MIDLAND CB MODEL 79-265 MOBILE EXTRA CHANNELS/CLARIFER MODS

[500kpot]

Thanks for the clarification on the 25 and 29.

The circuit I have on the "drawing table" has 3 buffered outputs, one for the RX mixer, one for the TX mixer, and a third for a frequency counter, if so needed. And, it can be used on radios that the owner didn't want traces cut on the PLL lines, just pull a few components, and put my board in place of the components.

Sounds really cool. I'm trying to visualize this in operation and I think I'm stuck on how to get more bands of 40 if the pll is tricked in to thinking its always in receive. Will your board replace the 15.36 megahertz signal? Or do you have some killer trick up your sleeve? I'm sure you do, so I'm eager to hear about it sometime!

Joe Truck Driver would want it so his radio to have the same channels as his friends with Galaxy type radios, not some odd 36 channel overlap or off frequency by 5 KHz.

I'm not sure about Joe Truck Driver, I'm not a truck driver. I would think a truck driver would just get a radio that does those funnies already? I only have a small base station with a home brew antenna. No extra power or funnies even.

I am under the impression that the radio is not off by 5 kilohertz because the slide is introduced on the 5 megahertz signal before the trippler not by manipulating the 10.240 . I think its the same concept as the Lescom 29 board with the 14.91 / 15.91 crystal as you have explained in an earlier post, but not on such a large change. Also there would be no overlap as it appears the frequency range is 26.605 - 27.505.

My take on the circuit in simple math terms:

(5 Megahertz x 3(15.00)) + 1.365 (Ch1 Transmit PLL loop) + 10.695 = 27.060 Indeed 5 kilohertz off.

If the slide does come from pulling the 5 megahertz signal and it was lowered to
4.9983 megahertz then

(4.9983 x 3) = 14.9949 +1.365 = 16.3599 + 10.695=27.0549

I arrived at this 4.9983 by

27.055 - 1.365 = 25.69 - 10.695 = 14.995 / 3 = 4.9983

(I did and experiment in class of pulling crystals and this pull amount is totally within the realm of possible.)

Close enough to 27.055 for transmit

And on receive
27.055 - 1.365= 25.69 -10.240 = 15.45 - 14.995 = .455 Megahertz


I believe this theory is correct as the frequency counter shown does indicate the slide which must mean its taken from the VCO + loop mixer and not the 10.240.

So friend I'll have to disagree with you respectfully that the radio is 5 kilohertz off or should I say the radio isn't 5 kilohertz off when the slider is used to lower the VCO 5 kilohertz to the 5's when using the transmit VCO + loop frequency. Granted the channels displayed are all messed up but IMHO the frequency counter makes up for it. On a personal note, I would rather have a VFO than a channel selector; I mean channels selectors feel so....CB.

I digress and I feel your getting burnt on answering my questions so I'll shut up and read more. And I'm really excited to see your PLL modification board.

And as always a big thank you to you for giving me a chunk of your time as I know your busy and I appreciate it.

PS. I guess your not to hip to the 858 switch modification for frequencies or the cobra 148 switch modification as those channels don't line up either, but you prefer a more organized approach. I'll bet your work bench, car and house are super clean and your work reflects the same.

500kpot back in the bush

 

[500kpot]

The circuit I have on the "drawing table" has 3 buffered outputs, one for the RX mixer, one for the TX mixer, and a third for a frequency counter, if so needed. And, it can be used on radios that the owner didn't want traces cut on the PLL lines, just pull a few components, and put my board in place of the components.

Hello NZ8N. I hope all is well going for you. Is there any traction on your PLL Modification board? Send the gerber off the get the PCB working? I'm so excited to get this circuit of yours installed and working..The way you describe it I can't help but to think of that in my cobra 29..

I was just hoping for an update.

Best Wishes

500kpot

 

[hpssb]

One of the local techs here did a Cobra 29 for one of their customers in the USA (they had to send a radio here 'cos cb radio down here is ssb only). Posted with all the technical details with pics (way over my level of understanding tho) on the trans1 forum. Cost was under $20 with currently available parts and no crystals needed. Gives high 40 band, low 40band, 10kc shifter and a 5kc shifter.

http://www.transmission1.net/viewtopic.php?f=53&t=56500

 

[nomadradio]

Just bought some of the eat-a-fruit clock-gen boards.

Gotta learn to talk to 'em.

73

 

[NZ8N]

Sorry, I haven't been on in a while, been busy with other things.
That particular transmission1.net posting would only work for a Cobra 29 or Uniden 76/78 etc... that uses a 15.360 mixer to downmix the VCO down to .91-1.35 MHz.
And, I do have Arduino boards and an Adafruit si5351 board to experiment with, but I wanted my own design, not use off the shelf items.

I do have a few designs I played with on my mixer board that had an onboard Atmel Chip for the brains, and an si5351 chip for the frequency generation, but that complicated the hardware design.
My system uses 2 mixers simultaneously, one to upmix the VCO to a higher arbitrary frequency range, then the second remixes that frequency back down to our target frequency.
One possibility is to break the VCO to the PLL input, but then you'd have to wideband the VCO, easy enough to do, but...
The other possibility is to break the VCO signal to the RX and TX mixers, leaving the PLL and VCO mostly unmodified, run the VCO signal through the custom add-on board mixers, then put the resultant signal to the RX and TX mixers.
There would be high pass, low pass and band pass filtering at each stage, and because the mixer I plan on using has gain, a small attenuator between stages to relevel the signal level. And 3 buffered outputs, isolated, so the radio mixers and/or frequency counter won't interfere with each other.
The first upmix frequency and the second downmix frequency are irrelevant, so long as they have the correct offset between themselves.
So, any RIT/Clarifier mods pass straight through, it will have the same plus or minus offset as if the board were not there.

 

[500kpot]

I like that post on transmision1. Looks fun to experiment with.

 

[500kpot]

Hello NZ8N! How are you today?


I hope you would be so kind as to help me understand.

My system uses 2 mixers simultaneously, one to upmix the VCO to a higher arbitrary frequency range, then the second remixes that frequency back down to our target frequency.

As an example say, channel 1 TX (Stock)
VCO = 1.365 + 15.36 +10.240 = 26.965 (Channel selector on Channel 1)

Does your mixer scheme add .450 to VCO output?
(1.365 + .450) = 1.815 +15.36 +10.240 = 27.415 (Channel selector on Channel 1)

Is this basic idea of how it works?

Thanks in advance

 

[NZ8N]

Ok, let's do some math.

AM only chassis for now, we can get into the SSB offsets later.

Ch-01 - VCO RX Freq 16.270, whether the Cobra 29 type (downmixes with 15.36 to get .91 MHz or Cobra 25 type that reads the VCO directly.

26.965 - 16.270 = 10.695 (the first IF). Anything else just confuses, so let's keep it simple. the VCO frequency being lower than the intended target channel frequency makes it a low side injection radio.

My first idea was, why not just upmix 16.270 with 21.39 (10.695 x2) to make the radio a high side injection, which comes out to 37.66 MHz.
26.965 - 37.660 = 10.695 (we are dealing with mixers, not math, so 10.695 above the intended channel frequency, or 10.695 below the intended frequency comes out to the same place, same channel. But, 21.39 might cause a "birdie" or an unintended carrier, being a multiple of the IF frequency. So, for normal channels, we just pass the VCO straight through, bypass the mixer.
21.390 - .45 (one band down) gives 20.940 MHz, and
21.390 + .45 (one band up) gives 21.840 to mix with the VCO then the 37MHz signal goes to the RX and TX mixers.
This would work on an AM only radio like a 29 or 25.
For an SSB radio (I have a Uniden PC122), the high side injection would cause the Upper Sideband to be off frequency by 5 KHz, and
Lower Sideband would be off frequency the other direction by 5 KHz.

BUT !!!

What if, we instead, mixed it twice? Upmix to get the VCO + upmix frequency to, say, 32 MHz, then downmix it (2nd mixer) with nearly the same as the first mixer, but offset in which ever direction we need to offset it to?

Like for instance,
16.270 + 15.360 = 31.630 (1st mix and High Pass Filtered in the add-on board)
31.630 - 14.910 = 16.720 (2nd mix and Low Pass Filtered in the add-on board)
16.720 + 10.695 = 27.415.

Or, to go the other direction,
16.270 + 14.910 = 31.180 (1st mix in the add-on board)
31.180 - 15.360 = 15.820 (2nd mix in the add-on board)
15.820 + 10.695 = 26.515.

If I use a balanced mixer at each stage, the fundamental frequencies would be balanced out, and only the sum and difference frequencies will be generated.

Here is the beauty of this idea, no matter what the VCO offset was for USB or LSB, and no matter what RIT/Clarifier mods were done, they will still have the same offset as before the add-on board.
I used 14.910 and 15.360 as examples, because they have the correct difference in frequency to each other to offset by one whole band, so long as they are .45 from each other, or what ever the offset you want to go above or below, 15.360 and 15.810 would work too, and by using 14.910 and 15.810, which are .900 apart, would cause the radio to go 2 bands up or down). The offset mixer frequencies are actually irrelevent, so long as they are offset from each other by the correct amount.
To go up or down by 5 KHz, one of the offset mixer frequencies has to be up or down by 5 KHz, and to go up 10 KHz, one of them has to be up by 10. To go up 5 KHz, use the -5 and the +10 at the same time, -5 on the 1st mix, and +10 on the 2nd mix.

This idea could be used for ANY radio, regardless of what PLL they throw in to keep the FCC from rejecting their new designs.

 

[500kpot]

I see. This idea is pretty cool.

I must admit it took me a bit to figure out how it works.

I assume that you can select any pair of crystals and I assume that there are 4 crystals, one used as the 10.695 transmit oscillator meanwhile the other three are 15.81, 15.36, and finally 14.91 and are used for mixing. I also assume your mod shuts off the transmit/receive shift and only the receive code is used or .91 - 1.35 megahertz.


Selecting Crystals
The first number is the first IF and Ch1 loop+vco is used 16.270 megahertz
Ill spare the math

A-G = 40 Channel Bands
d= Duplicates
o= Overlaps / Odd starting points from Ch1
VCO+Loop 16.270(rx) 16.725(tx) * -5hz on the first IF
14.910/15.360 = 26.515 - 26.955(B) 14.905/15.360 = 26.965 - 27.405(*C)
14.910/15.810 = 26.065 - 26.505(A) 14.905/15.810 = 26.515 - 26.955(d)

15.360/14.910 = 27.415 - 27.855(d) 15.355/14.910 = 27.865 - 28.305(o)
15.360/15.810 = 26.515 - 29.955(d) 15.355/15.810 = 26.945 - 27.385(o)

15.810/14.910 = 27.865 - 28.305(E) 15.805/14.910 = 28.315 - 28.775(F)
15.810/15.360 = 27.415 - 27.855(D) 15.805/15.360 = 27.865 - 28.305(G)

It seems the 15.360/14.91 using rx loop+vco is nothing but duplicates and using the 15.355/14.91 tx loop+vco creates odd channels in the sense that you explained about have Galaxy type bands.

I am under the impression if the stock 16 megahertz signal for legal 40 channels is used, the A channels will be unobtainable and the A channels will only be available when using the funnies. If I'm wrong, I apologize in advance.

The reason I'm under that impression is I could not find a combination of 15.36, 15.81 or 14.91 using the rx (16.270 - 16.710 megahertz ) which produced the legal 40. Since you have a 10.695 transmit oscillator then passing the stock VCO+loop gives the standard 40.

I did find a combination with the tx VCO+loop (16.725 - 17.615) using the 14.91 as the first IF and subtracting 5 kilohertz to 14.905 and the 2nd IF of 15.360 megahertz resulting in the legal 40 and A channels are the available.

It appears this mod would give 26.065 - 28.775, quite a healthy range.

The only questions I'm left with are and I'm sure easy for you to answer,

Are there any issues mixing signals that are very close to each other like 16.270 and 15.81?

Can you foresee and problems having two oscillators running that are close in frequency like 15.81 and 15.36 at the same time?

What type of circuit will you use to lower the fundamental frequency of the crystals 10 kilohertz?

How big is this board?

And finally the big one -
How much will it cost in a retail type setting?

Thank you again for your explanation of your frequency modification.

I hope I've got it all correct!

 

[NZ8N]

Start over.



Ignore the TX line for now, with a Cobra 29 or 25, the VCO in TX mode is still in the 16 MHz range, so it can be used normally, or if you wanted a frequency counter, then the TX line on the PLL would have to be disabled, hard wired to the RX mode, and add in the additional 10.695 TX oscillator to the circuit.

For an AM only radio like a Cobra 29 or 25, and the Uniden versions, I would probably just use the straight uppers and lowers without the + or - 5kc or +10kc.
On SSB radios, I'd use the full version, with the + or -5kc, +10kc switch.

Or maybe I'll just scrap the idea, and go with the si5351 and Arduino chip to take the place of the PLL, like the previous mentioned transmission1.net posting.

Like I said earlier, I have several different versions, with crystal oscillators, with Arduino and si5351, and various PLL's too, some even with a dual PLL chip like used in an older cordless telephone.

 

[500kpot]

When you stated to ignore the TX VCO+Loop

It was hard to ignore due to the normal 40 channel deal.

I accidentally hit the post button.

 

[500kpot]

or if you wanted a frequency counter, then the TX line on the PLL would have to be disabled, hard wired to the RX mode, and add in the additional 10.695 TX oscillator to the circuit.

I think this is the TT way in the video but hes kept the tx VCO+loop and created a -5 kilohertz slider on the 5 megahertz oscillator which allows operation on the 5s instead of 0s then slide back when using the rx VCO+loop.

Anyway, thanks again for the reply's. I learned alot by following your mod whether you actually make it or not, it served as an very nice learning platform.

500kpot

 

[NZ8N]

The top part of that spreadsheet shows how to derive -5kc, +5kc and +10kc from any band.
The second part of that spreadsheet shows how the bands would be set up, and the bottom 9 frequencies would be what would be needed to get the -5kc, +5kc and +10kc offsets for each band, I might not need all 9 frequencies, maybe 7 or 6 of them.

And, the frequencies do not necessarily need to be 14.910/15.360 etc, they need to be 450kc from each other, could use 18.000, 18.450 and 18.900 MHz, so long as there is a 450kc offset from one to the next.

 

[500kpot]

The top part of that spreadsheet shows how to derive -5kc, +5kc and +10kc from any band.

I understand the logistics of how to derive the +/- 5 kilohertz and 10. I was in wonder of what circuit you would use to actually pull a crystal that far from its fundamental frequency and still remain stable.

they need to be 450kc from each other,

And here I, again, was in wonder of running two oscillators so close in frequency and if there might be some kind of unwanted interaction which raised another question in parts procurement. Finding crystals that are 450khz apart may prove to be more difficult that building the mod although I haven't really looked.

Have an awesome day!
500kpot

 

[NZ8N]

A simple mixer takes 2 signals, mixes them together, and the resultant signals coming out of it are the original 2 frequencies, the sum of the 2 signals, and the difference, so 4 signals.
Example, the PLL offset mixer in your 29, the output of the mixer has 15.360, 16.270, 31.630 (the sum) and .910 (the difference). The PLL only sees the .910 signal to phase lock to, ignoring the rest.

A balanced mixer takes 2 signals, phase cancels the original frequencies, and only outputs the sum and the difference. It does not matter how close they are to each other, look at the balanced modulator (mixer) used to generate sideband signals. I am only interested in the high pass filtered signal in the 1st (balanced) mixer of my design, and the low pass filtered signal in the 2nd (balanced) mixer.

To be on the normal CB channel band, the board would either have a buffered bypass circuit to pass the original frequency (16.270) straight through, or use the same mixer signal on both upmix and down mix, net zero change.

As to shifting (pulling) the crystals to get -5kc, +5kc and +10kc, I found a supplier that makes custom oscillators to specific frequencies (square wave though), at $.80 each. So, it costs me $2.40 for 3 oscillators, and $7.20 for 9 oscillators. At that cost, there is no need to cap or choke a crystal to shift (pull) it, which reduces it's signal level. Which is one of the reasons why Expo 100 A+ kits seemed weak, crystals being pulled too hard from their original cut frequency.

The easiest and simplest way to expand your 29 is with an Expo 100 A+ kit clone, either DIY or bought off of eBay. This injects or overrides the 15.360 signal with 14.910 for lowers and 15.810 for uppers.