Differences

This shows you the differences between two versions of the page.

--- packet:transceivers_144pk [2025/07/08 09:21] – created g7taj
+++ packet:transceivers_144pk [2025/07/29 06:29] (current) – g7taj
@@ Line 1: / Line 1: @@
 ====== Wood & Douglas 144 PK 144Mhz packet transceiver ======
+A reverse engineering project by G7TAJ
+A post by Nick G4TAI on the OARC forums about a boot sale find, turned into a few discussions about it as he was wondering how it worked and if it would be possible to change the frequencies. It was on the old 1980’s 144Mhz frequencies. A few discussions later and I agreed, for a few beer coins, to take it and see what I could find out as a little interesting project (I always want to know how things work inside the box, even from a young age).
-A post by Nick G4TAI on the OARC forums about a boot sale find, turned into a few discussions about it as he was wondering how it worked and if it would be possible to change the frequencies. It was on the old 1980’s 144Mhz frequencies. A few discussions later and I agreed, for a few beer coins, to take it and see what I could find out as a little interesting project (I have always wanted to know how things work inside the box, from a young age).
+The 144 PK is a 4 channel ~10w 2meter transceiver made specifically for packet. It has no speaker, no RX light, just a TX light and some adjustment pots and a 4 channel switch.
-The 144 PK is a 4 channel ~5w 2meter transceiver made specifically for packet. It has no speaker, no RX light, just a TX light and some adjustment pots and a 4 channel switch.
 From the pictures Nick had posted online, it was already thought that an EPROM was at the heart of the controls. It had a few other chips on-board that were of interested.
-{{:packet:unit_front.jpg?400|}}
+{{:packet:unit_front.jpg?400|}}{{:packet:unit_back.jpg?400|}}
+{{:packet:front_1.webp |}} {{:packet:back.jpg?400|}}
@@ Line 18: / Line 19: @@
 The 4 channel switch is wired directly to some of the EPROM address lines, which Nick had already put to his multimeter.
-<switch state picture>
+{{:packet:switch_hl.jpg?400|}}
-Once in my hands ( a 400mile round trip back to my old QTH location, having just moved, which is coincidentally where Nick lives and I just happened to be there to visit family) and with all this information, it quickly lead me to work out that the binary counter driving the EPROM address lines and driving the synthesiser to setup the frequencies. What I didn’t know, is how.
+Once in my hands ( a 400mile round trip back to my old QTH location, having just moved, which is coincidentally where Nick lives and I just happened to be there to visit family) and with all this information, it quickly lead me to work out that it was the binary counter driving the EPROM address lines and driving the synthesiser to setup the frequencies. What I didn’t know, is how.
 I removed the EPROM and dumped it and noticed some very repeated patterns.
-<eprom.bin>
+{{ :packet:eprom.zip |}}
-The next phase was to trace out the digital circuit around the EPROM and these other chips to understand what was happening. This is where things started to get less straight forward.
+The next phase was to trace out the digital circuit around the EPROM and these other chips to understand what was happening. This is where things started to get less straight forward (this was a quick dump of tracing things. Most are surface mount and dont have any silkscreen identifiers).
-<<schematic>>
+{{:packet:schematic.png?400|}}
 Some points I noted about the circuit:
@@ Line 34: / Line 35: @@
   * The EPROM was clocked via the RC NOR gate oscillator
   * This oscillator could be stopped by D7 of the EPROM going HIGH
+  * D4 - D6 were used to send the DATA, CLOCK and ENABLE signals to the NJ88c22
   * The reset circuit was triggered each time the channel switch position was changed, which would start the counter again from zero
@@ Line 75: / Line 77: @@
 I modified my C program to account for the PTT line changes and found that on TX the frequencies were correct but at RX they were low. But they were all low by 21.4Mhz, which sounded very much like a standard IF frequency! So that answered that, phew!
 At this point, I was confident that I could change the data for a frequency I wanted.
 A quick modification to the C program made a new EPROM image and I was ready to test.
-On the bench, I plugged it into the dummy load and tested the channels according to what I had calculated from the EPROM image. They were all wrong :-?
+On the bench, testing with the original EPROM, I plugged it into the dummy load and tested the channels according to what I had calculated from the EPROM image. They were all wrong :-?
 So I double checked the switch connections and EPROM data lines and found that there was a jumper, JP5, that pulls A2 and A3 to ground if bridged, which it was. So I adjusted the address lines in the program and re-ran it and the frequencies matched, wahoo!
-With the new found addresses, I wrote a new EPROM image. I only had a 27c256 to hand and so I wrote it to both 128k sections. Pulled it out of the programmer and plugged pushed it into the socket on the board. I tentatively powered it up, watching for magic smoke, which there wasn’t any, thankfully!
+With the new found addresses, I wrote a new EPROM image. I only had a 27c256 to hand and so I wrote it to both 128k sections. Pulled it out of the programmer and pushed it into the socket on the board. I tentatively powered it up, watching for magic smoke, which there wasn’t any, thankfully!
 With a jumper wire for activating PTT, I had the H/H set to 144.950Mhz and went for it. And it only went and bloody worked! I quickly connected a set of headphones to the audio out and repeated the test from H/H to 144 PK on RX and it was working as well. JOB DONE!
 Well, not quite. I wanted to test it on packet, after all, that’s what it was designed for! So I unplugged my Motorola GM350 from the NinoTNC and plugged in the 144PK with the lead that came with it (5 pin DIN to standard 9pin TNC connector, as used by Kantronics and the standard that the Nino also uses).
-“C MB7NSC” I wrote at the node prompt. The TX light lit and the H/H screeched and “Connected to MB7NSC” appeared. It connected and on 5watts, over 63kms. Wow, impressive. I did a few more test over the air, getting nodes lists, going to GB7IOW and posting on this wall there etc
+“C MB7NSC” I wrote at the node prompt. The TX light lit and the H/H screeched and “Connected to MB7NSC” appeared. It connected and on 5watts (the brick PSU it was connected to was limiting the output power), over 63kms. Wow, impressive. I did a few more test over the air, getting nodes lists, going to GB7IOW and posting on this wall there etc
 <code>(07-Jul 17:32) < G7TAJ > Testing using 144PK Wood&Douglas TcRX with modified frequencies</code>
@@ Line 97: / Line 101: @@
   * The frequency of the EPROM clocking would be nice to know. My scope is in one of a million cardboard boxes from the house move, so I have no idea what its running at. The resistors are 223’s, so 22k. The capacitor is something I can’t measure. This would enable me to determine the startup timing from RX to TX and visa-versa. This may be <10ms or maybe far more, which wont make much difference at 1200baud (as my tests works just fine with no change on my normal TXDelays etc) but it might at a higher rate.
   * Its programmed for 25khz. I can change this to the NJ8822 but the front end will no doubt be hardware LC filtered for that. More investigation needed
-  * What is the audio path like for 3600 or 9600 ? I think it’s going to be restricted and may need some modification.
+  * What is the audio path like for 3600 or 9600 ? <del>I think it’s going to be restricted and may need some modification.</del> According to an advert in an old magazine, its good for 9600 !
+<sub>//Some stupid mistakes and time wasting caused by me not being able to read or count properly have been omitted from this article for readability :-P//</sub>