what is a rubbitin frequency standard?

True. They're currently selling for $180 to $400. Ouch.

However, I have an ulterior motive. I have about 200 CMS/Novatel AllStar 12 GPS boards that are looking for a project. APRS trackers are rather boring. I was thinking of a GPSDO. Lack of time and general lack of interest has brought that project to a halt.

I was just reading through the FE-5680A FAQ, which pointed to your web site. When I originally bought my FE-5680A, I didn't need to stabilize any microwave test equipment or radios. At the time, I was at 1GHz max. However, some of the stuff I traded for the FE-5680A goes to 4GHz and

I've looked at the Jackson-Labs GPSDO boards but could never find a dealer or price list. I think Symmetricom is now selling their products. yep: Still no prices or dealers, but that should be easy to find by calling or writing their order prevention department.

Thanks much.

What little I know, I've learned from fixing them. My score so far is

Yep. The magnetic field does have an effect on the frequency. Freq = 6,834,682,613. + 574 * H^2 Freq is in Hertz H is in Oersteds.

Look at the schematic. It's a few TTL parts and an SRD multiplier.

The Efratom manual talks about using a magnetic field to push the microwave resonance around a little. The 10 MHz output of ours is dead on.

What's fun is to get a scope, and triggger it from a caesium standard and look at the rising edge of the Efratom output. Crank it up to 10 ns/div and you would swear that the scope is internally triggered. Come back an hour later and it's moved on the screen a little.

On a sunny day (Mon, 19 Nov 2012 14:18:28 -0600) it happened "Vladimir Vassilevsky" wrote in :

It is the SiRFIII chipset, in a dealextreme module. I opend up the module, and it seems they connected the LED on front to the 1pps output of the chipset. It only starts flashing when locked on sats. There are several interssting signals that are brought out to little gold plated test points from that ballgrid chip, but not a real usable frequency. I think I can see the the phase comparator and maybe 2 bit signed ADC.

Yes, there also is supposed to be a UK long wave station that is synchronised, but I could not hear it on longwave. DCF77 at 77.5 kHz in Germany runs my radio controlled clock. At least I do not have top set winter-summer time on that one...

I wil report back how things work out with the rubbitinium

On a sunny day (Mon, 19 Nov 2012 14:45:12 -0800 (PST)) it happened " snipped-for-privacy@pop.net" wrote in :

Interesting link, yes the idea of a '2 step' system, driving a second crystal had occured to me. PLL loop bandwidth can then be very low. I could oven it too, as I replied in an other posting.

It is very likely that there will be a glitch of phase of 1pps signal when switching from one GPS satellite to the other. There could be small phase jerks at all time as they could use fractional dividers. Take a good frequency/period counter and check the 1pps from your GPS module against it. What kind of timing accuracy are you after? Getting to full potential accuracy of GPS is no simple.

Vladimir Vassilevsky DSP and Mixed Signal Consultant

On a sunny day (Tue, 20 Nov 2012 08:27:53 -0600) it happened "Vladimir Vassilevsky" wrote in :

Oh, I am satisfied with a few hundred Hz at say 1.5 GHz. starting directly from about 50255056.078356 Hz, and multiplying by about 30 that should work with the rubbitin unit.

I have no illusions about that GPS module 1pps, will not even try to use it. It wil be non-operatibve anyway once the project runs. Like a tiger byting its own tail. :-)

un

Neat, I was just going to seacrh the web and try to find that number.

I've got a broken SRS unit that John Williston gave us.

Hmm, I can't get into the absorption cell/cavity 'bit'. I was kinda hoping to see the step recovery diode. (Which I think is inside the cavity.)

Maybe I can still fire it up some day and see what works.

George H.

for

gnal

cal

Can someone tell me what the 1pps output (or input) is?

Thanks, George H.

plated test points

onal

ised,

ext -

Yikes, what a packaging nightmare!

ff

.highlandtechnology.com  jlarkin at highlandtechnology dot com

It actually came apart easily. But yeah it has PCB's on all six sides of the 'physics package', held together with multiple headers.

There's a resistor(?) stuck on thin brass posts right next to the lamp bulb... I don't know what it does.

There are some nice construction 'tricks'. The heater for the lamp is a voltage regulator (LM7808) and a transistor in TO220 packs. (I assume the transistor sets the heater current.) And underneath the two TO-220's are machined pockets that hold two thermistors. (I assume) (and why two?)

George H.

Incwww.highlandtechnology.com  jlarkin at highlandtechnology dot com

I have an SRS SC-cut 10 MHz oscillator. The oven design is terrible. They used a mosfet for the heater and located the temp sensor some distance away on the inner metal block. Obviously the thermal diffusion time constant was long, and the loop was unstable. So they went with a low-gain proportional-only loop, then added a second ambient temp sensor and did a feed-forward thing to correct for the finite gain.

The oscillator itself is pretty good. Jitter is around a picosecond per second, thousands of times better than a cheap TCXO.

off

it

ess

is

une

ww.highlandtechnology.com jlarkin at highlandtechnology dot com

Oh my boss suggested that the resistor (~500 ohms) on brass posts is the feedback 'tap' for the oscillator.

George H.

z off

l it

less

is

is

un

tune

t highlandtechnology dot com

.highlandtechnology.com  jlarkin at highlandtechnology dot com

Ahh we got the absorption cell open. I thought there would be two cells, first the 'so-called' hyperfine filter, (Rb85) and then the absorption cell Rb87. But there's only one. Maybe it's a mixture, I'll have to look at it next time I test a diode laser.

The step recovery diode is a through hole part. The only part number is an M on one line and 0024 next. (It's blue in color)

George H.

off

highlandtechnology dot com

Incwww.highlandtechnology.com  jlarkin at highlandtechnology dot com

Only companies that start with "M" are allowed to sell SRDs. MA/Com, M-pulse, Metelics, MicroMetrics. Everyone else had to get out of the business.

You might do ok with a directional antenna like a loop. I have been looking at LF loops to receive the WWVB signal and they have a deep null that can be steered to sources of interference. It can be a bit of a construction project to make your own loop antenna at LF, but at 10 MHz it might not be too bad.

Rick

range,

Fer damn sure. Some F**** idiot manager understood 1 pps and had zero understanding of proper clock synchronization. So they bought shit. Now it is the common thing, and essentially worthless.

?-)

intended

when

phase

it.

True. But 1 pps takes months to years to beat decent TCXO stability.