When you want to shoot a instrument calibration or when you want to make adjustments, I give you a tip that I learned from one old school repairman uses a normal mirror mounted in front of the display - carefully! image is reversed - and you can calibrate calm with no problems
The counter on the Rigol is deeply flawed as it appears to use a 2^n gate period instead of 10^n internally, and gives several completely meaningless digits at the end, which it should blank.
Rubidium oscillators are not primary sources like cesium oscillators are. This means you have to tune them on frequency. They are very stable, but they aren't "always bang on". And that's what the GPS board is for - it disciplines the rubidium frequency to match that coming from GPS (which is cesium based).
+Star Gazer What do you mean by 'disciplining' the rubidium frequency? You can't "discipline" a rubidium frequency. The only thing I can think of is - not disciplining, but synchronising. You sure you're not talking about disciplining a voltage controlled crystal oscillator? The other thing is..... I haven't been able to find anywhere the meaning of a 'disciplined oscillator'. I'm assuming it means 'synchronised' or 'phase-locked to'. Pretty weird how there appears to be no formal definition of 'disciplined oscillator' out there. Not good when these two words are used blindly.
+Kenny A "disciplined" clock/frequency source is one which is actively and almost continuously adjusted in both frequency and phase to match one or more better time/frequency sources, such that the disciplined device provides a very good extrapolation of the outside time source, which may only be available some of the time. For example, the GPS radio signal provides the time and date every 1.5s, but only to the extent it actually picks up enough satellite signals to compute the radio delay from space through atmospheric disturbances and past the background noise, as a close US ally, that GPS receiver board might even be one of the classified models that sync to the secret hires part of the signal. Those GPS satellite clocks up there are themselves disciplined by adjustment commands from the relevant US military team to closely match the official US IAT reference clocks. National time labs then compare their deltas from the GPS signal to help establish the weighted average which is the official International Atomic Time. (UTC is IAT minus a whole number of leap seconds).
The output of a rubidium oscillator is typically a 10MHz signal that is phase locked to the 6.834GHz excitation frequency that is tuned to resonance with the excited rubidium gas in the physics package. While this excitation frequency is very stable once set, the resonance frequency of rubidium is variable and must be tuned by applying a magnetic field to the rubidium chamber. You have to tune a rubidium oscillator to get it 'on frequency'. You normally do this by 'disciplining' the source by periodically or continuously comparing it to another (better) source such as a cesium oscillator via GPS, learning the rate at which the rubidium changes with age and temperature, and adjusting for those changes automatically. As a practical matter, there is a voltage controlled tuning input on rubidium oscillators that varies the magnetic field applied to the chamber, and thus the frequency of the 10 MHz output. Disciplined oscillator simply means an oscillator whose output is being corrected over long time scales by comparing it to another oscillator. It is more than simply synchronizing the phase or frequency. A DO learns the rate of change of the oscillator and continues making those corrections even when the reference is no longer available. So a GPS disciplined rubidium oscillator will continue to make corrections to its rubidium even when the GPS signal is no longer available to sync to, based on the change rate it learned while GPS was available. This generally increases the stability during the 'holdover' period when the oscillator is free running. The actual oscillator being disciplined in a GPSDO can be of many different types, such as quartz based, or rubidium. Quartz having better short term stability (seconds to minutes), and rubidium having better long term stability. If the GPS signal is always available to discipline the oscillator, then quartz is fine. If you need longer term holdover characteristics, such as for cell site synchronization, then rubidium is typically used.
Would be true if we're just talking about the GPS module. What you may have missed is that the 1 PPS out is from the atomic standard. And unless the 1 PPS is a loop-through from the GPS, which I highly doubt, the thing is darn accurate as it is derived from the atomic standard. All my atomic standards work that way.
You need to change the gate time on the counters to achieve a higher accuracy. Are you terminated correctly on both ends? And the rubidium reference *will* drift over time. The GPS timebase is *much* more accurate.
The reason the PCB is marked both 'Javad' and 'Topcon' is because a large portion of the high-accuracy GPS receiver industry can be traced back to... HP. HP developed a GPS receiver in the early 80's, but they shelved the project because it was unclear if the US Navy would get the funding required to complete the GPS constellation. In 1982, HP sold their breadboard prototype and stack of documentation to Charlie Trimble for around $50,000. One of Trimble's employees, Javad Ashjaee was the primary software developer for Trimble's first survey grade product, the 4000S. Javad left Trimble to form Ashtech in 1987, another producer of high-accuracy GPS equipment (acquired by Trimble in 2011). Then in 1998, Javad Ashjaee formed JAVAD Positioning Systems, which he sold to Topcon in 2000, forming the foundations for their GPS business. Javad's non-compete with Topcon ran out in 2007, and he formed JAVAD GNSS, which is still in business. Sadly, Javad Ashjaee passed away from COVID-19 earlier this year (2020).
Not likely. That kind of gear is likely to be consistently out in one direction or the other if it's out of cal. There's one oscillator for the whole thing, which is then divided or multiplied as needed. It's not a cheap piece of gear with multiple RC oscillators for the different ranges, like seen on some multimeters. Those things drift all over the place...
You've got high quality rubidium standard oscilator with distribution gear. Why didn't you just connect all of your stuff to one high-quality 10Mhz clock source? And as far as I remember you have second rubbidium frequency standard. Could you please compare them?
You need to lab recalibrate that HP unit. Please believe usRb (Rubidium frequency standard users when we say that IF the Rb unit is checked against a facility, college or government lab that has a primary atomic clock standard, the Rb (which are secondary frequency standards) are good and accurate out to 11 places right of the decimal point. Their drift when locked is less than .0001ppm per every 50 years give or take.
In the video about the HP5061A Atomic Clock we saw the cesium oscillator worked in conjunction with a ovenized crystal oscillator for short term stability. Does this rubidium standard need a crystal oscillator as well?
In theory somewhere in the board it could be a bnc attach scocket for the reference frequency entrance that you use if you buy that option. Try to connect to it the 10MHz refererence to that mini bnc connector and you got a calibrated instrument.
I was looking for a 53131 without options to replace my old PM6654, but after your video i am not sure that i should buy one of this devices :-) By the way, i do my measurements with some old but nice devices like the R&S CMS 52 for RF, my old PM6654 and for frequency reference i use a thunderbolt GPS, some OCXOS and an FE RB clock. At the moment try to repair my favourite device, a SCHOMANDL MES-1000. Thank you for your great and interesting videos !
I expect the frequency counter was always intended to run with the external frequency reference you got with it. You should thank your standards org for not wasting money on the counter version with the better internal reference.
I've got my hands on some old Polish "quartz frequency standard". Loud as hell, has some fail crusty custom switchmode power supply from a handmade transformer that drifts all over the place and eventually crashes if you don't put enough load on it. Oh boy, when the crystal was reaching the temperature and the oven switched off the whole thing crashed. Had to make special 12V rail for the crystal and put a dummy load on the old crystal rail to keep the old supply happy.. Good enough for Poland!
Great video Dave, it would be interesting to see HP counter and measurement accuracy with frequency standard you scored on eBay while ago. The one that was used in GSM base stations.
What you probably have is a GPSDO. Have a look at GPS_disciplined_oscillator at wikipedia for typical way GPS and Rubidium are combined. When GPS is available the rubidium is locked to the GPS 10MHz, the distributed 10 MHZ comes from the rubidium for low phase noise, but the GPS is used to drag the rubidium to the GPS 10 MHz.When GPS not available, the rubidium then is the only reference (which is still pretty good). This is called holdover (till GPS is re-acquired)
The 1PPS signal from the gps unit being out of time with the rubidium clock...doesn't that scream "I have a gps signal"? You don't need a full gps lock to get the time code, you need just one packet from one satellite; long before you have lock you get timecode.
I thought high-end frequency counters like this are supposed to take in a 10MHz signal from a disciplined standard. In't that the whole point to having a frequency standard on the rack?
Attempting to call a rubidium oscillator 'out of spec' while using a 10.5 digit counter (without a high stability option), and on the default 100 ms gate time... You need to set the gate time to 1 second. Set your instrument up properly, and then call it out of spec.
12:07 lithium-thionyl chloride battery, judging from the package. Wondering how this GPS module is performing bet its faster then off the shelf receivers.
Actually not true... at least from my own experiences, perhaps your gear works different... Until the GPS get 1PPS sync from the satellites cesium clocks it runs on the recievers oscillator, just an ordinary tcxo... As I said - at least on the gps's I use (for 1PPS control), and from what I've read.
congrats on the awesome deal. You may pay through the nose for software and the like over there, but at least you have the occasional bargain to offset that nonsense. I'm looking slightly green over here.
"Good enough for Australia", I hope thats a dig at our ability to tolerate cheap shit from other countries, Namely 'china'. And not the shit we make/design in Australia :P
mrfurball maybe not complete gibberish, but completely without context to the video. Also, your screen name is not actually being mentioned, that's "Fussball", the German word for soccer.
mrfurball actually, this is pretty much gibberish. I'll try to translate: "I have no video, just TV not really you know back then more soccer, skiing, bowling 3 sport now not anymore. More light mountaineering Gotti [which I presume refers to the name]". So yeah, pretty weird stuff.
The ERG board in the Agilent meter supplies the HV needed for the VFD display.
When you want to shoot a instrument calibration or when you want to make adjustments, I give you a tip that I learned from one old school repairman uses a normal mirror mounted in front of the display - carefully! image is reversed - and you can calibrate calm with no problems
Ionut Catalin Patitoiu that's the same way we used to do the color convergence adjustment on the old analog TV sets back in the day.
At 14:22 those Dallas parts are actually 1-Wire temperature sensors.
Your 1 PPS out does not depend on the GPS. The only thing lack of GPS does is that your 1 PPS pulse is not synced with the UTC time second.
The counter on the Rigol is deeply flawed as it appears to use a 2^n gate period instead of 10^n internally, and gives several completely meaningless digits at the end, which it should blank.
Rubidium oscillators are not primary sources like cesium oscillators are. This means you have to tune them on frequency. They are very stable, but they aren't "always bang on". And that's what the GPS board is for - it disciplines the rubidium frequency to match that coming from GPS (which is cesium based).
+Star Gazer much better an active maser :)
+Star Gazer What do you mean by 'disciplining' the rubidium frequency? You can't "discipline" a rubidium frequency. The only thing I can think of is - not disciplining, but synchronising. You sure you're not talking about disciplining a voltage controlled crystal oscillator?
The other thing is..... I haven't been able to find anywhere the meaning of a 'disciplined oscillator'. I'm assuming it means 'synchronised' or 'phase-locked to'. Pretty weird how there appears to be no formal definition of 'disciplined oscillator' out there. Not good when these two words are used blindly.
+Kenny A "disciplined" clock/frequency source is one which is actively and almost continuously adjusted in both frequency and phase to match one or more better time/frequency sources, such that the disciplined device provides a very good extrapolation of the outside time source, which may only be available some of the time. For example, the GPS radio signal provides the time and date every 1.5s, but only to the extent it actually picks up enough satellite signals to compute the radio delay from space through atmospheric disturbances and past the background noise, as a close US ally, that GPS receiver board might even be one of the classified models that sync to the secret hires part of the signal. Those GPS satellite clocks up there are themselves disciplined by adjustment commands from the relevant US military team to closely match the official US IAT reference clocks. National time labs then compare their deltas from the GPS signal to help establish the weighted average which is the official International Atomic Time. (UTC is IAT minus a whole number of leap seconds).
The output of a rubidium oscillator is typically a 10MHz signal that is phase locked to the 6.834GHz excitation frequency that is tuned to resonance with the excited rubidium gas in the physics package. While this excitation frequency is very stable once set, the resonance frequency of rubidium is variable and must be tuned by applying a magnetic field to the rubidium chamber. You have to tune a rubidium oscillator to get it 'on frequency'. You normally do this by 'disciplining' the source by periodically or continuously comparing it to another (better) source such as a cesium oscillator via GPS, learning the rate at which the rubidium changes with age and temperature, and adjusting for those changes automatically. As a practical matter, there is a voltage controlled tuning input on rubidium oscillators that varies the magnetic field applied to the chamber, and thus the frequency of the 10 MHz output.
Disciplined oscillator simply means an oscillator whose output is being corrected over long time scales by comparing it to another oscillator. It is more than simply synchronizing the phase or frequency. A DO learns the rate of change of the oscillator and continues making those corrections even when the reference is no longer available. So a GPS disciplined rubidium oscillator will continue to make corrections to its rubidium even when the GPS signal is no longer available to sync to, based on the change rate it learned while GPS was available. This generally increases the stability during the 'holdover' period when the oscillator is free running.
The actual oscillator being disciplined in a GPSDO can be of many different types, such as quartz based, or rubidium. Quartz having better short term stability (seconds to minutes), and rubidium having better long term stability. If the GPS signal is always available to discipline the oscillator, then quartz is fine. If you need longer term holdover characteristics, such as for cell site synchronization, then rubidium is typically used.
Was really hoping you were going to crack open the rubidium. Talk about how the liquid works and how they are calibrated.
I really enjoy your enthusiasm. Thanks for making these videos.
Would be true if we're just talking about the GPS module. What you may have missed is that the 1 PPS out is from the atomic standard. And unless the 1 PPS is a loop-through from the GPS, which I highly doubt, the thing is darn accurate as it is derived from the atomic standard. All my atomic standards work that way.
You need to change the gate time on the counters to achieve a higher accuracy. Are you terminated correctly on both ends? And the rubidium reference *will* drift over time. The GPS timebase is *much* more accurate.
The reason the PCB is marked both 'Javad' and 'Topcon' is because a large portion of the high-accuracy GPS receiver industry can be traced back to... HP.
HP developed a GPS receiver in the early 80's, but they shelved the project because it was unclear if the US Navy would get the funding required to complete the GPS constellation. In 1982, HP sold their breadboard prototype and stack of documentation to Charlie Trimble for around $50,000. One of Trimble's employees, Javad Ashjaee was the primary software developer for Trimble's first survey grade product, the 4000S. Javad left Trimble to form Ashtech in 1987, another producer of high-accuracy GPS equipment (acquired by Trimble in 2011). Then in 1998, Javad Ashjaee formed JAVAD Positioning Systems, which he sold to Topcon in 2000, forming the foundations for their GPS business. Javad's non-compete with Topcon ran out in 2007, and he formed JAVAD GNSS, which is still in business. Sadly, Javad Ashjaee passed away from COVID-19 earlier this year (2020).
Thanks for the U.S.A call out ! great vid ! never heard of these before your vids. Education at its best ! 8)
Not likely. That kind of gear is likely to be consistently out in one direction or the other if it's out of cal. There's one oscillator for the whole thing, which is then divided or multiplied as needed.
It's not a cheap piece of gear with multiple RC oscillators for the different ranges, like seen on some multimeters. Those things drift all over the place...
You've got high quality rubidium standard oscilator with distribution gear. Why didn't you just connect all of your stuff to one high-quality 10Mhz clock source?
And as far as I remember you have second rubbidium frequency standard. Could you please compare them?
You need to lab recalibrate that HP unit. Please believe usRb (Rubidium frequency standard users when we say that IF the Rb unit is checked against a facility, college or government lab that has a primary atomic clock standard, the Rb (which are secondary frequency standards) are good and accurate out to 11 places right of the decimal point. Their drift when locked is less than .0001ppm per every 50 years give or take.
In the video about the HP5061A Atomic Clock we saw the cesium oscillator worked in conjunction with a ovenized crystal oscillator for short term stability. Does this rubidium standard need a crystal oscillator as well?
That Dallas is a digital temperature sensor if I remember correctly.
Those are actually made by FE in the USA and they are quite good.
weird that those numbers just kept going up at the end. went from 4.1 to 6.5 in just under a minute.
In theory somewhere in the board it could be a bnc attach scocket for the reference frequency entrance that you use if you buy that option. Try to connect to it the 10MHz refererence to that mini bnc connector and you got a calibrated instrument.
"good enough for Australia", ok, but what about Austria?
I was looking for a 53131 without options to replace my old PM6654, but after your video i am not sure that i should buy one of this devices :-) By the way, i do my measurements with some old but nice devices like the R&S CMS 52 for RF, my old PM6654 and for frequency reference i use a thunderbolt GPS, some OCXOS and an FE RB clock. At the moment try to repair my favourite device, a SCHOMANDL MES-1000. Thank you for your great and interesting videos !
I expect the frequency counter was always intended to run with the external frequency reference you got with it. You should thank your standards org for not wasting money on the counter version with the better internal reference.
I've got my hands on some old Polish "quartz frequency standard". Loud as hell, has some fail crusty custom switchmode power supply from a handmade transformer that drifts all over the place and eventually crashes if you don't put enough load on it. Oh boy, when the crystal was reaching the temperature and the oven switched off the whole thing crashed. Had to make special 12V rail for the crystal and put a dummy load on the old crystal rail to keep the old supply happy.. Good enough for Poland!
I really wonder why the GPS module has an FPGA and a CF card slot
Yeah that would be interresting. I see loads of cheap chinese rubidium standards on ebay, and i'm wonder how 'bad' they are compaired to a proper one.
Great video Dave, it would be interesting to see HP counter and measurement accuracy with frequency standard you scored on eBay while ago. The one that was used in GSM base stations.
Only Dave could talk for 25 minutes about a box that makes an LED blink! Just kidding, interesting video as always :)
What you probably have is a GPSDO. Have a look at GPS_disciplined_oscillator at wikipedia for typical way GPS and Rubidium are combined. When GPS is available the rubidium is locked to the GPS 10MHz, the distributed 10 MHZ comes from the rubidium for low phase noise, but the GPS is used to drag the rubidium to the GPS 10 MHz.When GPS not available, the rubidium then is the only reference (which is still pretty good). This is called holdover (till GPS is re-acquired)
GPSDO without any rubidium, only with good TCXO can give 0.1ppb precision.
The 1PPS signal from the gps unit being out of time with the rubidium clock...doesn't that scream "I have a gps signal"? You don't need a full gps lock to get the time code, you need just one packet from one satellite; long before you have lock you get timecode.
I thought high-end frequency counters like this are supposed to take in a 10MHz signal from a disciplined standard. In't that the whole point to having a frequency standard on the rack?
45*C? that is pretty warm might want to think about mounting a fan or two inside that chassis.
Just sit right back and you'll hear a tale, a tale of a fateful trip, that started on this tropic port, aboard this tiny ship!
When can one get a Broadcom® BCM47755 GNSS Location Hub IC or UBX-G7020? Timing grade.
Attempting to call a rubidium oscillator 'out of spec' while using a 10.5 digit counter (without a high stability option), and on the default 100 ms gate time... You need to set the gate time to 1 second.
Set your instrument up properly, and then call it out of spec.
I bet that GPS receiver was designed for the marine environment originally.
12:07 lithium-thionyl chloride battery, judging from the package. Wondering how this GPS module is performing bet its faster then off the shelf receivers.
Are you going to hook up all your lab gear to the standard permanently?
Yes they do, but that's what the GPS is for - drift compensation by using the satellites cesium clocks...
I don't know what any of this is but it's pretty cool
It wouldn't need to be, since it's probably not used for positioning data, but rather for time syncing.
Wouldn't, say, 50MHz be off more now?
But it's pointless, because the crap stock oscillator is just going to drift more than the SS Minnow anyway
The ERG board might be the VFD driver?
Aaah I enjoyed this video! Interesting stuff!!
Could that GPS happen to be a military-grade GPS?
Those little modules have the crystal inside
@08:41 My imagination or are those bent IC pins? (The IC above/between the two power bricks.)
Anyway, thanks for the teardown. Fascinating stuff.
Believe in the force Dave! I know you can do better with calibrating that Agilent counter :)
Me either, I like to imagine if I found one of these how long it would take me to figure out what it's suppose to do
LoL the thumbnail!!! Nice video dave, keep'em coming
great video . I need to get me a frequency standard
Jesteś zajebisty ;)
22:22 - one eye, pop eye, tongue angle technique - be my reference point, please :)
since when is there a R in Data!
Wow interesting catch! Wonder if Dave has something to say about this. :P I know there is dodgy but that looks.... well dodgy I guess.
Yey! More geekporn!
Nice score! I liked your popeye imitation haha.
Actually not true... at least from my own experiences, perhaps your gear works different...
Until the GPS get 1PPS sync from the satellites cesium clocks it runs on the recievers oscillator, just an ordinary tcxo...
As I said - at least on the gps's I use (for 1PPS control), and from what I've read.
Nearly half an hr of high pitched stating the obvious STROOOTH!!!
Gilligan's Island reference, nice :D
Good ol' Philips quality. :)
congrats on the awesome deal. You may pay through the nose for software and the like over there, but at least you have the occasional bargain to offset that nonsense. I'm looking slightly green over here.
Niiiice thumbnail.
Mannn, that's a full server PSU :P
uh? Super Crappy 4700/35 electrolytics capacytors in the Rubidium cell ..
Sometimes I watch your videos and know not what the fuck is being said for 95% of it. I like it though.
I see the tongue in the right angle.. I click
19:32 i expected a hamburger in there
Nah, not enough pins for that.
made in USA USA! USA!
Cool vids! Your voice reminds me of this Stewie from Family guy bit: ?v=9D_cKZ_dhdo#t=11s. Just random observation.
22:22 OMG, I had seen ugly faces, but this is out of range !!!
i thought it is from china
lol... good enough for australia!
"Good enough for Australia", I hope thats a dig at our ability to tolerate cheap shit from other countries, Namely 'china'.
And not the shit we make/design in Australia :P
Delta make shit psu's. Agilent cheaped out.
Ich habe kein Video ja nur Fernsehen nicht besonders Damals mehr Fußball, Schifahren, Kegel 3 Sport jetzt nicht mehr . Mehr leicht Bergwanderin Gotti
If you write stuff in German to an English-speaking community, at least make it understandable to native German speakers!
+Joshua Wiedekopf (LtSurgekopf) so that's complete gibberish? I'm curious cause it seems my youtube nickname is mentioned, I wonder what that is about
mrfurball maybe not complete gibberish, but completely without context to the video. Also, your screen name is not actually being mentioned, that's "Fussball", the German word for soccer.
mrfurball actually, this is pretty much gibberish. I'll try to translate: "I have no video, just TV not really you know back then more soccer, skiing, bowling 3 sport now not anymore. More light mountaineering Gotti [which I presume refers to the name]".
So yeah, pretty weird stuff.
Also, I tried to be faithful to the sentence structure used by the OP, which is just horribly horribly wrong.
Ich kein Elektrik verschiedene Film Computer ich war jung 1935 kein Computer !
22:51 Did he say "Half a bee's dick"?! Awesome.