Make a very simple test for me

[_q12x_]

Hey, thank you all for the commitment and for debugging this problem with me.
You really helped me, especially by figuring out the 200kHz limitation of the osc.
Thank you very much !

 

[_q12x_]

Following your suggestion, using a dedicated chip..... I get involved into research and I find this:

Frequency divider - Wikipedia

en.wikipedia.org

And....I have a bright idea...
After reading about this process, I understand that division by 2 is done using a D flip-flop circuit.

I understand it is a digital logic operation.
And the bright idea is to actually build a D flip-flop circuit. Or use a chip that may already have some ready flip-flop pins.
I have a small list of personal chips I can put down for you to look over, and if you spot any IC that will fit this task, then poke me. I already seen one CD4013 x1 dual D flip-flop but I have only 1 of it.
See if you find others.
Until then, I will concentrate on reading how to make a D flip-flop from discrete components. And possibly make a circuit. Mmmmm....

 

[_q12x_]

Hmmm... I find Frequency divider using 555 timer !!!

Simple frequency divider

Electronic circuits, projects and tutorials. PCB details, microcontroller projects, power supply, high frequency and audio circuits.

www.circuitlib.com

 

[marconi]

q12x : Good evening. I have been occupied today to be of any help(?) but can see that some very able folk have responded.

What is your project goal now please? You have bought a frequency counter which arrives soon. You have some quartz crystals which you want to confirm work and their resonant frequencies. Thus, it seems to me you still need a test rig to put these crystals into to make them oscillate. And you need a way of connecting the test rig to the frequency counter to measure their resonant frequencies. I am happy to help but need some clarification on what next - the test rig?

But before all that - what do you intend to do with these crystals?

 

[7029 dave]

The finest profs on the EF
Just saying hello thats about my input.
Way over my little ole head just a mere installation electrician.
I will be the first to stand on one side of the corner.

 

[pc1966]

And the bright idea is to actually build a D flip-flop circuit. Or use a chip that may already have some ready flip-flop pins.

You can build them, but it is quite a few parts for a single stage and given the expected crystal frequency of 20MHz and your scope's limit of 200kHz, to see something approaching a square wave you need to show at least the 3rd harmonic, and ideally 5th, so you would want a final frequency of around 200kHz/5 = 40kHz.

So your divider would be 20MHz / 40kHz = 500 and closest binary option is 512 = 2^9 = 9 flip-flop stages.

I have a small list of personal chips I can put down for you to look over, and if you spot any IC that will fit this task, then poke me. I already seen one CD4013 x1 dual D flip-flop but I have only 1 of it.
See if you find others.

Most of your ICs are of the CD4000 series, they are only able to clock in to a MHz or so. You need to look at the 74HC series (or 74LS) to get 20MHz or more capabilities. Generally 74HC is easy to work with, some others are not so easy (restricted voltage range, or very fast switching so issues on breadboard use, etc)

The 74HC4060 is cheap, if getting one then buy at least 2 and check it is the DIP version, not surface mount package. It is finding a place that won't charge you a stupid amount of postage that is a challenge!

 

[pc1966]

Hmmm... I find Frequency divider using 555 timer !!!

Simple frequency divider

Electronic circuits, projects and tutorials. PCB details, microcontroller projects, power supply, high frequency and audio circuits.

www.circuitlib.com

That sort of a circuit only works well for a limited frequency range. A typical digital divider works from "DC" (clock stopped) up to the clock limit of the IC (tens of MHz input for the 74HC family).

 

[pc1966]

Way over my little ole head just a mere installation electrician.

Don't put yourself down, you contribute a lot on other matters!

 

[_q12x_]

q12x : Good evening. I have been occupied today to be of any help(?) but can see that some very able folk have responded.

What is your project goal now please? You have bought a frequency counter which arrives soon. You have some quartz crystals which you want to confirm work and their resonant frequencies. Thus, it seems to me you still need a test rig to put these crystals into to make them oscillate. And you need a way of connecting the test rig to the frequency counter to measure their resonant frequencies. I am happy to help but need some clarification on what next - the test rig?

But before all that - what do you intend to do with these crystals?

Hello my friend and salutations from Romania. I believe you speak in name of everyone and some clarification might be in order from my part. Indeed my frequency counter is on the road. That is plan B that is going to happen. But until it arrives, in 2-3 months, I think I can PUSH this thing a little bit. You know? That's my goal here. And by pushing, I mean to make something that works, even if I have to make it from different parts or even circuits. For example, a circuit made from discrete components that will make a single D Flip-Flop stage in the 10 stages that I calculated.
Here is my plan:

20000000 = 20M
2000000 = 2M
200000 = 200k
20000 = 20k =/1000

2^1 = 2
2^2 = 4
2^3 = 8
2^4 = 16
2^5 = 32
2^6 = 64
2^7 = 128
2^8 = 256
2^9 = 512
2^10 = 1024 =10 flipflops

20k or20000 means 20000000/1000 and 1000means 1024 in binary so 10 flipflops
Why 20k? because I think I have larger than 20M, I think Ive seen one at 50M.
Somewhat being in the safe side, and also easy to calculate if you roughly multiply by simply 1000 in your head.

I really don't think the "test rig" as you put it, is the biggest problem anymore. I think the "splitting" as I call it, is the bigger issue. Because as we discovered, my DSO138 is limited to 200kHz so we need to bring everything down to it's level of measuring. That's it. Simple.
And for this, I'll have to use what I have on my hands. Some chips if im lucky or the obvious way, build a couple of discrete D- flip-flops. I'll make a factory of production these days here in my room. Haha. You probably remember how I build my circuits, using cardboard and pieces of bended sheet metal that makes pads for soldering. Very shi..ty ones but good enough for the amount of experimentation I am doing.
What you are wondering so hard now, is why do I want so bad to PUSH it, instead of waiting for my new fv counter? Well, because why not.
If you are good enough to get involved into my small project here, because this will be hard to make anyway. Is to find for me a very good D flip-flop discrete circuit that I can build using my components I have in storage. Ideally is to build it from 555's because I have 100 both DIP8 and smd SOT-23. I am researching myself this possibility at the moment.
I hope is clear enough. If more clarifications needed, just ask me.
- Saying hello back to mister @7029 dave

....
Most of your ICs are of the CD4000 series, they are only able to clock in to a MHz or so. You need to look at the 74HC series (or 74LS) to get 20MHz or more capabilities. Generally 74HC is easy to work with, some others are not so easy (restricted voltage range, or very fast switching so issues on breadboard use, etc)

The 74HC4060 is cheap, if getting one then buy at least 2 and check it is the DIP version, not surface mount package. It is finding a place that won't charge you a stupid amount of postage that is a challenge!

Yes... I am new to this chips you are speaking here. I will start looking for them and probably even buy, if they are not expensive for 100pcs as I usually buy all my components. It is good that you tell me about them. Thank you.
But remember, I will start with whatever I have in hand at the moment.
I believe you are trying to tell me it will not work using discrete D flip-flops or the 555 version of it, because they will not be fast enough for the 20M (or more) speed. If this is the case, please be clear with me now, before I start doing anything ((stupid)ly hard and slow). Its a ton of work but in a few days I think I can do it.
Have a good night (to everyone) and see you tomorrow.
Thanks so far !

 

[marconi]

q12: Good morning. Thank you for your nice email. Gathering dust in my shed I have some 74LS273 octal D type flip flop ics.

https://www.ti.com/lit/ds/symlink/sn54ls273-sp.pdf?ts=1652639946774&ref_url=https%3A%2F%2Fwww.google.com%2F

I will post you some as a gift. Please private message me your address or somewhere to post them to from where you can collect them.

Each chip will divide by 2exp8.

https://www.electronics-tutorials.ws/counter/count_1.html

I will include some 74LS14 Schmitt triggered inverters chips so that you can input the 20 MHz sine wave and output from it a square wave to clock the 74LS273.

https://www.ti.com/lit/ds/symlink/sn74ls14.pdf?ts=1599590379543

You could use a circuit like this using two Schmitt triggered inverter gates as your test rig for the crystals which would in turn output square pulse to drive the frequency divider:

https://electronics.stackexchange.c...nverters-in-quartz-crystal-oscillator-circuit

I see that a later circuit uses 74HCU04 inverters. I will see if I have any. I know I have some 74LS04 ones.

https://www.infinite-electronic.kr/datasheet/54-74HCU04D.pdf

http://web.mit.edu/6.111/www/f2017/handouts/labs/74LS04.pdf



Regards

Marconi

 

[pc1966]

q12: Good morning. Thank you for your nice email. Gathering dust in my shed I have some 74LS273 octal D type flip flop ics.

https://www.ti.com/lit/ds/symlink/sn54ls273-sp.pdf?ts=1652639946774&ref_url=https%3A%2F%2Fwww.google.com%2F

Those are usually intended as bus registers and not so easy to turn in to dividers, though they are fast enough. However, you can covert them in to ring counters with the addition of an inverter:

Ring counters (Johnson Ring Counter)

Ring counters and John counters are special type of counters made of shift registers are being explained here in this article.

www.electronicshub.org

Those examples rely on a not-Q output that is inverted, but a 74LS04 or 74LS14 stage would allow the shift-register-feedback loop to be implemented.

I will post you some as a gift. Please private message me your address or somewhere to post them to from where you can collect them.

That is a kind offer of help

Each chip will divide by 2exp8.

https://www.electronics-tutorials.ws/counter/count_1.html

I will include some 74LS14 Schmitt triggered inverters chips so that you can input the 20 MHz sine wave and output from it a square wave to clock the 74LS273.

https://www.ti.com/lit/ds/symlink/sn74ls14.pdf?ts=1599590379543

You could use a circuit like this using two Schmitt triggered inverter gates as your test rig for the crystals which would in turn output square pulse to drive the frequency divider:

https://electronics.stackexchange.c...nverters-in-quartz-crystal-oscillator-circuit

I see that a later circuit uses 74HCU04 inverters. I will see if I have any. I know I have some 74LS04 ones.

https://www.infinite-electronic.kr/datasheet/54-74HCU04D.pdf

http://web.mit.edu/6.111/www/f2017/handouts/labs/74LS04.pdf

The LS versions are OK with high frequency crystals with low impedance but are far harder to make work reliably than the CMOS types. You generally can't use the single-inverter mode for parallel resonance as you need quite a low feedback resistor to pull the LS style of device in to its linear amplification region, so typically they are best used in series resonance using two inverters.

See page 164 of the linked book for circuit 11.3(a) as an example of forcing TTL in to dirty analogue activities.

Generally the LS14 and similar Schmitt triggers don't make good crystal oscillators as they never go "linear" but are great for cleaning up circuit outputs to get good square waves to reliably clock flip-flops.

 

[pc1966]

If all you have ar LS14 then using a single transistor (or the original two-transistor example) as oscillator and the inverter for clean-up is worth a go.

 

[pc1966]

I believe you are trying to tell me it will not work using discrete D flip-flops or the 555 version of it, because they will not be fast enough for the 20M (or more) speed. If this is the case, please be clear with me now, before I start doing anything ((stupid)ly hard and slow).

Typically the data sheets only show the 555 going up to 100kHz operation. They were typically seen doing audio and below speeds, with newer variants having additional dividers inside to allow very long time constants without crazy sized capacitors.

 

[marconi]

-q12- Good spot by pc1996. I may have a pile of jk flip flops and 8 bit shift registers instead of 273s but we can make 273s do the division. I have ordered some 20MHz crystals to work in parallel with you.

 

[_q12x_]

To mister @marconi
Thank you for your very nice and unexpected gift. I really-really appreciate it and is more than welcomed. I really like that you jumped into my crazy project. Its not an easy project and is very new to me with new challenges. That's why I am PUSHING it before the fv counter arrives. Its more like an experiment or an exercise. Hopefully everything will fit before the fv counter arrives and confirm the findings from this experiment, if successful. We'll see.
I will still PUSH with all I have, until your package arrives, and when is here, I will use your chips to simplify the entire experiment. At least this is the plan.
If your components (the 20Mhz OSX) will come faster than me doing my thing here, please do the entire Splitting circuit. Its a bit of a race with time what Im doing here. No matter from what corner of the world the results will come, the important thing is to get the results. So if you are faster than me in getting this thing working, please do it, because I am equally curious. If you can do it today, do it today, it will be awesome.
Thank you again !

 

[_q12x_]

Im looking on the 74HC4060 datasheet right now.
And from the look of it, it appears it has all these D Flip-flops linked , so pretty much for each advancing output pin, Ill get a greater signal splitting (in my case lower and lower frequency). It is very customizable. Is my observation correct? If it is, then this is a very nice chip !

And... I took a good deal of 3.77$ for 10pcs for SN74HC4060N.
Is always better to have more than less.
So mister @marconi if you happen to read this in time, you can option to not send me your D flip-flops chips. But if you do, is as much very welcomed as before.

 

[marconi]

I do not have any 74HC4060 to send you. I have a selection of other logic ics though which you may find useful. I have kept them all 74 series.

 

[_q12x_]

I do not have any 74HC4060 to send you. I have a selection of other logic ics though which you may find useful. I have kept them all 74 series.

No. What I'm saying is that I already bought 10pcs of 74HC4060 because like mister PC1966 said, they run at a bigger frequency and probably are newer IC's.
So if you want, you can or you can not - send me your IC's that are containing D-flip Flop inside them. You have 74LS273 octal D type flip flop. So now you have the option to send or not to send them. I will like them as well, so for me is still a good batch of IC's even if I will get these other ones. Like I said to you in particular, I have a grand list of IC's waiting for a long time. So every chip have its own characteristics that is always good to have a bit from everything. So im not saying no.

 

[_q12x_]

Could you give me some idea of what you will constructing next?

Yes, I have a line of engagement
1- In the immediate days, until all these packages arrive, I am very concentrated on how to make a D flip-flop from discrete components. And I would like very much you to help me in this little endeavor.
2- Then, I will try the 555 circuits for splitting the fv. See how those will do. Im planning something very simple to test, a very basic oscillation in a couple of Hz, like 10Hz or something, by rotating a wheel in front of a LDR or something fast to make. Or even another oscilator using the 555 again. I dont know. Then, after reading and being sure the real fv of the test device, I will try to read it through the 555 splitter circuit (splitter is faster to write than D flip-flop).
3-Next, if I get good results, I will try to make a couple of 'splitting' modules. Either from discrete or from 555's. All, hopefully until everything arrives.
4- after the IC's arrive, I will make some splitters with these IC's. I presume I will not be even close with my discrete components way, to finish anything important. But who knows, if I work hard enough.... we'll see.
5- in the very end when the fv counter will finally arrive, is check mate. I can test every OSX, put a label on them and mission accomplished. And the project is done.
Again this is more as an experiment or a test. Nothing really to remain in time. Well, only if I will make a video about it, if I make some important progress with all of this.

 

[marconi]

Here is my circuit to test the 20MHz crystals I bought. You can see the waveform and its frequency on my scope. I will draw out the single transistor circuit which uses an npn BC549.

The lower trace shows frequency in 25MHz horizontal divisions. You can see there is a second peak of lower amplitude at 40MHz to the right of the main peak which means the circuit is oscillating at two frequencies. The scope shows the waveform of these oscillations when they are added together.