Thanks! Those are perfect for upgrading my consumer-grade freq counters, and my ham transceivers, and at just over $4 each. Excellent. I just ordered 6 of them, my post-Christmas present to myself. Some people don't realize that the older a Xtal oven reference is the more stable it is. I'd pay more for a used one than a new one. Ten MHz is the most useful freq, although I wouldn't mind a couple that were 5 and 25 MHz. Of course, once you have a precision reference, with a bit of work those freqs can be generated, but then you're having to play with phase noise, and I'm lazy.
The price does seem to have crept up a little since last i checked, at least from the more 'reputable' ebay sellers. Or did you find them elsewhere for 4 bucks a pop? edit: ah, i found the seller. Yeah, insanely cheap.
When evaluating oscillators, one of the best practices is to calculate Allan deviation. John Miles' TimeLab software is free and can talk to most frequency counters. It would be great to see ADEV charts for these and the other oscillators being reviewed.
I used to work as a support technician for the Engineering department and second-shift production operations at CTS corp. Our product was TCXO and VCXO oscillators for Lockheed Martin and a couple other companies - these oscillators go through aging tests lasting days, weeks or months, each one would be tested at extreme temperatures both hot and cold, frequency data would be collected every few minutes the entire time and if they were off by (X) - PPM (of the part specifications) they would be considered a fail and sent to a production technician to be torn down, repaired, and then go through the entire process again. We even made the quarts crystal wafers that would be used in the oscillators, a very interesting process in itself.
Great video - lots of happy viewers. I have used these for several projects - I upgraded my HP 5315B from a TCXO to one of these OCXOs, also a Wavetek 288 function generator upgraded from a bare crystal oscillator. They work perfectly! On some of these modules the fifth pin is a well regulated reference output voltage to be used as the supply for the compensation pot. Regards, David
I bought something (bigger) similar years ago, I didn’t realise they were probably coming from cell towers. Way more complicated inside than I would have expected for a heating circuit
Nice parts, probably measured in ppb accuracy, rather than ppm. I've recovered some similar 20MHz OCXOs from some other telecom gear, I ought to make up a little board with some buffers and a divider down to 10MHz.
Cool part of the day! Could you do a video that takes one of these OCXO parts and a GPS to make a 10 MHz GPSDO? I'm thinking I'd divide the 10 MHz clock signal by ten million then use a microcontroller to line up the rising edge of the divided down OCXO output with the 1 PPS signal from the output of the GPS and vary the OCXO control voltage based on the difference in rise times. I know you can buy a GPSDO from China on the cheap however understanding one well enough to build one would be so much better.
I bought an old school OXCO out of some kind of HP microwave instrument and I calibrate that with a commercial Rubidium standard that I got on ebay for 85$ because the seller did not know what it was. I have been a freq. standard freak from time to time. I use the OXCO as an external standard for my counter and sig generator when I want better than the internal standards. I checked the Rub. std. informally with a Collins cal lab unit and yep, it is right on. But thus not officially cert. Those little oscillators you got there are a boon for ham radio or hobby work but I am still suspicious of stuff from China till proven.
I bought a couple of the OSC5A2B02 and was going to build a 10MHz reference. The OCXO's puts out a nice 10MHz sinewave. I would also like to have a square wave output. I'm thinking of using a 74HC04 to produce a square wave, but I don't think it'll be 50% duty. The VinH is 3.15 volts and the VinL is 1.35 volts. Any suggestions? Is there a better way to produce a square wave from a sinewave?
Some OCXO oscillators could be found for 3.3 Volts DC power, while TCXO variant is a bit more efficient, going as low as 1.8 Volts DC. TCXO and OCXO oscillators are quite good for as a source of precision clock source. Especially MEMS oscillator version which is much better than Quartz in terms of shock resistance.
yea! I come inside & play! thanks 4 breaking down the super freak tech! so the take away is, they R cheap? cause I prefer the tiny crystal osc! thanks 4 the peep show!!
Yeah, very stable devices. Mine puts out the center frequency (10 MHz) at 1.6 V input. I'm imagining that it used to be around 2.5 V when it was new, but don't really know. Maybe 1.6 V is normal to get the center frequency?
No, I didn't say I have a Rhodium clock. I currently have the Aurender A30 and an Aurender N20 I ordered yesterday of which I'm trading the A30 in for which can accept an external clock like the MC20, and yes it's 30k . The N20 has an OXCO oven controlled clock. In order to benefit from the MC20 that would be for a way more resolving system.
You mentioned a Rubidium oscillator. It doesn't look like you've done an episode on it yet. Can you show us that calibration process you were talking about?
Using regular IC's (not a microcontroller), what is the easiest way to divide the 10MHz down to 1 PPS (1 Hz)? I am trying to find an easier (and cheaper) way than using 7 divide by 10 counters. Thank you.
@@IMSAIGuyThanks, but I don't see how this chip can be configured to divide by tens, only by 2's. That's useful for a 32.768 KHz oscillator, but not one outputting an even number of Hz such as 10 MHz. If I'm incorrect, and it can be so configured, please let me know.
@@AB-zw5xx that's why I said might. you are asking for 10^7 counter which does not exist. if you can get there part way in 2^20 then get it back to powers of 10. the easy way is a $0.50 micro controller as you know.
They can have either sine or square wave output depending on the part number. The square wave ones that I have used have a very clean output when properly terminated. They appear to be designed for a 50 ohm load. David
So, did you manage to identify the IC? Looks like it's a microcontroller. Judging from the datasheet I found, this is likely an SC cut crystal, so yeah, good one!
the pot has a very narrow range, just about enough to comepnsate for ageing, but yes the pot should be good quality and itself kept at fairly constant temp.
The pot gets you close then the oven and crystal circuitry takes over for the fine stuff so the pot can move about a little but nothing the internals can't deal with....cheers..
True, and the supply voltage to the pot is also important. I use to low TC resistors and a 500 ohm pot mounted next to the module. By selecting resistor values the ten turn pot ends up having an adjustment range of 1 to 2 hertz total. I have checked them against GPS using an HP 5372A Frequency and Time Interval Analyzer and they are very stable.
@@davidv1289 Yep, that's the way to do it. These OCXOs do have a Vref output but the datasheet doesn't say what Voltage. The Vc input is specced at 0 to +4V so I'd expect Vref=4V. I've ordered a couple so I'll find out when (and if) they turn up. I feel a DIY GPSDO coming ...
I wonder was it output voltage is...? It seems like everyone wants to upgrade their 10 Mhz oscillator in various radios and instruments... It seems like one should take care to match the oscillator to their instrument.
Nice...I was curious...& looked it up...the output is HCMOS...square wave...load 15 pf...Voh greater or equal to 4.5 VDC, Vol less than or equal to 0.5 VDC It looks like you started an eBay run.... They are now going like hotcakes... Our VHF radio...used one..era 1982.. 3.2 Mhz...with +/- 1 ppm....from -55 to 125 degrees C...and was purchased for $550... Times have changed
Thanks! Those are perfect for upgrading my consumer-grade freq counters, and my ham transceivers, and at just over $4 each. Excellent. I just ordered 6 of them, my post-Christmas present to myself. Some people don't realize that the older a Xtal oven reference is the more stable it is. I'd pay more for a used one than a new one. Ten MHz is the most useful freq, although I wouldn't mind a couple that were 5 and 25 MHz. Of course, once you have a precision reference, with a bit of work those freqs can be generated, but then you're having to play with phase noise, and I'm lazy.
The price does seem to have crept up a little since last i checked, at least from the more 'reputable' ebay sellers. Or did you find them elsewhere for 4 bucks a pop? edit: ah, i found the seller. Yeah, insanely cheap.
When evaluating oscillators, one of the best practices is to calculate Allan deviation. John Miles' TimeLab software is free and can talk to most frequency counters. It would be great to see ADEV charts for these and the other oscillators being reviewed.
I used to work as a support technician for the Engineering department and second-shift production operations at CTS corp.
Our product was TCXO and VCXO oscillators for Lockheed Martin and a couple other companies - these oscillators go through aging tests lasting days, weeks or months, each one would be tested at extreme temperatures both hot and cold, frequency data would be collected every few minutes the entire time and if they were off by (X) - PPM (of the part specifications) they would be considered a fail and sent to a production technician to be torn down, repaired, and then go through the entire process again.
We even made the quarts crystal wafers that would be used in the oscillators, a very interesting process in itself.
Great video, thanks for showing the insides. I bought two and now have one built into a GPS disciplined oscillator. Works great.
Great video - lots of happy viewers. I have used these for several projects - I upgraded my HP 5315B from a TCXO to one of these OCXOs, also a Wavetek 288 function generator upgraded from a bare crystal oscillator. They work perfectly! On some of these modules the fifth pin is a well regulated reference output voltage to be used as the supply for the compensation pot. Regards, David
I bought something (bigger) similar years ago, I didn’t realise they were probably coming from cell towers. Way more complicated inside than I would have expected for a heating circuit
Five sold in the last hour on one sellers Ebay! Thanks for sharing! I always get something out of the "chip of the day". I hope Santa was good to you!
Nice parts, probably measured in ppb accuracy, rather than ppm.
I've recovered some similar 20MHz OCXOs from some other telecom gear, I ought to make up a little board with some buffers and a divider down to 10MHz.
Have you got a schematic for your test setup for this module? Thanks.
Yeah, they are cheap and, indeed, great oscillators! I have few of them.
Thanks for opening one up.. 👍
Man those are ginormous. I love big parts I ought to pick up a few.
I built my own GPSDO around this particular OCXO; I’ve found them to be very stable also!
Cool part of the day!
Could you do a video that takes one of these OCXO parts and a GPS to make a 10 MHz GPSDO? I'm thinking I'd divide the 10 MHz clock signal by ten million then use a microcontroller to line up the rising edge of the divided down OCXO output with the 1 PPS signal from the output of the GPS and vary the OCXO control voltage based on the difference in rise times.
I know you can buy a GPSDO from China on the cheap however understanding one well enough to build one would be so much better.
I made a nice external 10mhz source and it is spot on and very stable-great find!!!
As for the circuit I built for my OXCO, it was sine out. My counter accepts sine but sig gen wants a dang TTL signal so had to add extra circuitry.
I've used one to modify my cheap frequency counter. It is much more accurate now.
I bought an old school OXCO out of some kind of HP microwave instrument and I calibrate that with a commercial Rubidium standard that I got on ebay for 85$ because the seller did not know what it was. I have been a freq. standard freak from time to time. I use the OXCO as an external standard for my counter and sig generator when I want better than the internal standards. I checked the Rub. std. informally with a Collins cal lab unit and yep, it is right on. But thus not officially cert. Those little oscillators you got there are a boon for ham radio or hobby work but I am still suspicious of stuff from China till proven.
Thanks, I just ordered 6 for $10 shipped.
I bought a couple of the OSC5A2B02 and was going to build a 10MHz reference. The OCXO's puts out a nice 10MHz sinewave. I would also like to have a square wave output. I'm thinking of using a 74HC04 to produce a square wave, but I don't think it'll be 50% duty. The VinH is 3.15 volts and the VinL is 1.35 volts. Any suggestions? Is there a better way to produce a square wave from a sinewave?
use a 74HC14. there are very few products need a true 50%. usually they just need a 10 MHz repeating pattern
Some OCXO oscillators could be found for 3.3 Volts DC power, while TCXO variant is a bit more efficient, going as low as 1.8 Volts DC. TCXO and OCXO oscillators are quite good for as a source of precision clock source. Especially MEMS oscillator version which is much better than Quartz in terms of shock resistance.
Try shoving it into a block of polystyrene, to see how much will power drop.
With a high resolution counter, one can see the difference in frequency between horizontal and vertical position.. The effect of gravity?
yes, you can see changes due to the mass of the crystal being effected by gravity in different orientations. mechanical watch shows this too
yea! I come inside & play! thanks 4 breaking down the super freak tech! so the take away is, they R cheap? cause I prefer the tiny crystal osc!
thanks 4 the peep show!!
Us Audiophile love these 🌺 ♥️
Yeah, very stable devices. Mine puts out the center frequency (10 MHz) at 1.6 V input. I'm imagining that it used to be around 2.5 V when it was new, but don't really know. Maybe 1.6 V is normal to get the center frequency?
So you have an external clock with a Rhobidium clock going to your DAC? Who makes it? The one I've been eyeing is the MC20 by Aurender and it's $30k.
DIY using a used Fe-5680A off eBay
No, I didn't say I have a Rhodium clock. I currently have the Aurender A30 and an Aurender N20 I ordered yesterday of which I'm trading the A30 in for which can accept an external clock like the MC20, and yes it's 30k . The N20 has an OXCO oven controlled clock. In order to benefit from the MC20 that would be for a way more resolving system.
You mentioned a Rubidium oscillator. It doesn't look like you've done an episode on it yet. Can you show us that calibration process you were talking about?
CuriousMarc has a couple of videos on rubidium clocks
Using regular IC's (not a microcontroller), what is the easiest way to divide the 10MHz down to 1 PPS (1 Hz)? I am trying to find an easier (and cheaper) way than using 7 divide by 10 counters. Thank you.
SN74LV8154N might help
@@IMSAIGuyThanks, but I don't see how this chip can be configured to divide by tens, only by 2's. That's useful for a 32.768 KHz oscillator, but not one outputting an even number of Hz such as 10 MHz. If I'm incorrect, and it can be so configured, please let me know.
@@AB-zw5xx that's why I said might. you are asking for 10^7 counter which does not exist. if you can get there part way in 2^20 then get it back to powers of 10. the easy way is a $0.50 micro controller as you know.
no, it only gets you part way
Very nice. I just bought a bunch.
Thanks, what is the working temperature of the crystal ?
What does the output look like? I was surprised that an old TCXO I had has more of a saw tooth waveform instead of the expected square-wave.
They can have either sine or square wave output depending on the part number. The square wave ones that I have used have a very clean output when properly terminated. They appear to be designed for a 50 ohm load. David
1:11 I'm trying really hard not to make Inglourious Basterds joke here :)
nein! das ist verboten
Gaetano Stravenue
So, did you manage to identify the IC? Looks like it's a microcontroller.
Judging from the datasheet I found, this is likely an SC cut crystal, so yeah, good one!
Aren't they only going to be as stable as the external pot that is used to adjust it?
the pot has a very narrow range, just about enough to comepnsate for ageing, but yes the pot should be good quality and itself kept at fairly constant temp.
The pot gets you close then the oven and crystal circuitry takes over for the fine stuff so the pot can move about a little but nothing the internals can't deal with....cheers..
True, and the supply voltage to the pot is also important. I use to low TC resistors and a 500 ohm pot mounted next to the module. By selecting resistor values the ten turn pot ends up having an adjustment range of 1 to 2 hertz total. I have checked them against GPS using an HP 5372A Frequency and Time Interval Analyzer and they are very stable.
@@andymouse The 'internals' aren't doing anything smart. If the pot setting changes then the output frequency changes.
@@davidv1289 Yep, that's the way to do it. These OCXOs do have a Vref output but the datasheet doesn't say what Voltage. The Vc input is specced at 0 to +4V so I'd expect Vref=4V. I've ordered a couple so I'll find out when (and if) they turn up. I feel a DIY GPSDO coming ...
I wonder was it output voltage is...?
It seems like everyone wants to upgrade their 10 Mhz oscillator in various radios and instruments...
It seems like one should take care to match the oscillator to their instrument.
remember that adding a voltage divider to decrease the output or a simple amp to increase the voltage will not effect the frequency stability.
Nice...I was curious...& looked it up...the output is HCMOS...square wave...load 15 pf...Voh greater or equal to 4.5 VDC, Vol less than or equal to 0.5 VDC
It looks like you started an eBay run....
They are now going like hotcakes...
Our VHF radio...used one..era 1982..
3.2 Mhz...with +/- 1 ppm....from -55 to 125 degrees C...and was purchased for $550...
Times have changed
I have started to look for these oven controlled 10mhz oscillator modules.
Thanks.
De VU2RZA
Deja Inlet
Herta Lake
Are the crystals natural or synthetic?
they are man made. large crystals (salami sized) are grown and then sliced up.
Cool
Cheaper and smaller is death of technology.
HA!!!!!!
"ONE...of the good things coming out of china"
Agreed,....there's not much good come out of there,...but you found something.