The Icom IC-738 HF Transceiver
by Steve Katz WB2WIK/6
The IC-738 has been on the market long enough that it's already popular, as is its more pricey sibling, the IC-736. There are only two differences between these units: The IC-736 offers 6 meter (as well as HF band) coverage and has a built-in AC power supply, while the IC-738 lacks 6 meter coverage and requires an external source of 13.8 VDC power The two rigs share a common owners manual and most other features. Because the IC-738 lacks an Internal power supply, it weighs less (19 lbs) than the IC-736 (23.1 lbs) and may be easier to handle for field operations. As far as I can tell, the IC-736 has no provision for external DC power at all. making it a base station rig in every sense.
Despite the 738 s lighter weight and potential battery-power operation, it too, is really a base station radio because it is not small. Measuring 13" x4,4" x 11.2" (H x Wx D). the IC-738 has the look and feel of a real radio. Most of its controls are large and spaced far enough apart for even clumsy fingers, and its display fs large enough to be viewed from across a large room. The IconYs 59 panel-mounted controls are easy to use, with the possible exceptions of the RF-gain and power output controls.
The IC-738 comes well-packed and includes most items needed to get on the air quickly, save for power supply and antenna. Its instruction manual is reasonably complete and well written. The transmitter output power is rated at 100 watts SSB, CW. and FM, without any note regarding possible derating requirements for high-duty-cycle operation (like FM or RTTY). I donTt know whether Icom would consider it safe to operate at 100-W output continuously or not. Also the specifications discuss only SSB. CW, FM. and AM, and don't mention the digital modes except for brief connection instructions on page 16 of the manual, which state. "The transceiver does not have an FSK mode for RTTY, AM-TOR, packet, etc.; however, you can operate these using AFSK in the SSB or FM mode." I'm neither a Digital Demon nor a Binary Bimbo, but even so. Td expect more discussion of such a popular subject
A point in its favor, though, Is that the IC-738 passed my "Can I use this radio without opening the manual?" test with flying colors. I had the rig out of the box and on the air and completed my first 50 or so contacts, wittiout ever glancing at the manual. And that's good. This is important because, in an emergency, there may be no time to indoctrinate operators; they need to sit in front of the radio and use it immediately. The more intuitive the rig is to operate without instructions, the better its design, as far as I'm concerned. On a scale of 1 to 10, I'd give the IC-738 a "9" for intuitive use. I d have given it a *10f* except its built-in, two-port antenna switch (front-panel controllable) kept switching to the unused port as I switched bands, making me scratch my head about why signals seemed so weak on bands where I knew they should be strong! (The IC-738 remembers which antenna position was last used on each band, and reverts to it. As rt comes from the factory, who knows how it will be set up? You can override this feature, but it is not intuitive and probably requires an Instruction reference.)
The first thing I noticed about the IC-738 was how uncluttered its display is. On power-upr the display panel indicates VFO frequency, mode, tuner status, VFO selection (A or B). and channel number for memory operation. That's it! I like the tidy display, and only wish it omitted the channel number unless memory operation is actually in use. It would then display only what the operator absolutely has to know. The frequency indicator numerals are large (1/2" high) and very easy to read, being dark gray segments against a bright orange background. The display is readable even operating outdoors on a sunny day.
The second thing I noticed is how quiet the receiver is. Unless the PRE AMP is activated, the noise level on every band tends to indicate about "SO" (no reading) and the ambient hiss between signals is very slight. This tempts one to turn the volume up very high, to hear something. Then, when you tune across a signal, the sound jumps out at you from the rig's top-mounted speaker. The background noise is so low that you wonder if the receiver has the sensitivity we've come to expect from modem equipment. It does. The IC-738 tests as sensitive as any receiver I've come across. The small amount of hiss is due to advancements in synthesizer, IF, and audio stage design. The rig is deceiving. It doesn't sound sensitive until you need it to be, and then it's right in there with the best of them.
Within a few QSOs made, I was able to find a few shortcomings with the Icom—not in performance, but in creature comforts. For example, the analog meter indicates only one of three possible data on transmit: either power output (watts scale), SWR (1 to infinity, with SWR = 3 at center scale), or ALC activity (thick 'normal" range). However, there is no way to tell which scale is active. There is no ttiree-position panei-mounted switch to which to refer. The meter is switched by a smgle push-button, which toggles the meter mode between the three possibilities.
Another example is the power level control, which, like the receiver RF-gain control, is a small knobless shaft protruding through the front panel to the lower left of the main VFO tuning knob. While f don't find myself adjusting RF gain much. I do frequently like to change transmitter output power, This is a control thai gets a real workout in my station, as I've never been an fcAKTR- (All Knobs To Right) operator. If I work somebody very strong. J tend almost automatically to reduce my power If the station worked is very weak, I almost automatically turn it back up. Rapid-fire contesting notwithstanding, I adjust this control a lot, and know many others who do, too. The IC-738's control for this function is too small to be easily and repeatedly adjusted.
One other thing that struck me as inconvenient is the placement of the VOX controls, which are on the rear panel. It is inconvenient to have to reach around behind the rig to make adjustments. In many shacks, it might take quite a bit of effort, and maybe a flashlight, to do so! VOX controls, especially the DELAY control, need to be more accessible.
My fast gripe is that the IC-738 lacks transmitter mike gain, Tne factory-supplied, handheld mike is quite good and sounds fine on the air. but has rather low output level and requires the 738+s gain control to be run literally all the way up to make the ALC work proper-
The only other complaints I might have is that the CW sidetone level adjustment is internal and requires removing the covers. The rig has any CW offset you might want by using the ATx control (similar to a receive RfT, but works the transmitter offset), but the sidetone pitch always remains the same. As a true CW operator I like being able to adjust everything I can when using this mode.
Okay, enough gripes. Other than these small issues, the IC-738 works really great!
If you're already familiar with modern HF gear, you won't need any training in how to use the Icom. If this is your first modem HF transceiver, read the manual and try out each function one at a time, I like that the 738 has two key input jacks, one of them three-circuit for connection of a paddle to use Ihe rig's (standard) internal electronic keyer. and a separate two-circuit jack for use with a straight key. "bug," external keyer, or computer control, I also like the rig s use of a common RCA "phono" jack for connecting the key line to an external linear amplifier. (Note: This jack is labeled SEND, which is a bit misleading. It is connected in common with pins on ACC and ACCf the two accessory Dl hijacks, and is a dual-function I/O port. If grounded, as with a footswitch, the rig will transmit, If connected to the key jack of most ampliliers, it will ground on transmit and cause the amplifier to key.) I happen to fike RCA phono plugs and jacks, mostly because they've been around for generations, are likely to be in nearly everyone's junk box, and can be picked up inexpensively at a local Radio Shack in a pinch. DINs are fine, but are also the kind of thing you might forget to bring to a DXpedition; DINs are also difficult to assemble in the field.
My first 30 or 40 QSOs with the new IC-738 were on CW. my favorite mode. (I can hear the boos and hisses, but to each his own.) The rig had no trouble producing more than 100-W output on every band. The semi-QSK worked great and so did the full-QSK (break-in) mode, up to at least 45 wpm.
On SSB. the stock microphone lacks punch unless the mike gain is turned fuliy clockwise. I'll guess that Jcom might have a fix for that by now. I did try using one of my favorite desk mikes, an old American-made Shure 444 (big and clunky, but they usually sound good), and it produced far more punch than the stock hand-held mike and reports received were excellent. With the 444, I was able to turn the mike gain down about halfway and still have sufficient audio to get an ALC indication. The IC-738 has a built-in speech processor whose function is labeled COMP (compression) on the front panel. Its adjustment is also labeled COMP and is located on the rear apron of the radio, next to the VOX controls. The compressor worked pretty well, although stations contacted advised my audio sounded more natural with it off This is a pretty subjective issue, and I'm used to getting mixed reports about speech processors and compressors. Despite most stations* reports telling me I sounded better with it off. the compressor did its job of increasing PEP output power as indicated on my Autek WM-1 PEP computing wattmeter.
This is all one can ask for from a simple speech compressor circuit. The rig's IF bandwidth, stated as 2.1 kHz (-6 dB)t is very adequate for SSB work, but might be too wide for crowded CW operation. I liked the way the receiver sounded with the factory-equipped filter on SSB, It was tight enough to prevent much adjacent-channel interference, but wide enough to allow excellent fidelity of received signals, I thought it was just about the proper balance, and the receiver is one of the best Pve ever listened to. Although I have four modern HF rigs, my standards of comparison for HF receiver performance are older equipment that in many ways outperform the latest gear. I use a 1978-vintage Drake TR-7 (transceiver) and a 1958-vintage Collins 75A4 (receiver). Believe it or not, these old-timers offer better performance under some conditions than anything built since. If a modern synthesized transceiver can hold its own against these two radios, I'm usually astounded. The IC-738 comes very close.
The IC-738 is one of the few modern HF rigs I've used having a receiver that I tike listening to on SSB. I did not have at my disposal the optional narrow CW Miters on the review model, so I can t comment on them. But the standard SSB filter works very well, and the optional filters are plug-in, not solder-in (although installation does require removing the radio s covers).
The 738 s standard, built-in ATU (automatic antenna tuning unit) is fast and smooth as silk. I was extremely impressed with it. The ATU is activated and then implemented with a single push-button control (TUNER). A fast push activates it, and a longer push engages the tuner, which makes the transmitter operate at reduced power until the tuner finds a nearly perfect match point I could not find any point in any band where I could not achieve a nearly perfect match with the antennas I normally use. I could even get my 6 meter vertical to load up on 80 meters, as well as achieving success with my 160 meter dipole on every single band, including 10 meters.
As an experiment, I tried loading up a 12" long Radio Shack clip lead (#20-gauge wire with an alligator clip on each end) on 20 meters. I clipped the far end of the lead to a paper poster hanging on the wall (a contest award, actually!) and the rig found a match within a few seconds. I answered a strong W5 in Arkansas, and completed a contact with the clip-lead antenna. I should note that my shack is at ground level, enclosed within the house and having no windows to the outside. This is not a very objective test, and I have no way to measure the impedance of the clip lead on 14 MHz. but it goes a long way in demonstrating the effectiveness of this tuner. The ATU obviously has memories, and seems to know how to retune itself (once set
points have been established by prior use) as the receiver is being tuned across each band. The problem is that I don't know how many tuner memories it actually has. because the manual does not address this point. The ATU also retunes itself to the presets as one tunes around on the receiver which makes transmitting at full power on nearly any frequency as easy as pressing the PTT button or key; but the manual does not discuss this point either. This is a shortcoming of the manual: The rig has features that work very well but aren't even mentioned!
The receivers PREAMP works well to boost receiver sensitivity on the higher bands like 10 meters, but serves no useful purpose on the noise-laden lower bands. (This is typical of many modern HF rigs. Nobody needs a preamplifier when using any reasonable transmitting antenna to receive below 10 MHz ) The IC-738 also has an ATTenuator, which reduces signals reaching the front end and might be useful when receiving on or near the frequency of extremely strong signals
The IC-738 contains the normal complement of features and functions found in HF base station transceivers of the nineties. I've already mentioned most of the important ones, but a tour of the rig's panel controls would be incomplete without mentioning these as well.
The AGC switch is a push-button that toggles between FAST and SLOW, Although it lacks an OFF or MEDIUM mode, I found it satisfactory for most operations. The AGC performance is good. Concentric to the volume control (labeled AF) is a squelch control (SQL) that functions on all modes but is normally used only for FM work in the 29.6 MHz range. The rig has a pulse-type noise blanker activated by a push of the NB switch. It works as well as most I've used, and does a fine job of reducing ignition noise
The buill-m electronic keyer for CW work has a speed control (KEY SPEED) concentric to the mike gain (MIC) control and adjusts smoothly from 7 wpm to 41 wpm. The keyer works fine, but I usually use an external memory keyer for contest work Still, an internal keyer is handy in a pinch or for portable work. The previously mentioned antenna switch is an internal relay that is activated by a press on the ANT button, above which are mounted two LED indicators to indicate the antenna p" or u2") selected. The ANTenna selection data is normally automatically stored in band memory, but may be overridden at any time by a push of the ANT button.
The RF PWR control adjusts the rig's output power from less than 5 W to full output (100 W, except on AMt where the max is about 40 W). It works well and adjusts output smoothly, but I wish it had a larger knob to make for easier frequent use. Just below the main VFO tuning knob is a tuning brake adjustment screw, turned by a small Phillips-type screwdriver, to vary tuning dial tension. I found the factory setting just about perfect, but you can adjust this to your heart's content, from quite loose (allowing a spin across the band) to fairly tight (won't jar the knob even if the rig is bumped hard).
The main VFO tuning knob operates smoothly and has a weighted feel. Large enough to provide a comfortable grip, it also has a large circular finger-hole depression for rapid motion using a single finger. I liked it a lot. This is one of those features that many hams overlook, but can make a real difference in operating, especially for long periods (like contesting). I wish every rig out there had a tuning knob like this one. Just above the tuning knob is an electronic dial lock switch (LOCK), which prohibits the VFO from changing frequency. I never actually use this function, although most modem rigs offer it
The IC-738 offers 101 channel memories and six pages of the owner's manual are devoted to describing its operations. The memories will store frequency, offset, mode, antenna selection, and so forth. The memories can be of great assistance in a variety of ways, but I use them mostly for rapid band and mode changes, or to store repeater frequencies in the 10 meter FM subband. A very handy feature is the fC-738's scratchpad memories. These are five additional memories (extendible to 10 if desired) which do not occupy space in the normal 101 memory register, and can be instantly stored and recalled with a single push of the MP-W (Memo Pad-Write) or MP-R (Memo Pad-Read) button. If you want to store the frequency to which you re tuned, and then tune somewhere else, just punch the MP-W button and spin the knob. A single push of the MP-R button instantly recalls where you were before you tuned around, These memories stack up. and the factory default allows stacking five of them; if you add a sixth, the oldest one stored will drop out of the memory, and you'll be able to recall the latest five. This feature is very handy when DXing or contesting, and easier to use than the other 101 memories, which require more thinking and keystrokes.
Like any good, modem rig, the IC-738 has a frequency/band keypad, a set of 12 push-button switches that can be used to change bands instantly, or access specific operating frequencies without the need for spinning the VFO knob. Eleven of these buttons are labeled with band designators and/or the numbers "1w through "0." plus a decimal point. If you re on 80 meters and want to switch to 10 meters instantly, just push the "28" button, and you're there, right on the last 10 meter frequency used. If you want to dial up a specific frequency without turning the VFO knob, that's easy, too For 14.225 MHz„ you'd press FREQ INP frequency input), followed by 1-4-.-2-2-5, followed by ENT (enter).
Another way to tune around besides using the VFO knob is to use the UP and DOWN buttons, located below the keypad just discussed. If you press and hold either button, the rig will electronically tune up or down the band in programmed steps from 1 kHz to 1 MH2. Having this function on the front panel is almost superfluous, however. The function may be duplicated with the UP and DOWN buttons atop the handheld microphone (and also found on most desk mikes). That can be handy.
Operating "split" with the 738 is as easy as with most modern rigs, and involves using the SPLIT button in combination with the A/B (VFO "A" or HB" select) button. If you wish to check your transmit frequency instantly when operating split, a push of the XFC (Transmit Frequency Check) button will toggle the VFO from the receive VFO to the transmit VFO, If you press XFC and hold it down, the VFO knob may be tuned to change your transmitter frequency to anything desired, without changing your receiver frequency. When you release the button, your transmitter frequency will be whatever was last tuned. Very handy for working split-frequency pileups.
A Quick Tuning switch (TSt for Tuning Speed) allows changing the VFO tuning speed from normal (1 Hz or 10 Hz, user-selectable) to fast (1 kHz) for rapid band excursions. When the normal (slow) tuning mode is selected to allow 1 Hz tuning, a third numeric indicator to the right of the frequency decimal point illuminates; otherwise, when the normal tuning mode is 10 Hz. this digit is blanked and not part of the frequency display. Not a bad idea. Using the 1 Hz tuning mode can be painfully slow and is rarely necessary (can you really hear a 1 Hz change?), although it might be nice when using a sharp CW filter and the NOTCH function,
The mode selector buttons are easy to use and large enough to operate quickly. Located just to the left of the main VFO tuning knob, the mode switches are labeled SSB (toggles between LSB/USB); CW/N (toggles between CW wide and CW narrow if the optional narrow filter is installed); AM; and FM/TONE (toggles between FM and FM+CTCSS tone [encode] if the optional UT-30 tone encoder is installed).
The IC-738, like most recent Icom products for HF, has both RIT and XIT functions. RIT, Receive Incremental Tuning, is featured under various names on all HF transceivers made in the past two decades or so. XIT allows similar adjustment of the transmitting frequency, independent of the receiving frequency. Using either one or the two together, you can work split up to about 20-kHz separation, without needing to use the SPLIT function. The RiT and XIT functions are individually addressable but share a common tuning knob labeled RIT/ATX.
The rig also has two interference-fighting tools: PBT (PassBand Tuning) and NOTCH filtering. The PBT control has a center off detent, while the NOTCH control has its own separate on/off switch. I found the passband tuning PBT function extremely helpful under a variety of situations, but the NOTCH not particularly useful. Notch filters are mostly intended for reducing the intensity of an interfering single-frequency carrier, and don't help much in reducing interference from a nearby SSB station. Used in combination with some patience and skill, the PBT and NOTCH can complement each other to further reduce some types of interference. But I think the PBT will be of greatest help for most users, and the one in the IC-738 is good.
The unit has the usual scanning features that may be used to frequency-scan between memories or preset band limits. The scan modes are "programmed scan," "memory scan," and "select memory scan," and are described in two pages of clear instructions within the manual.
The IC-738 uses a triple-conversion receiver with "up conversion*' as has become the norm. Its receive first IF is at 69 MHz. second IF at 9 MHz, and third at 455 kHz. allowing the use of conventional bandpass filters. Following the signal path from the antenna jack, received signals are first routed through the antenna switching relay, then through the T/R relay, a low-pass fitter, and the receive attenuator switch before they reach either a low-pass filter for reception up to 1.6 MHz, or a set of diode-switched bandpass filters for coverage of 1.6 through 30 MHz* Seven such bandpass filters are employed, each covering an octave or less.
The filtered signals are routed to the receive preamplifier switch, where, if the preamp is engaged, they are amplified by a pair of 2SK2218 JFETs in parallel. Signals are low-pass filtered once again before reaching the first RF mixer, a pair of balanced 2SK2171 JFETs, then are bandpass filtered at the first IF of 69,0115 MHz before being amplified by the first IF amplifier, a 3SK121 dual-gate MOSFET. Received signals are then bandpass filtered once more at 69.0115 MHz before being mixed down to the 9.0115-MHz second (F in a diode balanced mixer, type ND487CIT LO injection to both mixers is provided by the PLL unit. The first mixer injection provided by the PLL tunes from 69.0415 through 99.11499 MHz, thus allowing continuous receive coverage of nearly the entire 30-MHz spectrum. This output range is achieved by four VCOs having outputs 69.0115 MHz above the received tuning range. (For example, VC01 produces 69.0145 through 76.0114 MHz; VC02 produces 76,0115 through 84.0114 MHz, and so forth). The second mixer injection at 60.000 MHz is provided by a frequency-doubled 30-MHz crystal oscillator, also located in the PLL unit.
The 9.0115-MHz second IF is filtered using another 2-pole bandpass monolithic device, after which the noise blanker gate is located. The noise blanker gate is a set of four MA77 diodes driven by a gate control circuit located on the output of the noise blanker loop. The noise input to the blanker is sampled at the 9.0115-MHz IF and is amplified and detected in a noise blanker AGC loop of its own prior to driving the gate control circuit. After the diode gateT the 9-MHz IF signals are amplified once more and then bandpass filtered by one of two 9-MHz crystal filter units. (The 2.1kHz unit is standard, but an optional narrow CW filter may be installed at this point as well. AM or FM signals continue through the diode filter switch without filtering and go on to the third mixer, whose description follows.)
The third mixer is an integrated circuit type uPC1037HA having LO injection at about 9.4665 MHz, generated by the crystal oscillator with some frequency shift afforded by the PBT control. This mixer's output at 455 kHz is filtered by one of four diode-switched ceramic filters, three of which are standard (SSB/CWt AM. and FM) and one of which is optional (CW narrow). The 455-kHz final IF signals are postamplified by a 3SK131 DGMOSFET and then directed to either of two additional IF amplifiers (2SK882 JFETs) before reaching the appropriate detector stage for the mode selected.
The detectors used are a UPC1037HA integrated mixer/product detector for SSB and CW; a set of diodes (HSM88AS) for AM; or a MC33570R integrated FM subsystem that acts as an IF amplifler/limiter and discriminator. Injection to the product detector is provided by a mixer (another UPC1037HA) used to provide the difference frequency of the 9.4665-MHz crystal oscillator previously discussed and a 9.0130-MHz "BFO" crystal oscillator. This difference frequency minus the 455-kHz IF, provides an audio frequency output from the product detector This AF output goes through the NOTCH switch and filter before being amplified by a set of cascaded 2SC4Q81 bipolar audio amplifiers, whose output is level adjusted by the front-panel AF (volume) control before driving another 2SC4031 and the final audio amplifier, a UPC1241H integrated audio amplifier rated at 2.6 W output.
AGC control is provided by an AGC detector driven by the last IF amplifier (455 kHz), then amplified by a 2SC4061 to control the IF amplifiers used at 69 MHz, 9 MHz, and 455 kHz. It is thus an IF-derived control loop that simultaneously adjusts the gain of all the IF amplifying stages in the receiver and is quite effective.
The circuitry is modern and well thought-out to provide a minimum of spurious or image signals. Most of the same circuitry is also used in the transmit signal path, as is typical of modern PLL-tuned transceivers, The primary differences are described below.
On transmit, the output of the 69.0115-MHz bandpass filter (already discussed in the receiver section) drives a 2SK882 JFET IF amplifier (which also provides "S" meter data to a meter amplifier) and then a balanced mixer using a pair of 3SK131 dual-gate MOS-FETs, whose output is high-pass filtered to drive a 2SC4673 bipolar RF amplifier. Injection to this mixer is provided at 69.0415 to 99-11499 MHz by the PLL Unit and comes from the same set of four VCOs discussed earlier for receive injection tuning of the first RF mixer
The output of the 2SC4673 RF amplifier is low-pass filtered and then attenuated to provide a fixed terminating load for this filter before being delivered to the PA Unit The PA
Unit contains a pre-driver (2SC1971 bipolar, operating common emitter with emitter peaking), driver (push-pull 2SC3133 bipolars, grounded emitter with adjustable base bias), and power amplifier. The final power amplifier (PA) uses a pair of push-pull 2SC2904S, grounded emitter with base bias regulation provided by a 2SD1406 series regulator.
Also located in the PA Unit is the fan control circuit, which appears to be driven by a thermostat and uses a 2SB909M bipolar fan driver There is no discussion of how this circuit operates, and because I am unfamiliar with some of the parts used, 1 can't say much about it. But it does work, and the fan in my unit came on within a few minutes of transmitting at full power and kept the unit quite cool to the touch. There are actually two fans used in the IC-738 and they appear to operate in parallel.
The output of the PA is routed to the Filter Unit, a set of relay-switched (12 relays used) low-pass filters using dual pi*sections to cover each band segment (except for the 10/14-MHz LPF, which uses three pi-sections). The SWR-detector circuit, which drives the panel meter to indicate forward output power and SWR, and also provides reflected power data for the control loop that reduces the transmitter's drive in the event of a significant mismatch (as well as data for the ATU control loop), is also located in the Filter Unit, as is the two-position (T or "2") antenna switching relay, RL16,
The ATU is located on its own board called the Tune Unit. It uses 12 relays to switch combinations of "F-networks to achieve the best possible match to a wide range of antenna impedances. While Icom's published specifications claim a tuning range of 16.7 to 150 Ohms (thus, it should match any antenna having a VSWR = 3 or less)T i found in actual use the tuner was better than this and successfully matched a clip lead on 20 meters. (This was discussed earlier.) I also used the ATU to match rain downspouts, window screens and other ridiculous loads with amazing success. The relay-switched matching network uses fixed Input and output capacitors and tuner-controlled variable capacitors across the fixed units, along with combinations of shunt "V inductors. Ifs a slick system that works better than most I've used.
In most cases, the ATU can find a successful match point for reasonable antennas within just a second or two. The ATU is in the RF path only when transmitting, which is normal. (It cannot work without a signal being transmitted, and we cannot transmit a signal outside the amateur bands. Thus, the tuner would be useless for the general-coverage receiver without having some way to adjust it manually.) The tuner may be bypassed entirely with a single touch of the front-panel TUNER button. When operated in the "bypass1* mode, the display screen indicates THRU. When the tuner is searching for a match, the display screen flashes a TUNE warning. When the match is found, the TUNE indicator stops flashing but remains illuminated.
I like the IC-738. It works very well and is easy to use. It meets its specifications with considerable margin, especially with regard to receiver performance, transmitter output power, and ATU matching range. If this were a perfect world, I'd get Icom to move the VOX controls to the front panel, make the RF power output level adjustment a larger knob, add a three-position rotary switch for meter function control on transmit, add a more readily accessible CW sidetone adjustment, and include sidetone pitch as well as volume, add more mike gain, and improve the owner's manual to more fully discuss some features and functions that are not well addressed.
As someone who has done a lot of technical writing, I'm more critical than most regarding operating manuals. And as someone who has done an awful lot of operating of amateur equipment over the past 30 years. I'm probably also more critical than most regarding control placement. After all this time, I know what I like, and no rig I've ever seen offers everything in one box. But Icom's on the right track. When all is said and done, on-the-air performance is more important than slick manuals and lots of knobs. And the IC-738 performs exceptionally well. Ifs a pleasure to use under most conditions, and I'll give it a thumbs up as a wise choice for a broad spectrum of users.