To clarify some slight confusion in the vid over terms for osc types : TCXO : temperature compensated - simple analogue circuit to compensate for ambient temp DTCXO : digitally temperature compensated - as above but potentially more accurate as can track xtal characteristic more accurately over the temp range,e.g. from a lookup table. OCXO :ovened - holds the xtal at a fixed temperature
I don't understand why you should cool or use a metal cover for the oven. It has a STABILIZED temperature of let's say 50 degrees celsius. 1. no plastic box will melt at 50 degrees 2. If you cool it or use a heatsink for it you will only increase the power consumption from the mains because the thermostat inside will try to keep the crystal inside at 50degrees. I believe it is best to keep it as isolated as possible from any temperature variation, may it be sink or source of heat.
@Zadster I've never seen that, and no one else reports it. As RUclips transcodes all uploaded videos, I suspect either everyone would see it, or it's just you?
I bought one of these years ago and used it to calibrate oscillators I used in experimental GSM cells. This is was over 10 years ago. Things were complicated back then, but fortunately now there are a lot of affordable GPS frequency standards available.
Brilliant! Love the videos and your style. I've discovered a lot of great things via your videos and this one came up when I was searching for info about frequency standards. Keep up the great work!
The info I saw seems to indicate the Rb is used in the gas phase, much like mercury in a fluorescent tube. At standard pressure, Rb vaporizes at 688 degrees, so heating something to that temp. electrically is probably what's taking SOOOO much current. Maybe the boiling point of Rb can be reduced some by lowering the pressure inside the lamp, much like boiling water at room temp. with a thick bell jar and a vacuum pump; I don't know.
Why is the 10 MHz wave so distorted? Of course it could be that this distortion does not play a significant role in measurements (or when another circuit is driven with this signal), but is this waveform the real output? Or is the sine wave form output influenced by wiring, stray capacitance or other external sources.
A couple of comments. These units are pulled from service when an internal monitor voltage drops to a minimum specified by the reliability folks, not when the cell tower is taken out of service. They get replaced every few years on a regular basis. That test voltage is an indication of how much of the rubidium has been burned off from the heated cup rubidium source. So if you pick one up surplus you should realize that it is like buying a used car. Don't just leave them turned on, thinking they are like a crystal oven and grow more stable with age. They are like a 1500 hour incandescent lightbulb and will burn out sooner if left on. While they are very stable they have high phase noise, noticeably worse than a crystal oscillator, so precision measurement comparisons should be done over several hours for maximum accuracy. The best option is warm it up, (a couple of hours,) then use it to calibrate a crystal oscillator, then make your everyday measurements with the Xtal oscillator.
Is the heat coming from the lamp? Does the resonating chamber take this into account and that is the reason for the delay in locking? If you do do a teardown, open it right after shutdown and check everything with a IR thermometer to see which parts are at what temperature.
What a coincidence! I bought one of these just 2 months ago. I'm currently planning to put it in a tidy box and stick a linear power supply in there. I'm also thinking of dividing the frequency and giving it a PPS output.
I have to ask, because Dave was soooo close when he ran it - How close to 10Mhz are these units when they have locked up - Has anybody *actually* measured one? I know that are supposed to be 10Mhz to something like 10e-11.... But measuring them would be interesting.
Hey Dave, I was wondering if you ever did a Sansa MP3 player tear down? Because I have one and I want the LCD, but I can't find any datasheets. Any tips?
I think Dave's caution about cooling is sensible. Yes, the temp control will switch on longer if ambient temp is low but there are inherently power consuming components inside (about 12 watts) so don’t wrap this beastie in a blanket!
The lamp cell is RB85 and the filter cell is RB87. The atoms dip when excited atoms interact with the buffer gas neon. The dip is transferred to a photo diode and is converted from a current generator to a voltage source that is partially fed back to a variable capacitor varactor . The VCO is pll since it has a DDS involved and not fll as he states.
There is a comb generation circuit that in coupled to the cavity tune high Q resonator . This is an antenna that multiplies the microwave frequency to interact with the 6.8 ghz transition.
What about using multiple regular temperature compensated oscillators? I was most concerned with the short time stability. Is it true that the SNR is proportional to the number of sources? Can I get 6dB better by using 4 oscillators at once?
Are you sure you want to "put a heat sink on it and keep it as cool as possible"? You are just making the heater for the Rubidium cell work harder, and make the unit draw more power. Some cooling is ok, but "as cool as possible" is counterproductive.
Sir can you please provide a document to study frequency standard like hydrogen and caesium..and what all other standard we have like time standard...plz help
I not sure if it's completely the same inside the "physics package", but we built a setup inn a lab at uni, which could show the hyper fine structure of Rb87 gas. There really are many of excatly the same part you have there (FE-5680A) on ebay for as low as under 30€. All coming from China though and some cheating by labeling their items "new".
Awfully simplified - who would not know that Rb87 refers to the isotope ? Dave Magic = Physics ! Says a gal who plays with Quantum Cosmology - a great hobby ;-)
@@andyhowlett2231 Andy, I do know that the LP1 unit I have has specific warnings that the case must be heatsinked. I guess to keep the clock at operating temp heats up other components excessively without a heatsink.
To clarify some slight confusion in the vid over terms for osc types :
TCXO : temperature compensated - simple analogue circuit to compensate for ambient temp
DTCXO : digitally temperature compensated - as above but potentially more accurate as can track xtal characteristic more accurately over the temp range,e.g. from a lookup table.
OCXO :ovened - holds the xtal at a fixed temperature
Yep, right around 9:20 it's all a bit incorrect (TCXO vs OCXO).
Nifty! I had no idea you could get this kind of accuracy for so cheap!
And now I know what a Rubidium Frequency Standard is:
Thanks again Dave !!
I don't understand why you should cool or use a metal cover for the oven. It has a STABILIZED temperature of let's say 50 degrees celsius.
1. no plastic box will melt at 50 degrees
2. If you cool it or use a heatsink for it you will only increase the power consumption from the mains because the thermostat inside will try to keep the crystal inside at 50degrees.
I believe it is best to keep it as isolated as possible from any temperature variation, may it be sink or source of heat.
@Zadster I've never seen that, and no one else reports it. As RUclips transcodes all uploaded videos, I suspect either everyone would see it, or it's just you?
I bought one of these years ago and used it to calibrate oscillators I used in experimental GSM cells. This is was over 10 years ago. Things were complicated back then, but fortunately now there are a lot of affordable GPS frequency standards available.
Brilliant! Love the videos and your style. I've discovered a lot of great things via your videos and this one came up when I was searching for info about frequency standards. Keep up the great work!
The info I saw seems to indicate the Rb is used in the gas phase, much like mercury in a fluorescent tube. At standard pressure, Rb vaporizes at 688 degrees, so heating something to that temp. electrically is probably what's taking SOOOO much current. Maybe the boiling point of Rb can be reduced some by lowering the pressure inside the lamp, much like boiling water at room temp. with a thick bell jar and a vacuum pump; I don't know.
Why is the 10 MHz wave so distorted? Of course it could be that this distortion does not play a significant role in measurements (or when another circuit is driven with this signal), but is this waveform the real output? Or is the sine wave form output influenced by wiring, stray capacitance or other external sources.
A couple of comments. These units are pulled from service when an internal monitor voltage drops to a minimum specified by the reliability folks, not when the cell tower is taken out of service. They get replaced every few years on a regular basis. That test voltage is an indication of how much of the rubidium has been burned off from the heated cup rubidium source. So if you pick one up surplus you should realize that it is like buying a used car. Don't just leave them turned on, thinking they are like a crystal oven and grow more stable with age. They are like a 1500 hour incandescent lightbulb and will burn out sooner if left on. While they are very stable they have high phase noise, noticeably worse than a crystal oscillator, so precision measurement comparisons should be done over several hours for maximum accuracy. The best option is warm it up, (a couple of hours,) then use it to calibrate a crystal oscillator, then make your everyday measurements with the Xtal oscillator.
The best option is... ...a GPSDO.
That is an interesting piece of hardware. Can't wait for the teardown of this and your phillips freq counter.
@randomgarfield Who else is seeing sync problems? I don't see it. I am using a new video editor, so if there is an issue, this will be it.
Is the heat coming from the lamp? Does the resonating chamber take this into account and that is the reason for the delay in locking? If you do do a teardown, open it right after shutdown and check everything with a IR thermometer to see which parts are at what temperature.
This is an awesome device! I can't wait to see you build a project out of it, like a function generator! iotw, PLEASE do that!
What a coincidence! I bought one of these just 2 months ago. I'm currently planning to put it in a tidy box and stick a linear power supply in there. I'm also thinking of dividing the frequency and giving it a PPS output.
I have to ask, because Dave was soooo close when he ran it - How close to 10Mhz are these units when they have locked up - Has anybody *actually* measured one? I know that are supposed to be 10Mhz to something like 10e-11.... But measuring them would be interesting.
Hey Dave,
I was wondering if you ever did a Sansa MP3 player tear down? Because I have one and I want the LCD, but I can't find any datasheets. Any tips?
$50 in 2012, $300 in 2020.
I think Dave's caution about cooling is sensible. Yes, the temp control will switch on longer if ambient temp is low but there are inherently power consuming components inside (about 12 watts) so don’t wrap this beastie in a blanket!
The lamp cell is RB85 and the filter cell is RB87. The atoms dip when excited atoms interact with the buffer gas neon. The dip is transferred to a photo diode and is converted from a current generator to a voltage source that is partially fed back to a variable capacitor varactor . The VCO is pll since it has a DDS involved and not fll as he states.
I want to replace the clock base in my old laserdisc player with a rubidium clock, is that possible?
There is a comb generation circuit that in coupled to the cavity tune high Q resonator . This is an antenna that multiplies the microwave frequency to interact with the 6.8 ghz transition.
@LukeeeeBennettPlus Yes it's what I recommend to most people.
What about using multiple regular temperature compensated oscillators? I was most concerned with the short time stability.
Is it true that the SNR is proportional to the number of sources? Can I get 6dB better by using 4 oscillators at once?
What do you think would happen if you got two and beat them together?
@CampKohler I think I mentioned that the lamp part was the hottest.
Are you sure you want to "put a heat sink on it and keep it as cool as possible"? You are just making the heater for the Rubidium cell work harder, and make the unit draw more power. Some cooling is ok, but "as cool as possible" is counterproductive.
@SwitchingPower Now that this video is out, will it drive the price up?
When the signalling line drops low what does that indicate ? The FLL has locked up ?
The output is independent of the hyper fine process since it is a DDS board driven frequency
Sir can you please provide a document to study frequency standard like hydrogen and caesium..and what all other standard we have like time standard...plz help
Thanks for the good explanation and useful , just in time I'm wondering why such big box for an "insignificant" 10 mhz oscillator , very clear !
I not sure if it's completely the same inside the "physics package", but we built a setup inn a lab at uni, which could show the hyper fine structure of Rb87 gas.
There really are many of excatly the same part you have there (FE-5680A) on ebay for as low as under 30€. All coming from China though and some cheating by labeling their items "new".
@EEVblog no audio sync problems here...
19:50 The frequency reading is drifting horribly much. Is that a problem with the standard or the scope, you think?
1020% the scope, lol
@@ivolol Wow, it's been a while since I left that comment, hasn't it...
Nice! Gonna hook one up to my AVR tinkering kit as an external oscilator! :P
could this be used to make audio dac's sampling-rate more stable?
Dave you planning anything for your 300th? It's right around the corner at the pace you're going :)
What's all this rubidium stuff, anyway?
@CampKohler I haven't seen the prices go up so far.
The multiple frequency is from the ocxo
guys what do you do with this ?
Check eBay for vintage, military grade "Efratom" rubidium units (black cubic things), they must be made much better than these OEM things.
Rubidium Frequency Standard = a low cost Atomic Clock
@eevblog the first 10 or so seconds, then it rights itself.
Thank you so much !
your videos are very helpful
You were so convincing that I just bought two of these. :D
@TheLaidukas I'll be sure not to spend my commission all at once! :->
Is it me or is the audio and video out of sync?
I've got a programmable one. Programmed it up to 10.0666666667MHz, to triple, to get 30.2MHz reference for my radio gear.
Pity, I don't think these are going for $60 anymore 😅
As a HAM, this is most interesting!
This is what I think all HAMs are like:
ruclips.net/user/shortskShiMo2rTfc?feature=share
Seems like some bastards bought all the supplies and now sell these puppies for 150 to 200 bucks a pop...
bastards...
love you dave , you are awesome
@EEVblog I'm not. It seems fine.
Awfully simplified - who would not know that Rb87 refers to the isotope ?
Dave Magic = Physics !
Says a gal who plays with Quantum Cosmology - a great hobby ;-)
Hey, where's the teardown? Stop teasing me!!
Awesome keep them coming....
@uGateKiller It's you.
@EEVblog I see it.
I just bought the same one on ebay for only €35 / $43 :)
If you are going to channel Bob Pease, you are going to have to grow a big bushy beard. :-)
Audio sync is japanese kung fu style! 5-6 frames I reckon.
i don't recognize any out-of-sync
@CampKohler Keep it family friendly. ;)
I want to build a clock with it :D
"don't bitch about the detail" lol
Interesting as HELL!
legend!
@envisionelec You naughty boy, you! Now we know what you do for a hobby. :-)
"stuff anyhow" yaaaaaa :D
You MUST cool these or they will burn out prematurely.
Rob
But surely if you cool them the heater will just be on more to reach the correct operating temp?
@@andyhowlett2231 Andy, I do know that the LP1 unit I have has specific warnings that the case must be heatsinked. I guess to keep the clock at operating temp heats up other components excessively without a heatsink.
@@mosfet500 Thanks for the info.
I keep clicking and nothing is happening! :-p Thanks Dave, can't wait for the next one.
@uGateKiller
no, it's utube ..., something sucks today, maybe G.W. Bush is sniffing again (Bush for president? no thanks)
Nice1