I'm also going to reprint this here and the original author can take credit or not, depending on...
In any case, during these modifications this is bound to come up. This is a better explanation than I could write in the next hour. (or two)
The definitive information is that changing the "Reference Modulus" pin (4) will make the PLL chip change frequency in 5 kHz steps instead of 10 kHz steps. The first problem here has to do with the filtering between the "Phase Detector" output and the VCO. The chip does not deliver smooth DC tuning voltage out of pin 5, it delivers pulses at a frequency of 10 kHz. Those pulses get smoothed out (filtered) into a steady DC tuning voltage by resistors and capacitors between that pin on the chip and the varactor (voltage-tuned capacitor) diode in the VCO (Voltage Controlled Oscillator). The pulses in the stock setup have a frequency of (you guessed it) 10 kHz. Changing this to 5 kHz requires that this filter be altered. It just won't be as effective as it was at 10 kHz. Rather than bore you with how to do this, we'll skip to the reason it won't matter.
The programmable divider inside the chip is where the binary inputs from the channel selector lead to. This divider will no longer deliver the same channel frequency when you choose 5 kHz as it did when the 10 kHz option was turned on. Depending on the particular radio you are looking at, the channel frequency will jump (usually) DOWN by one or two MHz when this pin is changed.
Pin 4 on the PLL02A chip contains an internal resistor connected to pin 1, the power supply pin. This way, if you leave it unconnected, it has a logic '1' on it, giving you the 10 kHz steps the radio is built to deliver. Saves them a resistor. Grounding the pin changes it to a logic '0', AND changes an internal divider from 1024 (10.24 Mhz divided by 1024 gives you... 10 kHz) to 2048. Dividing 10.24 MHz by 2048 now gives you 5 kHz INSIDE the chip, and this frequency becomes the new "reference" frequency.
The "programmable" divider inside the chip is where your voltage-controlled VCO frequency feeds in on pin 2. If you cut the internal reference frequency in half, changing it from 10 kHz to 5 kHz, the programmable divider will need a binary number TWICE the value used to give you channel 1 as it did with the 10 kHz setup.
This chip was used in a lot of radios, both AM-ONLY and SSB models. The internal binary arithmetic falls into about four categories. For three of these, you can change ONE crystal in the radio to match your new 5 kHz channel steps. Feeding a binary number that is TWICE the old one can be done by removing the channel switch wires from pins 10 through 15, and move them each DOWN by one pin number. The wire that went to pin 15 will now go to pin 14, the one that went to 14 now goes to 13, etc. Pin 15 will now become your "+ 5kHz" switch connection, on an added toggle switch. That will allow you to keep 40 channels. Consider that a 40-position channel switch would otherwise only cover 20 channels, at 5 kHz per click, if you leave them hooked up stock. The frequency needed for the NEW crystal depends on which of three frequeny schemes the radio used, the 10.05 MHz crystal, the 11.86 MHz crystal, or the other one I can't remember right now. This is a lot of trouble, and explains why the Digi-Scan toy was so popular in its day. All this junk was already built into it, just pop the PLL chip from the radio and hook four wires into the frequency synthesizer. This created a real problem with using an external amplifier. Putting all this sensitive circuitry into an external box made it easy for an amplifier to "bleed" the voltage-controlled oscillator and make the channel frequency "warble" in step with your modulation. This was seldom a problem barefoot, but who puts that kind of money into a radio just to run barefoot?
I hope that's "definitive" enough for you. You can usually tell a "definitive" explanation by how many of the audience it puts to sleep.
The digital "slider" for old Browning base stations that we're getting ready for market uses a chip that is almost identical to the PLL02A, the Motorola MC45106. That's the chip that millions of Connex/Galaxy/Superstar radios have used for almost 20 years. Our "slider" (stepper?) runs in 5 kHz steps, but then it was designed to do this in the first place. If I hadn't already had to work out the internal arithmetic for this toy when I designed it, I would have had to go crack a book or two for this long-winded story.
If you want to check the manufacturer's specs on the PLL02A chip, find an old (early 1980's) Motorola CMOS Data Book and look up MC145109. That's the original design that the Japanese company NPC licensed, and then sold as the "PLL02A". Their chip is just a copy of the original Motorola design. The only radio I know old enough to contain one with that Motorola number on it was the Dak Mark 9. Great chip, rotten radio.
Doing this to any radio is a really LARGE project, even if the labor is free. Explains why all the advice in print says to molest the crystal with a capacitor. Much simpler, just as effective.
73
Well, we know of another radio that contains the MC145109. Old (early 1980's) Motorola CMOS Datasheet attached
In any case, during these modifications this is bound to come up. This is a better explanation than I could write in the next hour. (or two)
The definitive information is that changing the "Reference Modulus" pin (4) will make the PLL chip change frequency in 5 kHz steps instead of 10 kHz steps. The first problem here has to do with the filtering between the "Phase Detector" output and the VCO. The chip does not deliver smooth DC tuning voltage out of pin 5, it delivers pulses at a frequency of 10 kHz. Those pulses get smoothed out (filtered) into a steady DC tuning voltage by resistors and capacitors between that pin on the chip and the varactor (voltage-tuned capacitor) diode in the VCO (Voltage Controlled Oscillator). The pulses in the stock setup have a frequency of (you guessed it) 10 kHz. Changing this to 5 kHz requires that this filter be altered. It just won't be as effective as it was at 10 kHz. Rather than bore you with how to do this, we'll skip to the reason it won't matter.
The programmable divider inside the chip is where the binary inputs from the channel selector lead to. This divider will no longer deliver the same channel frequency when you choose 5 kHz as it did when the 10 kHz option was turned on. Depending on the particular radio you are looking at, the channel frequency will jump (usually) DOWN by one or two MHz when this pin is changed.
Pin 4 on the PLL02A chip contains an internal resistor connected to pin 1, the power supply pin. This way, if you leave it unconnected, it has a logic '1' on it, giving you the 10 kHz steps the radio is built to deliver. Saves them a resistor. Grounding the pin changes it to a logic '0', AND changes an internal divider from 1024 (10.24 Mhz divided by 1024 gives you... 10 kHz) to 2048. Dividing 10.24 MHz by 2048 now gives you 5 kHz INSIDE the chip, and this frequency becomes the new "reference" frequency.
The "programmable" divider inside the chip is where your voltage-controlled VCO frequency feeds in on pin 2. If you cut the internal reference frequency in half, changing it from 10 kHz to 5 kHz, the programmable divider will need a binary number TWICE the value used to give you channel 1 as it did with the 10 kHz setup.
This chip was used in a lot of radios, both AM-ONLY and SSB models. The internal binary arithmetic falls into about four categories. For three of these, you can change ONE crystal in the radio to match your new 5 kHz channel steps. Feeding a binary number that is TWICE the old one can be done by removing the channel switch wires from pins 10 through 15, and move them each DOWN by one pin number. The wire that went to pin 15 will now go to pin 14, the one that went to 14 now goes to 13, etc. Pin 15 will now become your "+ 5kHz" switch connection, on an added toggle switch. That will allow you to keep 40 channels. Consider that a 40-position channel switch would otherwise only cover 20 channels, at 5 kHz per click, if you leave them hooked up stock. The frequency needed for the NEW crystal depends on which of three frequeny schemes the radio used, the 10.05 MHz crystal, the 11.86 MHz crystal, or the other one I can't remember right now. This is a lot of trouble, and explains why the Digi-Scan toy was so popular in its day. All this junk was already built into it, just pop the PLL chip from the radio and hook four wires into the frequency synthesizer. This created a real problem with using an external amplifier. Putting all this sensitive circuitry into an external box made it easy for an amplifier to "bleed" the voltage-controlled oscillator and make the channel frequency "warble" in step with your modulation. This was seldom a problem barefoot, but who puts that kind of money into a radio just to run barefoot?
I hope that's "definitive" enough for you. You can usually tell a "definitive" explanation by how many of the audience it puts to sleep.
The digital "slider" for old Browning base stations that we're getting ready for market uses a chip that is almost identical to the PLL02A, the Motorola MC45106. That's the chip that millions of Connex/Galaxy/Superstar radios have used for almost 20 years. Our "slider" (stepper?) runs in 5 kHz steps, but then it was designed to do this in the first place. If I hadn't already had to work out the internal arithmetic for this toy when I designed it, I would have had to go crack a book or two for this long-winded story.
If you want to check the manufacturer's specs on the PLL02A chip, find an old (early 1980's) Motorola CMOS Data Book and look up MC145109. That's the original design that the Japanese company NPC licensed, and then sold as the "PLL02A". Their chip is just a copy of the original Motorola design. The only radio I know old enough to contain one with that Motorola number on it was the Dak Mark 9. Great chip, rotten radio.
Doing this to any radio is a really LARGE project, even if the labor is free. Explains why all the advice in print says to molest the crystal with a capacitor. Much simpler, just as effective.
73
Well, we know of another radio that contains the MC145109. Old (early 1980's) Motorola CMOS Datasheet attached
