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


It is currently Jul Mon 22, 2019 12:49 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 32 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Wanted: Frequency Synthesizer Schematic
PostPosted: Apr Sun 28, 2019 10:00 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3491
Location: Lincoln City, OR
Greetings to the Forum:

I have not found much in the way of detailed frequency synthesizer schematics, so I thought I'd ask here.

I need a frequency synthesizer to generate the L.O. for an FM broadcast receiver (actually, a modulation monitor).

The exact parameters I need are:

1) a 109.4 MHZ output (98.7 MHZ receive with 10.7 MHz IF and high side injection) - frequency agility to cover the entire FM band desirable but not necessary (that would be approximately from 98 to 119 MHz).
2) external 5 MHz reference.... 5 MHz reference is approx 3.5 V P-P and will drive a TTL load.
3) output of +3 dBM minimum.
4) harmonics and spurs 50 dB below below output or better.
5) Minimum practical phase noise.
6) Simple to build; minimum component count; readily available parts; all through-hole parts design a plus.
7) Available voltages to operate synthesizer are +15, -15 and +5. (On-card regulation to derive other voltages from these possible but not desirable).

Anyone have a schematic lying around that could be tailored to fit these requirements? (Or a custom design to fit them exactly? :D )

Thanks,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Apr Sun 28, 2019 2:41 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
I've one.. it covers 88-108 MHz in 100KHz steps.

Coil L2 is unmarked. It's a BN-43-202 binocular ferrite core and power output is around 150mW. Here's PDF

https://docdro.id/XIJ9RPg"


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Apr Sun 28, 2019 4:09 pm 
Member
User avatar

Joined: Aug Tue 28, 2018 9:22 pm
Posts: 779
Location: Sanford Fla 32771
I am not the OP but Interresting schematic, thanks Dare4444.

_________________
Paul of Florida


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Apr Sun 28, 2019 5:50 pm 
Member

Joined: Sep Tue 30, 2014 6:08 am
Posts: 4230
Location: Norfolk, VA
Goes to 109.7 MHz:

http://67.225.133.110/~gbpprorg/lpfm/PLL_Exc.htm

I've got all the original parts to build one, but got sidetracked. Prescalers were a freebie - I got free samples from National Semi at my college mailbox. :)

Almost all CMOS construction, with the only hard to find item being the prescaler and the photomod. He discusses being able to sub the California Eastern Laboratories' UPB1506GV or UPB1507GV prescalers. I'll use an RCA (actually Magnetek) photomod, with 12V lamp, and 500K ohms dark resistance photoconductive cell. BB105 (MV2105) came from an RCA VHF TV tuner module from the 80s, or use an ECG611/NTE611.

Or, for minimum parts count, perhaps one of the Motorola chipsets like they used in their 2-way FM stuff? Or go modular with some MiniCircuits stuff?

_________________
Brian
"Capacitor Cosmetologist since 1979"
USN Retired 1984-2006 (Avionics/Cal)


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Apr Sun 28, 2019 9:26 pm 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3491
Location: Lincoln City, OR
Thanks, Guys!

There is definitely food for thought in both of these circuits. Unfortunately, both of them use on-chip oscillators and local crystals, neither of which is the 5 MHz that I need.

The L.O. circuit is used to provide a highly frequency accurate IF to a pulse-counting discriminator, which is in turn, calibrated by a set of pulses derived from the unit's 5 MHz ovenized xtal time base. This allows the unit to accurately measure the FM broadcast standard 75 KHz deviation to within 1%. In addition, the IF frequency is counted by a digital counter, again referenced to the unit's xtal time base which provides a measure of the station's carrier frequency error in Hz. Therefore, the L.O. has to be referenced to and locked to the same 5 MHz time base.

I'm not a good enough engineer to figure out what would happen if I jammed an external 5 MHz into one of the crystal oscillator ports in either of these circuits nor am I good enough to figure out how to reprogram the synthesizer for the desired 109.4 MHz output with a different time base frequency than specified.

However, the VCO and buffer circuits are interesting and may help me to overcome difficulties with the original circuit which at the moment are defeating me. I may post for trouble-shooting help under tools and test equipment if the moderator permits a second related but not identical subject.

Thanks again to all who responded.

Still looking,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: Apr Mon 29, 2019 1:49 am 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
The reference frequency is 1.5625 KHz being fed to the CD4046 comparator. Your 5MHz can be used to drive a CD4059 set to divide by 3200 and it will provide the 1.5626KHz reference to the CD4046 and this PLL design can easily cover your 88MHz - 120MHz requirement.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 24, 2019 6:13 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3491
Location: Lincoln City, OR
Greetings to the Forum:

Thanks to all who posted.

I have been analyzing the circuit at https://docdro.id/XIJ9RPg" and have come to the conclusion that, while it gives me a starting point, it is unable to meet my initial requirements.

The circuit uses a 6.4 MHz reference; my reference must be an external 5 MHz reference. The circuit uses a fixed power of 2 divider for the frequency reference; using powers of 2 as divisors, a suitable VCO frequency cannot be obtained since the divisors for the reference frequency and the operating frequency must both be integers.

Note that the requirement for my synthesizer is for high side L.O. injection (carrier plus 10.7 MHz) and therefore the divisors for 88 - 108 MHz are also unsuitable. This plus the change in reference frequency requires that all arithmetic associated with the frequency division process must be re-done from scratch.

Some work with a spread sheet has revealed that the circuit can be extensively modified to do what I need, but the CD4060 would have to be replaced with a second CD4059 which would require a second prescaler ahead of it to bring the 5 MHz reference down to a frequency it can count.

The first prescaler chip (ahead of the original CD4059) cannot be located, nor can any of the DIP chips that might replace it. There are a number of surface mount prescalers that might do the job, but I have not looked into their specs in detail. The original circuit calls for a divide by 64 in front of the CD4059; because of the higher frequency of the L.O., this is insufficient to bring the input frequency in range of what the CD4059 can handle so either a divide by 128 chip or two chips would be required here.

I still hope to salvage the original circuit as built in the device but if not, it looks as though a completely original synthesizer design will be required for my project. If I go that route and work up such a design, I will post it here.

Regards,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Sun 26, 2019 5:01 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
Hi Jim, the prescaler chip can be replaced by three cascaded 74AC74 or 74F74 to divide the output by 64. I believe the CD4059 will work fine up to 8MHz or so. Your reference frequency of 5MHz should drive another 4059 just fine. The divide rate would be 3200 to bring down 5MHz to 1.5625KHz reference frequency for the CD4046. The datasheet lists 4MHz frequency limit but at 12V I suspect it will go up to 10MHz. You'll have to test the 4059 for its upper frequency limit at 12V.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Sun 26, 2019 7:11 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
If you have an Ebay account you can order all the parts from China. The original prescaler chip used in the circuit is also available on Ebay for $5.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Mon 27, 2019 3:58 pm 
Member

Joined: Jan Tue 16, 2007 7:02 am
Posts: 2982
Location: Lexington, KY USA
The 74HCXXXX versions of the old RCA 4000 series parts will run quite a bit faster, and at 5V.

Not up at your LO frequency, but still might simplify things.

http://www.ti.com/lit/ds/symlink/cd74hc4046a.pdf

http://www.ti.com/lit/ds/symlink/cd74hc4059.pdf

Ted


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Tue 28, 2019 11:33 pm 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3491
Location: Lincoln City, OR
Greetings to the Forum:

Thanks for all of the replies.

I have already stumbled on the 74HC4059 chip and have downloaded its data sheet. It seems to only be available in a surface-mount package, but that's the way the cookie crumbles. The 74HC4059 will count to 18 MHz. This means that it can divide the reference directly without prescaling. There will be a prescaler required in front of the 74HC4059 in order to count the L.O. frequency, but a divide by eight should bring the frequency into the range of 14 MHz which gives the chip adequate head room as to frequency response. I have not yet researched a prescaler chip but I am sure they are out there.

I am attaching two spread sheets in case anyone is interested. One shows the minimum divisor for all of the FM broadcast channel frequencies assuming the synthesizer is intended to generate an L.O. signal 10.7 MHz above the received frequency. The second spread sheet shows the design requirements for an actual synthesizer for the frequency of interest in my case (98.7 MHz FM Carrier; 109.4 MHz L.O.).

Attachment:
PLL Divisor Chart.xlsx [15.08 KiB]
Downloaded 7 times


Attachment:
PLL Design Chart 3.xlsx [10.83 KiB]
Downloaded 7 times


Regards,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Wed 29, 2019 2:03 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
Jim, I can design the circuit for you. It will take the LO frequency as fixed reference and divide it down to 1.5625 KHz for the CD4046. I can also get the SAB6456 chip for you. Generating 98.7MHz carrier from 109.4MHz LO or reference frequency is easy. Will post the schematic for you once I get home. If you need a fixed 98.7MHz output frequency then the varactor can be replaced with a 16V zener diode. A FET or transistor buffer will be needed to isolate the LO source from SAB6456 prescaler. This prescaler chip has great sensitivity and needs very little RF input. Last time I used it I was feeding it by a 10K resistor connected to the collector of a low power single transistor FM oscillator.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Wed 29, 2019 2:33 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
One question. What are the design needs?

The synthesizer would be required to generate both the carrier and LO (10.7MHz above carrier) with a 5MHz external reference frequency?

Or is it that the LO of 109.4MHz would be taken from a receiver and fed to the synthesizer as reference frequency?

That's my only confusion. Once it's cleared then I can proceed with the design. Believe me it's easy.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Wed 29, 2019 4:33 pm 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3491
Location: Lincoln City, OR
Greetings to Dare4444 and the Forum:

The application is for what amounts to a precision FM receiver. The device is a modulation monitor, model TFT 723 designed many years ago by TFT (Time and Frequency Technology). The purpose of the synthesizer is to provide a stable and precise local oscillator signal to down convert a signal in the FM broadcast band to 10.7 MHz which is a very standard IF frequency and is the one used by this device. The receiver is designed to use high side injection. The station for which the monitor is intended is assigned a carrier frequency of 98.7 MHz. This requires that the local oscillator run at 98.7 plus 10.7 or 109.4 MHz. The actual requirement is for this particular frequency. Making the thing frequency agile would be icing on the cake.

The modulation monitor uses a pulse counting discriminator which is a very accurate discriminator which can be calibrated from a signal derived from the 5 MHz reference in the monitor. Thus it can accurately display the station's modulation level so as to ensure FCC compliance. In addition, the IF signal is counted by a frequency counter. The counter is pre-programmed with a value and this value is decremented by the IF frequency during the count interval. The resultant nixie display shows carrier frequency ERROR in plus or minus Hz from carrier. Both the frequency counter and the pulse counting discriminator are highly dependent on the IF being exactly 10.7 MHz, which is why the L.O. is frequency synthesized and phase locked to an internal 5 MHz ovenized crystal time base.

The original circuit is a non-agile frequency synthesizer. A suitable harmonic of the 5 MHz reference is generated by a pulse shaper and amplified by a three stage grounded base amplifier. The result is mixed with the buffered output from a VCO to provide a low frequency that can be counted by a set of TTL counters. These are programmed to produce the divisor needed to lock the loop when the resulting mix is at the correct frequency. Unfortunately, the circuit uses high-Q coils in cans.... three for the reference amplifier, one for the VCO and two for buffer stages... one buffer drives the PLL mixer and the other drives the receiver mixer. These coils are selected at the factory for the frequency of interest. These coils are no longer available; TFT has gone the way of the Dodo and I am having a very devil of a time getting the circuit to work on another frequency. The coils are all very high Q and the circuit simply doesn't develop enough signal to work with home made coils. I have a circuit for a grounded base tripler that I intend to try in place of the reference selector amplifier. I hope to find some higher gain devices for the buffer stages that will let me get enough signal into the PLL mixer for the circuit to work. If you wish to examine the old synthesizer circuit or the whole device in detail, the manual can be downloaded here: http://www.steampoweredradio.com/pdf/tf ... Manual.pdf

The synthesizer I am discussing here is a fall-back position in case I give up on the existing synthesizer. If I go that route, however, I think I would like to go with the 74HC series chips. I want to make the device as simple as possible with as low a component count as possible. Comparing at a higher frequency than 1.5 KHz should make the device a little better as to accuracy and it will simplify the board layout and component procurement. As I said in the previous post, a 74HC4059 divider can count to 18 MHz while a plain vanilla CD4059 dies at 1.5 MHz. I would have to prescale not only the L.O. but also the reference if I used CD series chips. If I used 74HC4059's for both counters, I can eliminate one prescaler entirely and the other only needs to divide by eight.

I hope I don't have to actually build this thing as it's going to cost me a fortune by the time I buy the chips and get the board made. If I can get the original circuit to work on the new frequency, that would be by far the best solution. If not, I will have to start from scratch.... but keeping the component count down and the circuit simple is a big plus for me.

Thanks for your offer of assistance. If you want to invest some skull sweat in the design, it is certainly appreciated. However, I hope to get the original circuit to work. If you have any suggestions there, they would also be greatly appreciated.

Thanks,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Wed 29, 2019 8:18 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
Jim, here's the block diagram. Reference frequency will be 12.5KHz for the 4046.

https://ibb.co/Qn59bJn

Addition of extra components cannot be avoided if want the reference frequency to be higher than 1.5625KHz.

Let me know and I'll draw the schematic for the above blocl diagram.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Wed 29, 2019 11:21 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
https://rover.ebay.com/rover/0/0/0?mpre ... 2431293732

74HC4059 DIP package


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Thu 30, 2019 1:31 am 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
https://ibb.co/fCzf3h6

New block diagram. It uses four ICs total. Reference frequency is now 25KHz for fixed 109.4MHz output.

74HC4059 will clock 40MHz so I removed one 74AC74 prescaler.

Let me know if you would like a schematic.

Will design the RF section separately. If you are okay with ugly bug construction then the prototype can be made very quickly. It saves time.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Thu 30, 2019 10:22 pm 
Member

Joined: Mar Thu 01, 2018 1:30 am
Posts: 195
Jim, please let me know.


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 31, 2019 3:24 am 
Member
User avatar

Joined: Nov Mon 02, 2009 7:01 am
Posts: 3491
Location: Lincoln City, OR
Greetings to Dare4444 and the Forum:

Thank You very much for your reply. I obviously mis-named the 4046 comparator in my previous posts, but I was going from memory. Sorry about that.

Please find attached a pdf of one page from a data sheet for the 74HC4059. (The whole document is to big to attach here.) It states that the maximum count frequency (if I read it correctly) is only 18 MHz in one chart, but gives 54 MHz immediately following. I do not understand the usage or which is correct.

Attachment:
CD74HC4059 CP.pdf [410.12 KiB]
Downloaded 8 times


I would not like to crowd its maximum parameters, so that's why I elected to use a divide by eight prescaler. The spread sheet below shows the minimum pre-scaling and appropriate divisors given the 18 Mhz limitation in my data sheet. In case of conflicting data sheets, what do we do? I really don't know, but it seems to me that it would be best to err on the side of caution.

Attachment:
PLL Design Chart 3.xlsx [10.87 KiB]
Downloaded 6 times


Dead bug construction is fine; easy to prototype. The only constraint on the VCO section is that it have enough power to drive the balanced mixer. Probably, it would be best to buffer it as in the original design.

I have not had any time lately to dive back into the original circuit; I would hate to have you spend a lot of time on something that may not end up actually being used. However, if you wish to proceed for the fun of it, by all means, go ahead.

Thanks,

_________________
Jim T.
KB6GM


Top
 Profile  
 
 Post subject: Re: Wanted: Frequency Synthesizer Schematic
PostPosted: May Fri 31, 2019 5:13 am 
Member

Joined: Jan Tue 16, 2007 7:02 am
Posts: 2982
Location: Lexington, KY USA
I believe that the 74HC4059 data sheet numbers mean that at the time the data sheet was compiled, about half of the production units would in fact clock at 54MHz. At 25C case temperature and 5.0V Vdd.

However, this number is not safe to use for design, unless you plan to test each individual part. And it is possible that you might not find any parts in a given batch that go this fast.

The fMAX numbers are pretty safe to use. If your equipment will be at a normal sort of room temperature, and the nominal supply voltage is 5.0V,. You can expect 100% of the parts to run at 22MHz. The 18MHz figure will be safe to use well past boiling water sorts of temperatures.

Note that close to the maximum clock frequency you will need to have nearly a 50% duty factor clock waveform, due to the tW specs.

There is every chance that current production parts are tightly clustered near the "good" end of the data sheet tolerances, since this product is very mature and has had the benefit of decades of process tweaking. On the other hand, sometimes even the best manufacturers have a bad day at the fab...

Ted


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

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 6 guests



Search for:
Jump to:  


























Privacy Policy :: Powered by phpBB