- Amateur Radio (Ham Radio) Community

Call Search

New to Ham Radio?
My Profile

Friends Remembered
Survey Question

DX Cluster Spots

Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement

[Articles Home]  [Add Article]  

IC-745 Service and Upgrade

(SV9DRU) on November 8, 2018
View comments about this article!

A "new" old rig IC-745, resurrection process. A long overdue production error correction.

Radio troubleshooting adventures and service tips by Marinos, sv9dru.

Early August 2018

Getting another "new" radio in the shack.

Last week I received a fixer upper IC-745 (in excellent cosmetic shape), for littel $$ due to low TX power. Had the problem located and fixed that same afternoon, proven to be nothing more than dirty RF power potentiometer. Spent the rest of the day well into late night hours, going through the service manual, tweeking, checking and peaking things, just to bring the old guy back to its factory specs.

In the next couple of days, I impemented all the recommended upgrades for the radio, including changing the plastic VCO trimmers and the 2nd LO one with ceramics, Upgrading the 2nd LO for better frequency stability and changing the RAM board original BA2325 battery (had started going soft) with a new one.

Finally checked the PLL board PS circuitry and changed the electrolytics close to the 7805 regulator due to the local heat stress present in the area. Did the same for the PS little module located just above the LOGIC board.

All went fine, BUT a stange RX distortion was present, especially noticeable when using headphones, mainly in AM mode, but to a lesser extent affecting SSB also.

The audio sounded distorted, just like if the NB was activated while receiving strong signals (with the NB off). for anything above S9.

Backing the RF gain just past the peaks (S-meter) for the specific signal, aleviated the problem. Even though there was the variable negative AGC voltage generated, depending on the signal strength of the station, it seemed to always be just one step behing the voltage needed (measured after reducing the RF gain to abolish the distortion) to prevent the audio from breaking up.

The IC-745 AGC Odyssey.

As most of you, who enjoy fixing your own radios know, troubleshooting is a highly addictive bussiness ! I personally can go for days and nights, spending most of my free time either thinking about the problem, formulating different scenarios as to what might the problem be, and then checking my hypotheses, by actually implementing the necessary actions to the rig and checking the results which prove or disprove the original hypothetical pathophysiology.....Just like dealing with an unknown illness in a new patient.

Reseating and cleaning the AGC connectors to the MAIN and Front panel boards, did not fix anything.

Checked all the AGC circuitry, and started changing things just to see if I could locate the problem. Changed the AGC electrolytics C122 and C123, the D4 and D5 Germanium diodes and C13 associated with rectification of the ACG voltage, Changed C2 and checked the values of all the resistors in the AGC detection circuit, around Q5, Q43 and Q4... Nothing

Finally changed the transistors themselves...... NOTHING still.

AM just sounded like garbage.

By now a full 10 days had gone without being able to solve the problem, Going through the schematics for one more time, started "poking around" the Q44 / Q45 area which are 2 swiches shown only at the expanded schematics (at the end of the service manual).

These are used to swich an additional RC paralel circuit while in AM mode composed of R250 (1M) and C123 (10uF) in paralel to the AGC time constant SSB circuit of R35 – C122.

Normally, the Q45 (2sc945) is a NPN transistor, which is controlled via Q44 a PNP one. Q44 (2sc1015Y) is activated by a (+) Voltage applied to its emitter when AM mode is selected.

When activated, Q45 is pulling the AGC voltage to ground through the R1250-C123 paralel circuit.

Measuring the voltages of Q45, did not show that. R250-C123 were still "floating" above ground potential in AM, and when I manually conected R250 to groung fixed the distortion !!

That was the first concrete evidence that I finally was close to discovering what the problem was.

First, looking at the Q44 (2sc945) pinout, I realized that, contrary to its silkscreen image,which showed it to be correctly installed, it should have been 180 deg shifted.

I thought aha !! now I get it...the silkscreening of the board had been wrong all along....Proud of myself, proceeded to removing Q45, verifying its functional status with my DMM and reinstalling it 180 deg away from its original location.

Turned the radio on, expecting to hear smooth AM signals, only to hear EVEN MORE DISTORTION !!!

Rats ! Now what ??

Had planned to exchange Q45 with a new one. Thinking that eventhough it had checked ok with the DMM, there could still be something wrong with it functionally, and while turning the MAIN board for the n-th time upside down, I was SHOCKED to actually see what was going on. R250-C123 instead of being connected in the printed circuit to the collector of the NPN Q45, was actually connected to the collector of Q44 (the PNP one) !!!!

Then finally all came together: There was a mislabeling on the board, which had Q44 exchanged with Q45. The silkscreen outline of the 2 transistors was correct, but the labeling of their numbers was inverted.

After exchanging them, finally the distortion was corrected !!

Or should I say it was corrected for the AM mode.... The sideband signals were still not very smooth sounding, regardless of the AGC time constant. Above S9 when in SSB, there was still a considerable amount of "overmodulation" type audio distortion at the peaks.

This was also aleviated by backing the RF gain to a position just above the highest signal peaks (S-meter), just as when dealing with the above mentioned AM distortion.

The AGC circuit details

The AGC negative feedback voltage is generated by sampling the output from Q1 (MAIN unit) which is a 455Khz 3rd IF AMP. This voltage is generated by Q5, rectified by D4 and D5, passing via Q4 as AGG and compared to the output from the RF gain control.\whichever of these two negative voltages is higher, is applied in 4 different stages of the Rx path as the AGC line.

In this radio, AGC is applied to the gates of four FETs in the IF chain.

At the two 1st IF amplifiers Q4 and Q3 of the RF unit, at Q12 of the IF unit which is a 3rd IF AMP and finally at Q1 of the MAIN board, which too is a 3rd IF AMP, one stage before the Balanced Modulator. This completes the negative feedback loop.

By backing the RF gain far enough, one can override the variable AGC produced by the above circuitry, thus providing a stable negative voltage to all the stages affected by it, and perform measurements at different points.

I did exactly that, discovering that from all 4 FETs accepting the AGC voltage in their gates, only the last one (Q1 of the main unit), actually received the whole negative voltage in its gate. For all the others, the AGC voltage at the gates, was slightly less negative, always lagging a few mV behind the AGC line voltage.

This was due to the presence of high value resistors in series between the AGC line and the actual gates of the AMPs for all the first 3 stages, but NOT for the final AGC controlled stage Q1, which is exactly where the AGC loop begins, by sampling its output. These are R17 (1,8M) for Q3 and Q4, and R76 (1M) for Q12.

For Q1 ( MAIN unit), there was a mysterious R243 shown at the page 10-1 of the service manual schematic, but not present either in the real circuit board, at the parts list of the MAIN unit or the detailed circuit diagram of the radio found at the end of the service manual !!!

So here we have a situation where we are trying to control the flow of a signal, applying the AGC voltage "brakes" depending on its strength, at different levels of its amplification path. We judge just how hard we have to hit the "brakes" based on the loudness of signal seen at the last stage of the series of amplification, BUT there are some "cushions" between the actual brake force applied to the first 3 stages of our amp chain, while LACKING for the 4th and final stage, which happens to be the one giving us information as to how much brake we need to apply.

Like trying to stop a running train by softly braking the back loaded carriages while abruptly stopping the locomotive is not going to result in a very smooth stop by any means...

As one can imagine, this is a perfect recipe for distortion in our circuit, since by not reducing the gain of the early stages by the same amount as the final one (or at least proportionally to the final stage), they can very well go into distortion which will propagate down the amplification path, and reach a hard stop at its final (Q1) link. There the signal level may be prevented from overdriving the successive stages of the circuit, but it has already been distorted from insufficient gain reduction in the stages before.

Thinking that the correct value for the missing R243 would have to be proportionally lower compared to the previous stage (R76 of O12), as R76 was lower than R17 of the RF unit, I calculated a prospective value of 555 K, very close to 560K, but not finding a 560K in my spare parts, I used a 470K one and installed it where it is shown at the page 10-1 of the schematic.

That addition fixed the distortion completely in SSB (AM was even smoother than before), but created another issue. When receiving very strong signals, or when tuning between them (above S 9+30 db), there were periodic AGC overshoots, which would blank the reception momentarily, pegging the S-meter.

These overshoots had no specific frequency or repeatability. They appeared at seemingly random intervals during AM reception. In SSB, they were abscent.

Seems like the Icom engineers were trying to prevent this phenomenon, ending up omitting R243 and accepting the audio distortion as a necessary trade off.

When reviewing the AGC circuit from the predecessor of the IC-745, the IC 740, one can see the almost identical implementation, using a resistor 1K at the place of R243. Nevertheless, in my IC-745, the 1K resistor had little effect in avoiding the distortion.

As a next step, tried paralelling the newly installed 470K (R243) with 10 – 100nF caps. Interestingly enough, that resulted in a similar effect like lowering the value of the resistor, preventing the AGC overshoots but proportionally allowing more distortion to be present.

During the process of searching for a lower value resistor, which might provide the best compromise between less distortion coupled with the lack og ACG overshoot, I stumbled on some forgotten 1N4148 diodes, and thinking I had nothing to lose, I installed one of them, cathode side towards the AGC line and anode connected to R1.

AND FINALLY...that did the trick !!! Not only it had absolutely no adverse effect on the audio distortion compared to the effect of the 470K (R243) alone, but it had ABOLISHED the AGC overshoot problem too !!

Talking about killing 2 birds with 1 stone....

The final touches for the mod and summary.

Installing the mod, the AGC line from near Q1 (white wire) at the MAIN board was unsoldered, and a tidy little paralel circuit composed of the 470K 1/8W resistor with the 1N4848 was installed between it and its original printed board trace. The end result just looked nice (hard to notice it is an "afterthought" modification when looking at it) and most importantly, IT JUST WORKS !

In summary, there are 2 dinstinct errors both affecting the AGC Rx performance of the IC-745 one being the mislabeling and wrong position of the 2 transistors involved in AM mode, the other and most important, being a design flaw in the implementation of the AGC line, by omitting R243 with all the resulting negative implications.

I even suspect these details being part of why the IC-745 got a bad reputation for its Rx being "noisy" compared to its predecessor IC-740. The original Rx audio indeed sounded so much hissy and distorted especially when using headphones, it is no wonder one could hardly tolerate listening to it for any length of time.

Actually after correcting the AGC, I can honestly say the IC-745 has one of the best Rx audio (inteligibility wise) from any of my rigs, and it is indeed very sensitive and a pleasure to use.

Comments - tips based on my experience with servicing the IC-745.

A) In the process of troubleshooting and preventive mainenance, I exchanged at least 15 – 20 electrolytics. ALL of them were checked to be well within specs (including the ones near the PS modules which are exposed to increased heat).

Initially I had in mind to replace all of the electrolytics in the radio, but after the above, I realized that it is actually not necessary.

B) When troubleshooting the MAIN board, if you have to unsolder things or work on its underside, it is not necessary or desirable to disconnect it from its multiple connectors. Such a task, is not only time consuming and difficult, but it runs the risk of damaging the connectors or their joints, creating even more trouble to be solved.

The main board can easily be flipped on its back, with all its connections in place, once the 5 screws holding it to the chassis have been removed.That is not only infinitely easier and faster, than disconnecting its connectors, but most importantly is SAFER.

C) In the process of changing the plastic trimmers of the VCOs, you have to gain access into their metal shield, which is soldered in its 4 corners. To take the lid off can be tricky and one must be carefull to avoid overheating the shield, which may lift its printed circuit supports.

Use a high wattage (40-60W) soldering iron with a thick tip. With a desoldering pump remove just enough solder from each corner of the lid as to uncover a small triangular space between the lid and its underlying metal frame. This space is just enough to fit the tip of a fine jewlers type screwdriver inside. This can be used to gradually pry each corner of the lid up in tandem, while briefly heating its respective solder joint. A fair amount of solder must remain througout the process in the joints, to ensure adequate heat transfer.

A similar process can be used for the underside metal shield, but there everything is much easier due to its higher elasticity, and easier access to its soldered tabs.

D) Probably to keep the frequency synthesis as clean as possible, ICOM choosed to incorporate a separate crystal controlled 2nd OSC to the synthesizer, instead of deriving the necessary 2nd OSC frequency from the 1st OSC one, as it did in its successor IC-751. The advantage was realized, as shown from the excellent phase noise characteristics of the 745

But it resulted in one more sourse of frequency instability and difficulty keeping the exact alignment between 1st LO, 2nd LO and BFO oscillators.

If you are a purist, you would like to see the radio being spot on in the lower part of its frequency spectrum (eg 80m band) with its alignment in the higher bands.

Of course, small differences can easily be corrected by the CAL pot on the top of the faceplate, but you have to do the calibration, everytime you change bands, even after the thermal drift has settled down.

There is a little trick I have found to align all 3 frequency oscillators in the radio, to well within a couple of Hz, and then at least you do not have to worry about recalibrating between bands.

1) As a 1st step, using a frequency counter, connected to R77 of the MAIN board per the service manual, you align the BFO USB – LSB frequencies as close as possible to the specified values.

Also align the 1st and 2nd LO frequencies per the service manual.

The following steps may or may not be needed depending on the accuracy of your frequency counter. In my case they do provide a useful fine tuning result.

Then, you find 2 known accurate frequency broadcast stations (In SV land, I use Radio China which transmits verified accurate and strong signals in both 40m band and 17650Mhz for several hours during the day and night) one transmitting in a lower band eg 40m and one in the upper part of the HF spectrum.

2) Use either LSB or USB mode while alternatively receiving these 2 BC stations, and detune by 1Khz the dial as to obtain a 1Khz audible tone from each. (If you use LSB you have to go 1Khz above the BC station's frequency and the opposite for USB)

3) Use an accurate AF generator to produce a 1Khz tone, eg from your laptop.

(a very good one is this )

4) Listen the beat of the 2 tones. No wiring between the laptop and the radio is required. Just listening to the audio of the 2 devices at a comparable level, will produce the distinct beat between the 2 tones, which when extinguished, the 2 tones are identical to within 1 Hz!!

5) Use the CAL pot on the top of the radio to zero beat the 1Khz tone in the lower band (eg 40m) and then check the higher one. Most likely you will notive a difference between them, even if you had set the 1st and 2nd LO of the radio with a frequency counter before.

6) If you indeed have a difference, you alternatively go between the 2 bands, and zerobeat the frequecy for the higher band fine-tuning the 1st LO ( adjust L2 in the PLL board).

Switch to the lower band BC frequency and zerobeat, fine tuning the 2nd LO with the CAL trimmer. (no need to readjust C5 on the actual 2nd OSC board itself).

By repeating the above steps (5) and (6) a few times, you can finally arrive at a condition when both the 1st and 2nd LO frequencies are exactly synchronized for the radio to be within a couple of Hz accurate in its frequency readout for its entire HF spectrum.

As a final step, after having achieved perfect alignment of low and high band reception, you can switch to the opposite sideband from the one originally chosen to do steps (2) to (6), to verify its exact BFO frequency. Go to the opposite sideband of the AM BC station signal as to produce the 1Khz tone, and zerobeat it to the 1Khz tone produced by the AF signal generator (laptop), fine tuning the respective BFO trimmer cap on the MAIN board.

The end result is a perfectly aligned radio to its entire HF spectrum.

Of course, you still have to live with the thermal frequency shift (which in my radio is about 80Hz from a "cold" summer start), but fortunately, once the different oscillators are precisely adjusted they more or less stay in sync during the warm up frequency drifting.

One can very well argue that it is of no practical use going through so much trouble to align the different oscillators especially when the end result is still subject to a much coarser temperature related drift, but at least it is a nice exercise and one step towards approaching perfection when dealing with an inherently imperfect situation....Just like everything else in life.

Afterall, when every action is deeply enough analyzed, it becomes more or less a matter of a philosophical view and personal preference.

Hope the above adventures will help other owners of this classic old IC-745 achieve its best potential, and finally function the way it was designed to do just about 35 years after its production.

Any comments – suggestions are welcome.

73 to all and enjoy the summer !

Marinos sv9dru / ki4gin

Member Comments:
This article has expired. No more comments may be added.
IC-745 Service and Upgrade  
by KB2DHG on November 8, 2018 Mail this to a friend!
WOW GREAT ARTICLE! I have a 745 that need the trim capacitors replaced and a good going over. Can you recommend a good service tech who I can send it to?
RE: IC-745 Service and Upgrade  
by SV9DRU on November 9, 2018 Mail this to a friend!
Hi Lou,
Living in Greece for the last 13 years, could not be of any help as to the US service places, but according to the reviews in this site, there are quite a few good people servicing these rigs.

I did all the servicing myself, and was not too hard when being carefull, of course it took me "forever" to locate and fix the AGC issues,due to them being part of faulty design - assemply rather than true malfunctions, but that is another issue.

Make sure to have these implemented when you send your rig for service, since they really do make a difference in Rx.

IC-745 Service and Upgrade  
by KB2DHG on November 9, 2018 Mail this to a friend!
OK Thanks. I love this rig I purchased it new in 1987, It has been a great performer all these years but now has developed an intermittent problem which I have figured it is the trip capacitors. My hands are not steady enough to perform the repair so I will send it out. I think this rig is worth the cost of the repair if I can get another 30 years out of it!
IC-745 Service and Upgrade  
by WC4R on November 9, 2018 Mail this to a friend!
Wonderful read. Super well descriptions and flow. I don't own the 745 but reading your analysis & thought process was a real pleasure.
RE: IC-745 Service and Upgrade  
by SV9DRU on November 9, 2018 Mail this to a friend!
Thanks for the nice comments Joseph, every such radio service "adventure" I think is worth sharing, just for the fun of it, and of course to help keeping our rigs in optimum shape.

RE: IC-745 Service and Upgrade  
by AF6AU on November 9, 2018 Mail this to a friend!
I had my old trusty Icom 751A worked on for the same plastic variable trimmer cap issue you might be having. The plastic slowly moves over the years and the moving capacitor 1/2 ends up with a crappy connection that messes up the oscillator. I took it to a local guy out here in Placentia Calif, named John Klewer (N6AX). He's good with Icoms, and even HRO out here sends stuff to him. However it was some time ago.... 714-993-0435... Give him a call (I hope he's still around, he's a repair artist!).

Another is a Icom specialist up in Washington state, Scott Malcolm, 360-864-8438. I know these guys are west coasters, but they do (did??) good work, worth a shot.
I Luv my Icoms....
751, 751A, 735(still has the little door).
J. M. Lenhert
RE: IC-745 Service and Upgrade  
by W5DXP on November 11, 2018 Mail this to a friend!
Thanks for a great article about a great old radio. Makes me wish I still had the one I sold when I bought my IC-756PRO.
73, Cecil,
RE: IC-745 Service and Upgrade  
by SV9DRU on November 11, 2018 Mail this to a friend!
Hi Cecil,
It sure is a remarkable radio with crisp full analog sound and after the AGC corrections a pleasure to use.
Maybe you will grab one of them again when it becomes available....

IC-745 Service and Upgrade  
by KC4AF on November 12, 2018 Mail this to a friend!
Fantastic Job. We need more articles like this. 73 Bob KC4AF
IC-745 Service and Upgrade  
by W3FAY on November 13, 2018 Mail this to a friend!
Your fine article reminded me of the IC-745 I owned about 30 years ago. It was a wonderful radio. I just wished that it was remote controllable like the IC-751 but I wasn't interested in buying an IC-751 just for that. At the time I was also taking a college "C" programming class and was interested in the idea of writing a rudimentary radio control program as a "show-n-tell" for my final project.

After comparing the schematics and pictorials in the service manuals of both radios, it appeared that the logic section of both radios were very much identical. When I opened up the radio, I saw that there were pads, traces and diagrams for a couple of 74xx TTL logic chips. There might have been a diode or two, as well. I don't recall. Anyway, I used the IC-751 manual to identify the absent components that I needed to install in the IC-745. I installed those as well as a cable for the computer connection which I was able to run out the back of the radio without any drilling. I tried out the radio with a radio control program designed for the IC-751 and it worked! Emulating an IC-751, I was able to do all functions that the IC-745 supported. I was then able to use this radio to code and test my school program, for which I received an "A". I later demonstrated this to my boss, an Admiral (USN), as a concept for quickly and cheaply deploying off-the-shelf computer-controlled radios. They deployed a ton of ICOM R-7000 receivers after that!

I later traded in my IC-745 for an IC-751 that showed up on the used shelf at the nearby candy store. I only did this because of my kluge setup with the computer cable dangling out the back of the radio and it was an affordable exchange. It didn't take me long to realize that the trade-off for the built-in functionality, was that I found the IC-745 much more enjoyable in day-to-day operation.

Now, whether anyone else would find it worthwhile to do this mod in this day and age, I can't say. However, I can say that it does work.

Again, great article! 73, Denis W3FAY
RE: IC-745 Service and Upgrade  
by NC8V on December 27, 2018 Mail this to a friend!
Scott Malcom is indeed still in action and recently brought up-to-snuff an IC-735 I picked up at a hamfest but hadn’t had a chance to test. Some years ago he did my IC-740 and I’ve been very pleased with his work. Not sure whether the ‘745 will go next, or the ‘211... or
Email Subscription
You are not subscribed to discussions on this article.

My Subscriptions
Subscriptions Help

Other Radios Articles
Little Boy -- A Water Cooled LDMOSFET Amplifier