PRO-43 ENHANCEMENT MEMORY EXPANSION & "FADE-AWAY" LIGHT Brett Bennett 4/12/94 bbb@kuhub.cc.ukans.edu Copyright 1994 B&D Electronics All Rights Reserved INTRODUCTION The following document details two enhancement modifications that are possible on the Radio Shack PRO-43 scanning radio. The first modification doubles the factory standard memory capacity, allowing you to enjoy an additional 200 channels of memory, plus 10 more monitor memories. After completing the first modification, it is a fairly simple matter to add a 7 to 10 second delay feature to the LCD back-light lamp circuit. I designed and performed these modifications on my PRO-43 and they do work. As I understand it, there (has been/is being) presented in WORLD SCANNER REPORT a similar memory expansion. I have not seen this modification and for all I know it is a better way of doing it. What is presented is a 'clean-room' design of my own that to the best of my knowledge is the only one to also offer a time-delay for the LCD back- light. Having said that, let me say: I ASSUME NO RESPONSIBILITY FOR ANYTHING THAT MIGHT HAPPEN TO YOUR RADIO SHOULD YOU OR ANYONE ELSE ATTEMPT THESE MODIFICATIONS. WHILE REASONABLE CARE HAS GONE INTO THE PREPARATION OF THE FOLLOWING DOCUMENT, I CAN NOT GUARANTEE THAT THERE ARE NOT ERRORS. The procedure described below is very complicated, and involves the soldering and desoldering of very tiny components and the cutting of printed circuit board traces. While the procedure is not irreversible, it will leave permanent scars inside your scanner. It should only be attempted by someone with moderate to advanced electronics experience. I don't think it should be attempted by, to quote someone else, a "weekend technician." If as you read through this you feel intimidated by the jargon and required procedures or if an electronic schematic looks like a strange piece of modern art to you, then this might not be the 'mod' you are looking for. The following document will assume you have obtained the materials listed below, especially the Service Manual, and you have studied the circuit in some detail. I will not describe in great detail where every component is or where every solder point on the logic board should be placed. Having the Service Manual, and having studied the drawings therein will allow you determine to where it is I refer. REQUIREMENTS 1 - PRO-43 Service Manual. (Tandy Parts # MS-2000300 - approx. $6.00) 1 - CD4053 - Triple - SPDT analog switch - Surface mount package (DigiKey #CD4053BCM-ND - approx. $1.50) 2 - 93C67 EEPROMs (Tandy Parts # MX-8050 - approx. $13.00/each) 1 - 100k x 5 resistor network. Common lead type. (DigiKey # Q5104) 1 - UN2111 PNP transistor (Q1 on Logic PCB. Tandy Parts # ) 1 - 1 uF 16v capacitor. 1 - 10 MOhm 1/8-1/16W resistor. Misc.: Wire wrap wire-30 gauge. (Radio Shack) Grounded soldering iron with approx. 1/32" tip or smaller. Desoldering Equipment. Digital Volt Meter. (Used to double check your VCC and Ground connections, pre-check switching assembly) Utility knife with fresh sharp small blade. (Needed to cut traces.) Strong hand held magnifying lens. (Needed to check solder joints.) +5-10 Volt bench power supply. (Used to check switching assembly.) Anti-Static work place and grounding wrist band. Very fine needle nose pliers, wire cutters. THEORY When studying the schematic for the PRO-43 one sees that the channel memory is organized using two 93C67 EEPROMS. As is common in microcomputer designs, the memory is all connected together on a common bus, with the individual memory devices being selected with a Chip-Select (CS) control line. The 93C67 are particularly simple, in that they are serial devices, only requiring 4 signal lines. (Data In, Data Out, Data Clock and CS) This inspired me to consider the possibility of adding two more EEPROMs, creating in effect a second 'bank' of 200 channels. One could then, presumably with the flip of a switch, choose between the original 200 channels, and a second bank of 200 channels. (In theory, this could continue to 600, 800, 1000 etc. channels. In practice however, space is very tight in the PRO-43, and getting anymore memory installed would be very tricky.) Switching two CS lines between two pairs of EEPROMs requires the equivalent of a DPDT switch. While it might be possible to do this with a mechanical switch, the physical implementation of this could prove to be very difficult. Rather than use a DPDT switch, I chose to use a CD4053 triple 2 channel analog multiplexer/demultiplexer. This CMOS analog 'mux/demux' can be viewed as a triple SPDT switch. Two sections of the CD4053 are wired in such a way as to act as a DPDT switch. The advantage being that a DPDT switch can be controlled with a SPST switch, and no actively switching digital signal lines need leave the area of the CPU. At this point one is still faced with the need of a mechanical switch to control the CD4053, albeit, only a SPST switch is needed. It was suggested to me by someone on the Internet, that the 'Keylock' switch might be a good candidate. And indeed it is! While it might be nice, once in a while, to be able to disable the keyboard, it is more than a fair trade to give this up for more memory. Thus I used the Keylock switch to control the DPDT switch made from the CD4053. Good design dictates that the state of all signal/control lines are always in a known state. The 93C67's CS lines are active high leading to the need for pull-down resistors on all 4 CS lines. This insures that the 93C67's not currently being used will have their CS lines pulled to ground. I accomplished this with 100K ohm resistors connected between the CS line and ground. Physically I implemented this using a resistor array (SIP package) containing 5-100K resistors all having one common lead. The 5th. resistor is used to pull-down the CD4053 control lines making up the DPDT switch. This is necessary when the Keylock switch is in the open position (unlocked) since the line must be driven low. This leaves a SPDT switch in the CD4053 not being used. Studying how the lamp circuit works shows that the keyboard light switch grounds the base of Q1 to turn on the lamp. Using the CD4053 to act as the switch to ground, and placing a large RC time constant circuit on the control line of the CD4053 will allow for a time delayed lamp modification. Because the input impedance of the CD4053 is extremely high, a small valued capacitor and large value resistor will work. In my design I used as 1 uF tantalum capacitor and a 10 MOhm resistor to form the RC time delay. The Pro-43's keyboard light switch is connected across the capacitor to discharge it when depressed. This causes the CD4053 to switch on, grounding the base of Q1. After the switch is released, the capacitor begins to charge through the 10 MOhm resistor. When this voltage reaches approximately 1/2 VCC the CD4053 switches off, and the lamp goes out. (One caveat: When the radio is first turned on, the capacitor is discharged and therefore a lamp timing cycle will begin with each power up.) PROCEDURE **REMEMBER TO ALWAYS WORK AT A STATIC FREE WORK- STATION, INCLUDING THE USE OF GROUNDED SOLDERING EQUIPMENT AND THE WEARING OF STATIC REMOVAL BRACELETS.** The first step is to build the switching assembly. To build this first tin the leads of the surface mount CD4053 and the resistor array. Following the connections shown in the schematic (figure 1), attach short (3/8" to 1/2") pieces of wire wrap wire between the 5 resistor leads of RA1-NEW and pins 1,2,3,5, and 9-10 on the CD4053. (You should always: Size, cut, strip, tin and trim the wire before soldering it to the CD4053. Doing this makes it fairly easy to get good solder connections.) Connect the common lead of the resistor array to pins 6,7,8 and 12 of the CD4053. If done correctly you will have the resistor array laying flat on its side with its leads facing the CD4053 as shown in figure 2. This makes the assembly very thin. Inspect the solder joints with a magnifying glass, and give them a gentle 'tug' to make sure they are securely soldered. The PCB board drawing on page 19 of the service manual is an actual size drawing. Check to see that your assembly will fit into the area just above IC1 and to right of IC2. Soldering to the pins of RA1-NEW, connect 4 - 2.5" lengths of wire to the resistor leads going to pins 1,2,3&5 of CD4053 (This makes up the 4 switched lines going to the CS pins.) Connect a 5" length of wire to the resistor pin connected to pins 9&10 of the CD4053.( This will go to the Keylock switch.) Attach a 2" piece of wire to the resistor lead attached to pins 6,7,8 & 12 of the CD4053. (This will be the ground connection.) Attach 3 - 4.5" long wires to pins 4,15,&16 of the CD4053. These will be the two poles of the switch and the VCC connection respectively. Finally attach 2 - 5" wires to pins 11 & 14 of the CD4053 if you intend to also add the lamp timer modification. At this point you should have a strange looking device with 11 wires sticking out of it. DOUBLE CHECK YOUR WORK AND SOLDER JOINTS AT THIS POINT. The next step is to check the operation of the switching assembly. Using a 5-10 volt power supply, apply power to the assembly. Using a DVM, check to see that there is continuity between pins 4&5 and pins 15&2 while the control line (the wire attached to pins 9&10) is not connected to anything. And also check to see there is NO continuity between pins 4&3 and 15&1. If that checks, attach the control line to VCC. There should then be continuity between pins 4&3, and 15&1 and no continuity between 4&5 and 15&2. Next check for continuity between pin 14 and ground when pin 11 is grounded. The resistance between 14 and ground should become infinite when pin 11 is then connected to VCC. After check-out put the assembly aside in a static proof container. Using extreme care bend all the leads, EXCEPT FOR PIN 3, of one of the 93C67 so that they are pointing as straight DOWN as possible . Bend pin 3 so it is sticking straight out from the body of the chip. These leads are very fragile and I suspect they will only accept a single bending before breaking,. After the leads are bent, carefully tin all leads. Put the EEPROM back in its carrier, and repeat on the second EEPROM. The next step is to prepare the scanner for the switch assembly. From here on out, "The cheese becomes more binding," so work carefully and think about everything twice before doing it! Remove the battery compartment lid and battery. Remove the 4 case screws and the back of the case. Disconnect the two wire connectors on the RF board. Desolder the antenna feed line and its ground connection. Remove the 6 screws holding down the RF board and remove it from the audio board. Disconnect the speaker wire from the audio board, and remove the 2 screws holding it to the front of the case. Separate this from the front of the case. Desolder and remove the RF-shield from the logic board. Wash your hands to remove as much skin oil as possible! Remove the screws holding the logic PCB. Turn the unit over, and with your hand supporting the PCB lower it from the case front. The reason for this is that the LCD assembly will stay with the board. If during this process the LCD and its backing glass do separate, handle them as little as possible. Also note the back-glass has a white frosted side. The frosted side, the light diffuser, is placed away from the LCD panel. Remove the rubber keypad, and locate the two traces running to the Keylock switch. Using a sharp utility knife, cut both of these traces near the switch, examine the cuts with a magnifying lens to make sure the switch is severed from the rest of the board. Avoid touching the keypad contact area as much as possible. Checking to make sure dust has not entered the LCD view area, put the logic PCB back in the front of the case and secure it with its screws. Locate IC2 and IC3 EEPROMs on the logic board. Add to each of the pins on both chips a small amount of solder. This will be necessary to attach the new EEPROMs. Of course you must make sure that you form no solder bridges. At this stage I suggest you re-install the audio board and RF board. (There is no need to solder the RF shield back on, nor should you resolder the antenna connections.) By holding the battery in place and slipping the battery lid into proper position, it is possible to power up the radio without bothering with the case back. Check to see that all keys are functioning correctly except for the Keylock switch, and that the display looks normal. You are now ready to begin the 'fitting' of your switch assembly into the radio. Locate the area just above IC1 and to the right (as being viewed from the backside of the radio) of IC2&3. There is a relatively open space here where the CD4053 and resistor array can lay. Cut a piece of thin card board stock to be slightly larger than the area taken up by the switch assembly. This will serve as an insulator between the logic PCB and the assembly. Keeping the assembly in its final position, start sizing the lengths of the wires. The wires connected to pins 1&2 of the CD4053 must reach IC3 pin 3, and the wires on pins 3&5 of the CD4053 must reach IC2 pin 3. Remember when cutting these down to leave enough extra wire for stripping, tinning and final trimming. The wire connected to pin 15 of the CD4053 must be long enough to reach the TRACE on the PCB going to IC3 pin 3. Likewise the wire on pin 4 of the CD4053 must reach the trace running to IC2 pin 3. The switch control wire (pins 9&10 of the CD4053) must reach down to outside most solder pad on the Keylock switch. VCC, pin 16 of the CD4053, is sized to reach the 82 ohm resistor R26 located next to Q1. The ground connection can be made on the large copper area under the switch assembly after scraping a small area of solder-mask off. After the wire lengths have been determined, strip and tin the ends of all the wires and set the assembly aside. Using a sharp knife, cut the trace running from IC1 to pin 3 of IC2. Repeat this for the trace running from IC1 to pin 3 of IC3. On either side of the cuts, scrape off a 1/8" area of solder mask and tin the exposed copper. Be careful not to damage any of the surrounding traces or components. Place the switching assembly in its proper location and solder the ground wire to the solder-mask free area you made near IC1. Attach VCC (pin 16 of the CD4053) to the top of R26. Attach the pin 5 wire to the cut trace that runs to IC2-pin 3, and attach the pin 2 wire to the cut trace running to IC3-pin 3. Next attach the wire on pin 4 of the CD4053 to the trace running to pin 18 (previously running to IC2-pin 3) of IC1. Likewise connect the wire on pin 15 to the trace connected to pin 19 (previously running to IC3- pin 3) of IC1. If done correctly you now have 4 wires running to the four 'new' connection points made by cutting the two CS line traces. Connect the wire tied to pins 9&10 of the CD4053 to the outside (left side) conductor of the Keylock switch. Attach a new piece of wire to the center conductor of the Keylock switch. Size and connect the other end of this wire to center lead of IC5. (IC5 is low battery detector for the CPU. The center lead of this is connected to 5 volts which is sufficient to switch the CD4053. Later we will need to attach another connection to R26, and squeezing 3 wires to this single point would be tight so I used the 5 volt bus for switching.) Place the insulator under the assembly and tape a corner in place so it will not move. Cut another cardboard insulator big enough to cover the switching assembly, and place this on top of the logic PCB. Bending the currently disconnected wires (pins 1,3,11,&14) in such a way that they don't short to anything (or each other) , install the audio board back in to the radio. Install the RF board and prepare to power up the radio. Making sure the Keylock switch is in the UNLOCKED position, turn the radio on. It should begin to scan normally. (Believe me you will breathe a big sigh of relief when it does!) Turn the radio off, move the Keylock switch to LOCKED position and turn the radio back on. You should see it either scanning frequency 000.0000 or the frequency field will be blank, with the channel counter incrementing. After a few seconds, the frequency number always blanked out and the am/fm indicators blinked randomly on my radio. If you are unable to make the radio scan correctly, there is something wrong with the switch assembly or your connections. The next step is to install the new memory. The leads on the new memory are not long enough to reach the leads of the factory memory. Trying to solder bridge the gap is difficult accomplish without also bridging adjacent pins. The best way I found to do this is to use tinned wire wrap wire. So the first step here is to tin about 6" of stripped wire wrap wire. Remove one of the new 93C67s from its carrier and place it on top of IC2, making sure that the extended pin 3 is over pin 3 of IC2. Put a droplet of solder on the tip of your soldering iron. While holding down the memory with a finger, apply the soldering iron to one of the corner pins of both ICs. The droplet should 'jump' over the gap between the pins and solder them together. Check both sides of the EEPROM to insure that all pins are accurately aligned. For the remaining straightened pins, proceed as follows: While holding the tinned wire wrap wire against the lower and upper pins, apply a freshly cleaned soldering iron tip to the gap area of the joint. The extra solder you earlier placed on the factory EEPROM will then be available to flow along the wire and solder to the pin of the new EEPROM. CHECK the joint with a magnifying lens. After a visual conformation of the joint, clip the wire close to the pin of the newly installed memory. If you like go back to the first pair you soldered and lay the wire into the molten solder. Repeat for IC3. Attach the wire connected to pin 3 of the CD4053 to pin 3 of IC2-NEW, and connect the pin 1 wire to pin 3 of IC3-NEW. Install the cover insulator over the switch assembly and prepare for a power up. With the Keylock switch in the Unlocked position power up the receiver. After insuring that the factory memory is still working correctly, power down, switch to the Locked position and power up again while holding down the 0 key and the Clear key. You should now see the radio scanning frequency 000.0000 for all 10 banks repeatedly. Press the PGM key to stop scanning and enter a valid frequency value and press enter. Press PGM, then the DOWN- ARROW, then PGM again. The frequency you just programmed should now be on the display. (At this point feel free to cheer! YOU have just added 200 more channels to your PRO-43. The next step will be the icing on the cake!) This next step requires you to lift the BASE of Q1 from the logic PCB. (I found this to be very difficult, in fact, I ended up destroying Q1. Thus the extra Q1 in the parts list.) Desolder and lift the Base lead of Q1 from the logic PCB. Solder together one lead of R2, the positive terminal of C1 and the wire from pin 11 of the CD4053. Also from this junction, connect a wire to the solder pad on the logic PCB where Q1's base was connected. The area to the right of the Keylock switch provides a space where these components can lay. Connect the base of Q1 to the wire connected to pin 14 of the CD4053 switch assembly. The other end of R2 is connected by wire to the top of R26, this provides the charging voltage for C1. The negative lead of C1 can be connected to the RF shield ground pad near the Keylock switch. (Check that the RF shield will fit in place over C1. If it will not, place C1 so that its body is just below the bottom of the shield. Insulate the R2-C1 combination with electrical tape. Tape the switch assembly's top insulator cardboard in place and install and solder the RF-shield back onto the logic PCB. Prepare the radio for bench power up. Turn the radio on, the lamp should come on. After 7 to 10 seconds the lamp should go out. Check in a darkened room to insure that the lamp completely extinguishes. Depress the light switch on the keyboard. The lamp should immediately light, and stay lit for 7 to 10 seconds after it is released. Install all the remaining screws for the RF board, solder the antenna lead and ground connections, and replace the case back. USAGE The light feature is self explanatory, and quickly finds itself useful in darken conditions. As noted earlier an unfortunate side- effect of the lamp circuit design is that the lamp will illuminate for 7 to 10 seconds every time the scanner is first turned on. (Maybe this is a feature to some.) Switching between active banks requires a little more thought. The safest way of changing banks is to: Turn the scanner off, switch banks, and then turn the scanner back on. But sooner or later you will switch banks while the scanner is on. When this happens you will see a curious event. The radio's CPU will continue to scan from the previously selected banks, but the frequencies scanned will now be those from the on-line EEPROM. This makes sense, since the CPU has no knowledge that its memory resources have been changed. Pressing the Manual button, followed by the Scan button, will cause the CPU to reload the scan banks that were last saved in the currently selected EEPROMs. Also any monitor frequencies that were stored are recalled. This allows you to set up two different frequency configurations, one in each bank of EEPROM. To change between them on the fly, just flip the Keylock switch, press the manual button, followed by the scan button. The only time I can foresee there being a problem is if some how one were to flip the switch while the CPU was doing a write to EEPROMs. Keeping this in mind you should never have any problems. There are a couple of 'gotch-ya' that can happen when switching memory banks without cycling power. The PRO-43 loads the list of all locked-out channels only when it is first powered up. (This makes sense if you think about it. While I don't know for a fact, I assume that the Lock-out information for each channel is stored together with that channel's frequency in the EEPROM. If the CPU didn't pre-load the list of locked out channels, it would have to constantly re-read each channel to determine its lock-out status.) The side-effect of this is that it can cause the radio to appear to be randomly locking and unlocking channels when switching memory on the fly. The only realistic solutions appear to be to either cycle power when you switch memory banks, or use Lock-outs very sparingly. The Priority channel assignment is also loaded only when power is first applied. This can make the radio start checking the wrong frequency when memory is switched and the priority feature is used. The best way around this problem seems to be to assign the same channel number to be the priority channel in each memory bank. CONCLUSION Performing the described modification will provide the owner of an already excellent hand-held scanner, the convenience of a total of 400 channel memories and 20 monitor locations, doubling the factory capacity for under $40.00. It also provides a "fade away" LCD back-light with a 7 to 10 second delay. The bank switching nature of the design allows for the user to program two unique 'personalities' into the scanner selecting between them with the Keylock switch. The modification while complicated can be performed by any person with moderate to advanced electronics experience.