Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Dec Tue 10, 2019 6:38 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 32 posts ]  Go to page Previous  1, 2
Author Message
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 31, 2019 9:56 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3852
Location: Lincoln City, OR 97367
Greetings to Ted and the Forum:

Thanks for figuring out why there were two sets of numbers. As I am always very conservative in my designs, I zeroed in on the 18 MHz number and chose the divisor that would be less than that. Since that works out to be just a bit shy of 14 MHz, I would expect that any chip I get that is fully functional should be able to handle it without difficulty.

Regards,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 31, 2019 12:28 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 201
I have used 74HC4059 in the past and they worked fine at 24MHz.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 31, 2019 1:52 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 201
If it won't be used much then build the original circuit only. It covers the whole of FM band. I'll suggest a few modifications in the driver and PA stage to make it more stable.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 31, 2019 2:26 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 201
https://ibb.co/PD5S8mx

My modification to the original circuit.

The RF stage is more stable now with cleaner output.
Use a resistor attentuator pad at the output to lower the power levels. The 100ohm resistor feeding the S9018 can also be increased to reduce the output power.

Also use a 7 pole low pass filter at the output for harmonic suppression of -50dBc or more.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Sat 01, 2019 3:52 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3852
Location: Lincoln City, OR 97367
Greetings to Dare4444 and the Forum:

Quote:
If it won't be used much then build the original circuit only. It covers the whole of FM band. I'll suggest a few modifications in the driver and PA stage to make it more stable.


I'm afraid that you have left me behind completely. I just don't understand this.

The unit, once I get it working, will be in continuous use, 24/7. The question remaining is whether I can get the TFT circuit to work on the new frequency or whether I will have to start over from scratch and build a completely different synthesizer.

The original circuit (I presume you mean https://www.docdroid.net/XIJ9RPg/const- ... mitter.pdf ) is not usable because it does not cover the frequency range of interest. The required frequency range is 98.6 MHz to 118.8 MHz if the unit is to be frequency agile. For a single frequency circuit (non agile) the required frequency is 109.4 MHz. This frequency is not within the U.S FM broadcast band.

I have not researched the CD4046 chip, but I would assume that any phase comparator has a margin of error. If that error is expressed in degrees plus or minus, then the comparison frequency should be as high as the chip reasonably permits because the phase error is multiplied by the divisor. Therefore, a smaller divisor and a larger compare frequency will result in less frequency error at the output. The 25 KHz comparison frequency that I suggested will have only 1/16 of the error at the operating frequency than would a division all the way down to 1.5 KHz.

For a simple FM transmitter as is presented at https://www.docdroid.net/XIJ9RPg/const- ... mitter.pdf it would be perfectly adequate for its intended purpose if it were within plus or minus 2 KHz at the operating frequency (this is the tolerance required of FM broadcast stations by the FCC). For my application, this magnitude of error would be totally unacceptable. The TFT time base is good for approximately 1 part in 10^8. This is approximately 1 Hz at the operating frequency. Since the device contains a frequency counter which is supposed to be used by station engineering personnel to monitor and correct if necessary the station's carrier frequency, it follows that it must be considerably better than the +/- 2 Khz allowed by the FCC. Therefore, IMHO, the higher comparison frequency is desirable for two reasons, component count and accuracy.

I thank you for your efforts, but I think it best that we put this project on hold until I have exhausted all efforts to make the TFT circuit work.

Thanks,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Sat 01, 2019 4:30 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 201
CD4046 worked fine with 750KHz reference in one of my earlier projects. With a high reference frequency, the TX won't cover the FM band in 100KHz steps but it will with 1.5625KHz


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Sun 02, 2019 5:35 pm 
Member

Joined: Jan Tue 16, 2007 7:02 am
Posts: 3251
Location: Lexington, KY USA
I would assume that any phase comparator has a margin of error.

Once the system is locked, the error is phase error, not frequency error. So the long-term frequency error will just depend on that of the reference frequency.

Ted


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Sun 02, 2019 10:32 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 201
Yes when the frequency is locked the error is only 1Hz or so with 4046.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Mon 03, 2019 7:03 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3852
Location: Lincoln City, OR 97367
Greetings to Usually Lurking and the Forum:

Back in the good old days, when it was not possible to deviate a master oscillator by very much, FM transmitters began with a crystal controlled or phase locked oscillator at a low frequency. This could only be deviated a small amount. However, by multiplying the frequency of this oscillator, the deviation was also multiplied. In some early FM broadcast transmitters, the signal was multiplied, hetrodyned down and then multiplied again to come up with the necessary 75 Khz of deviation.

Now, division is the inverse of multiplication. When a VCO in a synthesizer is divided by some number N to get the comparison frequency, the error in that frequency is also divided by N. The larger the divisor N, the smaller the absolute magnitude of the error. It may be that the error produced by the chip is so small that it does not matter anyway, but let us suppose that the error is 10 degrees. The divisor to get to 1.5625 KHz is 70016. This means that the phase error at the VCO frequency is 10 degrees times 70016 or 700160 degrees. If we divide this by 360 to get Hz, we end up with 1945 Hz. This means that the synthesizer can be off by this amount when "locked".

If we use my preferred higher divisor of 4376 to get a 25 KHz comparison frequency, then the same 10 degree phase error is only 43760 degrees, or 122 Hz. Hopefully, the chip is much better than plus or minus 10 degrees, but the same principal applies.

In other words, nothing is perfect... including frequency synthesis.

Regards,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Thu 06, 2019 5:33 am 
Member

Joined: Jan Tue 16, 2007 7:02 am
Posts: 3251
Location: Lexington, KY USA
I'm not sure what you are trying to explain (it's definitely too late in the day to try to figure it out); I hope I have not missed the point entirely. I do know that a properly designed phase locked loop can deliver an exact ratio between reference and output frequencies. There will be phase noise, or jitter, in the output, but this can be made small for most purposes. The PLL, once locked, is analogous to a mechanical gear train.

The real tricks are getting the loop to lock and stay locked, and producing a clean output.

Note that a constant 360 degree phase error is the same as a 720 degree or 360,000,000 degree phase error. It is modulo 360. And once locked, phase error at the phase comparator does not represent any frequency error, as long as it does not change.

Ted


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Thu 06, 2019 2:54 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 201
The CD4046 locks the frequency down to plus or minus 1Hz. I've checked it with a high precision frequency counter so my PLL TX will deliver a rock stable frequency output. A true phase lock is not obtained in my circuit, only the frequency is locked. The loop filter time constant allows the PLL to be modulated by + - 75KHz for high fidelity audio. The output frequency is as stable as the reference input to the 4046. The chip is a high performer.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Jun Tue 18, 2019 8:46 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3852
Location: Lincoln City, OR 97367
Greetings to the Forum:

Thanks again to all who responded. I finally managed to get the original synthesizer in the unit to move where it needed to go. What a struggle! Took months of trial and error and my brother's help with circuit analysis and some circuit modeling with LTSpice. However, it is now locking up at 109.4 MHz, which means the modulation monitor is able to receive the station it was intended for at 98.7.

I may wish to build one of the synthesizer circuits here (Dare4444's or the other) as a test and calibration signal generator for the thing. Mainly, I need a mono source that I can hit with the appropriate Bessel tone so that I can check the modulation monitor's calibration. All of the transmitters that are fully synthesized on eBay are either in the $300 range or use the BH1417F chip which cannot do mono as far as I can tell.

So, I may be back if I decide to build one of these and run into trouble.

Thanks again,

_________________
Jim T.
KB6GM


Top
 Profile  
 
Post New Topic Post Reply  [ 32 posts ]  Go to page Previous  1, 2

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 9 guests



Search for:
Jump to:  




























Privacy Policy :: Powered by phpBB