transistor on at the peaks. Then the output (collector) of the transistor will go low on each audio wave peak, that can drive your radio like your previous circuits or you could use a diode and cap on the output to mak eit constantly on, not just pulsing the wave peaks. That may not be needed, the impression i got from you previously was that the radio would tolerate an AC or pulsating input, but testing may be needed there.

 

[camerart]

For some reason, I'm having a lot of difficulty with this circuit, 3 popped optos and 3 popped transistors later.

I've attached a diagram. I've tried the transistor between 12v and opto too. I don't seem to be able to set the base at 0.5v. If R1 and R2 can be substituted by a pot, what value.

The input to the circuit is directly from the earphone socket.

The opto output does the keying ok.

 

[MikeMl]

The photo diode in the opto-isolator is a LED; it needs to have a current limiting resistor in series with it. That will make the NPN happier, too!

 

[camerart]

The photo diode in the opto-isolator is a LED; it needs to have a current limiting resistor in series with it. That will make the NPN happier, too!

Could you guess the value? I use a 9V Battery.

 

[ronv]

220 ohms would give you about 30 ma.

 

[camerart]

220 ohms would give you about 30 ma.

Thanks ronv, Do you think a 1K pot will replace R1 and R2?.

 

[ronv]

Yes, I think so, but you may still have trouble if you audio signal is only 200 mv peak to peak. You can try something like this circuit. It may be kind of touchey. If so we can probably use a comparator since you have added a voltage now.

 

[Mr RB]

Re your hand-drawn schematic; I would use the pot for R2, at the bottom, as you only need voltage adjustment under 1v. Then you can use a fixed resistor for R1. This will save the transistor in case of turning the pot too far and also give much better adjustability. Or just add a top resistor between 9v and the top of the pot in Ronv nice schematic. R1 should be about 5 ot 6 times larger ohms than the pot value.

R2 resistor on the opto collector in Ron's schematic might be a bit low, I would try 10k-47k range.

Also if you put a 0.47uF cap across the LED it will reduce the LED pulsing from the input frequency so the LED is "on" all the time although it's brightness will still ripple a bit. I don't think you have stated whether your radio needs DC input signal or can tolerate some pulsing?

 

[camerart]

Yes, I think so, but you may still have trouble if you audio signal is only 200 mv peak to peak. You can try something like this circuit. It may be kind of touchey. If so we can probably use a comparator since you have added a voltage now.

Thanks Ronv, I'm having problems with opening your attachments. I'll do a few more changes to the circuit before changing though.

 

[camerart]

As mentioned before I can't open the attachments anymore, which makes things a bit difficult.

I'm not sure we are all talking about the same circuit, so I hope I have all of the modifications on the attached jpg.

If there are any replies, can you stick to this circuit for a while please, till either it works or doesn't, then, if not I'll try another.

Thanks for being patient, sorry this is dragging on a bit!!.

Cheers, Camerart.