Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Driving a 4046 PLL input with Logic level , Transistor amp info?

Status
Not open for further replies.

EI3HG

New Member
HI All,

I am building a PLL circuit using a 4046 and and also a DDS chip, as part of a hybrid PLL/DDS synthesiser.
It all worked fine, then I spotted that I had accidentally used a 74HC4046 device, and the supply rail was 8 volts! No problem , I swapped it for a proper CMOS 4046. I t stopped working, so I tried another, and another ( of different types, Motorola, Phillips etc) Still no joy.

Then I put back in the 74HC4046, and it worked again, so I recconected the supply rail to to a 5 volt one. It stopped working!

The reference to pin 3 of the 4046 is from a 10 MHz crystal locked CMOS oscillator (running at 5 volts), and divided down from 10 MHz to 1.25 kHZ via a divider.
The DDS output is fed to the signal input of the 4046 PLL, is around 100mV RMS.
Some sources I found say that the 4046 will lock up with inputs down to 20mV, others, like the Phillips Datasheet say 200mV RMS.

The 4046 has a self biasing linear amp on the input (pin 14) for small signals, or logic level can be used, in which case the capacitative coupling should be omitted (according to datasheet research on the internet).

Further research uncovered that Don Lancaster in his 'CMOS Cookbook', (and Forrest Mimms in his Circuit scrapbook' re-quotes him) saying that "The linear amplifier operation of pin 14 is an unmitigated disaster when the wideband (type II) phase detector is being driven" and suggests not using it.

So then I need to create logic level signals from my 100mV sine wave from the DDS.

What I need then is either a class A transistor amp or a class c or D biased version to give logic levels.

My electronics knowledge (self taught) unfortunately has gaps in it.!I learnt using Valves (tubes) to make transmitters in the 1970's when RF and transistors didnt really mix that well, and progressed from there to audio op amps with some TTL and CMOS logic thrown in.

Shamefully, I sort of missed out on BJT transistors, and although I have tried to learn in the past, all the books I look in are either too basic or way too theoretical. FET's I get (they are voltage based like thermionic devices).
With valves (tubes) the databook used to give you suggested bias voltages and grid current for the varios classes of operating, (A, AB, and C). I dont see this sort of info or a web calculator anywhere.

So I need to know how to design a class A preamp, and / or how to design a transistor that can do class D. Do I need more than one device or a Darlington to get a logic level square wave?

Has anyone any good recommendations for reading material on here, either book based or internet?
Also anyone with experience of 4046 PLL?

Thanks,

Andy
 
Last edited:
HI All,

I am building a PLL circuit using a 4046 and and also a DDS chip, as part of a hybrid PLL/DDS synthesiser.
It all worked fine, then I spotted that I had accidentally used a 74HC4046 device, and the supply rail was 8 volts! No problem , I swapped it for a proper CMOS 4046. I t stopped working, so I tried another, and another ( of different types, Motorola, Phillips etc) Still no joy.

Then I put back in the 74HC4046, and it worked again, so I recconected the supply rail to to a 5 volt one. It stopped working!

Just use the 74HC4046 - it's NOT a TTL chip, it's a CMOS one (hence the C in it's name - there aren't any TTL 4046's).
 
Hi Nigel,
Thanks for the reply.

I know it's not a proper TTL chip, but it's maximum operating voltage is 7V, and if I run it at the recommended 5 or 6 volts the circuit stops working.
Running outside the specs is, I think, asking for problems, such as reliability or instability, it might stop altogether if the ambient temperature changes by much.
 
Without an actual schematic it's hard to tell.
If you have a scope, watch the input on pin 3. Use a couple high impedance resistors or a diode to bring the signal on pin 3 to an acceptable voltage window for the comparator of the 4046.
Maybe post a schematic-A picture is worth a thousand words
 
You can generate a digital signal from a 100mV sinewave with a comparator such as the LM339 or LM393. Be sure and add some hysteresis so it doesn't oscillate.

Note that most comparators, such as the LM339, have an open collector output and require a load resistor to the plus supply to generate the output signal.
 
Here's the schematic
Photo 27-01-2013 19 42 04 (Medium).jpg
 
From the AD9851 datasheet:

"The AD9851 contains an internal high speed comparator that can be configured to accept the (externally) filtered output of the DAC to generate a low jitter output pulse." "An on-board high speed comparator is provided to translate the analog sine wave into a low-jitter TTL/CMOS-compatible output square wave."

Figure 2 of the datasheet shows how to connect it.
 
I've used the 4046 a few times, there are big diferences bewteen manufacturers, the first time I used one it wouldnt work over a few m/c's, but different manufacturers produce different ic's.
4000 series cmos are usuall fairly compatible, but this chip does vary.
What ferq do you want to go to?
 
Could you be running into speed limitations?
The 74HC4046 is way faster than a CD4046
 
I noticed pin 15 is different between the 4046 and the 74hc4046.

I can't figure the schematic out. The block marked VCO is outside the loop. The steering voltage off the loop filter is capacitive coupled to the 4046 internal VCO control.
 
10 MHz is pushing the spec limits of the HC/HCT4046 phase comparator propagation delays, and the CD4046 is out of the question. It may work at 10 MHz, but it's not a reliable design. For the HC/HCT parts, at room temp, the signal input should not exceed 7 MHz and over the -55C to +125C temp range, the signal input should not exceed 5 MHz -- and that's assuming a square wave input. The comparator propagation delays are lower with higher supply voltage, which can explain the failure to work at 5V.
 
Last edited:
Probably out of spec but I've had the 'hc version operating in the hf band, around 30 megs.
 
From the AD9851 datasheet:

"The AD9851 contains an internal high speed comparator that can be configured to accept the (externally) filtered output of the DAC to generate a low jitter output pulse." "An on-board high speed comparator is provided to translate the analog sine wave into a low-jitter TTL/CMOS-compatible output square wave."

Figure 2 of the datasheet shows how to connect it.

That sounds promising! I didnt spot that!
Will give it a go tonight!
 
Dr Pepper: - 1.25 MHz

Fernando g: - possibly

Beerbelly, - yes sorry, its not really a great schematic, I did it in a bit of a hurry, only drawing what I thought were the relevant parts.
The VCO output feeds a sample to the DDS and this then divides it down to 1.25 MHz, so it is in reality a closed loop
 
10 MHz is pushing the spec limits of the HC/HCT4046 phase comparator propagation delays, and the CD4046 is out of the question. It may work at 10 MHz, but it's not a reliable design. For the HC/HCT parts, at room temp, the signal input should not exceed 7 MHz and over the -55C to +125C temp range, the signal input should not exceed 5 MHz -- and that's assuming a square wave input. The comparator propagation delays are lower with higher supply voltage, which can explain the failure to work at 5V.

Sorry if my schematic isnt clear - I have a 10 MHz XCO that is divided down to 1.25 MHz, this is used as the reference for the 4046 PLL
 
EI3HG, it seems to me you should easily be able to get one volt p-p sinusoid out of the DDS (and the output filter), so that is another avenue to look at. Would need to see more details of how the Rset pin on the DDS chip is handled and the detailed Output filter and both DAC output pin connections. The DAC is a current output DAC, so if there is no/incorrect load impedance on the DAC output (and I don't see one in your schematic) to give a path for the current to flow, there will be low/no voltage output.

Realize your schematic does not show a complete PLL as the VCO output is not being fed back for comparison. I am only concentrating on the issue with the input signal at hand.
 
Last edited:
CCurtis - I eventually managed to try the on-board comparator as per the datasheet as you suggested (after I managed to break the legs off a couple of examples of the AD9851! not good at $20 a pop!)

It worked perfectly though, after some slight adjustment to resistor values to alter the mark/space or duty cycle of the square wave. It still didn't reliably operate the 4046 though, as I am running the AD9851 DDS at 3.3v. This is because the controller I am using (an Atmel xMega256) runs at 3.3v.

I tried a simple grounded emitter transistor to step up the voltage from 3.3v to 5v, and although the square wave wasn't all that square anymore, the 4046 locked up perfectly.

I could use level converters between the Atmel and run the DDS at 5v, and benefit from slightly improved frequency response, or fiddle with the single transistor level converter. Although the latter sounds simpler, from looking araound the internet, it appears that I could lose phase stability and/or add phase noise using this approach if not done carefully. Your thoughts appreciated though.

I am using part of a Midnight solutions DDS 60 board, see:

https://midnightdesignsolutions.com/dds60/index.html#Schematic

without the on-board clock and output amplifier, and on-board 5v reg. Really, I am using it just a carrier board for the AD9851 and the output filter components. I am running the filter at 51 ohms in/out impedance.

The RSET pin is just grounded. I am getting about 200mV p-p sine from the filter output, which seems to tally with the datasheet and the DDS-60 notes.

I have ordered some HCT versions of the 4046 and also the 7046 (the 4046 with a lock indicator) to play with, as well as a couple more AD9851's and some stronger magnifying glasses!
 
Hi. I would wait for your HCT versions of the 4046 since you ordered them anyway. I can help with a fast level shifter, but ya, why bother with the added complication.
 
Hi CC
They arrived yesterday from Digi-Key (amazing service USA to Ireland is always only 48 hours with them) but my old HP 54601 scope decide tonight was the night to give up triggering waveforms! I will have to get my old analog scope from work, until I can either fix the HP or source something else on eb
Plan is to try the circuit with the HCT and see if it is stable with different supply voltages. I will post back when I have results
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top