Fake RTC 32768Hz (32kHz) crystal oscillators ? Or special setup for them to work?

Thread Starter

q12x

Joined Sep 25, 2015
1,601
Do Tell ME if what I did there is correct and if what you see is normal or I should do something special to them to work?
I want to show this movie to aliexpress to give me my money back ! But before that, I want to be sure my tests I made are the correct one and if there is some special trick involved for them to work? Or to be read by the fv counter? Special setup? Thank you.
This is a <6min video, very short.
 

Ramussons

Joined May 3, 2013
1,390
Maybe that oscillator will not work at KHz range, only MHz range.
You displayed the working with MHz Xtals that you have. Try it with KHz Xtals if you have any.
If not, fabricate a simple oscillator using NAND gates and test those 32 KHz Xtals before reporting that those are faulty.
Can you post the schematic of your oscillator?
 

michael8

Joined Jan 11, 2015
405
A 32KHz crystal is very different from a 5 Mhz crystal. The impedance of the 32 KHz crystal is closer to 40K ohms compared
to the 5 MHz crystal which is like around 40 ohms. The 74XX TTL logic has a very low input impedance so it's unlikely to work
with the 32Khz crystals.

Here's a simple 32 KHz oscillator circuit:

osc1.png

The 2n5089 isn't just any NPN. I don't know if it is required but it's a low level (& low noise) so it has high beta at low currents (uAs).
 

Thread Starter

q12x

Joined Sep 25, 2015
1,601
Maybe that oscillator will not work at KHz range, only MHz range.
You displayed the working with MHz Xtals that you have. Try it with KHz Xtals if you have any.
If not, fabricate a simple oscillator using NAND gates and test those 32 KHz Xtals before reporting that those are faulty.
Can you post the schematic of your oscillator?
From what I know, it doesnt have a setting for different ranges.
I thought I had a low fv crystal but I dont. All that I have is over 4MHz crystals, randomly scrapped.
Although I have some "unknown" ones that I dont really think they are totally busted, maybe its like with these new ones, not being able to read them? Hmmmm
I used these ccts from this page: http://www.z80.info/uexosc.htm
and also my own, adapted from the link:
These are my saved files on the matter. What I used in the past and now on the table (1 & 3).
1Crystal Oscillator with 7400 in series 1.jpg2Crystal Oscillator with 7400 in series 2.jpg3crystal Oscillator with 7404 in parallel.jpg4Crystal Oscillator with 7404 in series.jpg
 
Last edited:

Thread Starter

q12x

Joined Sep 25, 2015
1,601
to @michael8
I used your cct in the sim and this is the result, only when I reset the cct, meaning only when I put power to the cct, I get that waveform on the opamp output ! The rest of the time, is staying on 0V. -It is not oscilating all the time, but remaining on 0V all the time. How fast is that waveform oscilating...that I can not know. But is an interesting waveform to see it on the osciloscope I imagine.... Interesting circuit !
1700212989030.png
 
Last edited:

Thread Starter

q12x

Joined Sep 25, 2015
1,601
to @michael8
I made your cct and it is not working in reality. Although it is working fine in the simulator. I used bc548 for the tr and a LM358 for the opamp. Classic components.
I did remembered to take out and put back in the power, and on the osciloscope showed me a quick square wave from 0V while was unpowered to 5V when it was powered. The probe was on the output of the opamp ofcourse.
Im thinking maybe the pulse is too quick for my dinki osciloscope to register? it might be... but in conclusion, sadly, it didnt work. Hmmm.
 

michael8

Joined Jan 11, 2015
405
Don't use the circuit at the URL (analog devices), it's has an error in it in how the base is connected (the 1M resister in the base lead).

Also your version doesn't have the the transister biased into the active region (pin 6 on the analog devices version isn't ground).

Use the circuit I pasted into this thread instead.
 
Last edited:

Thread Starter

q12x

Joined Sep 25, 2015
1,601
Now, I know for certain I have a bad batch of xtals !!!!
How I know for sure? - I made a simple xtal oscilator that I build in my past, with 7400 NAND gates and it worked fine back then. I had to look back in my notes. As you can see, it is using 20MHz xtal. This particular project was made to test xtals, and the circuit involved some FF's to divide the fv down, and then I could read the fv with a LED output I think. The project worked fine on the breadboard, but in reality, the final product, I had some technical dificulties, with some connectors and now is working randomly because some wiring and very small connectors I know I had problems. Theoretically the cct is fine, but practically... meh.
So... I look back and see that I used back then this exact cct right here:
Crystal Oscillator with 7400 in series 2.jpg
And I re-build it on my breadboard and the idea was like this:
- If the counter will spin very fast, showing 888, it meas the fv oscilator circuit works, also in the same time, the xtal is good. So I inserted a 9MHz xtal, that was readable by my DL4YHF fv reading device in the movie, so I know this xtal is good. And, as expected, the counter spin very fast, showing 888. Good, so the cct is good and also is confirming the 9MHz xtal is good. When I switch with the 32kHz one, nothing, all was 000. This told me exactly, and I trust it now it is a reliable reading, that these motherfukers xtals are completly dead. There is no special reading for them !!! When I have 2 or more diferent testing sources that are showing the same result, then Im certain that something is good or bad. I am never satisfied from only 1 source, unless it is very trusty, which was not the case for my DL4YHF model_1.jpg
Another unreliable device is my dicky osciloscope DSO138model_1-DSO138.jpg
He literally shit itself. He was unable to help me in reading any xtal I have !!! It is useless for xtals ! This is what I certainly know now.
Now... I have to order another batch of 32kHz xtals. What an experience !
 

sparky 1

Joined Nov 3, 2018
754
It is an ultra low power crystal which requires an ultra low drive level.
Here is a datasheet that shows the drive level of a typical 32.7 kHz tuning fork crystal.
Looking at some of the circuits above I see they use a 10MΩ resistor,
It is a common mistake to bypass the drive level and how would average experimenter understand or measure that?

Eng 4v«¿í° TF.xlsm (citizen.co.jp)
 

michael8

Joined Jan 11, 2015
405
Now, I know for certain I have a bad batch of xtals !!!!
How I know for sure? - I made a simple xtal oscilator that I build in my past, with 7400 NAND gates and it worked fine back then. I had to look back in my notes. As you can see, it is using 20MHz xtal.

I have an ohm meter with a 1K ohm range, it mesures my 47 ohm resistor just fine but this 47K resistor it says is open,
must be bad...

The series resistance of the 20 Mhz (or 5 Mhz) crystals are in the 20 to 50 ohm range, the 32KHz crystals are in the
40K ohm range. It takes a different circuit to make them oscillate.
 
Last edited:

Thread Starter

q12x

Joined Sep 25, 2015
1,601
Tu cut the s*it, does anyone here have this dicky fv counter (or something very close to this model) I have?
Original name: DL4YHF Frequency Counter
but you will find it as
1Hz-50MHz Crystal Oscillator Tester Frequency Counter DIY Kits Meter with Case
at any online market.
So, if you have this model (or similar), and if you have a 32kHz crystal or any tens of kHz is fine, please make a measurement with these 2 and make a picture and post it here. My bet is this fv counter must read those tens of kHz from the crystal.
Why Im sure this fv tester is good both at high and low fv? because I read fv as low as 10Hz with it and as high as 40MHz(maximum for it) - at least my model.
model_2.jpg
 

DickCappels

Joined Aug 21, 2008
10,117
I found one microcontroller/crystal combination in which I could not get oscillation until I removed the load capacitors, seems the capacitors I though would work, about the same value as your were keeping the circuit from oscillating. You might want to try it with the load capacitors removed and just rely on the capacitances of the rest of the circuit.
 

Thread Starter

q12x

Joined Sep 25, 2015
1,601
I have a working 32kHz crystal oscillator buried somewhere. I will dig it up for you.
Thank you, and also a similar or the same fv counter as mine.
Although, it might not be necessary to disturb yourself too much because I made another video, testing with a lower fv, this fv counter device I have, and if these 2 extremes work, then the 32kHz crystal should have been read as well with no problem , if it was good!
 

Thread Starter

q12x

Joined Sep 25, 2015
1,601
I found one microcontroller/crystal combination in which I could not get oscillation until I removed the load capacitors, seems the capacitors I though would work, about the same value as your were keeping the circuit from oscillating. You might want to try it with the load capacitors removed and just rely on the capacitances of the rest of the circuit.
When Im testing, I test the sh*t out of it, and indeed I took out those capacitors and in my case it did no diference , either with them or without them. Probably they are needed for more stability in a pretentious circuit. But for my rough tests, they didnt matter. At least this is my experience. Thanks for trying.
Im editing now, another video with the definitive testings and my conclusions on these busted xtals !
I'll post it very soon.
 
Top