Cubic ASTRO B: a 100-125W 32 channel 2-24MHz USB/LSB/CW transceiver from the late 70s or early 80s. Cubic (The former Swan, of course) manufactured radios that were sold to government agencies, as I understand it. Information has been hard to come by for this radio. The 32 channels are programmed onto a Harris 7621 PROM. This PROM was widely used in arcade console video games. The frequencies stored in the chip are used to tune the VCOs and light up the external lights for mode and band. Even though the PROM lights up an LED to suggest a mode, you can use any one you want for any channel, but you must select the correct band or the Fault indicator lights up and audio is muted. Among the channels that came on the PROM with the radio are a few that are actually in 75 and 40m. Since this isn’t exactly a band cruiser with 32 channels and no frequency display, it would be cooler if it were programmed for the military radio nets we hear up here and I have been trying to figure out how to (a) read the PROM I have and (b) program a new PROM to the frequencies I wanted.

Cubic ASTRO-B

After I received the radio, I tested it and it worked well, so I bought a manual and located the PROM inside. The PROM was soldered to a carrier that plugged into a socket on one of the boards in the front inside of the radio. During my search, I located a slip of paper with the programmed frequencies and settings on the PROM! What luck! The image below is close to actual size.

List of Programmed Frequencies

I then wanted to learn how to read and program it. I located a PROM programmer and software and got it running on my PC. It was difficult to find a programmer that would burn the Harris chip. I then removed the PROM from its carrier so it could be inserted into the reader. That was somewhat difficult and I wrecked some traces on the PROM holder (Cubic calls it a channelpak). The programmer I have been using is a Data I/O Chiplab. This is what I got when I read the PROM:

:1000000009090402050D00000F0F0F0F0F0F0F0F4E

:1000100004040601050D050104040601050D050192

:1000200004080201050E000104080201050E00018A

:1000300004070201050E000104070201050E00017C

:1000400004060201050E000104060201050E00016E

:1000500009040101050C000109040101050C00015E

:1000600007040001050C000107040001050C000154

:1000700009090900050D000109090900050D000124

:1000800006090900050D050106090900050D050110

:1000900004090900050D000104090900050D00010E

:1000A00009090909040D00020F0F0F0F0F0F0F0FA1

:1000B00001030008040D050201030008040D0502F8

:1000C00004070907040E000304070907040E0003D0

:1000D00004070807040E000304070807040E0003C2

:1000E00004060807040E000304060807040E0003B4

:1000F00004040707040C000304040707040C0003AE

:1001000004030707040C000304030707040C00039F

:1001100009090607040D000309090607040D000379

:1001200004050607040D000304050607040D00037B

:1001300006030607040D050306030607040D050361

:1001400006020607040D050306020607040D050353

:1001500008000607040D050308000607040D050343

:1001600001000507040D050301000507040D050343

:1001700009090904080D00030F0F0F0F0F0F0F0FD0

:1001800005020001080D000405020001080D00042D

:1001900006010600080D050406010600080D050409

:1001A000050208080F0D0005050208080F0D0005DF

:1001B000010007080F0D0505010007080F0D0505D3

:1001C000020800050F0D0005020800050F0D0005CF

:1001D000090909040F0D00050F0F0F0F0F0F0F0F67

:1001E000050206040F0D0005050206040F0D0005AB

:1001F000010204040F0D0505010204040F0D05059D

:00000001FF

Here’s what I think it says:

:100

00 – Channel number in hex

00 - ?

0 - ?

09 09 04 02 – reverse this number and you get 2499. Take 2499 and add it to the desired frequency (2500 KHz) and you get 4999. This formula is true for all of the programmed frequencies below 5MHz. All of the programmed frequencies between 5MHz and 15Mhz equal 14999 when added to the value here- in reverse. Same is true for all of the programmed frequencies above 15MHz. They equal 24999 when added to the reverse of the value. At least I think this is how it works. This provides the KHz, any hundreds of hertz come later from a different VCO.

05 –This is a value that indicates which VCO to use- the radio uses three to generate the needed frequencies between 47 and 69 MHz. Trouble is, there’s four different values used in this spot.

value receive freq

05 <4999 KHz

04 5000 to 14???KHz

08 14??? To 15???KHz

0F 15??? To 24999KHz

0D – mode-

value mode

0C LSB

0D USB

0E CW.

00 - hundreds of hertz. This is the separate VCO.

Value Hertz

00 000

05 500

00 band- 0-5. the ranges below need to be refined:

Band 0 up to 3000 KHz

Band 1 3000 to 4500 KHz

Band 2 4500 to 7000 KHz

Band 3 7000 to 10000 KHz

Band 4 10000 to 15000 KHz

Band 5 15000 KHz and up

0F 0F 0F 0F 0F 0F 0F 0F – Transmitter frequency and band setting, same pattern as receive.

4E – No room on programmer interface for this number. And the PROM works without it.

I made a table of all of the values I wanted to program into the PROM. I thought I would be able to load this text file at first. Later I learned I had to thumb it in through Chiplab. To help me while I edited the file, I put the column headings at the top of each column. This is just a different way to break it down than I did above.

Except for the headers, this is how it looks in Chiplab (see below). The values appear in each column twice, once for receive (left half) and once for transmit (right half).

Column data (see above for range of values)

Chan channel in hex

Diffinrvrse difference of the VFO number and the desired frequency, in reverse

(four columns)

Vf which VFO to use

Mo mode

Hz hundreds of hertz in desired frequency

Bn band

chan diffinrvrse vf mo hz bn diffinrvrse vf mo hz bn ??

:10 0000 00 06 09 09 02 05 0D 00 00 06 09 09 02 05 0D 00 00 4E

:10 0010 00 04 03 09 02 05 0D 00 00 04 03 09 02 05 0D 00 00 4E

:10 0020 00 00 02 09 02 05 0D 00 00 00 02 09 02 05 0D 00 00 4E

:10 0030 00 07 01 08 02 05 0D 00 00 07 01 08 02 05 0D 00 00 4E

:10 0040 00 09 09 04 02 05 0D 00 00 09 09 04 02 05 0D 00 00 4E

:10 0050 00 04 00 00 01 05 0D 00 01 04 00 00 01 05 0D 00 01 54

:10 0060 00 03 00 00 01 05 0D 00 01 03 00 00 01 05 0D 00 01 54

:10 0070 00 02 00 00 01 05 0D 00 01 02 00 00 01 05 0D 00 01 54

:10 0080 00 09 09 09 09 04 0D 00 02 09 09 09 09 04 0D 00 02 A1

:10 0090 00 09 06 06 09 04 0D 05 02 09 06 06 09 04 0D 05 02 A1

:10 00A0 00 03 05 06 09 04 0D 05 02 03 05 06 09 04 0D 05 02 A1

:10 00B0 00 03 03 06 09 04 0D 05 02 03 03 06 09 04 0D 05 02 A1

:10 00C0 00 08 02 06 09 04 0D 05 02 08 02 06 09 04 0D 05 02 A1

:10 00D0 00 06 09 05 09 04 0D 05 02 06 09 05 09 04 0D 05 02 A1

:10 00E0 00 04 02 08 07 04 0C 00 03 04 02 08 07 04 0C 00 03 AB

:10 00F0 00 01 02 08 07 04 0C 00 03 01 02 08 07 04 0C 00 03 AB

:10 0100 00 09 01 08 07 04 0C 00 03 09 01 08 07 04 0C 00 03 AB

:10 0110 00 04 01 08 07 04 0C 00 03 04 01 08 07 04 0C 00 03 AB

:10 0120 00 09 00 08 07 04 0C 00 03 09 00 08 07 04 0C 00 03 AB

:10 0130 00 04 00 08 07 04 0C 00 03 04 00 08 07 04 0C 00 03 AB

:10 0140 00 09 09 07 07 04 0C 00 03 09 09 07 07 04 0C 00 03 AB

:10 0150 00 04 09 07 07 04 0C 00 03 04 09 07 07 04 0C 00 03 AB

:10 0160 00 08 06 07 07 04 0D 00 03 08 06 07 07 04 0D 00 03 AE

:10 0170 00 07 06 07 07 04 0D 00 03 07 06 07 07 04 0D 00 03 AE

:10 0180 00 06 06 07 07 04 0D 00 03 06 06 07 07 04 0D 00 03 AE

:10 0190 00 09 03 07 07 04 0D 00 03 09 03 07 07 04 0D 00 03 70

:10 01A0 00 03 00 07 07 04 0D 00 03 03 00 07 07 04 0D 00 03 79

:10 01B0 00 02 00 07 07 04 0D 00 03 02 00 07 07 04 0D 00 03 79

:10 01C0 00 09 09 09 04 08 0D 00 03 09 09 09 04 08 0D 00 03 D0

:10 01D0 00 07 05 06 00 08 0D 05 04 07 05 06 00 08 0D 05 04 09

:10 01E0 00 09 09 09 09 0F 0D 00 05 09 09 09 09 0F 0D 00 05 09

:10 01F0 00 02 04 08 06 0F 0D 00 05 02 04 08 06 0F 0D 00 05 C1

:00000001FF

At any rate, I think I had enough data to try burning a chip. The plan was to try a Signetics 82S131 chip that is supposed to be a substitute for the Harris chip. Harris ones are available on eBay for triple the money, no guarantee and a warning from the seller that they may have some programming on them. No thanks.

Here is the programmer:

I had originally edited a text file of what I had extracted from the PROM that came in the radio. Trying to get the device to recognize any of the edited files didn’t work at all. After groping around, I learned that you have to edit the data directly in the program. On 1-22-2011 I was able to burn a PROM successfully. I installed a 16 pin DIP chip socket on the radio’s PROM carrier card thing. Initial tests looked like neither the original or the newly burned chip worked, but when I probed the socket for continuity, I found a bad solder joint. See below:

There are two full screens to program. There’s nothing fancy about it, no copying and pasting to speed it up. I programmed 32 frequencies to test the burning process and the radio. These included several Navy and Coast Guard frequencies (for the Great Lakes and nationwide), several WWV frequencies, All 60 meters frequencies, and all of the mil radio ham frequencies I know of. The rig works OK, but output is low and LSB is rough. Also, I think it has a loose connection somewhere as I can’t get it to reliably read the chip- I get lots of different fault situations, like no mode and band lights, or all CW mode lights (Chip is entirely LSB and USB).

Work progresses on this radio. While testing my chip programming, I noticed that the radio is off frequency. So I bought a service manual and dove in. I wasn’t able to adjust the LO back to frequency. So I replaced it with an Epson SG8002 Oscillator. See the picture on the right. I trimmed the other oscillator to match the Epson’s output. The result is that when trimmed for 40 Meter frequencies programmed on my PROM, I am off on the 17 Meter frequencies- a lot off.

Above is the front view with the top open. In the middle you can see the PROM on it's carrier. In the photo on the right, you can see the oven- someone has been in there before with the duct tape. The photo below is the board with LM7805 and oscillator

The next step is to study the manual for the correct method of chip coding and revisit the frequencies the oscillators should be set to. I bet I’ll need to find a happy medium.

Main page