Crystals and Accuracy

[audioguru]

It is best to test your clock circuit after it has warmed-up. Its frequency will be stable and that's the way you will use it anyway. :lol:

 

[Screech]

It is best to test your clock circuit after it has warmed-up. Its frequency will be stable and that's the way you will use it anyway. :lol:

You mean warmed up as in temperature or
run it under power for a period of time for it to stabalize?

 

[audioguru]

Temperature.

 

[Screech]

Cool, I've been doing that and documenting the temperature of the room.

 

[Screech]

It's very interesting when you put 4 quartz clocks infront of you and you watch the second hands/display for a few hours

One of them, a no-named digital was a second slow only after 40 minutes.
Another, a seiko quartz analoge, was a second slow after a few hours.
compared to my pc clock.

Don't know the battery state of the clocks.
Interesting.

 

[Nigel Goodwin]

It's very interesting when you put 4 quartz clocks infront of you and you watch the second hands/display for a few hours

One of them, a no-named digital was a second slow only after 40 minutes.
Another, a seiko quartz analoge, was a second slow after a few hours.
compared to my pc clock.

Don't know the battery state of the clocks.
Interesting.

You should be aware that PC clocks are really poor time keepers (often out by minutes per day!), so you need to ensure it sets itself off the Internet before you take readings!.

If you have Teletext in your country?, try comparing it to the Teletext time (ensure you use the same channel every time though).

 

[Screech]

Nigel says

If you have Teletext in your country?, try comparing it to the Teletext time (ensure you use the same channel every time though).

I never would have thought of that.
Pitty, I don't have a teletex tv.

I agree with what you said about updateing of the internet before taking time readings (which I have been doing).
My pc's clock is within 1 second/day compared to internet time servers.

 

[audioguru]

If you have Teletext in your country?, try comparing it to the Teletext time

You can't trust them!
When I had a VCR it set its clock automatically to "teletext?" from a TV station. Then it screwed-up. I e-mailed them and they thanked me for pointing out their problem and it took about 1 month for it to be fixed until it screwed-up again a few months later.
On my new digital recorder I just highlight the program on the cable TV's guide channel and push the Record button on my remote. It even knows if my program is delayed or runs into overtime. :lol:

 

[Nigel Goodwin]

If you have Teletext in your country?, try comparing it to the Teletext time

You can't trust them!

You can in the UK, where it was invented! :lol:

When I had a VCR it set its clock automatically to "teletext?" from a TV station. Then it screwed-up. I e-mailed them and they thanked me for pointing out their problem and it took about 1 month for it to be fixed until it screwed-up again a few months later.

It doesn't sound like your TV stations are very reliable?.

On my new digital recorder I just highlight the program on the cable TV's guide channel and push the Record button on my remote. It even knows if my program is delayed or runs into overtime. :lol:

Is that using a kind of PDC? (Program Delivery Control), which is a Teletext based system - which has flopped pretty well in the UK.

Or is it a digital TV system?, which again must send some kind of start and stop signals to accurately record a program - and obviously relies on these signal being inserted!.

 

[audioguru]

It doesn't sound like your TV stations are very reliable?.

The station is an old non-profit American one. Canadian stations are newer and more reliable but don't broadcast the time signal.

Or is it a digital TV system?, which again must send some kind of start and stop signals to accurately record a program - and obviously relies on these signal being inserted!.

Yeah, it's digital cable-TV. They must have a whole army of many people keying in the stuff. :lol:

 

[JimB]

The crystal is tuned by altering the DIFFERENCE on the two sides of the crystal, so by adding to BOTH sides you're not really changing anything 8)

Not so, it is the capacitance in parallel with the crystal which changes the frequency.
If you have 27pF at each side of the crystal, you will have 13.5pF in parallel with the crystal. (The capacitors are in series and centre tapped to 0v). There will also be the stray circuit capacitance.

27 pF at each side of the crystal is not the same as 39pF at each side. In the 39pF case, the frequency will be lower.

JimB

 

[mechie]

If you have Teletext in your country?, try comparing it to the Teletext time

You can't trust them!

You can in the UK, where it was invented! :lol:

But we can't trust radio RDS time data!
One independent national radio station I listen to (UK 100.5FM) used to transmit time data about 20 minutes out - this kept resetting my hi-fi RDS reciever clock.
I emailed them to ask why this could be; they never answered but their RDS data now no longer carries the time data :!: they stopped transmitting it.

There is an "old Chinese proverb", it goes something like ...
A man with two watches never knows the time :?

 

[Screech]

Nigel Goodwin said

The crystal is tuned by altering the DIFFERENCE on the two sides of the crystal, so by adding to BOTH sides you're not really changing anything 8)

Then JimB said

Not so, it is the capacitance in parallel with the crystal which changes the frequency.
If you have 27pF at each side of the crystal, you will have 13.5pF in parallel with the crystal. (The capacitors are in series and centre tapped to 0v). There will also be the stray circuit capacitance.

27 pF at each side of the crystal is not the same as 39pF at each side. In the 39pF case, the frequency will be lower.

My tests confirm what you say JimB, it did run slower with bigger caps on each side of the crystal. see my graphic below.
What's our teacher gonna say?

I'm wondering , maybe the caps are not not identical (5% difference on tolerence)?

Mr. Nigel?

 

[JimB]

My tests confirm what you say JimB, it did run slower with bigger caps on each side of the crystal. see my graphic below.
What's our teacher gonna say?

I'm wondering , maybe the caps are not not identical (5% difference on tolerence)?

Never mind what the teacher says, he may learn something too.

The capacitors dont have to be identical, (say +/- 50 percent of each other, as a ballpark figure) they just have to give the correct capacitance to put the crystal on to the correct frequency.

JimB

 

[Nigel Goodwin]

Never mind what the teacher says, he may learn something too.

Always willing to learn 8)

 

[oliverb]

I have just finished making a Quartz crystal clock pulse generator for one of my clocks using a 32k Hz crystal.

It has just run for the last 3 days and has not lost or gained even a fraction of a second.

I set it up using a my old Maplin "Atomic" clock I made years back. This clock resyncs itself with an Atomic clock via radio every 60 seconds so is always correct. You can can get "Atomic" or Radio clocks for a few pounds now from Maplin/Argos etc. You can of course use "TIM" on your telephone.

All I do is sync the new clock with my ref clock then wait a day and recheck. I note how many seconds the clock is out then adjust the trimmer. Wait another day check and retrim again. Once you have got the clock running OK for a day just increase the time to 2 days or even a week. Don't forget by this time the adjustments on the trimmer will be very small indeed so I don't think with a 32K crystal this would be possible. The clock will also tend to drift summer and winter without temp. compensation so in the end you have to settle with a happy medium.

The setup takes a few days but in my opinion is the best.

 

[audioguru]

I like the key that is used to wind it. :lol: