[NLRS] Efratom Lpro101 Rubidium Standard

[Dr. Gerald N. Johnson]

There is some oven effect in the Rubidium Standard. And lots of heat to 
dissipate. The circuit will probably work at a higher ambient 
temperature, but with a shorter lifetime, I'm sure the maker's manual 
(on another computer here) has some suggestions about enclosing and 
cooling. Adding more cooling won't hurt, unless you bang a knee or elbow 
on a sharp end of a fin.

Some Chinese circuit designs for switching power supplies are very 
minimal and have no attention to low RFI or input/output ripple. There 
is an art to minimizing RFI by minimizing the area of high current 
switch loops and careful choke and capacitor input and output filter 
which cost money and circuit board real estate, reasons some makers 
leave them out to our detriment.

On 11/18/2012 1:54 AM, tosca005 at umn.edu wrote:
>
>
> Thanks to everyone who replied.
>
> It seems that the consensus is that in spite of KA7OEI's recommendation
> that the aluminum box alone is a sufficient heat spreader/heat sink, the
> addition of an additional heat sink is a good idea. Just to play it safe, I
> already picked up a couple of aluminum heat sinks from eBay, and will plan
> to use one on the project (and have a second as a spare, since it was
> pretty slow arriving from China!). It is nicely matched to the size of
> either the lid or the bottom outside of the aluminum box I am using.
>
> Doug, N0NAS: Yes, that was the web article I was referring to. I am still
> undecided about mounting the Lpro101 to the inside of the lid vs. the
> inside of the bottom of the box, but if the latter, I think I can grind off
> the worst imperfections with a Dremel tool. My box is not ribbed inside,
> just has some mold marks and perhaps a slight curved departure from true
> flatness. Your reminder about adding more filtering to the power supply is
> appreciated. I think that the capacitors and inductors of KA7OEI's design
> should do the trick.
>
> Chris, N0JCF: Thanks enormously for the tip on the thermal gap filler pads.
> This sounds like just the thing to make the heat transfer more reliable
> without resorting to a visit to a machine shop.
>
> Ron, K5LLL: I plan to use a pre-built boost regulator (from China via eBay)
> to raise the supply voltage from 13.8 to a valid operating level. So there
> will likely be at least a little extra heat from that. I think that the
> KA7OEI recommendation to only boost the voltage to the "minimum
> specification" level of 19 vdc instead of the "typical specification" value
> of 24 vdc makes good sense to minimize battery drain and heat production.
> But once the project is built, I will see how hot it gets, and as you
> suggested, adding fan(s) to the heat sink could be an easy upgrade if
> needed. And as I mentioned, Doug's admonition to provide good filtering is
> going to be taken seriously also to keep noise out of the 10 MHz output.
>
> Ben, NO5K: Sorry I forgot your name and callsign from the RMG meeting. This
> old brain of mine seems to leak like a sieve but practice should make
> better, if not "perfect". I also guess I mis-understood our ultra-brief
> conversation about which rubidium standard you are actually using, i.e. the
> FEI 5680A, not the Efratom Lpro 101. But I am sure that the same principles
> apply.
>
> Jim, K0MHC: I remember our prior discussion of how you use the OCXO for
> routine roving after re-checking it against the more accurate rubidium
> standard, and the benefits of using the OCXO out in the field. Thanks for
> confirming that I was not going crazy -- my recollection was that the OCXO
> needed to be put in an INSULATED container to keep the heat in, vs. the Rb
> standard which needs to GET RID of excess heat. I was a bit fuzzy on that
> before, but it is clear now. I have an enormous number of projects to work
> on, and even since retiring from my (often) 80 hour/week job (only paid for
> 40/week, unfortunately!), I still don't find myself twiddling my thumbs
> wondering how to keep busy. :) One of the other projects in my pipeline is
> a GPS-disciplined OCXO. I believe I have all the parts for that project
> except a PC board for the GPS module (so I will build it up on a piece of
> perfboard) and a GPS antenna (readily available, just have not picked one
> up yet). KA7OEI seems to dismiss the possibility of using the GPS
> disciplined OCXO for roving, unless you are operating mainly "stop and
> shoot" mode with ample time to re-gain GPS lock after moving to a new grid.
> This sounds overly pessimistic to me, since my experience with a good GPS
> unit is that it can easily keep and/or quickly re-acquire a lock during a
> grid-to-grid move under normal circumstances without a dense tree cover.
> (And who wants to operate a rover while parked inside a dense forest???) I
> guess if I get both projects built I can tell instead of guessing. Oh yeah,
> I was delayed in building that project because I thought to myself,
> "wouldn't it be neat if I added an LCD to display the Maidenhead Grid
> Square and other info like time, velocity, elevation, lat/lon, etc. right
> from the GPS that was being used primarily for its 1 PPS signal?" Oh how
> true that trying to make it "perfect" can prevent it from getting made at
> all! :( Sometimes, "good enough" is the only way to make it happen!

The performance of any oscillator is improved if the environment its in 
has fewer variations. That's why the old standard was a double oven. A 
fine control inner oven inside a thermos bottle or dewer and then a 
coarser control oven to give that inner oven less variation.

One maker of good oscillators got a crystal oscillator to parts in 10^-9 
short term stability over the rated temperature range by making at VCXO 
and then using a temperature sensor on the outer surface of the package 
and a custom EPROM'd correction chart to set the voltage on the VCXO 
line. Not very much worse than a good rubidium standard and far less 
power hungry. I've been working on a PIC chip version, but have nothing 
to show yet except a few parts collected.

On 10 GHz, my transverter has a thermistor on the crystal which slows 
the frequency changes from external heat by holding the crystal to a 
relatively constant frequency, and then I added chunks and chips of dow 
blue board around it and then I cut the IF drive power back to two watts 
or less and while I'm sure I moved around I didn't move enough to cause 
anyone to complain this year. I want to cut the drive back even further 
to reduce the heat changes in the trasnverter box while transmitting and 
that will improve the stability even more. While roving in September, I 
kept the transverter powered while driving to keep the crystal up to 
temperature and it seemed to work. You heard it and chased it, not very 
much chasing I hope.

73, Jerry, K0CQ
>
> Once again, my thanks to all of you.
> 73 de W0JT/5, John P. Toscano
> ______________________________________________________________
>