Crystal Measurements

[wa1fxt]

Bob,

In the past I have characterized crystals with an HP generator and Spectrum analyzer using the techniques described in EMRFD and elsewhere. While tedious, they have provided accurate and reliable results. I would like to streamline that process some by using the AIM.

Can you give me some insight into how your crystal measurement function is derived. It is easy enough to just "pick it off the menu". But, I would like to know exactly what is going on inside of the AIM/PC/software. As you might have guessed, I am using this stuff for crystal filter designs and the data must be accurate

Thanks es 73,
Bob WA!FXT

[Bob]

wa1fxt said:

Bob,

In the past I have characterized crystals with an HP generator and Spectrum analyzer using the techniques described in EMRFD and elsewhere. While tedious, they have provided accurate and reliable results. I would like to streamline that process some by using the AIM.

Can you give me some insight into how your crystal measurement function is derived. It is easy enough to just "pick it off the menu". But, I would like to know exactly what is going on inside of the AIM/PC/software. As you might have guessed, I am using this stuff for crystal filter designs and the data must be accurate

Thanks es 73,
Bob WA!FXT

Hi Bob,
The equations are in the manual starting on (or near) page 37. There's an example there too.

-- 73/ Bob

[wa3off]

On thing that I don't think is described in the manual is how the software performs the scan to locate the resonant frequencies. I've tested crystals that seemed to have reasonable resonances that I could find by scanning manually but that the Crystal Measurement function seemed unable to correctly identify. Obviously, if the resonances are not correctly identified, the rest of the results are meaningless as well.

Dave, WA3OFF

[Bob]

wa3off said:

On thing that I don't think is described in the manual is how the software performs the scan to locate the resonant frequencies. I've tested crystals that seemed to have reasonable resonances that I could find by scanning manually but that the Crystal Measurement function seemed unable to correctly identify. Obviously, if the resonances are not correctly identified, the rest of the results are meaningless as well.

Dave, WA3OFF

Hi Dave,
It may well be there's a problem in the routine. I haven't checked it with a wide assortment of crystals. Using a crystal for which it doesn't get the right answers, please send me the manual scan and a screen shot of the crystal data and I'll take a look at it.

73/ Bob

[wa1fxt]

Bob,

I am having the same problems as Dave. I have run a sample of ECS 10Mhz crystals thru. Then 100 FOX 10Mhz crystals and compared the results to the G3G3UUR methods and the 3Db methods and found the AIM 4170 all over the map. In fact if I made successive runs on the same crystal, I got data that was as much as 25% different on each run. I verified the G3UUR method data against a VNA and found it to be accurate to approx. 2%. So. I consider the G3UUR method my base line standard.

Also, a thingypit suggestion. When characterizing crystals for filters, one needs to record a lot of data. It would be nice if the AIM could at the very least save and/or print the crystal data from the crystal measurement function. If the data (once the above problem is fixed) can be saved in a .csv file, crystals can then be sorted and characterized for use in filters. Otherwise, the AIM is not saving us much time other the traditional methods such as the 3Db method and the G3UUR methods. And, as it stands right now, accuracy is in question.

Another anomoly.......My manual shows scans to four decimal place in the figures. When scanning thabove crystals, I could not get "zoomed in" closer than three decimal places whever I changed the Limits.

I can send you screen shots. And data files from a normal scan But, I don't think they would help you much.

Thanks for the help,

Bob WA!FXT

[Bob]

In the last few days I've made some major improvements to the crystal measurement routine. I'll send you a beta version to try. I'll take a look at saving the data too. Tell me what parameters and what format you would like.

73/ Bob

[wa1fxt]

Thanks Bob!!!!!!!

I was having second thoughts about this morning's EMAIL and previously didn't consider possible spurious crystal responses. I don't have a suggestion how to handle that yet....if it even affects the frequency identification.

I am anxiously waiting by the mail box for the new routine!!!!! When I am home from the company store that is!!

What to save? Well, most of the time you would want to see the series resonant frequency Fs, motional resistance Rm, RS or ESR, and Q. But there are times where average values of Lm, Cm and Co are helpfull. So, I vote for saving all the values in the current table. Now if you can use these individual parameters to build a table of multiple data sets and store that table to a .csv file. Then EXCEL can be used to sort the crystals, determine average values, standard deviations etc. Unless of course you would want to write more code to do that! <<grin>>.

If you change the current dialog box to include a field for the Crystal number identification at the top of the form, a CANCEL button, and a SAVE button at the bottom of the form. I believe the new form and the above changes should be all that one needs to design filters.

I have several filter design utility programs. All of which are very good. If I had the "new" .csv file from the AIM, I could combine the data from the AIM, the filter design equations from the books (or utility programs) I would really have a powerful tool. Alas, I don't have time from the company store to program and I am slow. Maybe you could hook up with Dan Maguire again and see if Dan could make up a wiz bang EXCEL program to that end. I like Dan's stuff because it is simple, powerful, and does the jobs that other applications fail to provide....all in one place. His extensions to EZNEC etc. are wonderful. I would be a very willing Beta tester for stuff for the AIM.

73 es TNKS, Bob WA!FXT

[wa3off]

Bob provided me with this beta version as well and I think folks will be quite happy with the improvement in the next release. The results on every crystal I have available to test appear to be very repeatable, within fractions of a percent.

Dave, WA3OFF