This section is part of the DSP-10 User's Manual, Chapter 3 - Background Information
Audio Processor
The programs used for the Audio Processor are the same as for the full RF hardware use. The features available in both are the same. There is even "tuning" at audio frequencies. The transmit funtions are available in audio form, suitable for modulating an external SSB transmitter. Some aspects, such as FM reception do not apply to the audio processor and while they may be available we may not have thought of a use! Eventually, you will want the RF hardware, and the audio processor gives an excellent introduction to the whole project.
Audio Processor Notes -
Using the stand alone audio processor requires almost no hardware construction. One line of the UHFA.CFG file MUST be changed to tell the program to not look for the RF hardware. Follow the link above, or look in the UHFA.CFG file, in the hardware section, a few lines down from the HARDWARE header, and edit the line to hardware=0.
You need to have an Analog Devices EZ-KIT Lite and some speakers or headphones. I use amplified speakers from Radio Shack that were intended for computer use. The audio input goes to the left channel of the DSP and the right channel is not used in the audio mode. Audio from most headphone jacks seems to be OK to a bit strong, so don't turn the audio gain on the receiver too high. If the bar graphs at the bottom of the screen are being displayed, the first red bar should light most of the time and the second bar should light occasionally. The EZ-KIT JP2 is jumpered to 'LINE' (1 to 3, 2 to 4).
Loading of the programs is identical to that described for the full transceiver.
The scale for signal amplitudes is all relative in the audio processor. Both the S-meter display and the spectrum displays use the same relative scales.
Functions having to do with frequency accuracy and transverters do not apply.
AGC - This is at the beginning of the audio processor, but after the FFT. So when it is on the signal level will be kept 'constant' regardless of the input. Your transceiver may already cover this function and it may be best to turn off the AGC in the audio processor.
Frequency - The frequency offset function replaces the RF tuning of the full transceiver. This is a "Hilbert shifter" (see Forrer, QEX Sept 96) that shifts all audio frequencies up or down.
Hardware - The DSP processor expects a manual key at the PF5 programmable flag input. It will droop to zero Volts and the key will come on for CW transmit, unless you place a 100K resistor between P3-30 and +5 Volts (P3-49.) See page 8-7 of the EZ-Kit manual for P3 details. Note: This is only a consideration if you are going to use CW transmit. The audio processor doesn't care. Everything else in the audio processor uses the unmodified EZ-Kit Lite.
Modes - No mode is locked out from the audio processor, but from a practical point of view, FM is not useful, leaving USB, LSB, CW, LHL-7, EME-2, PUA43 and LTI. Don't try to make much out of these names. Basically, CW allows transmission of CW as well as audio processing. In order to process the audio and run the spectral displays, it is best to use USB. The FM mode has the possibility of being used for power measurement.
Moon/Sun - Moon and Sun functions are available (see above.) Not having the RF hardware also removes the normal source of frequency information, that is supplied by a UHFA.CFG configuration file entry called f_eme. This is the frequency in Hz that is used for EME doppler calculations.
Audio Processor Transmit Function- The
normal
transmission output is the left channel. The right channel has
the sidetone and will sound about the same in the headphones.
Actually, there are some differences. Both are set to the
frequency 'cwoffset' set by Scrl C and Scrl V. The level of the
left channel is 'Xmit Pwr' set by Scrl O and Scrl P. The right
channel level is the sidetone level, set by Scrl S and Scrl D.
The cwoffset and sidetone level are displayed on the left between
'Mode' and 'CW', when in the CW mode. In CW transmit, the right
channel has key clicks, but the left doesn't. This is a quirk of
the way that the CW generation is done in the DSP. A square wave
keying waveform is run through a 500 Hz LPF to remove keyclicks,
and then this is used to AM modulate the CW carrier. This
processing is not used on the side tone.
CONNECTION TO A COMMERCIAL TRANSCEIVER -
Due to the DSP-10 sharing DAC
outputs, there are two levels for this interface. The simple one omits
both the stereo audio output and the voice transmitter input. This,
uses one ADC input for the receiver audio, one DAC output for the
receiver speaker and the other output for transmit audio input.
A second more complex hardware interface would use relays, or analog
switches, to switch either the receiver audio or the microphone voice
to the ADC input, and to switch one DAC output to either a speaker or
to the transmitter audio input. This is done as part of the DSP-10 RF
hardware, but I am not aware of anyone doing this as a stand-alone
project. This could be an interesting project, and I would be glad to
assist anyone that wants to take it on.
The switching for the audio processor can come from
the DSP board
(either type), or alternatively the LPT parallel port on the PC can be
used. In version 3.50 of UHFA.EXE, the capabilities of this control
have been expanded for this use. The LPT port, specified by lpt_port
(see the config file) corresponds to the outputs on the accessory plug.
There are three pins now of the LPT port:
Bit 1, pin 3 of the 25-pin D, ANT for antenna relaycontrol. 0V for
receive, 5V for transmit.
Bit 2, pin 4 of the 25-pin D, AMP for amplifier (or transverter)
control. 0V for receive, 5V for transmit.
Bit 3, pin 5 of the 25-pin D, RF for RF out. 0V for receive, 5V for
transmit.
There are four sequencer delays, specified in milliseconds in the .CFG
file:
dly_ant2amp The delay after the HOME key is hit until the AMP goes
high. ANT goes high immediately with the HOME key, and at that same
time receiver audio is muted.
dly_amp2xmit The delay after the AMP is high until the RF goes high.
When RF goes high, the DSP-10 can start generating transmit signals.
dly_xmit2amp The delay after HOME is hit again and RF is stopped until
the AMP goes low
dly_amp2ant The delay after AMP goes low until ANT goes low.
There is an additional 20 milliseconds of delay after ANT goes low
before the receiver is unmuted. This prevents a "pop" from the analog
switching of the audio leads. These delays are unchanged from previous
versions of the software, but are given here as they now apply to the
parallel port whereas they were previously only for the accessory plug
of the DSP-10 hardware. They continue to apply to the accessory plug if
the configuration variable is 1 (not 0) and will apply to the LPT port
if the address for the variable lpt_port is non-zero (see the KD7TS web
paper referenced 4 paragraphs above for more discussion of this,
including the choice of lpt_port values 632, 888 or 956).