What circuit for computer audio out to key rig?

[Marc Vaillant]

I recall W4BQF mentioning on his site somewhere that QRQ folks use sound card output to key their rigs. I'm wondering what circuit you guys use? Is the audio out just normal "sidetone" output?

Thanks,
Marc

[Chuck aa0hw]

FROM: http://wa8lmf.net/ham/tonekeyer.htm (picture edited for CW use)

NOTE: click on picture for full screen view

[Chuck aa0hw]

FROM: http://www.nonstopsystems.com/radio/article-audio-vox-G4JNT.pdf

NOTE: click on picture for full screen view

[Chuck aa0hw]

FROM: http://www.nonstopsystems.com/radio/article-hell-RTTY-interface.pdf

NOTE: click on picture for full screen view

[Joe]

Don't panic if you can't key the transmitter with Chucks simple transformer to rectifier to keying transistor set up shown at the start of this thread. I could not get it to key the rig at first, so I went to the 'Sound Effect Manager' in the Control Panel of my XP operating system. I fiddled around with the manager to maximize the output of the sound card between 500 to 1000 cps and cranked the sound card output to maximum and fed this to Chucks circuit. I can now use Fldigi to key the Corsair without problems. If my little Acer laptop soundcard has enough output to key the rig, then I bet any computer can!

Joe W3GW

EDIT:  we found out that the reason it would not key initially was that FLdigi was to set with too much attenuation - chuck your FLdigi attenuation to make sure it is set to ZERO - for mox power transfer -

ALSO, use an audio isolation transformer that is not 1-1.  In testing many different commonly available ratios, it was discovered that an 8 ohm to 200 ohm audio transformer provided the best ratio and power transfer.  

after further testing after lowering the FLdigi ATTENTION CONTROL( circles in red above) Joe had the following comments:

If I reset the manager to 'none' for envirenment and 'pop' for 

'equalizer' then I get keying of rig with Fldigi from 0 to -7.7db attenuation. I loose keying at -7.7db. If I go back and set equalizer to maximize the frequency response between 500 to 1000 hz like I did originally, I can increase the attenuation to -9.5db before I lose keying. So the equalizer seems to help sound card output, but only by a few db. I don't need to fool with the sound card manager if I keep Fldigi 'transmit 

level attenuator' at zero.    Joe  -

FURTHER NOTE:  - use the FLdigi QSK ON RT CHANNEL feature, and then you can set the attenuation to any value.  The RT CHANNEL will send out a tone at 1600 hertz at full volume, no matter what the FLdigi attenuation setting is set at.  The left channel will send out a pure sine wave cw audio tone with raised cosine edges at whatever parameters you set it to, and at whatever attenuation you desire.

ALSO, to control the cw element lengths when using an audio derived cw keying switch, use the FLdigi QST CONTROL's   POST-KEYDOWN TIMING and trim it to match the original cw element lengths:  HERE IS AN EXAMPLE:

[Chuck aa0hw]

Here are a couple of more hints for getting more power transfer out of your sound card:

 

1.  Use a  - stereo to mono -  audio adapter, so you get both channels of FLdigi audio output going into the audio derived cw keying circuit.( NOTE: this is NOT for FLdigi in AFCW mode )

2.  Use an  - 8 ohm - to - 200 ohm -  audio transformer -  for better power transfer between your sound card output and the circuits input.   the mouser part number is  42TU200

Here is an updated schematic of this simple audio derived cw keying switch

NOTE: click on picture for a better view

[Chuck aa0hw]

I did a measurement with my AMPROBE METER, of the "native" RMS AUDIO VOLTAGE  coming out of my laptop sound card (with no load)  -  without any equalizer - 

WITH FLDIGI SET TO ZERO ATTENUATION

I FOUND THE NO LOAD AUDIO OUTPUT VOLTAGE TO BE ABOUT

~1 VOLT

BY LOADING THE OUTPUT VOLTAGE WITH A 10 OHM RESISTOR THE VOLTAGE IS

~200 Milli Volts

- This is not enough voltage to activate the switch properly with a direct connection -

Then I hooked up an audio cable, and connected it to the input of an 8 ohm to 200 ohm audio transformer as the revised circuit shows above, and the DC VOLTAGE rectified by the full wave voltage doubler circuit in this device, when using this Mouser audio transformer  - 8ohm to 200ohm -  measured to be about 2.2 volts of rectified voltage.

I then tested the equalizer on my REALTEK HD AUDIO MANAGER, but with full volume already on FLdigi's output and my REALTEK SOUND cards output, the equalizer on my system only added an additional 2 tenths of a volt increase.  

The necessity of an audio transformer with the best ratio for power transfer is emphasized.  Try out some spare audio transformers in your spare parts bin.   You may already have a transformer that will get close enough to activate this switch.  We have had reports of even a  wall wart 9 volt transformer that worked !  [This Mouser audio transformer - 8 ohm to 200 ohm - just provided the best performance of all the transformers that I had available to test.]  

If you are only using FLdigi to key your rig in the typical cw mode, then you can also use a STEREO to MONO adapter and capture both channels of the FLdigi audio and it will add more power going into this audio derived cw keying switch circuit. may or may not make a difference

Another measurement was made,  testing the amount of resistance with my ohm meter between the ground and collector of the final transistor during a "key down".  Even with FLdigi set to minus (-) 15 db attenuation, the resistance through the transistor measured to be about 50 ohms.  With FLdigi attenuation set to zero db,  the resistance from ground to collector of the final transistor went down to 15 ohms.

chuck

[Chuck aa0hw]

Here is an article about using the QSK ON RT CHANNEL feature of FLdigi to key a rig  in AFCW MODE.  Improve the QSK of any rig using AFCW and the FLdigi QSK ON RT CHANNEL method.

I have used this method on my old ICOM 735 and it was amazing how great the QSK became after using this AFCW FLDIGI SETUP.   Also, the cw waveform on my ICOM by itself was not that great, and the FULL QSK  on it chopped off the start of each element,  but after converting to this method, my icom's QSK was far better, and it was sending out RAISED COSINES !!  and you can adjust the FLDIGI QSK FUNCTION'S pre and post controls, so that you get the most receiver audio possible between elements or words, and NO CHOPPING OFF elements now...

https://docs.google.com/document/d/1jt7npAl58t1uaKVCCdDD9g3tHAhxzD0...

[Chuck aa0hw]

Here is a picture of FLdigi's audio voltage(bottom) going into the switch compared to the actual key closure.  FLdigi's QSK function does a great job of keying the rigs cw jack and duplicating the exact cw element lengths as you can see from the picture

[Chuck aa0hw]

Here is a picture of FLdigi's audio derived keying voltage compared to the actual RF output of my rig.   FLdigi does a great job of producing the exact cw element lengths using its QSK FUNCTION for audio derived cw keying.

NOTE:  this is not AFCW mode,

this is typical cw mode, and my rig’s own cw waveform with a short rise time and exponential wave construction...the picture came out blurry but still shows the accuracy of FLdigi in audio derived mode keying the cw jack of my rig with the audio derived switch

[Chuck aa0hw]

HERE IS A GOOD EXAMPLE OF KEYING A CODE PRACTICE OSCILLATOR USING THE FLDIGI QSK ON RT CHANNEL FEATURE

This audio file demonstrates FLdigi's QSK FEATURE, on the RT CHANNEL, keying an old HEATHKIT CODE PRACTICE OSCILLATOR at about 75 wpm. The POST KEYDOWN timing, on the QSK CONTROL SCREEN, was set to minus 7 milliseconds [-7ms]. (You may notice a lighter keying from this setting.) The POST KEYDOWN TIMING is quite adjustable however, and you can make the "keyed" cw on the target hardware, whether it is your own RIG or an external CPO, to be as light or heavy as you desire. The text was taken from TOM's, W4BQF, famous qrq cw article, COPYING CW OVER 70 WPM. FLdigi does a great job of reproducing the exact, intended cw parameters by using an audio derived cw keying circuit.

[Chuck aa0hw]

Measure the input and output impedance of your own setup using this online calculator

http://www.sengpielaudio.com/calculator-InputOutputImpedance.htm

Here are measurements of the output impedance of some of my sound cards using this calculator:

BEHRINGER UCA222 headphone output           -               63 ohms

BEHRINGER UCA222 LINE OUT                     -              560 ohms

SOUNDBLASTER MP3+ USB LINE OUT         -              108 ohms

SOUNDBLASTER MP3+  headphone output    -                 83 ohms

AUDIGY PCI SOUNDCARD OUTPUT              -                 32 ohms

TURTLE BEACH USB OUT/has amp              -                 3 ohms

As you can see from these results, there is quite a variance in the output impedance of different sound cards.   

On my BEHRINGER UCA222, LINE OUT  ,  RCA jacks,  I was able to (barely) key the switch using a 1/1 ratio- 600ohm - audio transformer with FLdigi set to zero db attenuation.  However,  If I added even just one db of attenuation, the switch would not work.  

By far the best results I have had in using an audio derived cw keying switch circuit - is using a 2nd sound card,   a  USB SOUND CARD, that has an audio amplifier included in its design.  A SIIG SOUNDWAVE 7 did a great job for me and has plenty of audio power to work with this switch.  I used a radio shack 8ohm to 1k ohm audio transformer for the interface when using this sound card due to the impedance involved;  the SIIG USB SOUNDCARD impedance was very low at 3 ohms.

http://www.siig.com/it-products/multimedia/sound/usb/usb-soundwave-...

 

[Chuck aa0hw]

Just a note on how the audio volume effects the Audio Derived CW keying circuits output to KEY the cw jack of a rig...too much volume into the switch,  your cw elements will be longer than the original,  too little volume into the switch, your cw elements will be shorter than the original...this is a picture of the original  DIT sent by a cw keyboard(top), to the AUDIO DERIVED CW KEYING CIRCUIT SWITCH,  and  on the bottom is the sidetone output from the Elecraft K3S CW SIDETONE MONITOR,  showing an exact replica of the original length of the dit element....