Build a 10MHz Rubidium Frequency Standard and Signal Distribution Amp for my Lab 

This is the kind of video that inspires people to go to the next level and really get involved at an in depth level of electronics. Thanks so very much for sharing this

Your comments about enclosures really rings true! I have found a good technique is to find off-the-shelf enclosures then get some laser cut aluminium/stainless panels done for the connectors and controls.

A great video, great story and with a great result.....and just goes to show how resourceful electronics folks can be!

truly amazing gerry, kudos.its nice when every thing goes to plan and fits perfectly. you are truly inspirational. keep up the good work.

Great to see an end-to-end construction. Delighted to see a micro-controller make an unexpected cameo appearance too; I'm a +1 for any PIC or ARM videos btw. Also good to see a channel in English. Thanks for all your efforts. Subscribed.

Wow. It's been a long while since I've watched a technical video from end to end... but this was worth it! Thank you! I would've just been happy for a plain 'old signal distribution box!

Yes its on the way, I have been doing board layout for it and have a couple of mechanical contraction issues to overcome. Its definitely not on the back burner, but I like to have more than one project on the go - it keeps things interesting.

Yes it does a fantastic job, the unit basically runs just warm, the heat sink and fan combination I would say are just about right. Gerry

Your welcome, thanks for watching. Good luck with the build

Love the standby/ready light, what a nice touch.

Thanks Adrian, its always nice to have a sense of finished. The only thing I could have done better was to make a PCB instead of using the perfboard but, for a one-off I can live with that. Thanks for the feedback. Gerry

Gerry is one of the very best REAL engineers I have ever seen on RU-vid. There's no B-S, or annoying theatricality, but there IS a vast amount of real theoretical and practical electronic skills. Kudos to a genuine master of the discipline.

Love your videos! This one in particular was exceptionally inspiring to me. I've been running a GPSDO in my lab (I did purchase a couple of your TCXO boards for the HP/Agilent frequency counters), but because of this video I've purchased the Rubidium standard, and also a couple of the Extron D/As (two came in a rack mount duplex configuration). I've modeled the Rubidium standard, as well as a rectangular hole for a lighted rocker switch and CNC machined all the holes into the front panel. I also opted to not gang all 18 outputs together, but rather keep them in sets of 6. The top one I removed the BNC and installed a right angle BNC on the inside of the PCB so as to provide a connector inside the chassis to the standard. I then made two short RG-316 jumpers to jump the other two sets of outputs to the main section. Next I'll be CNC machining the fan holes and building the auxiliary PCB with the PIC for the LED control. Work in progress, but I owe you so many thanks for putting out this video.

I like your approach to making that! Thanks for sharing.

Indeed, its nice to be able to make use of them. Gerry

Really enjoyed this video and the process you took to repurpose old equipment to create a distributed 10 MHz reference for your lab. Valuable tool for any HAM radio operator or RF engineer.

Very nice project. Thanks. I have a few Q-Max cutters. They make a nice job of "big" hole cutting.

Hi Adrian, its a 12V fan and I have it running at about 7V so its running slowly and barely audible. I have been running it solid for two weeks and its not getting at all hot, I am really pleased with the thermal performance of it. Gerry

Thats useful to know, it certainly feels like they should be cooled somewhat. Having tested my RFS in its new case the front panel seems to be enough, but it wont hurt to have a heat sink on there. Thanks for posting the info. Gerry

Really interesting video Gerry, thanks for uploading.

Hi Gerry - if RU-vid had a 'vote for your favourite video ever' button - I would click on this video! All my hot topics in one session - synchronisation, electronics, practical stuff - truly a great video! Cheers!

I certainly could do that. I have been running it around the clock for a week and a half now and its rock solid and really is not getting above about 36 degrees. An over temperature shutdown would be a pretty good thing to add, I will bear that in mind next time I open it up and hack around. Thanks for the comments

looks really nice ! - great vid as always

A really nicly executed projrct Gerry, love it when a bunch of surplus bits get reused. Pity about the 80mm hole saw. The issue with the type you have is it relies on the 1/4" drill as a center which is OK in timber etc but the drill will flex when drilling metal in all but the most rigid drill presses or a milling machine. The best solution is to source a holesaw designed for metal work. These have very shallow depth of cut so the pilot drill cant flex.

Remember children. Don't do like your uncle Gerry and always unplug your mains power cable before sticking your hands inside any electronic device. :D

Thank you for your feedback. Gerry

Hi Patrick, You are welcome, thank you for watching. Gerry

Your welcome, thanks for the feedback. Gerry

Thank you for your comments. Gerry

Thanks for your comments. Gerry

For anyone else doing the 75 to 50 Ohm conversion. If you don't have 50 Ohm resistors, then a 150 Ohm resistor piggy backed onto the original 75 Ohm resistor will give you 50 Ohms.

BRavo!! Job Well done, thoroughly enjoyed it very complete! All the Best - Lloyd - DigiTek

I love the work you do. Thank

Hi Tom, its only the small physics package inside that needs to be kept at temperature, the rest of the electronics and the outer casing definitely does not - although the case does have some thermal attachment to the physics package. My aim with the heat sink is to help lower the temperature of the RFS outer case and the fan to keep the enclosures internal ambient temperature down. If the SMPS is run at 100 degrees plus it will die very quickly. Gerry

Thank you very much for the tip off about MPLAB X. I too am a Mac user, but I didn't know about this. I should now have a chance at getting back into PIC rather than using AVR exclusively.

Interesting, thanks for your reply, not something I had thought of!

Thanks, to be honest, I was inspired by the find of the Video Amp which I cannot myself take credit for, someone else pointed me at it.

@OtakuSamel it was easy for me to use the PIC, it took as long as it did in the video, and 5 mins to wire the ICSP to the chip on the pro to board - and i had more options :)

i have absolutely no use for one of these, but this was an interesting exercise in designing an enclosure, subscribed for more :P

Yes you are right, they are 70ohm, I had a look yesterday. However, it really does not matter for this application. The case was only £20 + £10 shipping (in the UK). for what you get I thought they were pretty cheap really. Thank you for the feedback. Gerry

I have a few Q-Max cutters too, although not an 80mm one. The problem is they are only rated at 16 gauge (about 1.5mm) max, the Aluminium on the case was 3mm thick. Gerry

Yes possibly, I will only know how well this works once I start using it. None the less, because of the switch mode I need airflow so the approach seemed to be the simplest and should work fine.

This is very interesting.

Hi Gerry, thank you very much for taking the time to produce this and provide this information. An excellent project 👍. Do you recall what size SMD resistors you used

To be honest, I was in a rush because I was on my way to Ibiza so I wired it but did not test it. The 1pps pulse that comes out on pin 5 of the DB9 connector is is a very narrow pulse and not easy to see at a glance, I will test that, I suspect it will be fine.

Had a look today and there really is not enough space on the front for connectors, but its easy enough to have a floating BNC cable plugged in the back ready for use on the bench. The 1Hz output is working fine, although it was noisy (I have fixed now), I will do a short video update on what I found and what I did about it incase anyone else runs into the problem. Gerry

@otakusanel - see my other comments a 555 would not have done the other things I originally planned. Messing about with caps/resistors to get the right rate and mark space is just too much like hard work. The micro was much easer to have up and running and I coulld be precise about the flash rate and duty cycle. If you find a 555 easier that's fine by me, for me though a PIC was what I chose to use

Excellent, enjoyed the video and idea, reuse of existing electronics for a useful test reference source. You got my attention Gerry. KC5WS

The manuals for these units always recommend heat sinking. They dissipate several watts, and if you just run it on the bench, not mounted to anything, it will get itself above the recommended max temperature pretty quickly.

Clear and upbeat presentation, but don't think twice of making retakes if you feel you can streamline the video in general (taking out that mental block for example). You have clear video quality, clean audio and all that jazz... Subscribed! Great to see another great electronics channel on RU-vid :)

thats true, but it was faster for me to use a PIC and write the code than it would have been to work out the component values for the 555 and the topology for the 555 in this application. Apart from that, I was originally going to generate a PWM signal to control the fan speed until I realised I had no 4-wire fan to hand. I was also going to drive the micro with the 10Mhz and derive a 1pps signal until I was reminded that there was already one on pin 8 of the DB9 connector.

Good choice on the uC for the LED. Funny how people think an 85 cent part is overkill instead of using 3 times as many parts and 2 times the price. Have you considered adding a thermister or thermocouple, and an overtemp LED and/or overtemp shutdown of the RFS? Seems to me you could add a MOSFET or relay to the 15V supply of the RFS and it's 5V buck regulator that you could switch it on or off depending on the internal temperature level (fan failure/blockage?)

Have a look on the related blog page (link above), I have indicated what you need to search for on e-bay. I will update the page with a reference to the PSU also.

off the topic i got arduino ported to use with atmega studio , but you can port small functions of arduino to normal compiler and debugger and get it working without a lot of effort .

bit of flat wood clamped behind the steel so the centre bit has something to continue to hold it, would help a lot, thanks for the interesting build

This is a great question and one I had myself. The answer is I am not entirely sure. They "look" more like 50R than 75R sockets in appearance and I could not see a part number to check. They work well with 50R plugs are not intermittent. By rights they should be 75R but, I suspect that the socket alone would not have much of an effect in the overall transmission line performance at video frequencies. I am happy to leave them as is for this application. Gerry

Hi Rob, I have a good drill press, Meddings, big heavy floor-staing lump, but I done a really crappy job fixing the work to the table, I relied on the density of a big vice which its self was just resting on the table. The vibrations were crazy and I knew it was not going so well but, limited time meant I just carried on :) It would be nice to find a better solution to cutting these kinds of holes.

Hi hide08, thanks for the comments. I would really like to do a PIC tutorial to balance up the Arduino frenzy that seems to be out there. Don't get me wrong Arduino is a nice platform but IMHO it was designed as an education tool, you really need to be writing in "C" (or dare I say even assembly) if you want to deal with real-time applications. To do it well would need to be aligned with one series of chips so it could include developing real working example applications. I know PIC's :)

That box was a very good find. Did you get the 1Hz output working? What I am missing is front connectors, i.e. one 10MHz and one 1Hz connector. That would make it easier when you need 10MHz on the desk.

Great video Gerry, keep up the great work. I'm sure it wasn't a requirement for your lab, but I was curious about the phase relationship of the 10MHz outputs. My guess would be that they're not all in phase, is this something you've looked into at all?

Yes of course , but the port to atmel studio don't change the difficult of the use of arduino library is the same exactly code but with more debugging options(emulator,isp,and professional dev boards)plus auto-complete etc so i consider it an enhancement , but anyway to make my point better clear i will upload my work to open source svn , and my atmel emulator for windows,i was working in usb atmega devices , stepper controllers for 3d printers and usb oscilloscopes,i will try to upload that .

@ligius I would have used a bi-colour LED myself but I did not have one to hand :) please remember that this was mostly a build with what I had operation :)

Thats pretty darned nice. Did you make two, I noticed a no. 8 and 16 hehe. Also reminds me that 8 pin MC's are well worth bothering with.

The only thing that needs to get hot is the physics package. My goal is to try and cool the outer case a little to stop the inside of the case turning into an oven. Not sure how well it will work, I will need to test it. I suspect that it will draw more current to keep the PHYP at the right temperature - not sure - as you say it might not be required at all. It is possible the BNC connectors are 75ohm but they look like 50ohm, I suspect thats what they are

As @elliott1787 mentioned (thank you), I could have and in years gone by I would have. However, its not the chip, its the extra components and all the messing about, and now days a micro controller is both easy and cheap and gives you many more options and less components.

Thank you for the sub. Gerry

Hi Ed, that sounds great, I would be interested to see it. Despite having my own comment spammed out by the Arduino fans out there, I don't know that much about using Arduino, all I know is a basic outline of its history and have played with it a little. Arduino is to micro controllers what .NET is to Windows, at least thats how I see it. At the end of the day it comes down to preference, and of course my comments reflect mine only :)

I think the main point is the Arduino project started as a development environment for simple programming mainly aimed at education which of course lead to the astonishing uptake by makers and hackers. Arduino is easy to get started with but once your needs get bigger and more professional then the limitation of the tools and development environment starts to become evident. I like what Microchip have done with their development tools, nice integrated and usable dev environment on multi-platform

Ib know this video is 10 ys. old and it's been 20 years since i worked with extron but I'll share my if the top memory. The green channel should be 315mhz bandwidth, red and blue somewhere between 50 and 80 mhz and i think 100 mhz for H & V. The green channel on extron hardware is designed to carry the composite video including chroma and luma and any sync info. Awesome to see you using it. It's been a long time.

I would have laid that out differently. Frequency standard flat on the floor. Power supply vertical on the front and the fan laid out vertical on the left had side sucking in air from drilled holes and expelling heat on the right hand side via more drilled holes.

great vid, could you tell me what heat sink you ordered?

Wonderful!

He mentioned that, but that particular PIC is about $1 and requires pretty much no outside components, actually saving time. It also gives him flexibility if he wants to do more with it later since there is plenty more I/O.

I would suggest a PicKit3 or ICD3, the PIC IDE I am using is MPLAB-X. Gerry

As much as I like throwing microchips at everything I do (and I actually do, a lot of 10F2xx) this is the only time I would have chosen a 2-color LED.. Also, I'm not so sure how hot the inside would get but having the heatsink outside takes space so I would either mount the standard outside and forego the heatsink or mount it inside but on the top and supply on bottom, omitting the fan. But yours is a higher quality build anyway.

Gerry i wanted to ask ypu if the device can be used to do experiments on light frequency, i don't really know much about these things, but i have seen a video that shows how light starts to dim when you increase the frequency up to 8 mhz. they say once the frequency reaches 9 mhz the light disappears. i was wondering if you can show us that in a video if it is in your field of knowledge. many thanks and God Bless.

Nice work! One quick comment, the 75ohm BNC connectors are different from the 50ohm version; the mismatch in not an issue here, but the 50ohm center pin is thicker and might damage the 75ohm connector over time.

Hey Gerry, I am no expert at all, and I really enjoyed your video, however having seen either @eevblog or @mikeselectricstuff take one of these apart, isn't the point of all the heat to keep the rubidium source at a set temperature. Essentially what I am asking is, does the heatsink and fan not just cool something that is trying to stay hot, like connecting an air conditioner to a electric heater? Interested to hear your thoughts. Tom

According to the spec sheet the 10Mhz output is 50 OHM at 500mV

Ye I stated that before it got to that part of the video. and I dunno I think using 3 more components is much easier/faster than dealing with a micro. Yes it has options to do more stuff but that's probably not going to happen so for something as simple as a 2 state indicator using a micro, to me at least, is complete overkill. It would actually be faster doing it with a 555 in this case i think.

When you have been writing code as long as I have you will never trust the compiler to do the right thing. Its hard to know if your code would be faster without looking at the generated assembly, but as you say, speed is not an issue in this specific instance so readability wins out IMHO. Gerry

Hello Gerry-When videoing try to prevent of moving items unnecesary in front of the camera so that the vieuwer have the chance to see a thing.

Where did you buy BNC blanking plate? Can't find where I can buy it ...

Nice video, but a quick question: Wouldn't the BNC connectors themselves be rated for 75 Ohm?

I posted my response on the related blog page (link just below the video)

Those case BNC connectors are most likely 75 Ohms used in video. They look very similar to the 50 Ohm ones used for transmission lines but usually the plastic insulator thickness is different (smaller for 75Ohm ones). Damn that case is quite expensive in ebay... more expensive than the rubidium clock :(

Hi mr Gerry Sweeney! I like all your video's but this one is amazing pls make more video's.....

Gerry I have one of these frequency standards could u give me the links for the enclosure and power supply? Thanx

continued... The other dodge would be to glue and/or screw some scrap timber so the hole saw is kept centered by the timber as the saw cuts through the metal. Had to laugh the Atmel fan boys have marked one your own comments as SPAM. Just like the holesaw it's tools best suiting the job at hand and what you are comfortable with.

does the fan keep the temperatures in check btw?

Question to do with energy consumption, thermodynamics, physics in general the purpose of the heater for the rubidium is to keep it hot actually very hot look up the specifications to confirm this over 80 C 90C ?. We do agree and know the power supply needs to be kept cool. The rubidium add a specified ambient temperature will work wire so many units of power watts to keep the BTUs of heat at a constant level within its containment compartment. The larger the difference of temperature inside the heated compartment through the metal surface of the walls to the outside Will cause the heater to cycle more often and stay on at its high output wattage use rate for a longer duration. For a climate controlled office space laboratory or interior of a residential house in a hobby room at 23C this may not be too much of a problem. For a hobbyist or do it yourself or in a basement unheated or outdoor shed in the backyard or upstairs in a an in insulated unheated attic space as their hobby room this will probably be a problem at 7 C. For every degree difference of temperature between two objects the speed or rate of energy transfer increases exponentially. Third video in a row I’ve watched of Yours liking every minute of it keep them coming for hobbyist like myself to learn how to retrofit and upgrade old equipment is a gift from heaven not being able to afford a brand new high dollar equipment. 👍

Thanks Rob, believe it or not I have a really heavy duty drill press, the problem is I did not have any easy why of clamping the work securely. If I would have had more time I would have probably done a better job :( Oh well, next time. Funny about the spam mark on the comment - I presume it was the one about education - maybe I have that wrong, perhaps the Arduino platform was not designed for education?? :)

No, I made one, I have reserved the second one for a Thunderbolt GPSDO variation. :)

Another thing, I find 555s to be actually more expensive than the 10F range from Microchip. It might be a shame on me but I don't remember using a 555 in the past 8 years. Also, the pics need absolutely no external components, except the bypassing cap.

Not sure why it would be faster. There's a handy graph that gives you a cap and resistors values for a given frequency and that info you could find out in only a few moments so not sure why determining components would be such a big deal(if you don't care about duty cycle). I know it's disabled state is high which is exactly what you wanted. If you were going to do the other stuff sure I would see a point but in this case think 555 would have been faster to implement and easier to deal with.

Please get rid of that non compliant plug which is inserted into the twin 13A socket. Your bench sockets are probably already fused down to 13A or less, but the cable on that thing is unlikely to be able to survive a fault.

Like it!

50Ω is standard for test equipment. But how do you know the impedance of the rubidium output?

What are you going to convert the second box into Gerry? You really need to keep a check on continuity lol and correct me if I am wrong but aren't 50 ohm and 75 ohm bnc sockets different ?

Would it be better to isolate 10mhz with styrofoam and cool the rest of space inside?