ICOM IC 7300 marks an important inflection point for the “big three” amateur radio manufacturers: The beginning of true software-defined RF processing from one end to the other.

As the first full SDR transceiver from “the big guys”, it is interesting to consider the price point, feature set, and performance both in light of the legacy analog designs, and as compared to what SDRs are known to be capable of.


First, the price point. My (first, I have two now) unit of the IC-7300 cost me $1500.00 in May of 2016, which put it above the ranks of an entry-level radio (which is more around the $500 to $750 mark) and well below those analog designs that are marketed as higher, and very high performance rigs.

The feature set is interesting. There are considerably more software features than hardware features, and as is my habit, I’m going to concentrate on the lacks, rather than the good stuff — most reviews go the other way, working hard to like something. I work hard to find the flaws, and while I enjoy the nice features, I leave it to others to sing their praises for the most part, unless they surprise me in some way.

Notably lacking on the hardware side are additional antenna jacks; ethernet connectivity; an external monitor port; and some very strange out-of-the-box omissions of basics such as a carrying handle and mounting bracket (the IC 7300 is a 12 VDC radio, and as such… if it had a mounting bracket… it would be easy to use in mobile fashion.)

There’s a USB port, but this is not really a good thing as compared to a real network port; USB support requirements are radically different between Windows, OS X and Linux, and sure enough, USB based support and maintenance software is pretty strongly tied to one OS or another (Windows in this case), which is really a shame. Especially considering that Ethernet hardware, which is easily and consistently supportable among all three major operating systems, is so inexpensive. Big thumbs down there. Perhaps Kenwood and/or Yaesu will school Icom on this one. The Kenwood TS-990 has an awesome network interface, for example.

There’s an SD/SDHC card slot in the front of the radio. This is used to save settings, images of the display, voice recordings in both TX and RX mode. It can also be used to update the radio’s software and FPGA programming.

The software feature set is also interesting, both for what it offers, and for what it does not. This is by far the most complex area to consider, so I’ll do that last.

Performance is a mixed bag. On receive, as long as the radio isn’t exposed to very strong signals, it does very well. As with any competent SDR software, adjacent channel signals only appear in your passband if they actually are in your passband. The slope of the passband is cliff-like, rather than softly curved as with an analog system. The value of such a bandpass cannot be overstated; it is not only easy to receive a weak signal next to a strong one, you won’t even know the strong one is there unless it is actually too wide — which is not usually the case for properly set up amateur equipment.

The IC-7300′s memory system leaves a lot to be desired. It stores the frequency of operation, and the filter bandwidth settings, but that’s about it. The AGC setting is not stored (something you’d almost certainly set completely differently for data and for voice); nothing to do with the waterfall is stored in a memory; the monitor setting is not stored; etc. None of this is configurable. All in all, a pretty darned feeble memory system; nearly useless for anything but the most mundane operations.

What you can do to run digital modes when your IC-7300′s firmware is lower than 1.30 using the audio codec is use the USB-D (or LSB-D) DATA mode instead of USB (or LSB); these are separately configured, and the IC-7300′s memory system distinguishes between the two. Thanks to one of my site visitors, Art, for that information — I missed this entirely on my own.

Update: If you have an IC-7300 with firmware 1.30 (or later, one presumes), ICOM has made two important changes. First, the transmit bandwidth of the data modes is now capable of full voice reproduction. Second, there is a new “USB and MIC” setting for the DATA modes under…

    Menu/Set/Connectors/DATA MOD/MIC,USB

…which will allow you to alternate voice and USB data fairly easily if you have appropriate receive bandwidths set up and ready to go.

One lovely “hardware surprise” was an under one ampere power draw at 12.6 VDC during receive operation. I can run this radio all day on a single car battery without any strain; it can pull up to 21 amperes on transmit, but at least in my case, I spend a lot less time transmitting as compared to how much time I spend receiving, so this is really a win for me. Just one 100 watt solar panel provides enough current to run the radio on receive and at the same time dump three to four amps into the batteries for evening fun.

The waterfall does an excellent job of revealing signals, even fairly weak ones, as long as you set it up with sufficient contrast to do so; this is done with the “Waterfall Peak Color Level” setup option, which I suggest you set to “Grid 2.” This will provide the contrast you need to some of the lower level signals (you can still hear some that won’t show up on the waterfall… they didn’t quite get the waterfall sensitivity right.) Careful adjustment of the REF setting to drop the noise floor down into a fine mix of dark blue and black is required as well, and this needs to be done any time you change the region the waterfall covers, as neither the VFOs or the memory system remember any of the spectrum/waterfall settings. As I write this, I can just barely hear a station in our audio group (hi, Larry!), he isn’t moving the s-meter at all, but I can see his signal clearly on the waterfall. That’s how it should work.

Waterfall useful, but...

Waterfall useful, but…
spectrum visually clipped

The spectrum, which is the waveform that is displayed above the waterfall, suffers from a rather severe design error: When adjusted so that the waterfall is most useful, large areas of the spectrum waveform disappear behind the waterfall.
Spectrum useful but...

Spectrum useful, but…
waterfall a rainbow mess

This is very disconcerting, and leads to a “flashing” effect where signals pop up and down like jack-in-the-box toys. There’s no adjustment to offset the spectrum vertical position; if you adjust the REF for the waterfall, the spectrum moves up and down, but that affects the waterfall colors such that if you have a good spectrum, you have a lousy waterfall, and if you have a good waterfall, you have a lousy spectrum. Oops.

Overall, the waterfall is far more useful than the spectrum could ever be, so I always adjust the reference point for that. It’d sure be nice if both worked together properly, though. Until such time, the following shows the best behavior I’ve managed to get out of the display. Details on the settings I use are in the caption:

averaging=4, Grid=2; REF adjusted so that baseline about 50% intersects the waterfall, leaving negative excursions below the visible line. This keeps the waterfall as dark as possible while still showing most weak signals.

averaging=4, Grid=2; REF adjusted so that baseline about 50% intersects the waterfall, leaving negative excursions below the visible line. This keeps the waterfall as dark as possible while still showing most weak signals.

When doing some AM broadcast band DXing, one very strange thing that happened was that when I was tuned to one end, for instance, 1.450 KHz, the signals on the other end of the spectrum, for instance 540 KHz, drop in amplitude on the waterfall as if there was a tuned front end. Which is just… weird. The claim is that this is a direct-sampling SDR system. As such, it should not only be able to see the entire band without variation, that’s one of the key benefits. It’s kind of mind blowing that they’d put this together in such a way as to make it intentionally blind. This is so weird, I want to call it a bug, but it acts like there’s analog filtering going on in there, and if that’s the case, it’s not a bug, it’s an outright design error. (It turns out that yes, there are filters in there. Too bad.)

I would have liked the ability to control the color palette on the waterfall. The colors chosen are generally adequate, but not by any means great. The waterfall goes from black, to bright blue (it should be a dark blue), to bright aqua, to yellow, to red. There’s no way I have found to turn off the fill under the spectrum waveform, but I was able to set the fill to black and the waveform to green, which produces an easily interpreted display. I expect that this is a feature I’ll see in a higher end Icom later on. It’s just that kind of “value-added” software feature they can charge for without costing themselves much.

There’s some user interface weirdness with the audio scope / spectrum / waterfall (which, in and of themselves, are great) where if you switch to various menus and features, these go away — the radio completely forgets that you had them turned on. For instance, I have the waterfall and spectrum up; I press MENUMENU Button, then VOICEVOICE Button; pressing EXITEXIT Button now brings me back to the main display, but the waterfall is gone. The area below the frequency readout is blank. It’s a little annoying, but the waterfall can be brought back fairly easily, just a long-press of M.SCOPEM.SCOPE Button will do it. A problem with having the audio displays on is that the waterfall cannot be adjusted in this state. The lack of a user menu (or the ability to add items to the “quick menu”) shows strongly in situations like this.

Now, about the features. There’s one notch filter. It is an RF filter, which is good. It has an automatic mode and a manual mode. Both work reasonably well, but the notch is too wide in the “narrow” setting, absurdly wide in the “wide” setting, and both are applied after the visual RF processing instead of before it, so they don’t remove the interference on the spectrum or the waterfall, which is bad.

Likewise, when the noise blanker is on, the waterfall does not change at all. Icom is taking the waterfall display RF before the noise blanker processing; that’s a design error in my view. Turning on the blanker should be able to remove it from the waterfall as well. Because this was not done, in the presence of impulse noise, the waterfall often becomes useless, a far worse problem than with the notch processing.

The noise reduction is a very poor performer, but that’s also standard. No one has put a good noise reduction system on the table with any radio design, SDR or analog. So no points off for this; it’s state of the art, it’s just that the state of the art is terrible.

Back to the spectrum and waterfall for a moment. One irritating shortcoming is that in fixed mode, even in general coverage, the band sections really are fixed (and there aren’t very many of them.) If you step outside of them, you get no spectrum / waterfall shift; the display just sits there dumbly showing part of the RF domain you’re not working with. That’s a pretty serious limitation, and it’s totally a software issue with an SDR system, so I’m not inclined to be forgiving about it. When you move out of the displayed region, the waterfall and spectrum should move with you in “pages.” That’s how all the good SDR software does it.

So in order to see what’s going on, you have to go with center display spectrum, but of course when you do, signals slide around as you tune and a good bit of the advantage of having a waterfall / spectrum display goes right out the window.

There’s a menu item for SWR, which opens a display that contains an SWR graph. Using this involves pressing what looks like a “play” icon (a right-pointing arrow), then pressing transmit or the mic transmit button for each point on the graph. You can save the resulting graph by holding “recall” for one second. This function seems to be informational only; it does not set the antenna tuner, or at least, I don’t see how one might do that, which I found… odd. With the ability to look across the band in steps, paying attention directly to the SWR, it seems like the most natural thing in the world would be to set the internal tuner using that information.

On the subject of the antenna tuner, it’s good for up to 3:1 SWR compensation, and it remembers up to 100 tuned locations. As you tune, it will switch to the nearest tuned location if that location is within 1.5% of the previously tuned point, otherwise, it turns itself off. This means you can arrange for the tuner to understand a fairly wide band segment, which is a very nice feature indeed. You can also use an external antenna tuner if you prefer.

The printed manual also talks about a “1/4 tuning function”, indicating that there is a button in the function screen to turn this on; this only works (or appears) in the CW, SSB-D, and RTTY modes.

The RTTY capability… it’s minimalist to a fault. RTTY itself is fairly uncommon now, so this isn’t exactly a deal-breaker, but still — this isn’t so much “just barely acceptable” as it is “just barely there.” The radio only supports 170 Hz shift, and no other modes are available. No PSK, etc. I don’t think this will be of much use for most people. The manual does talk about supporting other digital modes via external software, but of course, then the whole “standalone” capability Icom touts in the IC-7300 advertising goes by the board.

This is an SDR review

One thing I want to make very clear is that this is an SDR review, not an analog radio review. In comparison to an analog radio, the IC-7300 wins on almost every point, hands down, with the exception of ability to receive in the presence of extremely strong in-band local signals.

Just prior to purchasing the IC-7300, I had a Kenwood TS-990, an expensive, high end analog radio. I also had a Yaesu FTDX-3000. I sold both of those because they couldn’t match the receive performance of a $500 SDR, and neither Kenwood or Yaesu had lifted a finger to seriously address their respective radio’s shortcomings in the time I’d owned them, over a year for both. Once I purchased my first IC-7300, I found its receive performance was on par with what I expected from a good desktop SDR (which means way better than the TS-990 and the FTDX-3000), and I bought a second IC-7300 for my mobile trailer setup.

As an SDR, the IC-7300 could be better, sure. As I stated in the main body of the review, there are lots of little shortcomings, and a couple of real screwups, too. But I wouldn’t buy an analog radio any longer. I see the IC-7300, and presumably its successors, as the end of the line for analog. If you want a great SDR radio, I’d say wait. If you just want a radio that’s better than any analog radio you can buy, with the caveat about local signal strength in mind, the IC-7300 is that radio. Just the first of many, I am sure.

One last thing: If you’re thinking you might still want to buy an analog radio, I feel I ought to warn you that I am quite convinced that these SDR radios will adversely affect the resale value of any analog radio on the market. If the IC-7300 isn’t high-end enough for you, my advice is to wait, rather than buy analog now.

Trying to get the radio into a useful state for operating RTTY, I attached a RTTY keyboard using a USB-A to USB-B adaptor. The radio failed to recognize the keyboard. That leaves transmit capability limited to memories you set up in advance. So if you want to respond with correct signal reports, for instance, having a memory for each report (5-7, 5-8, 5-9 etc.) will take up most of the RTTY transmit memories, leaving insufficient room for CQ, equipment reports, identifiers, etc. Keyboard support is, in my view, pretty much a “must have” here. Perhaps it’s just a bug; there is a menu option to turn an external keyboard on in the SET –> CONNECTORS menu, but I was unable to figure out what it did. I was hoping it would change the USB connector mode from client to host so it could talk to a keyboard, which is the only use I have for a USB connection on the radio as far as RTTY goes in the first place. If this is a bug, I hope they fix this ASAP, as I do happen to like operating RTTY and it would be very useful to me to be able to do so without having to add an external computer to the mix of gear, as I do a lot of portable operation.

Note that the USB connection is also intended to act as a USB sound card; this allows a host computer to do digital modes with it, which is nice in and of itself, but takes away from the standalone RTTY capability which is not nice. It seems to me that the whole point of built-in RTTY in a radio such as this is its ability to operate sans computer assistance. If you want an SDR radio that has to be plugged into a computer to run RTTY, Flex and Anan have offered that for years now — and they’re much better radios than the IC-7300 is.

Deep menu surfing to do common tasks becomes one of the worst oversights in the radio’s design when combined with the fact that there’s no way to set up one button access to deeply buried, but often needed, settings like these.

The radio’s VOX circuitry doesn’t activate when USB modulation arrives. So you have two choices here, barring more USB complexity: one, you can push transmit manually, then trigger your AFSK source, or two, you can (I can’t believe I’m even suggesting this, but…) send your AFSK output to an audio line-level output on your computer (you might need to add a sound card for this, or a USB sound “card”), wire the resulting audio the IC-7300′s ACC jack audio input (which is connected to the VOX trigger), and then the radio will work in a reasonable fashion. Of course, now you have analog signals running around, and RFI issues become much more likely. The USB connection / CODEC would have been so nice, if they’d just thought it out better. As it stands, it’s mostly just annoying, and requires (under OSX) a completely separate driver just to allow the RTS signal to be raised via the USB port, which is the only way I’ve found yet to actually key the radio from a computer without yet more hardware / cabling.

The last thing I’d point out as a lack is the bar graph S-meter. They really could have done better as far as meter display goes; the bar graph meter is pretty minimalist. The most I can say for it is that it works… which isn’t really saying much. Compare it to the metering on a Yaesu FTDX3000 or my own SdrDx software and you will see what I’m talking about.

Transmit performance is excellent. Good reports all around.

I’d say that this radio is priced on the high side for what it is, and what it does. I can get higher performance from free SDR software for metering, noise blanking, spectrum, waterfall, notches, and a fairly long list of other features, with a receive-only SDR for about $500. Which means that in terms of expenditure, you’re paying $1000 more just because this radio offers a transmitter — and getting poorer receiver performance. ICOM was fairly bold, I think (I’m intentionally being kind, I hope that’s obvious), selling this at a $1500 price point. But at this time, there is no competition from Yaesu and Kenwood, so Icom can pretty much do what they want. When that changes — and I am quite sure it will — it is my hope we will see price competition that has some sane relationship to features, user interface, and performance. Right now, this is a very interesting radio to own, even a collector’s item in some sense, but I would not say it is reasonably priced for what it is.

IC-7300 Rating as an SDR

 0 - Truly awful
 1 - Bad... could be worse
 2 - Just barely usable
 3 - Reasonably usable
 4 - Usable performance
 5 - In-class (average) performance
 6 - A little better than average
 7 - Good
 8 - Very good
 9 - Excellent
10 - Truly awesome

I give the radio overall a 6/10, while reserving the option to change that when / if they come with a software update for the radio that addresses some of its shortcomings. By 6/10, I mean: better than average, with lots and lots of room for improvement.

Also, this review is subject to update and change in order to produce the best accuracy I can and/or increase detail about features, shortcomings, etc. Also, typos. I hate typos. They get fixed when I learn of them.

The IC-7300 reviewed here was purchased by me directly from AES at full retail price. The transceiver was supplied to me (and reviewed) at the following software / firmware levels:

Main CPU: 1.12
Front CPU: 1.01
DSP Program: 1.06
DSP Data: 1.00
FPGA: 1.11