what is a rubbitin frequency standard?

Do you mean

Are you really the Jan Pantelltje we know and sometimes love?

Paul

On a sunny day (Mon, 19 Nov 2012 04:08:16 -0500) it happened "P E Schoen" wrote in :

Only sometimes huh? :-)

BTW I have been watching Dave's EEV blog 2 entries on the rubbitin frequency standard. and I have to compliment Dave on an excellent presentation. Not only did he get it working, but also went into much detail as to the chips etc in it (AD9830 frequency synthesizer etc). I found some detailed serial programming info for that unit on the web too, as Dave found his has actually a serial interface, not sure if he ever got around to programming that (for the FE5680A). Mine is on its way from China.

Watching mains frequency ... mmm it moved left, it moved right, versus a crystal, maybe all that wind power. I do need a very precise reference, else I get out of band in the GHz range, so this thing seems the right solution. I tried getting a 10 MHz crystal beating against the 10 MHz WWV shortwave and that sucks, too much interference here, probably from all the ADSL unscreened copper in the area.

I have also soldered a MSOP10 SW-439 by hand, so all that makes me look at electronix in a very relative way. Its a 2 GHz switch.. Faster than a 4016 :-) I have also been watching F1 in Americca yesterday, so they seem to have plenty electronics problems,anybody here working on that?

It's a rubidium standard made from a Rubick's Cube physics package.

No, you measure it infrequently.

Y're welcome. It's 3:30AM and I'm a bit short on humor.

Oh-oh. Watch out for problems with the rubidium physics package on the EFRATOM FE5680A series. Some are so old that they take forever to lock. It should lock in less than 10 mins. I bought one really cheap. It lasted about 3 months. I was able to raise the dead with a checklist of common failures: Mine had a cooked R6. I traded it last month for some non-working test equipment.

I'm switching to a GPSDO (GPS disciplined oscillator). The trick is finding a GPS board that has an output other than 1pps.

that sucks,

area.

ADSL2+ quits at 2MHz. More likely you're hearing 10baseT ethernet. The carrier frequency is at 5MHz for alternating zeros and ones, and

10MHz for all ones or all zeros. Looking at it on a spectrum analyzer, I see rubbish between 1Mhz and 30MHz, with the amplitude varying with whatever is being sent. Some of the local hams have gone to an all wi-fi network in their homes, in order to avoid hearing RFI from their networks on their HF radios. 100baseT moves the fundamental data rate 10 times higher, which raises the rubbish generated to roughly between 5 and 250MHz with a wide peak at 31.5MHz. You'll get less interference at 10MHz with 100baseT than with 10baseT.

Incidentally, the worst offender is 10base2 (Cheapernet) but only if the shield connection falls off one of the BNC connectors.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Good luck with that. Pretty sure that 1pps is the standard.

On a sunny day (Mon, 19 Nov 2012 03:27:42 -0800) it happened Jeff Liebermann wrote in :

I do not know if it is EFRATOM, I bought ebay nr 280655233263

In the EEV blog it locked in a few minutes or less IIRC. That pdf seems useful it shows how to rejuvenate a lamp.

Right, I only have the 1pps, you can PLL to that, but the multiplying factors are a bit big :-) Actually I want to grab the 50.xxx MHz and feed it into my spartan FPGA board that now has a 50.XXX MHz oscillator module. All other frequencies are derived from that.

that sucks,

the area.

I went with a battery powered shortwave receiver, and switched of every piece of equipment I have, starting with the computahs... And tried at different times of day too, 5 MHz, 10MHz, and 15 MHz, and also the Russian references that are just next to those. I can hear WWV, at 10, but very weak, moving the receiver around the room finds interference from (on) mains wires etc.. Maybe an outside antenna would work better. But this thing should give me a good reference any time of day. Maybe I will calibrate something, use my tritium decay experiment oven next year when the experiment finishes as a crystal oven. That has been running now for more than half a year, 6 month to go.

Rub it in (frequently)

Regards,

Boris Mohar

Got Knock? - see: Viatrack Printed Circuit Designs (among other things)

void _-void-_ in the obvious place

"

..

Hmm weird, So Efratom, becamce datum and then went out of buisness. (For a while we bought the lamp package from them.) The FE5680A is a Rb clock made by Frequency Electronics on Long Island. (I interviewed there, but I couldn't stomache the idea of living on long island.)

George H.

Some are so old that they take forever to

a

nge,

e and that sucks,

in the area.

ann

Neat, will you open it up and post pictures when it arrives... or should I check out the eevblog?

George H.

a

ange,

tors are a bit big :-)

oard that now has a 50.XXX MHz oscillator module.

ve and that sucks,

r in the area.

iece

quoted text -

package.http://en.wikipedia.org/wiki/Rubik%27s_Cube

Ditto, and ditto. Ragweed and traffic jam capital of the world.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

standard.

Here's the manual of an old Efratom unit.

The stuff that synthesizes the 6.834687 GHz microwave frequency is very simple.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

oen"

..

p://en.wikipedia.org/wiki/Rubik%27s_Cube

Grin... It was the housing prices that were the final nail. I remember going to realtor.com and putting in $250k as a maximum price... there were no houses listed for that price, I needed to get above $350k or so to find a shoe box. (This was years ago before the housing meltdown.)

George H.

.highlandtechnology.com  jlarkin at highlandtechnology dot com

The Rockwell, Conexant, or Navman Jupiter GPS engines have a 10Khz outputs. There are a few others, but you're correct, 1pps is the standard and most common. It's possible to build a phase lock loop with a 1 Hz reference, but I really don't want to deal with lock times measured in days.

I have a large pile of CMS / Novatel Allstar 12 receivers that are looking for a sellable project. Too bad they only output 1pps.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

I've got one or two spare Jupiters with the 10 Khz output. I can part with them.

oscar sierra romeo @ohio.net

change phonetics to 3 letters.

Steve

It's EFRATOM. Note the diamond logo. Looks like you have a 14 day return warranty. No details from the vendor on the condition of the specific physics package. The one I bought (also from Hong Kong) arrived looking somewhat different from the eBay picture. Instead of the DE9P connector, it had a wire pigtail. If it locks in 5-10 minutes, it's good. If it takes longer, or every time you turn it on it takes longer, it's bad. Ask the vendor if they did any refurbishing of the physics package. That's become almost standard among Rb clock vendors on eBay, but it doesn't hurt to ask.

The spec is That pdf seems useful it shows how to rejuvenate a lamp.

Yep. It was VERY useful in fixing the one I purchased.

are a bit big :-)

that now has a 50.XXX MHz oscillator module.

See the comments in the first two paragraphs on the choice of reference frequency at: The first 1Hz GPSDO that I built took 3 days to settle down and about an hour to recover from a loss of GPS lock. It was a clip lead and breadboard abomination, but it worked. However, it had some jitter on the 1PPS output that ended that idea and made me look for something faster and better.

You might want to read through: for additional clues and info.

Ah, conducted radiation. I'll guess(tm) a Plasma TV, negative ion generator hi-v supply, noisy wall wart power supply, or high power photovoltaic solar controller. I've seen them all locally.

It's difficult to move around with an outside antenna. Walk around the neighborhood as it's probably coming from something nearby.

Nice. I just needed 10Mhz to keep my pile of antique 1970's test equipment on frequency. The big problem is that I live in a rather dense forest, which tends to block my view of the GPS birds. I'm seriously considering hiring someone to climb one of the 60 meter trees and installing an amplified antenna near the top: I'm staying home today to finish cleaning up that mess. I found two dead mice under the bench last week.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Try Google or Blekko images:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

That's a different model. For the FE-5680A see:

More detail:

Simple? Surely you jest.

With the FE-56880a, the mixing goes something like this: 20 MHz * 3 => 60 MHz 20 MHz / 4 => 5 MHz 5 MHz / 16 => 312.5 KHz 5 MHz * 0.312500 MHz => 5.312500 MHz 60 MHz * 114 - 5.312500 => 6834.687500 MHz

The problem is the 6834.687500 Mhz result of the math is 4.89 KHz off frequency from the rubidium resonance at 6834.682612 MHz. To nail it exactly, the 20MHz reference is offset slightly, which is why you don't want to use the 20MHz oscillator as a reference output, unless you really want a 19.984452 MHz output.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Oops(tm). That should be roughly: 20MHz + (4.89Khz / 342) = 20.000014 MHz output. which is close enough for most applications.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

If it is not specifically stated that particular GPS module is intended for timekeeping, you can expect inaccuracies and irregularities with 1pps signal from it. Using GPS as accurate frequency/time reference requires good local oscillators and sophisticated software algorithm to control.

There are also American 60kHz and Russian 66.666... kHz references. The LW and MW AM radio stations have their carriers synchronized to national standards; that could be used for timekeeping also.

Vladimir Vassilevsky DSP and Mixed Signal Consultant

As someone else mentioned below, it's better to use a GPS module that's designed specifically for use as a frequency standard.

has a good selection of economical GPSDOs, while on the eBay/surplus market, you can't beat the bang:buck ratio of a Trimble Thunderbolt.

The FE-5680As came in several different configurations, some of which have a high noise/spur floor (see

, where I pitted several small rubidiums against an HP 5065A.) If you're going to multiply to the GHz range you probably do not want one of the

5680As that are currently floating around, although their overall stability characteristics are good.

-- john, KE5FX