Poor Man's Rb Atomic Clock

okie dokie class is in session thank you everyone for coming and spending your time especially the last talk of the day here and hopefully we'll wander down one of the many paths and sharon experience together of what my mind can do against everything around us but uh in the vein of that kind of a crappy opener is that show up okay ish whatever now Eddie yeah everything else so my name is Russ hahndorf I've been around in this plane of existence for quite some time my contact information is Russell at hand or calm or org and my twitter handle is don't look behind you I volunteer extensively in the wireless village so if you see me around at conferences

that's typically where you would have found me where I run and design some of the software defined radio challenges when they work and other than that from those life experiences I try to move them over and two other practical components exploring new technology and new directions and things that are a little bit wonky and this is exactly one of those moments where things get a little bit wonky for me so I've recently finished defending my PhD as a so if anyone ever wants to play doctor I'm qualified but one of the big problems that I had while going through that was that little kind of attention deficit squirrel moment where there will be these random projects or random ideas

that just show up out of nowhere that drove me away from what I was exactly supposed to be working on and this turned into one of them so over the last year I'm I've got about eighty percent of my carputer rebuilt from scratch I built a PDP eight out of a raspberry pi that's done you know I got about an acre of land I got about eighty percent of an irrigation and system installed myself that was awful I'm never done with the wireless village just can't quit zero this talk is about one of those tangents which was building an atomic clock at home as poorly and quickly as possible and then there's another project which i

call I gave a talk on last year called Soho SIGINT where I'm collecting signals of interest around my house and using pattern matching to identify people well tell me when a friend has shown up at my house or not versus the postman or someone who's going to proselytize or prostitute or whatever anyways so I can gauge the level of nakedness that I'm going to answer the door and but regarding Soho SIGINT personal tangent thanks to drag horn I think he's uh put around in that project with his new revamped version of kismet because it now does everything Soho Sagan was starting to do so tip of the hat to him for doing that but anything else that

kept me from having to write my dissertation was fantastic and I do have a daughter she was a distraction in that as well but that's that's a responsibility I just can't leave her on her own I can just pick up a soldering iron and go to town on something else that's my own self detriment towards finishing that well finally I'm done so I have a void of time to fill future projects our experiments with geophone some things along those lines about whatever security haunts us by day you got to have something by night that helps you rebalance yourself well in this particular project I managed to wrap in one of my friends John krina who

likes to explore these wild random ventures every once in a while together we've made a musical Tesla coils nixie you name it clocks caller ID displays all that sort of random stuff plasma speakers just really I think a sub mission is to see how long I could leave a project on the kitchen table unfinished to see how quickly my wife gets pissed off but he joins me in these ventures and building and atomic clock at home was one of the ones that interested them so rubidium oscillators 101 this slide pretty much describes exactly at a high level the mechanics of how they work so in the you have this thing that is beautifully named the physics package and inside

that physics package you have a chamber that has one isotope of rubidium that's been gasified and then you have a rubidium lamp that's another isotope and of that same isotope of the lamp is a light detector and when you get your physics package excited and the rubidium starts generating to the tempo of a very specific temperature it oscillates and it oscillates very very controlled in a very very predictable predictable manner whereas to the point where if you use it as a time source the error in seconds is measured in hundreds of thousands of years it's not a cesium clock which costs a whole lot more to build and requires a lot more equipment and add

zeros to the the end of the cost for the build for you to do it at home and is also not accurate as accurate and this sorry is far more accurate than a rubidium one but I'm going to be long dead and this junk is going to be something for my kids to have to handle anyways so a hundred thousand years is good enough for government work so the the easiest fastest way to do this is to purchase this device is called a fe 568 OA and they come if you're going to take this particular route i'm going to give you some warnings about this they are they all have the exact same model number there are no other distinguishing

characteristics from something that you can search but they are all very different from each other all not all of them are very different from each other there's families of them that are very different from each other some of them it may only output a pulse per second some of them may be calibrated for a 10 megahertz sine wave and nothing else some may have any combination of those fine features of being able to set the frequency that is going to oscillate at on its way out or be a PPS generator or anything want along those lines and in addition to that it may have different interfaces on it as well so in the leftmost image you

can see that there's a db9 connector on it but in the right image there's some SMA connectors off the back so depending if you're going to go out and purchase one of these pay very very close attention to what it looks like and what the specified description is about it and you can buy them off of ebay without any problem and what you are ideally looking for off of the ones on ebay are ones that give you a pulse per second out one that also gives you a 10 megahertz signal out sine wave and then even more ideally one that has serial input output on it that way you can do my new calibration on the rubidium

oscillator itself this picture on the right that has everything outlined look for anything that has descriptions for pens one two four three and seven and if you see those you are golden on having to purchase that particular oscillator so what does that physics package and what does this mystery device look like on the inside well once you open it up you have some discrete electronics but you can see that there's two pieces well there's one piece of yellow at the top which is soft foam that's actually a really crappy thermal isolation from the actual physics package which is underneath that like expandable foam that you would use for insulation that is the thing that you absolutely have to

make sure keeps warm and all that sort of stuff but if it gets too hot it can create error in one of the other sub oscillators and the other component on the board now on the other ones if they don't have the PPS output or anything along those lines when you do open them up you can see that there's that blue wire that comes down to a uml connector on the motherboard sometimes you can just do that yourself and save yourself the headache or you could open it up and follow the trace is very very they're wonderfully labeled but you can follow the traces to directly connect to the data out signal out or serial out

off of the the box out of the box itself now the thing that's I got a little bit ahead of myself but i'll go ahead and explain it real quick these things actually and i'll go back to this slide to explain that part where these come from typically are recycled cell phone tower transmission sites they get sent over to China dismantled and then they somehow show up on eBay and then ship back into the US somehow that southpark episode of jewelry moving around the world kind of comes back in my head brigade but this time regarding electronics these are typically used to provide the signal and clock sourcing for TDMA and cdma-based hours they're also in GSM as well but that is what

their actual commercial uses for there are military uses for these as well you can kind of guess what might require precision guidance controls directions that can be disposable so you don't want to lose something that costs a couple hundred to low millions of dollars if it's a one and done sort of thing a couple hundred bucks is disposable so that that's another use out of these things well the but what's more interesting to me is like going it's not so much the electronics piece it's like learning how these suckers work and inside of them inside that physics package when you cut off the gold container this is exactly from that previous overall diagram I showed you

what is inside on the top there you have the photo detector in the middle you got your gas chamber and if you can see it all right there's a little beat vidiyum inside of there that's nice and cool so when it heats up its going to gasify and and the bottom is the lamp so that's that's what it looks like on the inside yes they typically use cesium based oscillators however you can I'm going to show you a comparison between this and GPS in a minute because ntp made the decision for me so my initial build objective was to sarcastically be able to answer the question from my wife do you know what time it is or at least

that's one of the most sub motivating goals was but more importantly it's I just kind of like to have one of these things at the house and once I understood how to make it work build another one and be able to like possibly do some very simple time experiments and demonstrations with my kid like you can leave one at home and then take the other one with you to the mountains for a long weekend and then come back and you show the time is different between the two because the effect of gravity on the rubidium oscillator which is a very measurable thing to be able to demonstrate and then for anyone who's ever seen the madness at my house I also

wanted to make this fit inside the rack and then have rest of my other components pull off of it so I ordered one what late one night and it took a while to get through customs and the reason why I took a while was because of the rubidium piece of it but it eventually made it through and it was labeled as amateur radio parts and the reason for that is a lot of amateur radio operators like to use these things will like quotations to use these things primarily because of the the use of the times the 10 megahertz freak sine wave frequency that comes out of it 4 gigahertz communication so anything that's like 12 gigahertz and high oh

sorry one point giga hertz and higher you really need to have a good TCO tcxo in order to drive that to be the oscillator to drive that radio these are really good cheap solutions for that well I mentioned on one the previous slides you have two voltage inputs one is a 5 volt and the other ones like a 17 to 19 volt input well the folks who mailed this over were kind enough to also send a free power 5 volt regulator with it that you just solder yourself which is just that one chip that's there and then they also gave like this little stuffed bear thing anyways i'm not sure if it fell in the

box by accident and the the packing month or not this is how mine showed up and it was fantastic it had everything that I needed it had a chopped db9 connector on it which was great had nice foam padding on the side which told me it was handled with tender loving care when it was removed and they were kind enough to throw in like a really inexpensive 5 volt regulator and these things like a few cents but it's the thought that counts well I mentioned that Cree know my friend went on this adventure as well and he bought one from the exact same supplier his did not show up as an such good condition if you can

see in the lower right hand image it looks like they there's grommets and rivets that they used to put it onto a heatsink plate and they just like cut him off angrily and he he contacted the ebay seller and then they sent him another one for free nice on them we got this one working again so we got two for one so essentially I just wanted to see what would happen if you know let's just go ahead and solder it up plug it in turn it on fingers crossed let's see if it works and the the things you got to watch out for is that you got your two different voltages so you got to apply

the right ones to the right pens otherwise you're going to have a sad a moment there is a pen that is high that goes too low when the physics package is able to receive a sense that the 10 megahertz frequency is stable because what is doing is is constantly watching itself that sine wave and if it's more than 10 megahertz it's starting to cool itself off and if it's less than 10 megahertz is starting to warm itself back up so once it's exactly at 10 megahertz it tries it's best to keep the temperature under control and it goes hey I'm locked onto a good signal source life is peachy and then once luck is achieved the pulse per

second pen starts firing out a pulse once every second so I was estatic to see that I didn't release the purple smoke there were no funny smells I got a fantastic perfect 10 megahertz a signal out of it the pen the lock pin went low and it totally worked as advertised out of the box this was going to be easy or so I thought but that was the good news and what I did notice out of it was a few other things this sucker gets hot the temperature on that meat thermometer says it's just below 110 degrees I took that picture when it didn't receive lock yet so it gets a lot hotter I measured

it at 180 degrees when it got locked when it locked onto the 10 megahertz frequency and regarding that I was kind of curious of can I prove even further that the temperature heating cooling mechanism is working correctly by forcing it to go in and out of loch by you know just dropping the heat sink on it and then secondarily to that knowing that there's some discrete electronics on the inside that are really really sensitive to the amount of heat that this oven can generate inside of it is there something I can do to control that because this thing is mounted typically on a giant sheet of metal and that's its way to convict the heat off the exit

seat so I just went through the parts been found low profile when you heat sink with a fan plopped it on it within like a few seconds it went out of lock and then I just found it moved it to a good spot where then went back into lock but now the average temperature of the whole package externally to the touch is about a hundred degrees I can live with this now burn my house down so the next part of the experiment was alright I i now have a working physics package I like saying that and my next thought was alright well I've got this pen that's one pulse per second I should just be

able to you know throw some bubble goo bubble gum & goo & gum and hope and dreams directly to the GPIO pin of a Raspberry Pi and start counting those receiving that pulse per second and life will be great and that was definitely not the case so the before you connect anything up you always got to measure what you're seeing and what you're receiving so I measured what my pulse per second look like out of this thing and the characteristics were is that the pulse was one micro second in length so that's really really really really quick it typically peaked out at about five volts for its pulse per second which is fairly high and then the I took a look

at the specs of what the raspberry pi can did today can detect on his GPIO pins and typically it wants a signal to last it's kind of like a imagine a bell ringing if it goes tink it might not hear it but if it goes gong and just rings out a whole lot longer there's a higher chance or a higher probability that the pie is going to pick that signal up when its internal clock goes through and goes and what's the status of me who that tickles and registered a tickle and so the pen must have received a pulse the other last part was the raspberry pi will only take on his data pens a max voltage of three volts so I

was frustrated this was not going to be as easy as like take this thing and plug it in there and life was going to be great so I amended my build objective list to not cook my raspberry pi so that was injected directly as a new thing but I needed something to compare it against so I had another GPS that did work directly off the GPIO pins of the Raspberry Pi it had a 3 volt max and it also did a PPS output so I did a quick comparison between what my rubidium clock was generating and what a GPS generates so the GPS is the image on the right you'll recognize the one from the

rubidium oscillator on the left and if you can't read the values at the bottom each interval here is one microsecond each interval here is a hundred milliseconds so significantly longer but it was exactly at three volts DC it proved my theory of like it needed to be about 100 milliseconds in length and but also at the same time I also looking at it I was like well it my previous one was a nice little as square wave as you can get for a pulse this thing isn't so that tells me that the Raspberry Pi isn't picky enough about that for its signal n for the GPIO pins alright so it doesn't have to be perfect fine I can

live with this this is going to make that part at least hopefully easier so leave it to a blonde to turn a DC signal back into an AC signal for a pulse per second and what what was happening was the pulse would hit the diode and the diode would resonate up and down up and down and then slowly get quieter but the good news was is that I saw that the diode was resonates trib increasing the length of that resonation so I was like okay I was able to draw out that that pulse to an acceptable window but damn it I now have an AC signal this isn't cool and I was a the other thing that I

noticed eventually after I fixed that part of the problem was the Raspberry Pi was now able to detect the PPS but it was picking it up on these time intervals like way way way too quickly so I'm intervals were in milliseconds of each other so it would the DDD DDD DDD DDD over and over and over so anyone who works with radio what do you think I was getting mmm what happens if you have power tripping down a line and it hits the end of that line remember then that you have to put a terminator yep in order to keep reflections from coming back oh it up elections up the pole second though huzzah I didn't cook the

pie it was picking the signal itals pickin it up too much so this is the other hint as to the solution that I ended up doing to find out really of off of the pie and not offer the rubidium oscillator itself their ground references go through not directly to the ground pins all the time through the GPIO bus so in short when I connected all this up I actually had to throw the ground reference directly back to the power supply that was powering the entire kit as opposed to being lazy about and saying uh ground is ground I could just pull it off of the pie or pull it off of the oscillator nope not

in this case I had to actually just pull it directly off of the power supply itself so now I have that nice pulse within the right voltage and it's extended into the window where things were hunky-dory peachy keen instead of seeing like that massive reflection I it would pick the signal up twice so I still had a bit of a reflection issue I got lazy and just threw it REM pod on to fix that so once I did that I was able to sew the reflection would go down it will come back and then hit back down to the PI that second one was a weaker one and the i used the zener diode to just throw anything below

that voltage level directly to ground life is peachy so all of a sudden now I am now getting pulse per second and the raspberry PI's picking it up I let it run for weeks on end I didn't get any blue or purple smoke the GPIO bus on the pi didn't fry I didn't have to move to another pin for like whacking it over and over again because if you want to think about in some regards don't do this but if you were to turn to the person next to you and just thump them once every second it gets a little irritating and the way the GPIO bus on the pies are engineered in some way is

that strong thump it they'll actually I parallel it to like getting bruised the electronics wear out from it so this was still within the tolerances of it and everything was fine so I had an intermission at this particular point in time I lost about three weeks of focus for this project due to work travel kids thinking thinking about actually writing on my PhD which I didn't but the I'll still kind of like in the back of my head I got it to extend a close to 100 milliseconds but not exactly there and I'll still worried that that was going to be a problem so I threw a tweet out nasta for some and some friends

responded there's a guy named Tim who has been a player in our wireless village he's the guy who's rocked the dog shock collar challenge anyone who wants to play that fair game and then there are some other folks that have worked in the space that responded back to me like Rob from New on and drag horn guy wrote kismet and they they gave me some suggestions of some other stuff like one of them was to use a 555 timer in order to try to get the PPS from the oscillator to trigger the 555 for it to then in turn flip the light switch on for a time interval and then turn it off so I tried a variety of these things and

they worked but when I started having to rely on other components in order to do that time delay the ultimate measure for this was well I might be ringing a buzzer or having something else reading the buzzer for me but what is ntp think the actual application ntp think of these sources so i slung it together I got ntp to receive the PPS and compare it against what the GPS was giving it so without the the oscillator the raspberry PI's internal clock jitter was worse than the GPS is but the gps is jitter was at a 15 micro seconds with the rubidium oscillator the pulse per second was point 0 1 micro seconds for jitter

so that tell if ntp goes yes this thing is stable and awesome then that meant that i now had the functions of an atomic clock using a Raspberry Pi in this external oscillator so the other thing that was interesting was later in life getting the boot up process to behave correctly and it turns out that GPS isn't also always the most reliable source for getting a pulse per second because weather phenomena can delay it your GPS receiver has a firmware in it that tries to account for that error so that's like a band-aid to a broader problem that i experienced solely because of being able to compare it against this other source well next thing was hearing from Kirino and how he

was doing with his he had very similar success and stabilizing the pulse per second getting his piety and put it we we worked in isolation but slightly pink each other solely because we wanted to see what the other solutions were possible and all that sort of stuff so we compare notes and it turns out that we were very very close to the exact same solution from a circuit standpoint but software wise we were exactly the same and we both had ntp pick up on things wonderfully so we're for me personally like I I don't I'm aid a kernel hack in order to fudge some of the razz pies detection of the PPS i don't really like

doing that ideally i'd like to use some sort of a circuit to reliably trigger that for me the ones listed there are ones that i've built and tested minus the i still haven't done the multivibrator yet where's render man when you need them but the the other component of it is is I have an external as to like do I have 10 megahertz signal or not ideally I want to use the NB 506 internally already on the same bored that i'm building for this and then the last stage is to just cram it all in a 1u rack case and ultimately like just a little bit before this presentation I was like you know what let's make

another Internet of Things thing and just quickly hack together something that gives me like a status what it uses GPS for is to go what time is it mostly oh it's 1231 all right pulse per second directly after 12 31 and then every once in a while at queries gpsd again in order to see does my 1231 still match yours and when I was solely relying on gps d being the time source for it the ntp would get angry and start saying your GPS is not precise enough for me to use it as a stratum 0 or stratum 1 time source it wanted to use GPS as stratum 2 or stratum 1 but stratum 2 preferably

because the amount of jitter that it was picking up an association with it so just quickly threw together something that would visibly render the status of the appliance yes sir it was about a hundred and eighty dollar GPS from Garmin so it's the the no no no it was a it may have had mostly due to maybe noise from the cable because it was a serial device as opposed to USB and passing data that way but there was something specifically about it that it just didn't like that I could replicate on another device as well may have dropped at one too many times yes sir rubidium oscillator on ebay you can find them anywhere between

eighty dollars to two hundred dollars and it's mostly because some people don't realize what they have and what they're selling so you may get lucky with that that's the $80 end of the spectrum the ones that really go yeah I know what I got and it's actually interesting they'll sell them for about 200 bucks when you buy them brand-new they can be a little under a thousand bucks I've seen the manufacturer sells them I've seen them with the variety of features that they have under the exact same part number starts from like 500 bucks and goes up to about a thousand bucks and these these are the devices that are and the cell phone towers that

are allowing multiple customers to share the same frequency because of tdma or cdma hey I think the TNT DMA means what anyways so it's fairly important for those sort of functions and this is just a fun little reuse of something else that's out there now the other while researching this thing I should also note that there are some other videos and other people that played with this device but it's kind of the way I described some of the write-ups is like these people bought this thing in the mail and they just wanted the lazy plug it in turn it on and everything works and then they wrote a skating renting table flipping Rivi you of the device that they got without

actually trying to understand what it was or how it worked and so so that's that information is on the end of the spectrum of the people that don't like it and then on the end of the spectrum are like some of the ham radio operators that have taken this to use in contests for gigahertz radio links and it's good enough for that where they're able to have multiple transmitters operating right next to each other without them stepping on each other with this cheap little $80 oscillator I've seen other people use them in their work benches for radio experiments and there's this one guy who's actually made this like really crazy steampunk style atomic clock that screams at you what the time

is that looks it really looks steampunk ish but his works just fine and its really really neat looking but he enjoyed playing with this device my personal opinion of reviewing some of this stuff is that people just weren't really sinking their teeth in understand what they got they got a little bit lazy with some of them or write ups or even their own research on it it isn't the that's out there but it works well enough to be able to do this other stuff if you can figure out a band-aid to apply to it if you need to and like I said they may be they are used you're getting them off of ebay so your mileage

may vary from it but don't give up or be frustrated with it the sellers will try to work with you in order to get that favorable you know 5 out of 5 star review but you can repair them you can experiment that's Alton oh so now that i got thing at home and i'm starting to do time sink in my house my my ultimate goal is to no longer have the mismatched time settings on devices or the microwave of the oven and my cell phone have like wildly different times set on them but you know the law a regular oven with a clock on as well but you know that devices since they don't take an

external time source I might be additional warrant a near future so at this point I can show you through a interesting rendering of the web page of what the current status of the system is otherwise I'll take questions if anyone has them at the moment yes sir you got to take time off yeah yes yeah I did see that on so now you can actually you're careful you know what you're looking for you yeah and it's just a little bit lucky on that yeah this is as close as easy sort of thing that you can get so like I I threw some questions on the time nuts mailing list and even back from those people and digging on the

history of it you still had widely like some people really really liked it and other people really really hated it yeah yeah if he wants to get rid of 1 i'll be having these were he forms the third and so one of the things that will help you even though you can't measure it yes take your pack your box and put it in an environment where it stays breezy yes absolutely ya know the temperature since the way the device you're absolutely right because of the way the device functions is based upon an oven that controls the vibrations of the rubidium inside of it no temperature stability is absolutely key all inside of his lab the glass wall

yeah a bunch of custom air conditioner oh yeah not like a I've an air-conditioned server room that's on its own spot I've got that I've got that covered as well I've got that covered as well there's one individual in this room that has seen the walking us of the hand or household of insanity but actually that might turn into another talk of the rabbit hole sort of situation yes sir

these devices are are forgivable so that's why I like there they are nowhere near as accurate as like a cesium clock but they are still accurate enough to within a second of error within a couple hundred thousand years so if you go with like a cesium clock you're going to be adding some zeros to the end of that hundred thousand years such a lot of zeros so regarding the temperature sensitivity of it the the operational range for it is for the oven if I remember correctly it's somewhere between 185 to 190 degrees is where it ideally wants to be within so that five degree temperature drift is more like way overly forgiving for an environment

that is typically housing one of these devices so inside inside a cabinet for a cell phone tower they do have air flow and some refrigeration that's occurring inside of it which controls it but it's also on that giant skid so based upon the thermal dynamics of the way that they engineered that cabinet they know by default they're just going to be shedding X amount of heat off into the chassis so that it just runs and sits in the Sun because of all the thermal protection between the electronics inside and the external cabinet I don't know if you've ever had the opportunity open up and see inside a cell phone tower cabinet thick doors they look

really big on the outside but there's not a lot of electronics on the inside and it's you have the Nina standard for ratings of boxes for devices that can be put outside I'm oversimplifying that but they are of a standard that can withstand like a lot of moisture a lot of temperature a lot of dust and grit yeah

so the way that those cabinets are built is that it's occupying a very small spot inside a broader space that's already air-conditioned and or doing a bigger amount of passive heating and cooling already so that's another discussion that I'll have with you if you'd like oh yeah you know it's a it's a part of the engineering of their actual installation in use because it's not it's not like you took a server rack and placed it outside there's a lot of thermal protection on those cabinets as they're sitting in the Sun I've opened one up and gotten cool air pouring out of it in the middle of the heat so it's their way more controlled on the inside than you

would imagine

so what these things are used for for the towers themselves is being able to know when how long a second is in order to support the current load of the customers off of that individual tower itself the handoff from one customer to another tower is an entirely different protocol but it doesn't matter what the time pulse is between one to the next when it does the handoff it only cares about the pulse per second off of the active transmitted transceivers off of that one tower itself in order to be able to squeeze more customers into narrower time slots but when you hand off to another one you're going to have a brand-new negotiation and new time

slot mechanism a new time source when you're over there as well cell phone towers typically pull their time off of the t1 or t3 or ds3 lines from the signal from there if the internal clocks inside those towers fail on that they also have GPS but because neither of those mechanisms are good enough for pulse per second for being able to cram like a couple hundred subscribers off of one radio so for that you need to have something that can give precision error correction and that's what these are used for and that it's instance problem having worked in that industry for a period of time cool any other questions queries posers yes if you can get to step three you

survived the blood loss yes yeah I I rant about these things on my personal website every once in a while which is hand org yeah haan do are fo RG there's a sub component of it and rhetoric / code has all the really weird random home projects some radio some other stuff as well but yeah that that's where you'll find my ramblings or just email me and i'll give you the bill of materials and here's my schematics and here's the gerber to send to ash Park for them to print you a board to solder some components on all that sort of that's a fun enjoy yes sir how did that come to that conclusion ntp just

triggering on that over and over and over and over again was the symptom and essentially when I jammed the oscilloscope on the exact same pen you would actually see the signal a boop boop boop boop boop boop boop so that was in the time period of one pulse having gone out and then there's multiple pulses still for that one pulse going out and I was like that's not supposed to be like that there's only supposed to be one of those so that was the diagnostic process that i was getting reflection so the reason why is that i'm cheap and that was the whole point of this you can see i just used the original db9 connection and just

sloppily soldered on some extensions and then just straight wires and to breadboard so there's like signal noise and loss out of the breadboard itself when I got to the point where I was burning PCBs for this a a lot of that noise just straight up goes away because you're no longer dealing with like alligator clips coming off of a lead that has maybe like four strands out of the 24 individual the copper just kind of braided out waukee like my hair is so like that that's not a good conduct conductor point so by building the PCB I got a whole lot of the background nonsense well but it's it's a really stupid simple circuit so far right now

I'm trying to figure out I've got like five of them and I'm now trying to get to the last point I'm going I want to get all of these onto one board as opposed to having like five different mount points inside the one you case I just want it to be one thing and then nicer cabling and all that sort of stuff currently right now the way that I'm detecting the the 10 megahertz signal if I pull if I throw the oscilloscope directly on the pen it says 10 megahertz exactly if I measure it off the lead that comes off to the microcontroller that is doing the comparison there is noise introduced there because that lead

and that run of the wire is also right next to the pulse per second wire which is also right next to some of the voltage wires so it's getting additional noise right next to it so my ideal solution with this board is that it slides directly into the db9 point port but one of the traces that comes off goes to an SMA connector or some sort of radio specific cable that can then go to anything external on the box if it's going to be on the board with an SMA connector or anything along those lines I haven't finished that part yet am I had mostly because hey actually it probably could is I free time but yeah

that's the there is someone who has made a bit of a daughter board that slides into it where you give it the 19 volts DC n and it's got break out for serial an LED to display the high the lock state and a two pens one for the pulse per second one for the we'll source I wish they made a made the signal source an actual SMA connector or some equivalent actual RF connector that would have made that a nicer board itself but I think they sell it for like 35 bucks if you it's a very easy google search to find any other questions queries posers alrighty well thank you for your time go out and do weird stuff

stop doing security when you go home do weird stuff