• 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.

Big dissapointment, Need your help.

Status
Not open for further replies.

alphacat

New Member
We've received our pilot boards from the manufacturer factory and we have a major problem.

We have on board two crystal that are connected to the MCU - 32MHz and 32.768KHz.
Both of the crystal show unsteady frequency at their outputs.

I'd like to ask you please 2 questions:

1. If you had experience with these phenomenon, Could you please write here the reasons that it occured and how did you solve it?
What have you experienced that caused the crystal to output unsteady frequency?

2. How can i know what is the voltage that should be input to the Crystals?
It isnt mentioned in datasheet:
http://www.abracon.com/Resonators/abs25.pdf (32.768KHz)
http://www.citizencrystal.com/images/pdf/m-cs325.pdf (32MHz)
Should both pins of each Crystal receive same input voltage?

Thank you very much for any help.
 

blueroomelectronics

Well-Known Member
Make sure the correct caps are on the crystals (if any, it depends on the oscillator you're using)

Don't apply DC to a crystal, they don't like it.

PS post your schematic next time, you do like to keep us guessing.
 
Last edited:

felis

New Member
We've received our pilot boards from the manufacturer factory and we have a major problem.

We have on board two crystal that are connected to the MCU - 32MHz and 32.768KHz.
Both of the crystal show unsteady frequency at their outputs.

I'd like to ask you please 2 questions:

1. If you had experience with these phenomenon, Could you please write here the reasons that it occured and how did you solve it?
What have you experienced that caused the crystal to output unsteady frequency?

2. How can i know what is the voltage that should be input to the Crystals?
It isnt mentioned in datasheet:
http://www.abracon.com/Resonators/abs25.pdf (32.768KHz)
http://www.citizencrystal.com/images/pdf/m-cs325.pdf (32MHz)
Should both pins of each Crystal receive same input voltage?

Thank you very much for any help.
I suggest hiring an engineer.
 

alphacat

New Member
Thank you for your advice.
As you could guess, I cant post here the exact schematic becuase my company wouldnt allow me to do so.

The MCU is CC2430
Here is the crystals part of the schematics.
(The crystals are XTAL1, XTAL2).


I have measured the DC voltage of both pins and it was 1.8V in each pin of the two crystals.
I must add that the MCU has an internal 1.8V regulator (supplied by the 3.3V VDD PSU) that according to datasheet is supposed to supply the analog part of the MCU.
Does the analog part contain both crystal?
Meaning is it logical that the each crystal's pin receives 1.8V?


Thank you.
 

Attachments

Last edited:

blueroomelectronics

Well-Known Member
I've used plenty of watch crystals in my projects, they all work fine. You can't use a scope on them as it will load the circuit.

And of course you used the correct caps and recommended PCB layout with ground ring for the crystals.
 
Last edited:

blueroomelectronics

Well-Known Member
Some of those watch crystals use 6.5pF load caps. Some probes are 10 to 15pF
Either way we don't know how the OP arrived at his conclusion about the frequency error.

PS the internal oscillator will use the digital supply. The datasheet will often recommend matching crystal types.
 

Nigel Goodwin

Super Moderator
Most Helpful Member
Some of those watch crystals use 6.5pF load caps. Some probes are 10 to 15pF
It would be a VERY poor x10 probe if it was :D

A x10 probe increases the input impedance by a factor of ten, and decreases the input capacitance by the same factor of ten - and also obviously decreases sensitivity by a factor of ten.

I've use scopes on crystals many times over the years, it's a common thing to check for - x10 probes have worked in every single case, including numerous 32KHz ones. Incidently, it's almost always the crystal faulty (and not the chip) if it's not oscillating.
 

alphacat

New Member
Hey.
I used 1x probes, but it has an optional 10x probe so i can use it tomorrow and try it out.

The crystals in the prototype board (the one which we manufactured before the pilot) are working fine, and when i measured its crystals with the same scope and probe, then i received nice 32KHz and 32MHz measurements in the scope.

Then one thing that you say is that maybe the manufacturer has soldered wrong capacitors.
Did you experience any other faults in your past?

Thanks.
 

alphacat

New Member
I've used plenty of watch crystals in my projects, they all work fine. You can't use a scope on them as it will load the circuit.

And of course you used the correct caps and recommended PCB layout with ground ring for the crystals.
Cant I measure frequency (By using AC coupling) of crystals?
I've done it with the prototype (not pilot) board's crystals and it showed me the right frequencies.
I just dont understand what makes the crystals go crazy, since both protoype board and pilot board have the same MCU schematics.
What do you think?
 
Last edited:

Sceadwian

Banned
Verify your board traces are correct, obviously it's not the same as the prototype or it'd work =)
 

alphacat

New Member
The traces are connected the same, But there is a difference in the way that the traces go from source to destination.
Can it be a problem that will cause the Crystal to bounce and not keep a steady frequency?

Moreover, could you please tell me by this table, what voltage is supplied to the crystals?


 

Attachments

Sceadwian

Banned
alphacat, you don't seem to understand how crystals work, they're not fed a steady voltage, they're used in the feedback loop of an amplifier. Have you tested the circuit boards or are you just assuming they're not going to work? What types of caps are you using on the finished boards and what were you using to test the setup before? I'm guessing it wasn't in a breadboard.
 
Last edited:

alphacat

New Member
alphacat, you don't seem to understand how crystals work, they're not fed a steady voltage, they're used in the feedback loop of an amplifier. Have you tested the circuit boards or are you just assuming they're not going to work? What types of caps are you using on the finished boards and what were you using to test the setup before? I'm guessing it wasn't in a breadboard.
Answers to your comments.

1. So when I measured the DC voltage on the crystal pins and received 1.8V, wasnt it a bias voltage? where did it came from if not from the XOSC_Q1, XOSC_Q2 pins of the MCU?

2. Yes I've testet both boards.
Both pilot's and prototype's MCU sections are the same, have the same components (same capacitors), the only different is the way that each trace goes from A to B, but it is the same A and the same B for each trace.
What could make both crystals not working? it seems something that bothers both of the crystal, do they have dependent on each other?

I'd appreciate it if you could refer '1' and '2' separately.

Thank you.
 

Mr RB

Well-Known Member
Does your finished PCB have a ground plane? Are your tracks between the MCU and the XTAL longer than 10mm? You may have excessive PCB capacitance.

And the way I measure XTAL freq is to program the MCU so it toggles an output at a divisor of the freq (like /256) then measure THAT freq with a freq meter (or a CRO). This allows the XTAL to work exactly as it would in normal operation, ie no CRO leads attached.

And I kindof agree with felis... ;)
 

Mikebits

Well-Known Member
It would be a VERY poor x10 probe if it was :D

A x10 probe increases the input impedance by a factor of ten, and decreases the input capacitance by the same factor of ten - and also obviously decreases sensitivity by a factor of ten.

I've use scopes on crystals many times over the years, it's a common thing to check for - x10 probes have worked in every single case, including numerous 32KHz ones. Incidently, it's almost always the crystal faulty (and not the chip) if it's not oscillating.
What kind of probe you using Nigel? I have a pretty good 10x probe, TEK P6137, and it has 10.8 pf .
 
Last edited:

Mikebits

Well-Known Member
Answers to your comments.

1. So when I measured the DC voltage on the crystal pins and received 1.8V, wasnt it a bias voltage? where did it came from if not from the XOSC_Q1, XOSC_Q2 pins of the MCU?

2. Yes I've testet both boards.
Both pilot's and prototype's MCU sections are the same, have the same components (same capacitors), the only different is the way that each trace goes from A to B, but it is the same A and the same B for each trace.
What could make both crystals not working? it seems something that bothers both of the crystal, do they have dependent on each other?

I'd appreciate it if you could refer '1' and '2' separately.

Thank you.
I take it that your a tech on the project. Don't be afraid to consult the design engineers for help (Send an email if out of reach). They are often happy to assist you in learning the product, and your will learn much from them... :)
 

Mikebits

Well-Known Member
Here is a good app note on scope probes.
 

Attachments

Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top