An older offering from Icom. For us UK hams we get HF/2m/70cm including 5MHz and 70MHz. It's architecture is the older IF DSP design so it's hackable to include IF / RF taps if you want to add a panadapter. KV5R has done a great write up on IF tapping the radio. The IF is apparently pretty high at 124.487 MHz, making it slightly less than trivial to use a buffer amplifier, but his process appears to give a reasonable IF output without dragging down the stage too much.

UK Diode Matrix, 5MHz + 70MHz
6308 6316 6324 6342
1    1    1    0
1    1    0    0
0    1    1    0
0    0    0    0
1    0    0    1
0    0    1    1
1    0    0    1
0    0    0    1
6301 6309 6317 6325

UK Diode Matrix, 5MHz
6308 6316 6324 6342
1    1    1    0
1    1    0    0
0    1    1    0
0    0    0    0
1    0    0    1
0    0    1    1
1    1    0    1
0    0    0    1
6301 6309 6317 6325

North America Diode Matrix
6308 6316 6324 6342
1    1    0    0
1    1    0    0
0    1    1    0
0    0    0    0
0    0    0    1
1    0    0    1
1    0    0    1
0    0    1    1
6301 6309 6317 6325

Japan Diode Matrix
6308 6316 6324 6342
1    1    0    1
1    1    0    0
0    0    1    0
0    0    1    0
0    0    0    1
0    0    0    1
0    1    0    1
0    0    1    1
6301 6309 6317 6325

A document published by Icom available here describes the diode mods for MARS/CAP as follows:

Remove D6315 and D6316
0.1-199.999999MHz
400-470MHZ

Remove D6315
1.6-25.999999 MHz
28-29.999999 MHz
50-54 MHz
144-146 MHz
430-440 MHz

Remove D6316
1.6 - 54 MHz
68 - 74.799999 MHz
118 - 174 MHz
430 - 470 MHz

Remove D6310
5.255 - 5.405 MHz ONLY

The RF connectors on the radio are trivial to remove. There is a clip on the PCB which holds the centre pin of the RF connector, so swapping a connector is as simple as unscrewing it, pulling it out, and putting in a new one. The two-hole N-Female connectors sold by M0MAT on eBay are high quality PTFE insulated jobbies and fit perfectly. It's probably worth removing the top cover when you do this just to ensure that you are sliding the pin home correctly and not mashing the little clip.

On 70cm, into a quality dummy load, my transmit SWR was ~1.6:1. After swapping the connector the transmit SWR was close enough to 1:1 that the radio didn't display anything on the meter.

M0UKD has done a great write up on this with some good photos. Not a scary mod!

On close inspection the radio I have had a little bit of “gunk” between the little clip and the connector pin. This is no bueno, so I would recommend soldering the pin down.

Some of these radios appear to have a problem with their ALC circuit (and / or possibly the compressor) causing them to put out low-ish power on transmit. The radio I have also seems to have this issue (putting out only around 60% desired PEP on normal speech, with the MIC gain, compressor and EQ set up). Some observations from playing with this radio on the bench:
- The radio puts out a full power CW / FM carrier
- Modulation on FM / AM is good'n'strong, so there's no issue with the microphone / MIC gain controls
- Injecting a 1.5KHz audio tone from a siggen into the MIC socket produces full power at ~10mV
- Tones generated from WSJTX into the USB audio system produces full power
- Whistling, or pumping a tone into the microphone produces full power

So in a testing environment the radio does put out the power one would expect. However, on real speech, it wasn't nearly as consistent:
- When hollering “aaaaaaudio” into the microphone, driving the ALC to the top of the “normal” range and engaging 5-10dB compression, the radio only puts out ~40W when set to ~100W. Doing this on my 857d (for example) gives ~80W.
- There appear to be some hidden time constants built into the radio which push and pull the ALC during normal speech - one syllable in one moment might produce 70W PEP, then 30W PEP, then 50W PEP. Very difficult to pin down, but in the dynamic environment of “real speech” the power output is somewhat inconsistent.

There are a number of solutions which have been mooted, so I tried 'em:
1. Upgrade the firmware to E5, there are reports that Icom fixed this by changing ALC / Compressor parameters
CPU 1.11
DSP 1.03
The radio I have was sold in 2024 and came shipped with the E5 firmware installed.

2. Solder in a capacitor to slow down the ALC response
TRX Lab on Youtube shows a modification which dramatically increases the average and PEP on SSB by slowing the ALC response. I tried this and it “works”, but there are a number of issues:
- Vicous overshoot. With this modification speech peaks are ~150W
- Dramatic ALC overshoot below full power. When set to 20W the radio sill produces ~150W peaks
- ALC “pumping”. During the initial syllable power climbs slowly, then drops again after peaking leading to a QSB-like signal changes.

3. Use the setup menu to adjust power output for a given audio input signal
A mods.dk article suggests that Icom have made an error in their setup, and claims to completely solve this issue. The Icom service manual calls for a 30mVrms 1500Hz tone to be injected into the microphone socket and the TX power alignment to be performed. The article on mods.dk noted that other Icom manuals from the era, using the same model of microphone, specify this tone to be 1mV 1500Hz. He then set up the TX using 1mV audio instead of 30mV and claims to have solved the issue completely. I tried this and it doesn't work. Observations:
- When setting up the TX gain path the radio breaks HF into 160m, 20m, 10m. 6m / 4m (and also 2m / 70cm) are set up separately.
- On HF, one injects the audio tone and adjusts the radio to put out 50W. This first stage of adjustment happens in the 20m band. With this tone still being injected the radio then automatically sets the gain for 160m and 10m with a few subsequent button presses.
- When injecting a 30mV audio tone this automatic setting on 160m and 10m works just fine
- When injecting anything below ~15mV the radio cannot complete the auto-adjustment on 10m

Therefore I'm not sure how this person managed to successfully align the TX using a 1mV audio input - and 1mV is rather too low anyway for a condenser mic. Who knows… YMMV.

All this being said, these sorts of power differences don't make too much difference to intelligibility on the air. Readability and signal reports are, if they can be believed, just fine. My take is just use the radio as intended, and don't worry about it. Perhaps don't use it to drive a linear amplifier.