I want ask where I can find some information about power fed to XT1 and XT2 crystal in different configuration of XT1DRIVE/XT2DRIVE bits. In datasheet is information about rising current consumption but this is a total consumtion as I think (fed to crystal and oscillator) not equal to power fed only to crystal as I think. I ask about this because some of crystals need to power 0,1uW other 1uW or even 10uW...On the other hand too much power is not good for crystal.
Sorry for my english.
This information is indeed not in the datasheet. However, it is not as easy as it might seem.
An ideal crystal with ideal load capacitors won't require any power at all. However, capacitors have an ESR and the crystal also has an ESR, that means that for pushing the charge back and forth requires some additional energy. This energy is required for each movement of the charge (whose amount is the same independent of frequency and determined by crystal and load capacitance).
So with increasing frequency, the losses per second increase with the number of oscillations per second. And to compensate for the losses, the driver needs to push harder, as the ESRs stay the same. And pushing harder usually means increasing the driving voltage. That's actually what the driving strength setting does: increase the oscillation voltage, so the amount of charge flowting from capacitor through crystal to the other capacitor stays the same. And of course does this increase the produced heat: in all components, Crystal, capacitors and driver. But the crystal takes only a fraction of this.
You can calculate the values by your own: the datasheet contains typical current consumption for different crystal speeds (driving strengths). Assuming a linear increase of current though the crystal, the table in the 5438 gives me ~7.5µA/Mhz for the frequency-dependent part. The rest is sort of a base current. But even the frequency-dependent part isn't all for the crystal. A large portion goes to the driver and much of it to the capacitors.
Unfortunately this doesn't really help for the power calculations, since the driving voltage is still unknown. It's below 3V, definitely :)
From own measurements on a crystal, I think I remember that 60mV p-p are not enough to be detected, while a typical range was 0.6V p-p. But I'm not sure of these values, just a dim memory, and it was for a 1611 processor anyway, which did not have a driving-strength setting at all. Based on these values I'd get a driving power of 13µW on 8MHz. Well, sounds reasonable.
Sylwester NowocienOn the other hand too much power is not good for crystal
_____________________________________Before posting bug reports or ask for help, do at least quick scan over this article. It applies to any kind of problem reporting. On any forum. And/or look here.If you cannot discuss your problem in the public, feel free to start a private conversation: click on my name and then 'start conversation'. But please do so only if you really cannot do it in a public thread, as I usually read all threads. And I prefer to answer where others can profit from it (or contribute to it) too.
Hi Jens-Michael,Thank You for comprehensive answer. I know that this is not easy to measure, that's way I'm asking:).You have absolutely right Vp-p is something about 0.5V and current we can get from data sheet (this was my first thought) but I'm afraid that this simple calculation is an oversimplification even that sound good.I agree with You that faster crystal and with larger ESR usually need more power but if I understand You correctly. Driving current is as high as need to get something about 0.5V Vp-p oscilation but what if I take crystal which need 1uW (this is optimal power for it) and set drive strength bits to max?.This is of course rhetorical question but that's mean that data in data sheet is measured for a particular crystal with particular ESR and some specific capacitors. So the question is what were the conditions of measurement of current from data sheet and how does this apply to my system with my particular crystal and capacitors?Do you have any idea except direct current consumption measurement by my system in different configuration of drive strength bits and in different enviromental condition?Or setting drive strength as low as possible to need to correct crystal work and belive that this is not to much for crystal and that ESR will not change to much with temperature and time (even that is not true)?... Thank you for your interest...Best regardsSylwek
Sylwester NowocienDriving current is as high as need to get something about 0.5V Vp-p oscilation
Sylwester Nowocienwhat if I take crystal which need 1uW (this is optimal power for it)
Sylwester Nowocien that's mean that data in data sheet is measured for a particular crystal with particular ESR and some specific capacitors
Sylwester Nowocienhow does this apply to my system with my particular crystal and capacitors
I mus tadmit that I'm not really a crystals expert. This is a wide and specialized field. And the manufacturers datasheets aren't always helpful. As you probably noted yourself, soem basic parameters are given, such as frequency and required load capacitance. But for the rest liek ESR, optimum power, and many more, only part of them is noted in changing constellation, if at all. If you need a high safety margin or highest precision, you might discuss this with a manufacturer directly. There is a reason why crystal manufacturers have more than one type of crystal with same frequency and package in their portfolio.
Personally, I begin with try and hope (well, of course after looking at the standard datasheet). If it works without any suspicious flukes, fine. If not, I proceed to experiment and measure. And only if this isn't successful, I start digging deep into datasheets and additional documents. Sometimes it helps, sometimes not. Last chance then is callign an expert at the manufacturers side. Luckily, it isn't necessary often :)
You have right with CMOS technology, this is obvious. I'm not quite clearly in the words (it's still my problem with the language - I'm sorry).
I know that this is real engeener work, this is not my first time:)... But in that project I need very high long term clock stability and I'm trying minimalize all factor which increase aging efects include "too much power". Since it is difficult to reasonably measure and I don't have much time, hence my question.
I see you're an engineer with a reasonable, rational approach. Have you any idea how can I speed up that process? I mean crystal aging process associated with the excessive power? Of course drive them more much power but this is bad idea because relationship between power and long term crystal stability isn't linear and as You also noticed, much power increase crystal mechanical damage faster than in normal works (let we say by 10 years) so I'm afraid that get any reasonable results in 3-4 weeks are impossible:(...
Thak you for your interest and comprehensive posts.
Sylwester Nowocien I'm not quite clearly in the words (it's still my problem with the language - I'm sorry).
Sylwester NowocienI know that this is real engeener work, this is not my first time:)
Well, long-term clock stability is quite a challenge, independently of the driving strength. There are so many influence which will make the crystal drift (besides aging, there is temperature, even the supply voltage of the oscillation driver). It's a good thing if you can connect to a network time server form time to time and do some adjustments :)
Well for aging effects, a typical approach is to heat the device unde rtest. Increased temperature increases aging. I think there are tables telling you how much aging speeds up when rising the temperature by x degrees.None of the manufacturers lets e.g. a harddisk run 10khrs just to say that the MTBF is 3khrs. When one device fails after the MTBF, it is usually long out of production already. All those 'values' are usually determined by operating the device a few days or weeks at increased temperature and scaling the result.
However, I have no idea how this scales to the mechanical wear of the crystal. And also, the temperature change of course lets the crystal drift, so this has to be taken into account when analyzing the results.
Well, all I can suggest is contacting the crystal manufacturer. If yo're lucky they have own experiences with MSPs, and even if not, they maybe able to clear up some doubts regarding the mechanical stress.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.