XTL5000 W4 - to - Kantronics KPC-3+ TNC for 1200 Baud Packet Radio

[ka9rxk]

I have seen some Astro Spectras connected to Kantronics KPC-3's on eBay via their rear connector, so I thought why not try it on an XTL5000 W4? I would like to use this for Winlink 2000 email.

Here is what I have:

XTL5000 W4 VHF - M20KSS9PW1AN - FLASHcode: 100001-000000-7
CPS Version: R20.01.00
Codeplug:
DSP: R19.50.01
Firmware: R20.50.05
UCM: None
Tuning: R01.00.04
PSDT: R04.00.00
Bootloader: R07.00.00

Kantronics KPC-3+ - Version 9.1 - 512K

Connections are as follows (pin numbers in parenthesis):
KPC-3+- - - - - - - - - - - - - - - - - - - - - - - - XTL5000
TX Audio (1) - - - - - - - - - - - - - - - - - - - - AUX Mic (23)
PTT (3) - - - - - - - - - - - - - - - - - - - - AUX PTT (16)
RX Audio (5) - - Audio Isolation Xfmr - - SPKR + (20)
Ground (6) - - Audio Isolation Xfmr - - SPKR - (26)
Power In (7) - - - - - - - - - - - - - - - - - - - SW B+ (24)
Ground (9) - - - - - - - - - - - - - - - - - - - Ground (1)

The radio / TNC receives and decodes packets just fine.

The problem is that even though the TNC keys up the radio via AUX PTT and the radio transmits the packet audio, nothing ever decodes it on the other end. I have experimented with both 80mv and 300mv settings in the AUX Transmit Sensitivity on the XTL5000, and have tried different settings of TXDELAY and XMITLVL in the TNC.

The packets that are being transmitted sound just like the other packets on the air to my ear. They are the same volume and duration. Unfortunately I don't have a scope or deviation meter.

Is there something within the XTL5000 that is filtering the transmitted audio which is making my packets undecodable? I'm ready to give up on this project! :wtf:

 

[rainbowpenguin]

Try this with XMITLVL on the KPC-3+: turn it down all the way. Increase level until you don't hear it getting any louder, then back off just a bit. I've gotten several tinytrak3 to work well with this method.

Not exact science, probably a bit off, but without the proper equipment it seems to work.

Other thought: does the xtl/kpc-3+ have a DC bias on the mic/tx audio line? Perhaps it needs a blocking capacitor.

 

[Mars]

Make sure TX AGC is disabled. Also check the channel deviation to ensure it is 5 kHz. Also disable TX emphasis.

 

[ka9rxk]

Make sure TX AGC is disabled. Also check the channel deviation to ensure it is 5 kHz. Also disable TX emphasis.

I tried your suggestions: TX AGC is disabled, channel deviation is 5kHz, and TX emphasis is disabled. I then tried a variety of XMITLVL settings on the TNC. The transmitted packets sound good to my ear, but still having the problem of nobody on the other end decoding me.

What about that 80mv vs. 300mv setting for AUX transmit sensitivity? I'm currently set to 80mv.

 

[MojaveJoe]

You will want to make sure TX preemphasis is turned on on your transmitter. This TX preemphasis is counter acted by the other end receivers de-emphasis. Most tnc receive connections are made at the speaker leads, which applys deemphasisized audio to the TNC. By turning off the transmit preemphasis on your end, the received tones at the other packet station will be skewed by 6db to its tnc the decoder. Thats quite a bit of difference and makes packets very hard to decode. The decoder chips used in TNC's were meant for use on wireline circuits - which doesn't have any skew between the high and low frequency tones.

Transmit tones should be 3.3 Khz dev max. Most limiters start compressing at a tad over 4 Khz. With preemphasis, your high tone (2200 hz) will be approximately 5 db higher than the lower tone (1200 hz). The difference of 6dB is seen on an octive difference - i.e. 1000 hz vs 2000 hz, or a doubling of the frequency.

As a simple deviation check, with a second radio and a analog voltmeter with a dB scale, connect the AC meter across the speaker leads on the second radio. Measure the single (higher tone) tone with the AC meter, and turn the TNC transmit level up (on your XTL5000 W4) until the meter doesn't read any higher (this is full deviation, and the radio is in clip). Then note the dB reading on the scale. Then turn down the transmit level (in your tnc) until the AC dB meter drops by 6 db. This should make your transmit tone be around 3-3.3 Khz dev. Use the higher of the two tones for this reading. I added a picture of a Simpson 260 showing the dB scale on the meter.

As to the AUX connections, I'm not sure if this port also does preemphasis or not. It sounds like its not. A trick we used to do, in these cases, is to insert a .001 uf cap in series with the transmit audio. The cap allows the high tone to pass through at a higher level than the lower tone. In a sense, it introduces preemphasis to an other wise flat transmit
audio input.
If I get a chance at work next week, I'll investigate the aux port on an XTL5000 on the service monitor.
MJ