The article is a nostalgic reminder of when I was first interested in electronics and got interested in quartz resonators for projects. Grinding down old war surplus FT243-style crystals and other types so as to resonate at frequencies I needed was a common practice with hobbyists back then.
I used many similar techniques to the same end of removing quartz which raised its frequency. Grinding materials included abrasives such as jeweler's rouge, cerium oxide, commercial polishes such as Brasso and Silvo and even HF solution. I'd place the quartz on a small section of plate glass and slide it through a slurry of the abrasive periodically testing its frequency until I'd reached my target.
There's an art to this that's too long to mention here except to say abrasives were used strategically, course grinding would get me near the desired frequency and I'd finish off with a fine abrasive. Then there was the job of re-aging the crystal after its recent abuse to increase its stability. Other techniques were involved such as not lowering its Q factor, etc. which I'll not cover here.
The most desired crystal cut was from the XT-plane (being the most stable) but it was generally difficult to get as it's only a small section of the quartz crystal (also each cut oscillates only over a limited range of frequencies). I used to have a book that explained these cuts in detail, their frequency ranges and electronic properties along with the basic crystallography which I lost years ago. A quick glance at the book would have shown that a great deal of science, engineering and skill is involved in the selection of quartz and its manufacture into useful resonators.
BTW, the mentioning of HF will likely horrify chem-phobic readers. We were well aware of its dangers and took special precautions never to come in contact with it.
This awesome. Could you perhaps remember the title of the book you lost?
> BTW, the mentioning of HF will likely horrify chem-phobic readers
I would not say I am chem-phobic, but yes indeed that stood out. HF is nasty stuff, and yes requires some care I suspect.
The other details are fascinating, though - the intersection of mechanical, crystallographic, and RF (?) properties of a crystal that you can adjust through abrasives and selection of the cut.
Working with HF was extremely difficult before WWII, but it has become much easier after the invention of Teflon.
Teflon is not affected by HF, so if you use only vessels of Teflon to hold the HF solution and tweezers made of Teflon for handling anything that you submerge in the solution of HF, it works fine.
Besides using Teflon for anything that is in contact with the HF solution, you must do all work under a hood that evacuates the vapors of HF emitted by the solution, otherwise handling a HF solution would be very dangerous. It is good that gaseous HF is lighter than air, so after being evacuated it will continue to rise in the air, while becoming more and more diluted.
So is this art of selecting right crystals is applicable for synthetic crystals also or these can be grown perfect each time to specifications?
That sinking feeling of knowing you've ground off just a bit too much...
I've just spent the weekend tuning brass reeds from an organ. It sounds like a very similar process, except you can grind both ends of the tongue to raise and lower frequency.
And, if you're sneaky, you can add solder.
Oh that is so cool. I played the one in Liepaja, Latvia for a bit and it was absolutely amazing. It's love/hate for me (like the harpsichord), I love the instruments but I usually do not like the music that is played on them because of the grating effect. I have pretty bad tinnitus which really spoils a lot of music for me, extremely annoying.
Adding solder has also been frequently used to correct the resonance frequency of quartz crystals that have been ground too much, and I mean during industrial mass-production, not only in a home-lab setting.
How are people sticking stuff to quartz? I know less than nothing, but the pieces of quartz you find in rock don't look like they'd take a solder bond.
I'd assumed that with piezo crystals etc there was a mechanical connection rather than an electrode bonded to the crystal?
But if you can add solder presumably there is some kind of molecular connection with the metal?
I remember when I worked at Motorola some decades ago, there was a little section in the factory area in Schaumburg that made quartz crystals. It's all long since been sold off (I think all the equipment went to China) but I remember all the signs for cyanide alarms, presumably due to some step in the manufacturing process.
The quartz crystals themselves were grown with carefully controlled levels of specific impurities (like scandium) in order to reduce their temperature sensitivity.
Nice to see this get more love, I had a fun time going through that site, there are lots of gems there.
> according to [2] a train crashed in 1972 due to a badly designed crystal oscillator spontaneously jumping to its third overtone
New fear unlocked
More interesting: amazing sleuthing to figure out that that was the root cause.
Believe or not, when I explained to many non techies how quartz watches work and how any computers’ hardware clocks work in the same principle, they were all surprised how elegant and how efficient the mechanism is. I was also impressed when I first learned about it. True science and engineering beauty.
Have a look at the company Silicon Time, they make the most amazing little devices.
Could someone take the equivalent circuit of the article and put in in CircuitJs to form an oscillator? Thanks!
With a crystal of the right dimensions and correct input voltage could one feel the crystal vibrating?
There are piezo actuators you could certainly feel but not sure they are made of quartz, maybe PZT.
You mean could you get it down to a low enough frequency? Hmm
I guess you can get down into audio frequency but maybe the amplitude will be tiny, you probably want a piezo mechanism that will give you more of a rumble.
You can get quartz to resonate down to upper audio frequencies with certain crystal cuts and manufacturing techniques but it's difficult. Typically, the lowest frequency in common use is with its use in watches with a frequency of 32,768Hz (that's about the lower limit where manufacturing and frequency combine to make a useful product).
For electronic circuits such as frequency reference markers where frequency stability is important the lowest practical frequency is 100kHz with 1Mhz preferred, and where frequency tolerances are tight 5 and 10MHz are much preferred with operation in a temperature stabilized oven to minimize frequency drift.
The most frequency-stable crystal cuts are at those frequencies, as frequencies increase (say >10MHz to 100MHz), which at the highest frequencies require the crystal to operate in overtone mode, frequency stability again tends to decrease.
As mentioned in the article there are a lot of common watch crystals that oscillate at 32.768 kHz, but they are tuning fork crystals rather than bulk mode (or modern SAW), which have much higher frequencies. It would be challenging to lower the frequency of these, but perhaps you could evaporate Au or W onto the tips and reseal them to get into the audible frequencies. Much easier would be to get two crystals which beat against each other in the audio range. Even a couple of 6MHz crystals 100ppm off would be ~6kHz (temp sensitive), and you would need some driving circuits, but you might be able to hear the beat by driving an electret microphone or a really tiny tweeter coil speaker (probably need an amp though).
It's much easier to build/buy electrical rather than mechanical LC components to hit audio frequencies ~100Hz-10kHz.
There are piezo buzzers and beepers, of course.
Yes, but those normally just convert the electrical oscillations to sounds, they are not parts of the oscillators that determine the audible frequency.
In the past, audio RC oscillators or LC oscillators were used, with the former being preferred as the latter required too bulky inductors to reach so low frequencies.
Nowadays, it is usually simpler to not use any audio oscillator, but to use some microcontroller that divides the frequency of its clock until reaching the desired audio frequency.
Those can make sounds at other frequencies but they are very loud on their resonant frequency. Loud enough to damage your hearing if you're unlucky.
Nice. I'm an EE by education. We weren't taught how such a key piece of the puzzle worked.
The article is a nostalgic reminder of when I was first interested in electronics and got interested in quartz resonators for projects. Grinding down old war surplus FT243-style crystals and other types so as to resonate at frequencies I needed was a common practice with hobbyists back then.
I used many similar techniques to the same end of removing quartz which raised its frequency. Grinding materials included abrasives such as jeweler's rouge, cerium oxide, commercial polishes such as Brasso and Silvo and even HF solution. I'd place the quartz on a small section of plate glass and slide it through a slurry of the abrasive periodically testing its frequency until I'd reached my target.
There's an art to this that's too long to mention here except to say abrasives were used strategically, course grinding would get me near the desired frequency and I'd finish off with a fine abrasive. Then there was the job of re-aging the crystal after its recent abuse to increase its stability. Other techniques were involved such as not lowering its Q factor, etc. which I'll not cover here.
The most desired crystal cut was from the XT-plane (being the most stable) but it was generally difficult to get as it's only a small section of the quartz crystal (also each cut oscillates only over a limited range of frequencies). I used to have a book that explained these cuts in detail, their frequency ranges and electronic properties along with the basic crystallography which I lost years ago. A quick glance at the book would have shown that a great deal of science, engineering and skill is involved in the selection of quartz and its manufacture into useful resonators.
BTW, the mentioning of HF will likely horrify chem-phobic readers. We were well aware of its dangers and took special precautions never to come in contact with it.
This awesome. Could you perhaps remember the title of the book you lost?
> BTW, the mentioning of HF will likely horrify chem-phobic readers I would not say I am chem-phobic, but yes indeed that stood out. HF is nasty stuff, and yes requires some care I suspect.
The other details are fascinating, though - the intersection of mechanical, crystallographic, and RF (?) properties of a crystal that you can adjust through abrasives and selection of the cut.
Working with HF was extremely difficult before WWII, but it has become much easier after the invention of Teflon.
Teflon is not affected by HF, so if you use only vessels of Teflon to hold the HF solution and tweezers made of Teflon for handling anything that you submerge in the solution of HF, it works fine.
Besides using Teflon for anything that is in contact with the HF solution, you must do all work under a hood that evacuates the vapors of HF emitted by the solution, otherwise handling a HF solution would be very dangerous. It is good that gaseous HF is lighter than air, so after being evacuated it will continue to rise in the air, while becoming more and more diluted.
So is this art of selecting right crystals is applicable for synthetic crystals also or these can be grown perfect each time to specifications?
That sinking feeling of knowing you've ground off just a bit too much...
I've just spent the weekend tuning brass reeds from an organ. It sounds like a very similar process, except you can grind both ends of the tongue to raise and lower frequency.
And, if you're sneaky, you can add solder.
Oh that is so cool. I played the one in Liepaja, Latvia for a bit and it was absolutely amazing. It's love/hate for me (like the harpsichord), I love the instruments but I usually do not like the music that is played on them because of the grating effect. I have pretty bad tinnitus which really spoils a lot of music for me, extremely annoying.
Adding solder has also been frequently used to correct the resonance frequency of quartz crystals that have been ground too much, and I mean during industrial mass-production, not only in a home-lab setting.
How are people sticking stuff to quartz? I know less than nothing, but the pieces of quartz you find in rock don't look like they'd take a solder bond.
I'd assumed that with piezo crystals etc there was a mechanical connection rather than an electrode bonded to the crystal?
But if you can add solder presumably there is some kind of molecular connection with the metal?
I remember when I worked at Motorola some decades ago, there was a little section in the factory area in Schaumburg that made quartz crystals. It's all long since been sold off (I think all the equipment went to China) but I remember all the signs for cyanide alarms, presumably due to some step in the manufacturing process.
The quartz crystals themselves were grown with carefully controlled levels of specific impurities (like scandium) in order to reduce their temperature sensitivity.
Nice to see this get more love, I had a fun time going through that site, there are lots of gems there.
For instance:
https://www.pa3fwm.nl/projects/sdr/
The longer you read, the more amazed you will be.
> according to [2] a train crashed in 1972 due to a badly designed crystal oscillator spontaneously jumping to its third overtone
New fear unlocked
More interesting: amazing sleuthing to figure out that that was the root cause.
Believe or not, when I explained to many non techies how quartz watches work and how any computers’ hardware clocks work in the same principle, they were all surprised how elegant and how efficient the mechanism is. I was also impressed when I first learned about it. True science and engineering beauty.
Have a look at the company Silicon Time, they make the most amazing little devices.
Could someone take the equivalent circuit of the article and put in in CircuitJs to form an oscillator? Thanks!
With a crystal of the right dimensions and correct input voltage could one feel the crystal vibrating?
There are piezo actuators you could certainly feel but not sure they are made of quartz, maybe PZT.
You mean could you get it down to a low enough frequency? Hmm I guess you can get down into audio frequency but maybe the amplitude will be tiny, you probably want a piezo mechanism that will give you more of a rumble.
You can get quartz to resonate down to upper audio frequencies with certain crystal cuts and manufacturing techniques but it's difficult. Typically, the lowest frequency in common use is with its use in watches with a frequency of 32,768Hz (that's about the lower limit where manufacturing and frequency combine to make a useful product).
For electronic circuits such as frequency reference markers where frequency stability is important the lowest practical frequency is 100kHz with 1Mhz preferred, and where frequency tolerances are tight 5 and 10MHz are much preferred with operation in a temperature stabilized oven to minimize frequency drift.
The most frequency-stable crystal cuts are at those frequencies, as frequencies increase (say >10MHz to 100MHz), which at the highest frequencies require the crystal to operate in overtone mode, frequency stability again tends to decrease.
As mentioned in the article there are a lot of common watch crystals that oscillate at 32.768 kHz, but they are tuning fork crystals rather than bulk mode (or modern SAW), which have much higher frequencies. It would be challenging to lower the frequency of these, but perhaps you could evaporate Au or W onto the tips and reseal them to get into the audible frequencies. Much easier would be to get two crystals which beat against each other in the audio range. Even a couple of 6MHz crystals 100ppm off would be ~6kHz (temp sensitive), and you would need some driving circuits, but you might be able to hear the beat by driving an electret microphone or a really tiny tweeter coil speaker (probably need an amp though).
It's much easier to build/buy electrical rather than mechanical LC components to hit audio frequencies ~100Hz-10kHz.
There are piezo buzzers and beepers, of course.
Yes, but those normally just convert the electrical oscillations to sounds, they are not parts of the oscillators that determine the audible frequency.
In the past, audio RC oscillators or LC oscillators were used, with the former being preferred as the latter required too bulky inductors to reach so low frequencies.
Nowadays, it is usually simpler to not use any audio oscillator, but to use some microcontroller that divides the frequency of its clock until reaching the desired audio frequency.
Those can make sounds at other frequencies but they are very loud on their resonant frequency. Loud enough to damage your hearing if you're unlucky.
Nice. I'm an EE by education. We weren't taught how such a key piece of the puzzle worked.
Obligatory: https://youtu.be/wHenisSTUQY?si=oNOHR9QRCG3GYvf2
Just realised this video is referenced in the article! Didn't read it carefully.
The account of the youtube video linked in the article has been terminated apparently. Anyone know why or if there's an alternate link?
https://web.archive.org/web/20160422054642/https://www.youtu...
Now that was nice. I liked that a lot thank you for sharing.
[dead]