The optical encoder unit is very sensitive and only works properly when everything is aligned and in order. Simply by undoing a few screws on the encoder itself you can put it out of alignment and even warp the internal masks making it useless. If it still works and if the problem is intermittent but bearable - leave it alone! Only attempt the repairs if you are prepared to accept the fact that the problems could get worse and that you may have to find a replacement encoder that is quite rare and expensive.
This article covers the repair of the optical encoder for the Icom IC-735. It may be relevant to other radios that use the same encoder assembly. Symptoms, reasons for failure and solutions are covered for each case.
Figure 1 shows my radio that has suffered from many problems, the latest being that the optical encoder refused to work until the radio warmed up. This didn't really become an issue until it needed 5 to 10 minutes to start working.
I spent all of the evenings of a week and the weekend to repair the radio, it was very frustrating, fiddly and it took 8 or 9 attempts at various solutions before it worked enough to be usable. So be prepared to get annoyed. When you get tired and frustrated, put it down and start again the next day, otherwise you may make a mistake and risk damaging the radio further.
The problem with my radio was tracked down to a phototransistor that had failed. I thought it wouldn't be too much of a problem replacing it and started to take it all to pieces. This is where I fell into a number of traps waiting for the uninformed. I probably fell into all of them, which is why this article is fairly comprehensive! My pain is your gain - don't make the mistakes I did and you should get your encoder working 100%. A good rule to follow here is
DO NOT DISMANTLE ANYTHING UNLESS YOU ARE SURE WHAT YOU ARE REMOVING IS THE PROBLEM.
This will save you much frustration in realigning everything in the encoder.
The symptoms of encoder failure are quite varied, but obviously involve some problem with tuning the radio:
Only the top cover needs to be removed to perform the inspection and repairs. There are 12 screws to be removed, two on either side of the radio, 4 on top and 4 around the speaker grille. When this cover is removed, you can see the speaker and the heat-sink part of the final unit as shown in Figure 2.
Before removing the final unit, remember to disconnect the small phono patch lead at the rear otherwise when the final unit is removed the phono connectors will be tugged and you may already know what sort of problems this can cause. The final unit is fixed by two countersunk screws at the front of the radio and two screws at the back. The servicing location for the final unit is as shown in Figure 3.
The optical encoder flying lead is connected to the main circuit board at the bottom left of the radio, as shown in Figure 4. Disconnect the lead from the board.
The VFO knob can be removed with an Allen key inserted into the hole in the knob edge that contains a grub screw as shown in Figure 5. The encoder unit itself can be removed from the radio by unscrewing the 4 screws on the encoder unit and by carefully threading the flying lead and plug through the access hole inside the radio.
Figure 7 shows the rotary encoder unit as removed from the radio. Mine has already been heavily modified and parts of the lead cut - yours will be nice and pristine.
Case 1 : Alignment of PCB.
The PCB that contains the IR LEDs, phototransistors and switching transistors is mounted on top of the encoder as shown in Figure 8. It is attached via two small screws. The alignment of this PCB is important, as it has some play around the fixing holes which will shift the location of the phototransistors inside the unit.
A typical problem caused by this is losing the ability to tune, or only tuning up or down. These problems occur when one of the optoelectronic pairs stops working properly for some reason.
Case 2 : Alignment of phototransistors.
Before removing the PCB board from the encoder, be informed that the optoelectronics inside are positioned around a set of very delicate and sensitive foil masks that can easily be bent and ruined by the PCB board being twisted when inserted inside the unit. This is Trap 1 that I fell into. DO NOT DO THIS IT WILL CAUSE YOU NO END OF WORK AND PROBLEMS.
Place the encoder in a secure place for testing, if it falls while the board is inserted but unattached, you will bend the foils particularly if you are holding the flying lead at the time. With this in mind, carefully remove the PCB.
Typical problems caused by this are losing the ability to tune, or only tuning up or down. Again, these problems occur when one of the optoelectronic pairs stops working properly for some reason.
Figure 9 shows the IR LEDs on the right, marked DS1 and DS2, and the phototransistors on the left marked Q1 and Q2. For the encoder to work, the lenses on the sides of the devices must be pointing directly at each other. Also, the devices must respond identically to changing IR light conditions so that the quadrature phase (i.e. the square waves outputted by the encoder for each of the two optical pairs) is correct. You may find that slightly offsetting the phototransistors or IR leds may improve performance, but it is best to have the lenses pointing at each other when aligning other parts of the encoder.
Case 3 : Alignment of static mask or rotary mask.
Figure 10 shows the rotary mask inside the encoder unit. It is fixed by three screws and a metal plate that holds the mask flat against the plastic part of the rotor behind it. At the bottom of the image, you will see two screws and washers either side that fix the static mask to the body of the encoder. The static mask has two sets of five slits cut into it and the rotary mask has a circular set of slits cut around its circumference. Typical symptoms are tuning in the opposite direction to rotation, tuning up and down randomly or poor performance when tuning slowly.
Do not loosen the screws attaching the static or rotary mask unless you are sure they are the problem and they have to be removed. Do NOT lubricate the masks - this is Trap 2 that I fell into. The masks on my unit were warped and lubricating them caused the masks to lie perfectly flat on each other due to the surface tension of the oil. This appeared to improve performance, but in the end proved to be the cause of tuning up and down randomly due to fluctuations in how the IR light passed through the masks. They can be cleaned by removing them carefully and washing with switch cleaner and lint free cloth. Read case 4 if you are going to do this. Remember that any loosening of the screws (Argh! - Trap 3!) will require time consuming alignment of the masks.
To align the masks, first align the rotary mask. Rotate the shaft and look closely to see if the disk edge wobbles even slightly at the edge of the disk. Loosen the screws and tighten them until it looks perfect. Do not attempt to touch or move the disk. Use only the action of loosening and tightening the screws at the centre to align the disk.
What really affects the behaviour of the encoder is the positioning of the static mask. This is quite fiddly to align. Attach the PCB to the encoder in the most central position you can and attach the encoder to the radio. Loosen the static mask fixing screws and with a mix of loosening and tightening and moving the mask with a fine pick or screwdriver find the best location for the mask. In some locations, you will find the encoder tunes up when it should tune down and vice versa. When it is in the right place, you will find that it tunes in the right direction with accuracy when tuning slowly or quickly in both directions. Good luck!
Case 4 : Warping of static mask or rotary mask.
Figure 11 shows the masks sitting on top of each other from the edge. The optoelectronics fit around the disk. The masks must lie against each other and not be warped in any way. Even the slightest amount of warp will cause problems. In the image, it looks like they are pretty flat, but they are not flat enough, or in contact enough to work! From what people are complaining about on the internet it looks like this is a common problem.
If there is a warping problem with the masks, then a symptom is that only a certain portion of a VFO rotation will work, or the radio tunes up and down randomly when the encoder is rotated. However if they are badly warped then the symptoms are as for cases 1 and 2 as one of the phototransistors will be permanently switched on due to IR light getting through the masks due to the gap in between them.
It is possible to repair the warp providing it is not too severe. If the masks are only very slightly warped then it is possible to close the tiny gap between the masks by folding a small piece of A4 paper in two and placing it in the gap between the case and the static mask. This pushes the static mask against the rotary mask. Figure 12 shows such an arrangement which solved many of the problems I had with my encoder. Make sure that the piece of paper does not cover any of the slits in the mask. Do not use glue to affix the paper. The spring from holding the masks together is sufficient to keep the paper where it is located.
If the masks are badly warped, then they will have to be removed. Note that which side of the mask faces forward is important for both masks. The masks will not work correctly the wrong way around. Using two strong, perfectly flat surfaces and some monitor screen wipes, sandwich the mask inside the wipe and compress with the two perfectly flat surfaces. Striking with a hammer may aid flattening the masks. Once this is done, then the masks can be refitted, a paper spacer inserted as described earlier and the masks aligned.
Case 5 : Failure of phototransistors or IR LEDs.
This is what happened to my encoder. The symptoms were no tuning until warmed up. When warming up, approximately ¾ of the way through the encoder would intermittently start working, sometimes tuning randomly, sometimes in only one direction. In my case, the reason the phototransistor failed was because when it was being mounted, the device leg inside the case moved and stressed the internal connections. This led to a fault that appeared much later on in service. Upon examining the device, it was noted that the central leg was ever so slightly loose due to the bending of the legs required for soldering. It would not surprise me if many of the failures out there were due to this.
In figure 9, the transistors Q3 and Q4 (labels covered by the components) are 2SC3399 devices which have their own internal 47k biasing resistors. This explains the lack of resistors on the board itself. The image shows replacement transistors that I have put in, your PCB will have the original devices in a ZTX series transistor style package.
Exact replacement phototransistors and IR LEDs are not available. The original phototransistors have quite a high gain and generate quite a square waveform. Transistors Q3 and Q4 square up the waveform into a nice square wave ready for the radio to process. I experimented with some phototransistors out of a mouse, which did not generate a nice square switching waveform and were a pig to get the collector current right so that they worked. It was while trying these that I managed to warp the masks. I had already misaligned the masks by checking them.
After much experimentation, frustration and discovering of the problems in the other cases, soldering and de-soldering, making small veroboards to fit in the flying lead to further condition the waveform I finally found the solution.
If failure of the phototransistor is the problem then you will need to carry out all of these modifications. You cannot just change one of the phototransistors as they need to be of exactly the same type to maintain the timing of the quadrature output. Everything will need to be replaced apart from the series resistor for the IR LEDs and the capacitor and the IR LEDs if they can be proved ok.
To check if the IR LEDs are working, measure the voltage across both of them and each of them. There should be approximately 1.4V across each of them and around 3V across the pair. If they do not measure this, then there is a faulty device or poor connection. If you have a camera with IR night vision capability then this is ideal for checking that the devices are illuminated. If they cannot be directly checked like this then just replace them.
Figure 13 shows the original circuit, and Figure 14 shows the replacement circuit.
Replacement IR LEDs are Sharp GL480E00000F devices and are available from Rapid Electronics in the UK, order code 58-0962. Replacement phototransistors are Sharp PT480E00000F devices and again are available from Rapid, order code 58-1062. The replacement transistors are of Darlington type so that the processed waveform is as square as possible - I used MPSA13 devices. Order several of each device as it is easy to make a mistake. The IR LEDs in particular need soldering with a low iron temperature or the case will melt and the lead separate from the device! I was as surprised as you probably are reading this when I removed an LED because it wasn't working for some bizarre reason and the leg came off in my hand!
The new circuit needs to be built on the little PCB - keep everything as short, tidy and low profile as possible as there is not much room to work with when replacing the encoder back into the radio. Figures 8 and 9 should give some clues on how to fit the components into the space available. Note that the phototransistors are 2 leg devices which need to be attached to the left and centre pads on the PCB when viewing from the foil side. The right pad is a ground connection, make sure the leg of the device does not touch this pad.
Hopefully this has explained some of the problems that can occur with the optical encoder on the Icom IC-735, and helped avoid further damage to the unit that can occur by accident. My encoder is now fully functional which is a miracle in itself due to the amount of work, and unfortunately damage, that has been done to it. However the journey has led me through all what is likely to go wrong with one of these encoders.
The 735 is a favourite of mine due to its excellent audio and small size. It is typically used for portable and alternative location operation. Good luck keeping yours alive.
Webmasters' Note: I've found a text document hidden away in a Yahoo group, that I belong to, which deals with this same repair. I sent it to Dan for his thoughts. Below is the text document (posted in May 2007) and Dans' reply:
Repair of the IC735 tuning sensor Tuning sensor failure can write off a perfectly good radio. The failure is caused by one of the opto devices dying and Icom can no longer supply the original parts. I have sourced alternatives and made a good repair of my beloved 735's tuning sensor and would like to pass on my knowledge. For the LED emitter use Sharp device GL480 For the Schmitt receiver use Sharp IS486. Replace all four components as a set The new type emitter is too bright for proper switching so replace the LED dropper resistor R1 150ohms with a 1K. The sensor works ok with as much as 4k7 in this position. I fixed my radio about 12 months ago and its still working fine. Hope this is of use 73 Alan G0KMC
Dans' reply: I've had a look at the text file and the datasheets. It certainly explains why I thought the photo-transistors had a high gain or had an additional switching transistor in - they've got Schmitt triggers inside! The pin that connects to 0V was not a base connection for external bias, but a 0V connection for the IC inside.
Interesting note about the LEDs too. I did not come across this problem, however it could easily be an issue.
I will update the document soon, in the meantime put the text file and this email at the end of the html document on the website.
Dan further contacted me with this email to add from a satisfied repairer:
I hope you are the same Dan White, the author of "Repairing the IC-735 optical encoder".
I am writing to thank you for the informative writeup. Thanks to the information in it I was able to successfully repair my failing optical encoder without even removing the cover of the radio. I removed the knob, removed the four mounting screws, removed the IR LED pcb and discovered the slightest warping of the moving disc. The paper shim did the trick!
The fix seems to be holding up well. I have used the rig for two days now with no problems.
Thanks very much!
Mike Flicinski K2UXE
Round Rock, Texas USA (near Austin)
Thanks Dan, always good to know that people are making good use of the information.
I found the best way to monitor the functions of the encoder is with an oscilloscope with a DC-coupled probe connected to the output lead of the optical detector.
When functioning correctly, there will be a nice 5 Volt square-wave (with reference to ground, of course) generated that has a period that follows the rate the tuning dial is turned.
Each of the detector pairs should be tested using this method to determine if the problem is with the encoder or with the logic circuitry that follows it.
The Sharp Opto devices that Alan, G0KMC, used are also no longer available. I've cross referenced Honeywell devices that are near direct substitutes for the original ICOM parts and have successfully repaired an IC-735 encoder that had a faulty Optoschmitt detector:
Change both the emitter and detector as a pair - be careful with the heat sensitive board traces! I changed only one of the emitter-detector pairs and the encoder works fine. The only change to the original circuit is the addition of a 390 Ohm pull-up resistor between VCC (+5V) and the output pin of the detector. See the SDP8371-001 datasheet regarding the pull-up resistor if necessary.
I've included two photos (edited into a single image - webmaster) of the repaired encoder for reference, and a video of the repaired encoder in operation is on flickr at: http://flic.kr/s/aHsjzoJyeN, along with these photos and a few more.
I hope others will benefit from my update as much as I have from Dan's article.