I just bought an Efratom 100318

[tkamiya]

Hi,

I just bought an Efratom 100318-001.  (Rubidium standard)  Lamp voltage = 9.8V

I see Efratom had a various models, including FRS-A, FRS-C, LPRO-101, and the last one I can't remember.  100318 is visually very close to FRS-C.

Is 100318-001 same as FRS-C?  How does it compare to other Rubidium standards?  I did look over google results but this one appears to be least mentioned and I didn't gain any useful/tangible info.

[Johnny10]

There is a manual online for the FRS unit 100318-001.

http://www.ko4bb.com/getsimple/index.php?id=download&file=02_GPS_Timing/Efratom/Efratom_FRS-C_FRS_N_Rubidium_Oscillator_CO100138B.pdf

Explains everything about this model including versions of 100318, FRS-C and FRS-N
I think the 100318 is newer version of FRS-C.

Output is 10MHz Sine wave.
Very easy to connect to that 10-pin male connector.

You can get Frequency,Lock Monitor which is TTL on that unit 0.06 = Locked 5V = Unlocked, Lamp Voltage and Ground, Power to Heater and + power is on one pin.



 

[tkamiya]

Wow!  Thank you! 

I went to BB's site but completely missed it.  Wonderful resource with circuit diagram included....!

[tkamiya]

I ended up going directly to the D-sub connector.  For whatever reason, the header type connector is missing few signals I wanted.  Due to lack of heatsink, I could only test for a while but it locked quickly and output was remarkably on-the-spot.

[tkamiya]

I have a heat sink mounted and RB running on lab power supply.

I'm seeing 6x10^-11 with occasional jumps of magnitude 10.  I guess when going for this kind of precision and accuracy, one can not go cheap.  I am using a cheap 10 turn register for trimming.  I am guessing it is affecting stability as it didn't move this much when trim was not present.

I was trying to avoid spending 15 dollars (minimum) for name brand but oh well. 

Heatsink temperature has stabilized at 43.5C. 

I have a question for the group.  My initial plan was to encase everything in a case including the sink.  But if I do this, I'm sure temp will jump.  What is the consensus on using a fan?  All I need is a little bit of air movement.  I wouldn't even consider this on OCXO but what about Rubidium?  As I understand it, the absolute max is 65C at base plate, according to what I can dig up.

I can also go larger on heatsink or mount it on thick case bottom to radiate via conduction.  Manual doesn't specify what is and isn't recommended.

[Johnny10]

Bigger heatsink.

Without a fan I get mine to 37 C after 6-8 hours with larger heatsink covering entire baseplate.
I tried with fan and it stayed about 30 C.

[edpalmer42]

Don't go crazy with the heatsink.  A Rb standard has two or three ovens that have to stay hot.  Crazy heatsinks just force these ovens to work harder.  Most of the time, all you have to do is bolt the Rb standard to the outer case and make sure that air can flow around the case.  Commercial units use external heatsinks so that they can run in high ambient temperatures.  Does your workshop get up to 50C? 

Ed

[tkamiya]

I wish I could find out what the optimum temperature is....  So moving to Alaska is out?  (I'm in Florida)   I understand cooling is not for the oven'd part but for the electronics around it.   

My lab stays around 25 to 28C.  (77F to 83F)...  Oh I wish everyone standardizes on metric system!

I guess I should be happy frequency is not based on presidents' whatever bodyparts....

[Johnny10]

These 100318 units were mounted in a closed metal case.
Labeled CTS-101
With a buffer board providing a SMA connector out the side and a larger connector with 4 sine wave 10MHz outputs.

I asked about the buffer board a few months back.

Pictured below.

[edpalmer42]

How was the Rb mounted inside that case?  Was it screwed to one side?  Did it use any form of thermal mounting material?

Ed

[Johnny10]

The box is about 3 times larger than the oscillator case.

Just screwed to case.

[Towger]

15 x 6.5 x 2.5 cm is the size of the heatsink which came with my FRS-C.  Which I have been told came from a decommissioned satellite ground station or uplink.  It has a 10Mhz square wave output.

[edpalmer42]

I have a PRS10 Rb standard mounted in a box.  It's bolted to the rear of the case.  No thermal compound was used.  The PRS10 is similar in size and form factor to the FRS-C.  The PRS10 has a steady-state power drain of 14.4W @ 25C.  They don't say whether that 25C is ambient or base plate temperature.  The FRS-C is rated for 12W @ 25C base plate temperature.

After running for a few hours in a 22C ambient, the base plate temperature of my PRS10 was about 45C.  No further cooling is required.  Even in summer, the base plate temperature won't come near the 65C limit.  If I was concerned, the first thing I'd do is mount an external heatsink on the back where the PRS10 is bolted.  Orient the fins vertically for best air flow.  A fan isn't required.  For comparison, The Efratom PRFS Rb frequency standard contains an FRS standard.  It has no fan.  On the rear, there's a heatsink with a small number of thin fins.  The unit is rated for operation in a 50C ambient.

Towger, are you sure that heatsink is original?  The only way you'd need such a large heatsink is if you were running in a very high ambient temperature.  The fins hanging out past the ends is also suspicious.  In your pictures I can't see any holes, clips, etc.  How would it be mounted originally?

Ed

[tkamiya]

Read the manual (for my Rb) again and concentrated over power supply requirement.  It states 1mV ripple max is necessary, otherwise spectrum purity will suffer.  It states "If non-harmonic at the supply ripple frequency (50, 60, or 400Hz) are not a consideration, the ripple can be higher"

These days, I don't know anything that uses linear power supply.  I am wondering now, how critical this is for my particular use.

Also read into various technique where careful filtering of switching power supply can bring ripple (typically 100mv?) can be lowered to match that of linear supplies.

FRS* has two input pins.  One goes to electronics and the other goes to heater.  Latter is not further filtered but former is regulated and filtered internally.

I wonder if they came up with 1mV p-p as "matter of routine?"

[Towger]

It came from eBay as spares or repairs, as the seller had not tested it.  I was told it  was part of a larger decommissioned satellite ground station system.  I am assuming the heatsink was from its use in this system.  The fins are professionaly cut out in the location of the factory mounting holes, you can still see pen marks at the side.  The seller did not know the make and model, they were covered by the heatsink.  He was worried it was near/at the end of its life as (he thinks) it was powered up for the life of the satellite.   But when I tested it, it locked quickly and voltages were well in line with the manuals specs. 
It is another work in progress project to mount in a nice case etc...

@tksmiya  Thanks for that. I tested mine on a linear bench supply supply, but picked up a decent 24v SMPS for it.  Even still i have my doubts about using a SMPS it.  I'll keep a look out for a linear one.

[tkamiya]

I guess we have to be careful in thinking if "some" is better, HUGE is more better.... 

I found a statement in user manual that said bottom heat plate must be attached to a case better than 0.005".  I think that's a bit thicker than 1mm?  But didn't mention how big.

I have a few switching power supplies that came out of HP equipment.  I'm going to put some load on it and see how well it performs.  I really don't want to use linear power supplies.  Even high-end PLL signal generators today use switching power supplies.

[edpalmer42]

I guess we have to be careful in thinking if "some" is better, HUGE is more better.... 

I found a statement in user manual that said bottom heat plate must be attached to a case better than 0.005".  I think that's a bit thicker than 1mm?  But didn't mention how big.

That's not the thickness, it's flatness.  And according to my calculator, 0.005" = 0.127 mm.  That's closer to heavy duty aluminum foil than an aluminum case suitable for mounting anything.  But you're right, they don't say anything about how big the surface should be.  That's up to you to determine.  The only thing they care about is that the temperature doesn't exceed 65C.

I have a few switching power supplies that came out of HP equipment.  I'm going to put some load on it and see how well it performs.  I really don't want to use linear power supplies.  Even high-end PLL signal generators today use switching power supplies.

If you're not really careful, switching supply noise will leak into the output.  It often shows up on phase noise plots.  When manufacturers use switching supplies, they are very, very careful to make sure that doesn't happen.  If your application isn't sensitive to phase noise, it might be okay.

Linear supplies are bulky and inefficient, but very quiet.  IMO, they're easier to repair than switchers, but I haven't had much experience repairing switchers so take that opinion with a few grains of salt.

Ed

[tkamiya]

I guess wrapping my Rb with aluminum foil is out then.

[Johnny10]

I really don't understand the disdain for linear power supplies.

You can get such fantastic deals.
Not Chinese manufacture but HP, Lambda, etc.
Easy to use for hobbyist.
I have a bunch of HP 6263A power supplies I paid 11 dollars each for !

I am always repairing SMPS have never had to repair a linear.

[tkamiya]

Well, I have seven older rack mount HP power supplies.  I probably paid $10 for each on average.  I wanted to one light package.  I know about those modules from old days.  Well, I didn't want to get those again.

Oh, why do I have so much STUFF that don't "just fit"!

I am exactly the opposite.  I had so many linear supplies to repair.  I've just thrown away switchers when they fail. 

[Johnny10]

Off topic but the area we live in has such an abundance of surplus equipment.
Both for price and ability the ability to pick heavy items up from sellers make this a hobbyist paradise.

High tech to low tech within a fifty mile radius, surplus from all the space related companies.

I love it !!

[tkamiya]

I'm near Orlando, Florida.  It used to be much better but all the electronics surplus stores consolidated to just one.  I also have other local sources.

Yes, such a wonderful time we live in (as far as our hobby is concerned).  A bit older QUALITY equipment are abundant available at such a bargain prices.  Last  year or so, I just went NUTS and went from just two multi-meters to lab full of equipment.  Most of them purchased broken and repaired. 

If I were to do this during 80s, it would have cost me my life time worth of earnings and more. 

I also have a friendly local cal house which helps me out a lot.

I am leaning more toward linear supplies, either internally or externally.  I also came up with a neat "housing" idea.  I have lots of non-working-and-hopeless HP sweep generators.  I am thinking of putting in really nice power supply and distribution amp, etc into the mainframe part.  Then use the plugin shell to have time standard.  Perfect for experimenting.  Just about everything already at home.

With good cleaning and nice coat of paint, it's as good as new.

[Johnny10]

If you want one of the boxes similar to the one I am showing with the output plug having a few 10Mhz outputs.
Drop me a line.

[FriedLogic]

I have a heat sink mounted and RB running on lab power supply.

I'm seeing 6x10^-11 with occasional jumps of magnitude 10.  I guess when going for this kind of precision and accuracy, one can not go cheap.  I am using a cheap 10 turn register for trimming.  I am guessing it is affecting stability as it didn't move this much when trim was not present.

I was trying to avoid spending 15 dollars (minimum) for name brand but oh well. 

Heatsink temperature has stabilized at 43.5C. 

I have a question for the group.  My initial plan was to encase everything in a case including the sink.  But if I do this, I'm sure temp will jump.  What is the consensus on using a fan?  All I need is a little bit of air movement.  I wouldn't even consider this on OCXO but what about Rubidium?  As I understand it, the absolute max is 65C at base plate, according to what I can dig up.

I can also go larger on heatsink or mount it on thick case bottom to radiate via conduction.  Manual doesn't specify what is and isn't recommended.

The external frequency adjustment range is usually quite small, so you don't need anything special for the trimmer. A really cheap cermet would probably be fine, but a $1-$2 branded one might be more certain.


In the 'Environmental Specifications' part of the LPRO Manual ( there is one at http://www.ham-radio.com/sbms/LPRO-101.pdf ) it gives the MTBF for 'ambient' temperatures from 20ºC (381,000 hours) to 60ºC (126,000 hours)

There's normally fairly limited thermal coupling between anything that is meant to be hot and the base plate, so you're not likely to have an 'over-cooling' problem at anything close to normal room temperatures.

[tkamiya]

Wow....  I was *just* picking 15 dollars pieces from Digi-key's offerings.....  Thanks!

I am going to get some Burns kind for 2 dollars a piece.  These are 20 turns, so tuning should be little easier.  While using cheap ones, I was noticing changes in value weren't linear.  At several spots, downward changes of frequency suddenly reverses course for a bit, then goes back to downward changes again.  I hope better kind will have less of this type of issues. 

I am finding the module works from 18 volts and doing so will significantly reduce power consumption without anything negative.  I would think heat generation will be less as well.  (Yes, I know I'm just shifting where the heat will come from, power supply or rubidium module)

It will be an interesting experiment.