Original page: 27 June 2004:
Updated : 1 July 2004: Change background
Updated : 1 July 2004: Add Day 2 of Grinding
Updated : 2 July 2004: Add Day 3 of Grinding
Updated : 8 July 2004: Add Day 4 of Grinding
Updated : 22 July 2004: Add post grinding analysis
Updated : 16 August 2004: Add crystals 2, 3, 4, 5, and 6
Updated : 25 August 2004: Add crystals 7, 8, 9, 10, 11, 12, and 13
Last Updated : 14 October 2004: Add crystals 14 and 15


Cute little horsey

Preface

Recently, Brian sold me a quantity of Quartz crystal blanks, along with the mounting plates. These were mostly in the range of 5 to 6 MHz. Unfortunately, these are mostly useless for amateur radio operators, since they lay outside of an amateur radio band. However, it is possible to grind these crystals in order to raise the frequency so that it falls within the 40 meter amateur radio band (7.000 MHz to 7.300 MHz).

The basic procedure for accomplishing the frequency changing grinding operation is covered in the "Radio Handbook, Twenty-Second Edition" by William I. Orr, W6SAI (1981, Howard W. Sams & Company, Inc, Indianapolis, IN 46268), on page 11.7:


Crystal Grinding Techniques

Crystals may be raised in frequency by grinding them to smaller dimensions. Hand grinding can be used to raise the frequency of an already finished crystal and the can be accomplished without the use of special tools or instruments. In the case of the surplus FT-243 style of crystal, the blank may be raised in frequency up to several hundred kilohertz, if it is a fundamental-frequency cut.

A micrometer is required to measure the crystal thickness and grinding is done on a small shett of optically flat glass. A piece of plate glass will suffice for the home work-shop. A grinding compound composed of carborundum powder and water is required. A few ounces of #220 and #300 grits are suggested.

Before grinding is started, the crystal should be checked in an oscillator to make sure it is active. Activity of the crystal can be rechecked during the grinding process to make sure that the faces of the crystal remain parallel.

One face of the crystal is marked with a pencil as a reference face. All grinding is done on the opposite face in order to maintain a reference flat surface. A small amouint of #400 grinding grit is placed on the glass disc and enough water added to make a paste. The unmarked side of the crystal is placed face cown on the disk and the blank is rubbed in a figure-8 motion over the disc, using just enough pressure from the index finger to move the crystal.

After about a dozen figure-8 patterns have been traced (depending on the amount of frequency change desired), the crystal is washed with water and wiped dry. The crystal is then placed in the holder for a frequency check. The process is repeated a number of times until the crystal is gradually moved to the new frequency.

For larger movement of the crystal frequency, the #220 grit may be used. Additionally grit should be added to the glass plate as the compound gradually loses it's cutting power with use.

If crystal activity drops with grinding, the blank should be measured with a micrometer to determine the degree of flatness. Normally, the corners are one to three ten-thousandths of an inch thinner than the center of the blank. A thick corner will tend to reduce activity. Grinding the edge of the crystal will restore activity in some cases.

When reassembling the FT-243 holder make sure that the raised corners of the top electrode press against the blank; these are the only points of the electrode that make contact with the crystal.


Additional information on Quartz crystals may be obtained from http://www.bliley.net/XTAL/docs/eng_bulls/E6/E-6_Bulletin.pdf .

A good discussion of Quartz crystal grinding techniques can be found at http://s88932719.onlinehome.us/ham/rokgrind.htm and http://www.bliley.net/XTAL/docs/misc/XTAL_grinding/grinding.html .

A good history of Quartz crystals can be found at http://www.ieee-uffc.org/fc_history/bottom.html .

Crystal One

Preparation

For the first test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 5.850 MHz [1].

[1] All frequency measurements were conducted with a Roadstar model FC250 frequency counter.

In order to test the resonant frequency of a Quartz slab, it is mounted between two plates. These plates have raised corners such that the only positions that they touch the Quartz crystal at are the very corners. The crystal and mounting plates are then mounted in a holder made from two small sections of printed circuit board material. The two pieces of printed circuit board material are held together by two machines screws and nuts, with a spacer between the screw farthest away from the crystal. This forms a triangular mount for the crystal, where the crystal (between the plates) is held in place by the printed circuit board material. This allows electrical contact to be made to the plates holding the crystal. Additionally, the natural springiness of the printed circuit board material allows a firm yet moderate mounting pressure for the crystal.

The mounted crystal is then connected to a Pierce configured oscillator, built using a J-FET transistor. The oscillating frequency is read from a frequency counter.

A sheet of thick (3/16 inch) glass was obtained, which was approximately 4 inches wide by 13 inches long. Half of this sheet was roughened by etching it with a paste of Ammonium Bifluoride [2]. The thought was that a rough substrate would hold the abrasive material better than a smooth glass surface.

[2] Ammonium Bifluoride paste is used for etching glass, and is commercially available as "Armor Etch" at Michaels stores (as well as other outlets).

Grinding - Day 1

The crystal was first abraded on a carborundum sharpening stone. 10 figure eight laps were made on the sharpening stone, while holding the crystal by it's corners. The first 10 laps produced a frequency change of approximately 1 KHz (e.g., The resonant frequency was raised from 5.850 MHz to 5.851 Mhz.). A further 20 figure eight laps increased the frequency by an additional 2 KHz (5.851 MHz to 5.853 Mhz) [3].

[3] All grinding to the crystal was carried out by grinding the same surface, leaving one of the surfaces as the original cut. While grinding, the crystal was held by two opposite corners, and periodically throughout the grinding process, the crystal was rotated so that it was held by the different corners in order to equalize the grinding.

At this point, the glass sheet was sequentially covered with a variety of abrasive materials, and the crystal was rubbed in figure eight patterns. After each grinding session, the crystal was washed in tap water, dried, mounted between a pair of plates, and tested.

The first material tried was Comet scouring powder, formed into a paste with water. Comet scouring powder is reported to have been one of the favorite materials used by early amateur radio amateurs to grind crystals. However, note that Comet is now advertised as being "scratchless", which may indicate that the abrasive material has been reformulated. The first 25 laps produced an increase in frequency of 4 KHz (5.853 MHz to 5.857 MHz). An additional 100 laps produced a further increase of 2 KHz (5.857 Mhz to 5.859 MHz) [6].

[6] It will be noticed that the frequency change is not necessarily proportional to the number of laps. There may be several explanations for this. It is possible that a different pressure was used throughout the grinding operation. It is also possible that the abrasive was worn out after a limited number of laps. Finally, it is also possible that the glass plate was polished such that its ability to hold the abrasive in place was reduced. After the first day's grinding operations, it was noticed that the etched portion of the glass plate was noticably smoother than it was immediately after being etched.

The glass plate was cleaned, and the next material was tried. This material was Barkeeper's Friend cleaning powder. 100 figure eight laps were performed, which produced a frequency increase of 6 KHz (5.859 MHz to 5.865 MHz). An additional 100 laps were performed, which produced a frequency increase of 9 KHz (5.865 MHz to 5.874 Mhz).

The glass plate was then cleaned again, and the next abrasive material was tried. This material was Wright's Copper Cream . The first 100 figure eight laps resulted in a frequency change of 20 KHz (5.874 MHz to 5.894 MHz). A further 100 laps resulted in a frequency change of 15 Khz (5.894 Mhz to 5.909 MHz).

The glass plate was again cleaned, and the next abrasive was tried. This material was Crest toothpaste. Toothpaste was another favorite of early crystal grinders. However, the first 100 figure eight laps produced a frequency change of -1 KHz (5.909 MHz to 5.908 MHz, or a 1 KHz lowering of the frequency) [4].

[4] It is possible that the crystal wasn't cleaned adequately since the addition of a small amount of debris on the surface of a Quartz crystal is known to cause the lowering of the resonant frequency. See the work done by Hans Summers on the subject of crystal penning , and several reports concerning "Crystal Penning" in 1930s editions of the magazine QST.

The glass plate was again cleaned.

The crystal was then abraded on the carborundum sharpening stone again. 100 laps on the stone resulted in a frequency change of 4 Khz (5.908 MHz to 5.912 MHz).

The crystal was then coated on one surface with the Ammonium Bifluoride solution, and allowed to sit for 10 minutes. After 10 minutes, the Ammonium Bifluoride solution was washed off of the crystal. This caused a frequency shift of 9 KHz (5.912 MHz to 5.921 Mhz).

As this point, additional abrasive laps were performed. However, instead of the firgure 8 patterns, a linear motion was used, coupled with frequent rotation of the crystal to equalize the abrasion. After each series of abrasions, the crystal was examined under a magnifying glass to check for any abnormal wear patterns. 500 linear strokes, using Barkeeper's Friend scouring powder, produced a frequency shift of 18 KHz (5.921 MHz to 5.939 MHz). A further 400 strokes produced a frequency shift of 9 KHz (5.939 MHz to 5.948 Mhz).

At this point, the figure eight pattern was resumed, due to the observation of longitudinal abrasion patterns on the abraded surface. An additional 200 figure eight laps, using the Barkeeper's Friend scouring powder, produced a change of 7 KHz (5.948 MHz to 5.955 Mhz). An additional 400 laps produced a change of 11 KHz (5.955 MHz to 5.966 Mhz). An additional 400 laps produced a frequency change of 40 KHz (5.966 to 6.004 MHz). At this point, grinding operations ceased for the night.

At the end of one night's worth of grinding, the frequency of the 5.850 MHz crystal has been shifted 154 KHz, to a new frequency of 6.004 MHz.

Grinding - Day 2

The crystal was again ground on the glass plate using Wright's Copper Cream as an abrasive. 400 laps were ground, which changed the frequency by 20 Khz (6.004 to 6.024 MHz).

An additional 200 figure-8 laps were ground, using Wright's Copper Cream, producing a frequency change of 37 Khz (6.024 to 6.061 MHz).

In order to determine if the abrasive material was necessary, the crystal was lapped 100 times in a figure-8 pattern with no abrasive. This was accomplished by performing the figure-8 pattern on the moistened glass plate. This produced no observable change in frequency (6.061 to 6.061 MHz), although did put an optical polish on the ground surface of the crystal. The crystal wanted to adhere to the glass plate, and it required considerable force to move it at times. No futher bare grinding was attempted due to the lack of a frequency change, and due to fear of cracking the crystal or welding it to the glass plate.

An addition 200 figure-8 laps were ground, using Wright's Copper Cream, producing a frequency change of 28 KHz (6.061 to 6.089 MHz). The optical polish on the ground surface of the crystal was seen to disappear after this round of grinding.

An additional 200 figure-8 laps were ground, using Wright's Copper Cream, producing a frequency change of 34 KHz (6.089 to 6.123 MHz).

An additional 250 figure-8 laps were ground, using Wright's Copper Cream, producing a frequency change of 25 KHz (6.123 to 6.148 MHz). At this point, grinding operations ceased for the night.

Grinding - Day 3

The crystal was again ground on the glass plate using Ajax cleaner as an abrasive, which was another favorite of grinders from ages ago. 200 laps were ground, which changed the frequency by 2 Khz (6.148 to 6.150 MHz).

At this point, grinding resumed using Wright's Copper Cream, and an additional 400 figure-8 laps were ground, which changed the frequency by 74 KHz (6.150 to 6.224 MHz).

An additional 200 laps were ground, using Wright's Copper Cream, which resulted in a frequency increase of 30 KHz (6.224 to 6.254 MHz). At this point, grinding operations ceased for the night.

Grinding - Day 4

The crystal was again ground on the glass plate using Dremel Polishing Compound as an abrasive. This stuff is great!

200 figure-8 laps were ground, using the polishing compound, which changed the frequency by 253 Khz (6.254 to 6.507 MHz).

Another 200 figure-8 laps were ground, still using the polishing compound, which changed the frequency by 142 KHz (6.507 to 6.649 MHz).

Another 200 figure-8 laps were ground, using the polishing compound, which changed the frequency by 208 KHz (6.649 to 6.857 MHz).

At this point only 150 figure-8 laps were ground, since the desired frequency was being approached. Unfortunately, at the end of these 150 figure-8 laps, the crystal experienced its first loss of activity. The corners were lightly beveled, using 20 linear strokes on each corner, but this produced no increase in activity. The edges were lightly ground, using 20 linear strokes, but this also produced no change in activity. At this point, 60 figure-8 laps were ground, and the activity returned, although the oscillations were not as strong as before. These activities produced a frequency change of 248 KHz (6.857 to 7.105 MHz). Unfortunately, the desired target of 7.030 MHz was overshot, although the crystal is still usable since the 7.105 MHz frequency falls into the novice portion of the 40 meter band.

No further grinding activities are planned for this crystal. Unfortunately, since the target frequency was overshot, no chemical etching is planned for this crystal either, since this would shift the frequency higher, and possibly out of the 40 meter CW bands, which would make the crystal unusable. This should be a good testcase to determine how much frequency shift occurs to to debris shedding over a period of time.

Post Grinding Analysis

The crystal was sent to a friend to measure (due to the current lack of a micrometer, which will soon be rectified). The measurements prove that a crystal need not have both faces parallel in order to oscillate. Two corners are .015 inches thick. The middle is .014 inches thick. The other two corners are .009 inches thick, for an error of .006 inches! Amazingly, the crystal was still able to oscillate.

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
10 Figure-8 Carborundum Stone 5.850 5.851 .001
20 Figure-8 Carborundum Stone 5.851 5.853 .002
25 Figure-8 Comet Scouring Powder 5.853 5.857 .004
100 Figure-8 Comet Scouring Powder 5.857 5.859 .002
100 Figure-8 Barkeeper's Friend Scouring Powder 5.859 5.865 .006
100 Figure-8 Barkeeper's Friend Scouring Powder 5.865 5.874 .009
100 Figure-8 Wright's Copper Cream 5.874 5.894 .020
100 Figure-8 Wright's Copper Cream 5.894 5.909 .015
100 Figure-8 Crest Toothpaste 5.909 5.908 -.001
100 Figure-8 Carborundum Stone 5.908 5.912 .004
10 Minutes * Ammonium Bifluoride Paste 5.912 5.921 .009
500 Linear Barkeeper's Friend Scouring Powder 5.921 5.939 .018
400 Linear Barkeeper's Friend Scouring Powder 5.939 5.948 .009
200 Figure-8 Barkeeper's Friend Scouring Powder 5.948 5.955 .007
400 Figure-8 Barkeeper's Friend Scouring Powder 5.955 5.966 .011
400 Figure-8 Barkeeper's Friend Scouring Powder 5.966 6.004 .038
400 Figure-8 Wright's Copper Cleaner 6.004 6.024 .020
200 Figure-8 Wright's Copper Cleaner 6.024 6.061 .037
100 Figure-8 Bare Glass Substrate 6.004 6.004 .000
200 Figure-8 Wright's Copper Cleaner 6.061 6.089 .018
200 Figure-8 Wright's Copper Cleaner 6.089 6.123 .034
250 Figure-8 Wright's Copper Cleaner 6.123 6.148 .025
200 Figure-8 Ajax 6.148 6.150 .002
400 Figure-8 Wright's Copper Cleaner 6.150 6.224 .074
200 Figure-8 Wright's Copper Cleaner 6.224 6.254 .030
200 Figure-8 Dremel Polishing Compound 6.254 6.507 .253
200 Figure-8 Dremel Polishing Compound 6.507 6.649 .142
200 Figure-8 Dremel Polishing Compound 6.649 6.857 .208
200 Figure-8 Dremel Polishing Compound 6.857 7.105 .248

Crystal Two

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.949 MHz.

Grinding

The crystal was ground 100 figure-8 laps using the Bartender's Friend powder. This increased the frequency by 3 KHz (6.949 MHz to 6.952 MHz).

The crystal was then ground 100 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 173 KHz (6.952 MHz to 7.125 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
100 Figure-8 Bartender's Friend Scouring Powder 6.949 6.952 .003
100 Figure-8 Dremel Polishing Compound 6.952 7.125 .173

Crystal Three

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.925 MHz.

Grinding

The crystal was ground 75 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 41 KHz (6.925 MHz to 6.966 MHz).

The crystal was then ground 40 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 29 KHz (6.966 MHz to 6.995 MHz).

The crystal was then ground 40 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 32 KHz (6.995 MHz to 7.027 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
75 Figure-8 Dremel Polishing Compound 6.925 6.966 .041
40 Figure-8 Dremel Polishing Compound 6.966 6.995 .029
40 Figure-8 Dremel Polishing Compound 6.995 7.027 .032

Crystal Four

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.899 MHz.

Grinding

The crystal was ground 100 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 81 KHz (6.899 MHz to 6.980 MHz).

The crystal was then ground 50 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 13 KHz (6.980 MHz to 6.993 MHz).

The crystal was then ground 20 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 13 KHz (6.993 MHz to 7.006 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
100 Figure-8 Dremel Polishing Compound 6.899 6.980 .081
50 Figure-8 Dremel Polishing Compound 6.980 6.993 .013
20 Figure-8 Dremel Polishing Compound 6.993 7.006 .013

Crystal Five

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.800 MHz.

Grinding

The crystal was ground 120 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 111 KHz (6.800 MHz to 6.911 MHz).

The crystal was then ground 100 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 136 KHz (6.911 MHz to 7.047 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
120 Figure-8 Dremel Polishing Compound 6.800 6.911 .111
100 Figure-8 Dremel Polishing Compound 6.911 7.047 .136

Crystal Six

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.950 MHz.

Grinding

The crystal was ground 100 figure-8 laps using the Dremel Polishing Compound. This increased the frequency by 120 KHz (6.950 MHz to 7.070 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
100 Figure-8 Dremel Polishing Compound 6.950 7.070 .120

Crystal Seven

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.850 MHz.

Grinding

The crystal was ground 100 figure-8 laps using carborundum powder. This increased the frequency by 53 KHz (6.950 MHz to 6.903 MHz).

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 107 KHz (6.903 MHz to 7.010 MHz).

The crystal was ground 40 figure-8 laps using carborundum powder. This increased the frequency by 22 KHz (7.010 MHz to 7.032 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
100 Figure-8 Carborundum powder 6.850 6.903 .053
200 Figure-8 Carborundum powder 6.903 7.010 .107
40 Figure-8 Carborundum powder 7.010 7.032 .022

Crystal Eight

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.815 MHz.

Grinding

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 38 KHz (6.815 MHz to 6.853 MHz).

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 65 KHz (6.853 MHz to 6.918 MHz).

The crystal was ground 100 figure-8 laps using carborundum powder. This increased the frequency by 26 KHz (6.918 MHz to 6.944 MHz).

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 62 KHz (6.944 MHz to 7.006 MHz).

The crystal was ground 100 figure-8 laps using carborundum powder. This increased the frequency by 27 KHz (7.006 MHz to 7.033 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
200 Figure-8 Carborundum powder 6.815 6.853 .038
200 Figure-8 Carborundum powder 6.853 6.918 .065
100 Figure-8 Carborundum powder 6.918 6.944 .026
200 Figure-8 Carborundum powder 6.944 7.006 .062
100 Figure-8 Carborundum powder 7.006 7.033 .027

Crystal Nine

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.800 MHz.

Grinding

The crystal was ground 300 figure-8 laps using carborundum powder. This resulted in a loss of activity.

The crystal was ground 60 figure-8 laps using carborundum powder. This resulted in a frequency increase of 336 KHz (6.800 MHz to 7.136 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
300 Figure-8 Carborundum powder 6.800 ? Loss of activity
60 Figure-8 Carborundum powder ? 7.136 .336

Crystal Ten

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.800 MHz.

Grinding

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 129 KHz (6.800 MHz to 6.929 MHz).

The crystal was ground 200 figure-8 laps using carborundum powder. This resulted in a loss of activity.

The crystal was ground 50 figure-8 laps using carborundum powder. This increased the frequency by 398 KHz (6.929 MHz to 7.327 MHz). Whoops!

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
200 Figure-8 Carborundum powder 6.800 6.929 .053
200 Figure-8 Carborundum powder 6.929 ? Loss of activity
50 Figure-8 Carborundum powder ? 7.327 .398

Crystal Eleven

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.776 MHz.

Grinding

The crystal was ground 320 figure-8 laps using carborundum powder. This increased the frequency by 302 KHz (6.776 MHz to 7.078 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
320 Figure-8 Carborundum powder 6.776 7.078 .302

Crystal Twelve

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.775 MHz.

Grinding

The crystal was ground 300 figure-8 laps using carborundum powder. This increased the frequency by 187 KHz (6.775 MHz to 6.962 MHz).

The crystal was ground 100 figure-8 laps using carborundum powder. This increased the frequency by 49 KHz (6.962 MHz to 7.011 MHz).

The crystal was ground 40 figure-8 laps using carborundum powder. This increased the frequency by 14 KHz (7.011 MHz to 7.025 MHz).

The crystal was ground 40 figure-8 laps using carborundum powder. This increased the frequency by 27 KHz (7.025 MHz to 7.052 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
300 Figure-8 Carborundum powder 6.776 6.962 .187
100 Figure-8 Carborundum powder 6.962 7.011 .049
40 Figure-8 Carborundum powder 7.011 7.025 .014
40 Figure-8 Carborundum powder 7.025 7.052 .027

Crystal Thirteen

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.725 MHz.

Grinding

The crystal was ground 340 figure-8 laps using carborundum powder. This increased the frequency by 179 KHz (6.725 MHz to 6.904 MHz).

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 86 KHz (6.904 MHz to 6.990 MHz).

The crystal was ground 100 figure-8 laps using carborundum powder. This increased the frequency by 37 KHz (6.990 MHz to 7.027 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
340 Figure-8 Carborundum powder 6.725 6.904 .179
200 Figure-8 Carborundum powder 6.904 6.990 .086
100 Figure-8 Carborundum powder 6.990 7.027 .037

Crystal Fourteen

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.500 MHz.

Grinding

The crystal was ground 400 figure-8 laps using carborundum powder. This increased the frequency by 344 KHz (6.500 MHz to 6.844 MHz).

The crystal was ground 200 figure-8 laps using carborundum powder. This increased the frequency by 206 KHz (6.844 MHz to 7.050 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
400 Figure-8 Carborundum powder 6.500 6.844 .344
200 Figure-8 Carborundum powder 6.844 7.050 .206

Crystal Fifteen

Preparation

For the next test, a Quartz crystal was selected at random, which proved to have a resonant frequency of 6.500 MHz.

Grinding

The crystal was ground 400 figure-8 laps using carborundum powder. This increased the frequency by 270 KHz (6.500 MHz to 6.770 MHz).

The crystal was ground 400 figure-8 laps using carborundum powder. This increased the frequency by 296 KHz (6.770 MHz to 7.066 MHz).

Results

Count Type Material Starting Frequency Ending Frequency Frequency Shift
400 Figure-8 Carborundum powder 6.500 6.770 .270
400 Figure-8 Carborundum powder 6.770 7.066 .296

Future ideas

Some of the future work involves using alternative abrasives, such as sand (Silicon Dioxide), and Aluminum Oxide/Emery [5][6].

[5] Quartz is a very hard material, exceeded in hardness by very few materials.

[6] Quite a few of the "abrasive" materials that have been used at this point appear to be based on Calcium Carbonate as the abrasive element. This is somewhat surprising since Calcium Carbonate is a relatively soft material.

A tip I received was that the Ammonium Bifluoride paste that I was using may have been too thick. This may have allowed the part in immediate contact with the crystal to etch it, but once this layer had reacted, no further etching occurred. It was suggested that I dilute the Ammonium Bifluoride paste to a thinner consistency. This may allow more of it to be exposed to the surface of the crystal, especially if some type of agitation is applied.

Some discussion has taken place regarding the construction of automated crystal lapping machines.

One idea involves the use of a round glass substrate as a rotating turntable to perform the abrasion with, while holding the crystal against the turntable with a slowly rotating clock driven holder. In operation, the rotation of the turntable would perform the etching, while the rotation of the holder would cause the grinding to be even across the crystal, and prevent uneven etching.

This is somewhat similar to a historic approach based on a drill press. In the literature describing that system, the substrate was an iron plate which was coated with an abrasive powder.

Another idea was to use a reciprocating mount for the crystal formed by attaching a rod eccentrically to a rotating wheel. This linear motion would be turned into a figure-8 motion by attaching another eccentrically mounted rod to the holder. By the proper adjustment of the speeds of the two motors driving the rotating wheels, a figure-8 patern could be achieved.

An alternative to the previous idea that may be mechanically simpler involves using attaching the second reciprocating arm to the glass substrate. Thus, the crystal mount would be pushed and pulled in one dimension, while the glass substrate was pushed and pulled in the other dimension. Again, by the proper synchronization of the drive motors, a figure-8 pattern could be made to occur.