[net.ham-radio] ICOM 22S mods

rjr@mgweed.UUCP (Bob Roehrig) (05/02/85)

Someone asked about mods for the Icom 22S. I have been well pleased
with mine, except for one minor problem, which I will mention later.

I don't know where this document came from that I will quote from
and I assume that the mods in mine were made from this dope.
(I got mine 2nd hand and it already worked below 146).

"to operate below 146 mhz:

1. locate IC1 on synthesizer board. remove diodes connected to
pins 7 and 8 going to pin 13, and the 10k resistor from pin
13 going to ground.

2. Connect pin 13 to pin 16 of IC-1.

3. locate IC-10 on the PLL board. Cut pin 2 close to the
board and bend upwards. Cut the path between pins 2 and 5.
Connect a 10k resistor from pin 5 to ground. Connect cathode
of a silicon switching diode to pin 2. Connect its anode to
pin 5 of IC-8. Connect a jumper between IC-10 pin 2 and 5.

4. Connect another diode on IC-10: cathode to pin 5, anode
to pin 6.

These mods still have the radio operating on 15 kHz channels.
The ./. N factor for the channel you want is figured
as F-144.39 / .015

For example, 144.990 is divide by 40, and 145.005 is divide
by 41.

I found that the RF stages in both the rcvr and xmtr needed
repeaking to use below 146 MHz. There is no way the 22S will
cover 144 to 148, RF-wise. You can get 3 out of the 4 MHz
and that is about it. 

My 22S block diagram shows that a reference frequency of 7.5
kHz is used, derived from a 7.68 MHz osc. I would assume that
to get a 20 kHz spacing between channels, one would need to
use a crystal that would derive a 10 kHz reference - I have
not explored this.


The problem I mentioned was intermittent locking of the PLL.
The problem was caused by corroded connections in the plug
on the matrix board. Cleaning did no good, so I removed the
pins and the board connector and "hard wired" the matrix board
using ribbon cable.

/-/-/-/-/-/-/-/ additional changes /-/-/-/-/-/-/-/-/-/-/

I tripled the number of channels the 22S will operate on
by adding one switch on the rear of the unit. This switch
is now wired as the power switch. The original front panel
power switch is now labeled C, E (center), and D. The original
duplex switch is left alone - A, S (simplex) and B.

All diodes were removed from the matrix board and new ones were
installed as shown below, except to rows 0, 6, and 7.

The C-E-D switch is now wired as follows:

1. ground the common contact.

2. connect the C contact of the switch to the D7 wire of the
matrix board.

3. connect the D contact to the D6 wire of the matrix board.

4. Connect two 2.7k resistors from +9 volts to the C and D lines.

5. Install diodes in positions D1 thru D5 as follows:

CHAN D5 D4 D3 D2 D1 FREQ      CHAN D5 D4 D3 D2 D1 FREQ

 1   x     x  x     146.01     11              x  146.34
 2   x     x  x  x  146.04     12           x     146.37
 3   x  x           146.07     13           x  x  146.40
 4   x  x        x  146.10     14        x        146.43
 5   x  x     x     146.13     15        x     x  146.46
 6   x  x     x  x  146.16     16        x  x     146.49
 7   x  x  x        146.19     17        x  x  x  146.52
 8   x  x  x     x  146.22     18     x           146.55
 9   x  x  x  x     146.25     19     x        x  146.58
 10  x  x  x  x  x  146.28

NOTE: I have put no diodes in row D0. This row is only
      needed for frequencies ending in "5".

      Position 24 on the selector is blank and is used to 
      obtain 146.31, which requires no diodes.

With the modifications as outlined, the following channels
are available:

			SIMPLEX CHANNELS
			________________

CHAN	FREQ		CHAN	FREQ		CHAN	FREQ
____________		____________		____________

			14D	146.43
15C	145.50		15D	146.46		15E	147.42
16C	145.53		16D	146.49		16E	147.45
17C	145.56		17D	146.52		17E	147.48
18C	145.59		18D	146.55		18E	147.51
19C	145.62		19D	146.58		19E	147.54


			REPEATER CHANNELS
			(outputs shown)
			_________________

	CHAN	FREQ		CHAN	FREQ
	____________		____________
	(Dup A)			(Dup B)

	1C	146.61			
	2C	146.64		2D	147.00
	3C	146.67		3D	147.03
	4C	146.70		4D	147.06
	5C	146.73		5D	147.09
	6C	146.76		6D	147.12
	7C	146.79		7D	147.15
	8C	146.82		8D	147.18
	9C	146.85		9D	147.21
	10C	146.88		10D	147.24
	11D	146.94		11E	147.30
	12D	146.97		12E	147.33
	13D	147.00		13E	147.36
				14E	147.39

	24D	146.91		24E	147.27

 I wired the following other channels as follows for local
 use:
	CHAN	FREQ	DIODES	Sw pos.

	20	145.635	0-1-4	  C
	21	145.575 0-1-2-3	  C
	22	145.800	1-2-3-4	  C
	23	147.375	0-1-2	  E

brian@sdcsvax.UUCP (Brian Kantor) (05/03/85)

Since the synthesizer in an IC22S uses 8 bits to set the frequency, it
is quite easy to drive it with a microprocessor.  I use exactly this
scheme in a remote base, where the touch-tone commands are converted
by the Z80 system controller into the synthesizer input to put the radio
on any frequency in the band.

But a more practical idea:  since the endings of all frequencies at .015
spacings are unique in the top two MHZ of two meters, you can easily use
something like a 2716 to translate 9 bits (two BCD thumbwheels and a 0/5
toggle) into the synthesizer input.  That way you can dial up any
frequency the radio will do directly.  There are 11 address bits on a
2716, so you can even use the other 3 512 word partitions for special
effects - or (this is nifty), switch one of the address lines with the
transmit line, which gives you the ability to have custom offsets for
repeaters and such - since the transmit and receive frequencies are
separately stored and separately loaded into the synthesizer from the
prom.  At about $6 for the switches, and $3 for the chips, its a cheap
mod!

	Brian Kantor, WB6CYT	UC San Diego

	decvax\ 	brian@ucsd.arpa
	akgua  >---  sdcsvax  --- brian
	ucbvax/		Kantor@Nosc