Audio Response: Kenwood TM-V7A

The Kenwood TM-V7A is a very popular radio for packet use, largely because it can operate both 2m packet and 2m voice (almost) simultaneously. It has a data port, and I tested both the 1200 and 9600 baud inputs and outputs at that port.

The test results are a surprise and a disappointment, because the audio performance does not match what I'd expect from a radio designed with packet in mind.

Transmit

The biggest problem is 1200 baud transmit response. Although there's a very modest preemphasis at the lower frequencies, the audio response is basically flat from 1kHz to 3kHz -- there's no preemphasis at all, and the data transmitted by this radio will suffer from many of the problems of an over-driven radio (minus the clipping distortion). The audio input is also very sensitive and can be easily over-driven; it only takes about 14 millivolts RMS input for 3kHz deviation of the 2200Hz tone.

Changing the "Data Transfer/Data Speed" menu to 9600 baud changes both the input sensitivity and the frequency response. In the 9600 baud setting, the performance looks pretty reasonable for 9600 baud operation -- the transmit audio response is flat within 1dB out to 6kHz, and the drive level is a bit more reasonable at about 1.3 volts RMS for 3kHz deviation.

Fixing the 1200 Baud Transmit Audio

You can fix the 1200 baud transmit audio response by adding an RC network between the TNC audio output and pin 1 of the data connector. A minimal approach is to simply put a 0.01uF (or thereabouts) cap in series with the audio. This will provide about 3dB of preemphasis; that's not as much as we'd like to see, but much better than nothing.

You can do a dB better, and elminate some roll off above 2.5kHz, by adding a 6.8kOhm shunt resistor to the radio side of the capacitor. The audio response with that circuit added is shown below with the other charts. This setup also has the advantage of requiring a bit more drive from the TNC, making the audio setting a bit less critical. Here's an ugly, hand- drawn schematic of the circuit I'm talking about:

Receive

On receive, the story isn't quite so bad, though again it's not what I expected to find.

Although the radio offers separate data outputs for 1200 and 9600 baud, I'll be darned if I can find any difference between them. The frequency response, and overall output level, of the two outputs track each other within a fraction of a dB.

The response is basically flat out to 4.5kHz, with a very gradual rolloff beyond that. There's no trace of deemphasis. This setup should work OK for 1200 baud, though I have some doubts about its performance at 9600, given the rolloff starting at 4.5kHz.

Changing the Data Transfer/Data Speed menu between 1200 and 9600 baud doesn't seem to have any effect on either of the audio outputs. I'm frankly not sure why Kenwood bothered to put parallel outputs on the data connector; it would have made more sense to separate the data input lines!

For what it's worth, here are charts of the audio performance on both 1200 and 9600 baud transmit and receive. Note the chart scales -- they have some variation.

1200 Baud

Here's the receive audio response (at pin 4 of the data port). The audio output level is about 250mV RMS for a 2200Hz tone modulated to 3kHZ deviation.

Here's the transmit audio response (Data Port/Data Speed set to 1200 baud).

This is the 1200 baud audio response with an RC filter added in the transmit audio line as described above.

9600 Baud

This is the 9600 baud receive performance, measured at pin 5 of the data port. It is virtually identical to the 1200 baud one, except that I extended the frequency plot.

And, this is the 9600 baud transmit performance, measured with audio going in to pin 1 of the data port and the Data Port/Data Speed menu set to 9600 baud.