-
You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
L21 Frequency mod for MB8719
- Thread starter Tolteca
- Start date
Interested in what you find about the RCI variants.
There are other things to keep in mind. First is very short lead lengths around that pin.
Some people like to run amplifiers and sometimes those amplifiers are a little dirty and on top of that the radios might be clipped and snipped.
Since we are talking about a tenth of a volt change that might either cause that inverter to flip stray RF becomes a concern. I just did it to my test radio with thruhole parts (1" leads) and I was able to create enough stray RF to flip it. When I shortened the leads to exactly what is needed this was nearly eliminated. Bypass caps eliminated the rest of the issue.
Why does this matter? It might not.
I know a lot of people here work on radios and more often than not you cannot control what the operator is doing so keeping the leads short here does a lot to prevent weird issues. Bypasses are just good practice.
You see these guys running 8X cable and throwing 1kW into 200 feet of it with 20 feet of it spooled on the ground under the desk. Who knows what quality of cable...
There are other things to keep in mind. First is very short lead lengths around that pin.
Some people like to run amplifiers and sometimes those amplifiers are a little dirty and on top of that the radios might be clipped and snipped.
Since we are talking about a tenth of a volt change that might either cause that inverter to flip stray RF becomes a concern. I just did it to my test radio with thruhole parts (1" leads) and I was able to create enough stray RF to flip it. When I shortened the leads to exactly what is needed this was nearly eliminated. Bypass caps eliminated the rest of the issue.
Why does this matter? It might not.
I know a lot of people here work on radios and more often than not you cannot control what the operator is doing so keeping the leads short here does a lot to prevent weird issues. Bypasses are just good practice.
You see these guys running 8X cable and throwing 1kW into 200 feet of it with 20 feet of it spooled on the ground under the desk. Who knows what quality of cable...
some one here at one time had the channel mods to take you below the traditional 8719 mods and I think it was unit_399. if it is him maybe he will chime in here. I have it saved to a disk some where but finding it would be next to impossible right now.
the mod I saw would give you 27.415 to 2745 I think it was. it gave you all the skips channels some guy's want . I use to be in a group that used 27.425 on SSB a lot.whoever it was here had the mod.
just remembered to get all the channels you would need 3 switches for most channels then a 4 th switch for the 10KC jumps.
the mod I saw would give you 27.415 to 2745 I think it was. it gave you all the skips channels some guy's want . I use to be in a group that used 27.425 on SSB a lot.whoever it was here had the mod.
just remembered to get all the channels you would need 3 switches for most channels then a 4 th switch for the 10KC jumps.
Last edited:
here is a link to the uniden XL site for the channel mods. not sure if this is the one to give more channels or not. you can check it out. been too long since I did the mod to get 27.415 to 27.465 out of the 8719 PLL. remember you will also need a 4th switch for the 10kc jump to get the alpha channels
http://www.geocities.ws/cobra269/files/Cobra-148-Grant-XL-mods-PW.pdf
http://www.geocities.ws/cobra269/files/Cobra-148-Grant-XL-mods-PW.pdf
Sonoma,
I think you have identified the chl. mod on my 148. I have been searching, in order to figure out the stability of it. Mine will hit that 27.415. Pin 9, pin 8, and pin 6?
I may have it backwards tho? Looking at wires attach to pin out solder side, front of radio.
Last edited:
That was a joke, man.
Just cut the trace to pin 14, and bridge with a 4.7 k resistor. Then, depending on your switching scenario, you can make pin 14 (n=4) go low or high. This will give you your alpha channels, and combined with the standard poor man’s mod, restore 41-44, too.
Givens:
Cobra 148 with Fujitsu branded MB8719...
When I try this mod to pin 10 and change N to (N+128), will the frequency increase or decrease?
Seems to me that if it INCREASES in FREQ by 1.28 MHz, this would be ideal to have the radio bone stock, but flip the switch and have a "band" of frequencies in the Tech Voice Portion of 10 meters...
28.245 - 28.695 ??
ButtFuzz:
This mod will make it go up another bank of 40, so it will increase, if you use the original 11.325 loop mixer crystal. That being said, you can cover 10M SSB voice if you go N+128.
~Cheers~
This mod will make it go up another bank of 40, so it will increase, if you use the original 11.325 loop mixer crystal. That being said, you can cover 10M SSB voice if you go N+128.
~Cheers~
Just tried this with a "Made In The Malaysia" 148 GTL. MB 8719 PLL. No Go.
I will try another generation another day.
All y'all are goin' 'bout this all bass ackerds.
First, Pin 10 is tri-state as Andy mentioned in post #12 earlier.
Pin 10 has an internal pull up resistor, else it would need to be strapped to either VCC or GND on the board, hence the isolation (not tied to anything) (148/2000/Grant/Madison) on the PCB board.
So, let's do some math.
11.3258 * 3 is 33.9775 MHz, which when "dialed down" by L23 (AM) takes it to 33.975 MHz
Pin 10 when high (or isolated), adds 64 to the N code from the channel selector. N=15 for Ch. 1, so the PLL sees 79 when on Ch. 1.
33.975 + .79 = 34.765. 34.765 - 7.8 = 26.965. So that checks out.
When you ground Pin 10, it turns off the +64 bit, and through the internal inverter, turns on the +128 bit.
Now, to have both +64 and +128 bits set (+192), the pin needs to be hard biased to about 1/2 VCC. This tricks both pins into being "on" at the same time, it is just high enough to turn on the +64 bit and just low enough to "ground" the input to the inverter going to the +128 bit. Remember to take into effect the internal pull up resistor, though.
Tying it to L21/R138/C111 through a resistor, C111 (.01uF) filters out the RF in the coil, so you are introducing a DC component, not the AC from the loop mixer. And the voltage is the emitter voltage from TR20, the loop mixer, so close enough to 1/2 VCC.
It also "sort of" works with the F-In (Pin 17) as it is also biased at 1/2 VCC, but may load the F-In signal.
What I used to do is put a 1uF@16v electrolytic cap from Pin 10 to ground, so the voltage is stable going into this pin. I also used a divider network of 2 10k resistors in series from VCC to GND, to set up the 1/2 VCC voltage, and a small resistor (1k?) to pin 10 and the 1uF.
Isolated, or N/C, normal (N+64).
Grounded is N+128
1/2 VCC is +192 (+64 and +128).
I haven't worked on a 148/Grant etc... radio since 2000, this is all from memory and looking at the schematics.
Also, R95 (1.5k) from VCC to the output of the 2nd Active Filter output and input to the VCO, this resistor pre-biases the VCO so it can not go below about 1.5-2v or so, if you increase it's value to 4.7k or 10k, this allows the VCO to be fed lower voltages when going down, you would then readjust the VCO coil to just go into lock at 26.815 at around .7 - 1.0v, this gives you a little more head room for the upper frequencies.
First, Pin 10 is tri-state as Andy mentioned in post #12 earlier.
Pin 10 has an internal pull up resistor, else it would need to be strapped to either VCC or GND on the board, hence the isolation (not tied to anything) (148/2000/Grant/Madison) on the PCB board.
So, let's do some math.
11.3258 * 3 is 33.9775 MHz, which when "dialed down" by L23 (AM) takes it to 33.975 MHz
Pin 10 when high (or isolated), adds 64 to the N code from the channel selector. N=15 for Ch. 1, so the PLL sees 79 when on Ch. 1.
33.975 + .79 = 34.765. 34.765 - 7.8 = 26.965. So that checks out.
When you ground Pin 10, it turns off the +64 bit, and through the internal inverter, turns on the +128 bit.
Now, to have both +64 and +128 bits set (+192), the pin needs to be hard biased to about 1/2 VCC. This tricks both pins into being "on" at the same time, it is just high enough to turn on the +64 bit and just low enough to "ground" the input to the inverter going to the +128 bit. Remember to take into effect the internal pull up resistor, though.
Tying it to L21/R138/C111 through a resistor, C111 (.01uF) filters out the RF in the coil, so you are introducing a DC component, not the AC from the loop mixer. And the voltage is the emitter voltage from TR20, the loop mixer, so close enough to 1/2 VCC.
It also "sort of" works with the F-In (Pin 17) as it is also biased at 1/2 VCC, but may load the F-In signal.
What I used to do is put a 1uF@16v electrolytic cap from Pin 10 to ground, so the voltage is stable going into this pin. I also used a divider network of 2 10k resistors in series from VCC to GND, to set up the 1/2 VCC voltage, and a small resistor (1k?) to pin 10 and the 1uF.
Isolated, or N/C, normal (N+64).
Grounded is N+128
1/2 VCC is +192 (+64 and +128).
I haven't worked on a 148/Grant etc... radio since 2000, this is all from memory and looking at the schematics.
Also, R95 (1.5k) from VCC to the output of the 2nd Active Filter output and input to the VCO, this resistor pre-biases the VCO so it can not go below about 1.5-2v or so, if you increase it's value to 4.7k or 10k, this allows the VCO to be fed lower voltages when going down, you would then readjust the VCO coil to just go into lock at 26.815 at around .7 - 1.0v, this gives you a little more head room for the upper frequencies.
Last edited:
Thanks for that. Someone do it, show it works, and I'll buy it. Putting a relatively highly regulated 1/2 vcc should do it... which is what I did, and It didn't work... So until someone does it, captures it with photos xmitting on said 1.28 MHz higher frequency, it simply isn't worth the time to me... as I have radios that xmit from .5 - 54 MHz continuously.
That is all.
That is all.
Buttfuzz:
You have to have anywhere from 3.65-4.2V (or roughly 1/2 VCC) to pin 10 to get the +192 state. That's why I made a circuit with a LM7805 and a 3.6V 1W zener diode, with a couple caps and a variable resistor, to dial in the 1/2 VCC necessary to get the +192 to work. I incorrectly stated that in my previous post as N+128, when I meant to say N+192 as stated by NZ8N. (Thanks to NZ8N for that corrrection.)
However, starting with the Malaysian radios, they started doing away with the Fujitsu MB8719 PLL IC's, and started going with "copies" of that PLL. Some of these radios have an MMB8719, and some of them also had some early variants of the RCI8719 PLL.
When trying to mod the newer PLL's for N+192, it will not work, as the silicon substrate "setup" in the PLL is not the same as the old Fujitsu PLL's. That being said, in testing the MMB8719's simply will not toggle both N+64 AND N+128. it will only toggle one, depending on voltage. On the early RCI8719 variants, if you tried this method, you'd destroy the PLL by "damaging" pin 10 so it is no longer useful. That's likely why it didn't work on your "The Malaysia" radio.
This N+192 method will ONLY work on the Fujitsu MB8719's.
(Edit) Also, NZ8N also brought up a good point about R95. Changing that value slightly will improve bandwidth "reach" to where you can lock the N+192 channels.
~Cheers~
You have to have anywhere from 3.65-4.2V (or roughly 1/2 VCC) to pin 10 to get the +192 state. That's why I made a circuit with a LM7805 and a 3.6V 1W zener diode, with a couple caps and a variable resistor, to dial in the 1/2 VCC necessary to get the +192 to work. I incorrectly stated that in my previous post as N+128, when I meant to say N+192 as stated by NZ8N. (Thanks to NZ8N for that corrrection.)
However, starting with the Malaysian radios, they started doing away with the Fujitsu MB8719 PLL IC's, and started going with "copies" of that PLL. Some of these radios have an MMB8719, and some of them also had some early variants of the RCI8719 PLL.
When trying to mod the newer PLL's for N+192, it will not work, as the silicon substrate "setup" in the PLL is not the same as the old Fujitsu PLL's. That being said, in testing the MMB8719's simply will not toggle both N+64 AND N+128. it will only toggle one, depending on voltage. On the early RCI8719 variants, if you tried this method, you'd destroy the PLL by "damaging" pin 10 so it is no longer useful. That's likely why it didn't work on your "The Malaysia" radio.
This N+192 method will ONLY work on the Fujitsu MB8719's.
(Edit) Also, NZ8N also brought up a good point about R95. Changing that value slightly will improve bandwidth "reach" to where you can lock the N+192 channels.
~Cheers~
Last edited: