I purchased, possibly the same ocxo's a few years ago, with the intention of making a pulse generator. I haven't had a chance to try them yet. Thanks for sharing, very useful information.
I had considered getting a gpsdo as an exxternal reference for my scopes and function generators. Having seen this and the price , that idea is out the window and it's hello ebay and CTI. Thanks for that.
Thanks for the hints. I'll certainly takes those into an account when I have an application that really needs them. This was just supposed to be the $3 look into the $3 oscillator.
If you get a neo-7m / neo-m8n GPS module you can program them to spit out 10mhz as well, buffer them etc. and use to compare against the OCXO. They'll generally need an active antenna if inside. GPS + OCXO + MCU = your own diy GPSDO
Matt You have to be careful when using the NEO-7 (and probably the NEO-8) in this way. The on-board GPSDO runs at 48 Mhz and dividing down to 10 Mhz results in a pulse train of asymmetric pulses - on an oscilloscope it appears to be jitter but is in fact a recurring pattern. Division by an integer ( ie 4 for 12 Mhz or 6 for 8 Mhz) results in a normal 50 % duty cycle output. Depending on your application, the "jitter" may cause problems. Regards, David
Not bad for such a cheap reference. I'm quite partial to the HP 10544A and HP 10811A references myself, however being so large they require lots of enclosure space.
The Vectron C4550A1 series OCXO costs about the same and is in the same package but also includes a stabilized voltage reference output (pin 4) that can be used to supply the frequency adjustment voltage for even better temperature stability. Regards, David
More accurately, some of them do - I have a bunch of those Vectron OCXOs, all with the same -0213 suffix and a have a couple with fairly low serial numbers (< 10000) and those have the Vref output and a bunch more with higher serial numbers (40000 to 176000) that don't. So a reasonable guess is that the reference output was not part of the spec and was removed at some point between 10000 and 40000. It would be interesting to know if your oscillator is another -0213, and if so what it's serial number is, since that might help to narrow the range.
@@TrimeshSZ Good information, thanks. It seems from reading the spec sheet on these that the Vref output is an option and the device can be ordered with or without it.
@@TrimeshSZ Interesting. I have five Vectron C4550A1-0213 OCXOs - three installed in test equipment and two tested and awaiting projects. All of them have internal reference supplies. I also checked the latest spec. sheet from Vectron and it says the C4550 series have the internal reference. My units are s/n 11656 (0908), 23675 (0936), 178480 (0951), 186032 (1029) and 189616 (1032). Maybe you got some damaged ones? Regards, David
@@davidv1289 I don't think they are faulty - I chopped one of the "No Vref" parts open with a Dremel and the components feeding the Vref pins were simply not installed. I also just checked again to make sure I wasn't hallucinating and got the same results #9584 (0908 - Has Vref), #41240 (0943 - No Vref), #41818 (0943 - No Vref), #176570 (0950 - No Vref), #192451 (1034 - No Vref) - I also have a couple more low serial number parts that are in equipment and have the Vref output.
Chunky, power hungry ? Try running a Rubidium oscillator 😂 I bought a bunch of these oxco's as a substitute for my rubidium oscillators especially now that electricity prices are going thru the roof! Nice video!
@@uni-byte I have several that I bought years before everyone got on the bandwagon and the prices really went thru the roof. My main one which has distribution for 8 instruments draws about 50-60 watts. It's so warm that I have a fan blowing across it and it is attached to a large copper heat sink using a large piece of double sided flexible heat sink material. I also have a 15MHz Lucent unit that consumes significantly more power - I rarely power it up because of that.
For size and lower power consumption I would agree with you, but for stability and accuracy an OCXO has a TCXO beat in a big way. As a frequency standard the OCXO is much better. A rubidium time base or a GPSDO would be better still, but you are not going to get either of them for $3.00.
TCXOs essentially drop cycles according to an internally programmed temperature compensation look-up table. Sudden phase and/or period drops are not a good thing. OCXOs are just a very very stable crystal oscillator. No phase or period jumps like a TXCO. Horses for courses. For lab use, who would care about a few watts of power consumption? In many applications, yes, a TXCO is the obviously choice. Blanket "better" is not the correct description.
I purchased, possibly the same ocxo's a few years ago, with the intention of making a pulse generator. I haven't had a chance to try them yet. Thanks for sharing, very useful information.
Thank you!
I had considered getting a gpsdo as an exxternal reference for my scopes and function generators. Having seen this and the price , that idea is out the window and it's hello ebay and CTI.
Thanks for that.
You're welcome. But please be aware that you need some way to calibrate these. Once that's done they are rock solid and teh bargain of the decade.
i put 1 in my Counter with a 2.5V reference voltage to a trim pot works good!
They are pretty awesome!
I think there are a few major things to measure additionaly - CW power, phase noise and 2nd harmonic
Thanks for the hints. I'll certainly takes those into an account when I have an application that really needs them. This was just supposed to be the $3 look into the $3 oscillator.
@@uni-byte yep, i understand , for 3 bucks it looks very good .
Thanks, that was interesting
You are very welcome!
Some of the Chinese OCXOs are from government owned factory. They are made for military or science purposes.
Interesting. That would explain their high quality.
On bottom right, it listed the 54th Institute which is China Electronics Technology Group Corporation No.54 Institute. （CTI）
If you get a neo-7m / neo-m8n GPS module you can program them to spit out 10mhz as well, buffer them etc. and use to compare against the OCXO. They'll generally need an active antenna if inside. GPS + OCXO + MCU = your own diy GPSDO
I actually found a GPSDO for under $100 and it's on it's way.
Matt You have to be careful when using the NEO-7 (and probably the NEO-8) in this way. The on-board GPSDO runs at 48 Mhz and dividing down to 10 Mhz results in a pulse train of asymmetric pulses - on an oscilloscope it appears to be jitter but is in fact a recurring pattern. Division by an integer ( ie 4 for 12 Mhz or 6 for 8 Mhz) results in a normal 50 % duty cycle output. Depending on your application, the "jitter" may cause problems. Regards, David
Not bad for such a cheap reference. I'm quite partial to the HP 10544A and HP 10811A references myself, however being so large they require lots of enclosure space.
Everything from HP is the right stuff .. but it has to HP written on it. None of that Agilent or Keysight rubbish!
The Vectron C4550A1 series OCXO costs about the same and is in the same package but also includes a stabilized voltage reference output (pin 4) that can be used to supply the frequency adjustment voltage for even better temperature stability. Regards, David
Sounds interesting. I'll have to look into them. Thanks!
More accurately, some of them do - I have a bunch of those Vectron OCXOs, all with the same -0213 suffix and a have a couple with fairly low serial numbers (< 10000) and those have the Vref output and a bunch more with higher serial numbers (40000 to 176000) that don't. So a reasonable guess is that the reference output was not part of the spec and was removed at some point between 10000 and 40000. It would be interesting to know if your oscillator is another -0213, and if so what it's serial number is, since that might help to narrow the range.
@@TrimeshSZ Good information, thanks.
It seems from reading the spec sheet on these that the Vref output is an option and the device can be ordered with or without it.
@@TrimeshSZ Interesting. I have five Vectron C4550A1-0213 OCXOs - three installed in test equipment and two tested and awaiting projects. All of them have internal reference supplies. I also checked the latest spec. sheet from Vectron and it says the C4550 series have the internal reference. My units are s/n 11656 (0908), 23675 (0936), 178480 (0951), 186032 (1029) and 189616 (1032). Maybe you got some damaged ones? Regards, David
@@davidv1289 I don't think they are faulty - I chopped one of the "No Vref" parts open with a Dremel and the components feeding the Vref pins were simply not installed. I also just checked again to make sure I wasn't hallucinating and got the same results #9584 (0908 - Has Vref), #41240 (0943 - No Vref), #41818 (0943 - No Vref), #176570 (0950 - No Vref), #192451 (1034 - No Vref) - I also have a couple more low serial number parts that are in equipment and have the Vref output.
The adjustment pin is not 0-5Vcc... It is 2V +/- 2V... So, the max is 4V. See the data sheet to confirm. A negative adjustment is below the 2V center.
All very true but it was handy to just bodge the pot to the 5V.
How accurate is the counter?
It has a GPSDO as a time base. So it should be quite good.
I have a jar of them 😅
I must be lost in the past since TXCO was a big deal when I was a tech years ago.
Still are. Not as accurate as an OCXO, but a TXCO is cheaper, much smaller and is suitable for battery powered projects.
Chunky, power hungry ? Try running a Rubidium oscillator 😂 I bought a bunch of these oxco's as a substitute for my rubidium oscillators especially now that electricity prices are going thru the roof! Nice video!
Those rubidium standards are expensive in more ways than one. That's for sure. About 60W or so?
@@uni-byte I have several that I bought years before everyone got on the bandwagon and the prices really went thru the roof. My main one which has distribution for 8 instruments draws about 50-60 watts. It's so warm that I have a fan blowing across it and it is attached to a large copper heat sink using a large piece of double sided flexible heat sink material. I also have a 15MHz Lucent unit that consumes significantly more power - I rarely power it up because of that.
Awesome video! Like your setup! Hope to win the HP freq counter, JK!! :) ~Jack, VEG
Thanks. That HP counter is a real nice bit of kit. I just need to find a 7-segment LED display module to fit it. I'll keep looking.
Annabelle Plaza
Not quite sure I understand.
6699 Fausto Views
TCXO would be better.
For size and lower power consumption I would agree with you, but for stability and accuracy an OCXO has a TCXO beat in a big way. As a frequency standard the OCXO is much better. A rubidium time base or a GPSDO would be better still, but you are not going to get either of them for $3.00.
For accuracy, no it wouldn't. For power draw, yes.
TCXOs essentially drop cycles according to an internally programmed temperature compensation look-up table. Sudden phase and/or period drops are not a good thing.
OCXOs are just a very very stable crystal oscillator. No phase or period jumps like a TXCO.
Horses for courses. For lab use, who would care about a few watts of power consumption? In many applications, yes, a TXCO is the obviously choice.
Blanket "better" is not the correct description.