Scullcom Hobby Electronics #38 - Design & Build a GPS locked Freq. Part 3

[@rayreese2]

WOW...You've once again pushed the 'hobby' up to a much higher level!!!

[@scullcomhobbyelectronics1702]

+Ray Reese Thanks.

[@G0WZB]

HA it works!!!... Local chap who's also been following this article just turned up with one of his membraine keypads and that one works So i've got 2 new ones and neither work, There both either faulty or wired different somehow. Thanks to Dave for popping around and putting me out of my misery lol

[@scullcomhobbyelectronics1702]

+Brian Burdis Glad to hear you solved the problem.

[@charlesbickenheuser8281]

WOW... Smashing Upgrade to the Project. I always learn something from You. The application of the FET's was most interesting.

[@scullcomhobbyelectronics1702]

+Charles Bickenheuser Thanks.

[@voltlog]

Nice to see you improve you projects. Thanks!

[@scullcomhobbyelectronics1702]

+VoltLog Thanks.

[@DJSolitone]

Excellent video as usual. Great idea to make your own level converter I was quite impressed being used to get an of the shelf board for that purpose ! But indeed there are just 2 MOSFET for this circuit ! Neat !! One suggestion would be to plug both the rubidium standard AND the GPS onto the oscilloscope to easily track any delta in the frequency between the 2 sources. Dave and/or Shariar both did the trick I believe at some point in the past to check accuracy between different non aligned frequency sources. Now for the phase noise appearing in the MHz range, I'd be interested to know how you would solve this issue. I don"t know about the Schmidt trigger, but I guess that you chose a component order of magnitudes above the 10 Mhz BW. I'd be interested to see a follow up video on how to clean up the output signal in order to make a good (I mean better) quality frequency standard. Also following my previous messages on Video 2, I'd like to know what is the maximum locking frequency of this GPS module. Did you try to push it into the 20 MHz area ? Anyway thanks again for the incredible quality of your videos, really among the best on KZbin. Be prepared to get A LOT of audience in the coming months !!

[@scullcomhobbyelectronics1702]

+DJSolitone Thanks for your comments. The Schmitt Trigger is the 74AC14 which is a CMOS high speed version which can handle frequencies well over 100MHz so should be OK. I will test and see if the output frequency can be pushed up to 20MHz. The phase noise seems not to be an issue below about 2 or 3 MHz. At Frequencies approaching 10MHz phase noise clearly effects the waveform and that is something I am still working on. As the project now stands it is still an accurate frequency reference source at a low cost for the hobbyist. Thanks for your kind remarks about my You Tube channel which I simply do in my spare time to share my ideas with others.

[@roelandriemens]

Great project. I learn a lot the way you teach. Thank you.

[@scullcomhobbyelectronics1702]

+Roeland Riemens Thanks.

[@ericgee6585]

Hi Lewis. New PCB made. All working so far. Still waiting on a couple of bits to finish it. Just a note, the circuit shows 2x10K resistors as the battery voltage divider, the sketch shows the correct values of 100K and 47K.

[@scullcomhobbyelectronics1702]

+Eric Gee Glad to hear you now have the new PCB. Yes the original circuit did use 2 x 10K resistors for the battery voltage divider but I later changed them to 100K and 47K so as to reduce the load on the battery. I have been busy of late but hope to film an update to this project soon. I have been working on the software to add a new feature which will show number of satellites locked on to. I will supply the new software in the next part to this project. I will also illustrate the output frequency response on a spectrum analyser and look at some filtering options.

[@scullcomhobbyelectronics1702]

+Eric Gee Eric, I forgot to mention that I have amended the circuit download to show the battery voltage divider resistor changes to the 100K and 47K. If you click on the circuit download link you will now get the latest circuit diagram. You may have to press F5 key on you keyboard as your PC may have cached the old circuit.

[@ericgee6585]

+Scullcom Hobby Electronics Thanks. Excellent news about the satellites.

[@allanpierron2296]

Hi where can i find the circuit board for the gps freq standerd? please write back. Allan

[@ericgee6585]

Bugger Lewis I've just finished the original PCB s! It's a good job I'm retired. Back to the etchant. This is a much cleaner design, so I'll start again. It is a hobby after all. Many thanks to all concerned. I miss seeing the number of locked satellites, a software tweak perhaps?

[@scullcomhobbyelectronics1702]

+Eric Gee Sorry about that Eric. I did consider adding the extra few components for the logic level converter circuit and the keypad socket on the original display PCB which would be fairly easy to do (this is still an option). However, I thought I would have another go at trying get all the components on one standard size PCB which also allowed some improvements to track layout to reduce noise. The new board also has extra pins for future options which will make it easy to add new features or upgrades later. Since we now have a keypad any future features or improvements can be easily managed by simply downloading new software updates. I will make any software upgrades available as they develop. With regards to seeing number of locked satellites I am still looking at that as a possible software upgrade. As the project stands at the moment if there is no locked satellite signal the output frequency is set at a fixed 3.33KHz, which makes it quite obvious (although this frequency can be changed in software) and can not be changed by the keypad. As a result you will only be able to change frequency with the keypad entry when the unit is locked on to a satellite. So in essence that tells you you have a locked frequency. With regards your finished PCB's, the main board with the divider IC's on we may use as an add-on later to give some additional frequency when it is not locked on to a satellite, so do not dump them in the bin! I am currently working on a construction manual for this project (which will include details on the software) and when finished will make it available in a PDF download file. Regards, Louis

[@ericgee6585]

+Scullcom Hobby Electronics Thanks for the full reply :) You were right about the Arduino software, it would not compile. The latest version, no problem.

[@scullcomhobbyelectronics1702]

+Eric Gee Yes Eric. I had to upgrade my Arduino IDE to the latest version which currently is version 1.6.7 and once I did that it compiled with no problems. That's why I though I would mention that in the video as I think many people may still be using an old version of Arduino IDE software.

[@zx8401ztv]

What a smashing frequency generator :-). i like your clever use of fets for the logic conversion, using the clamp diode was just genius :-D. It seems like there is only the final output to tinker with and it will become part of your test equipment, loverly job sir :-) Does it make tea as well ha ha :-D

[@scullcomhobbyelectronics1702]

+zx8401ztv Thanks. The logic level converter idea has been around for some time but not often used. So I can not take full credit for that. I remember from my working days when I used to deal with the Philips R & D Engineers at Eindhoven in Holland, they had some issues with logic levels problems with their I²C Bus technology and there engineers came up with the idea of using a 'N' Channel Mosfet shift the logic high levels (this was around 1997). At the time developments in IC technology was improving which resulted in the clearances on the silicon chip substrate were reducing to less than 0.5μm which limited the logic high level and supply voltages to around 3.3 volt. With regards getting the unit to make tea, nice idea - LOL. I still remember the Teasmaid (made by Goblin) I had in the 1970's which use to wake me up with a cup of tea. Wonder what happened to them!!

[@zx8401ztv]

+Scullcom Hobby Electronics Good point, i havnt seen a tea's maid for years, i suppose people grew out of the idea, it happens :-( I still think the fet with clamping diode being used for a reverse path very clever, you just found a use for it again :-D Nothing in electronics is really new, just reworking of other clever people.

[@GADELHAS82]

Great, as always!

[@scullcomhobbyelectronics1702]

+GADELHAS82 Thanks.

[@PelDaddy]

Nice upgrades. Thanks for sharing.

[@scullcomhobbyelectronics1702]

+PelDaddy Thanks.

[@dan3460]

Superb, this is a much cleaner board than the previous one. I like it. I have two questions:1- I would like to clean the output at 10Mhz, what can be done to accomplish this, what are your ideas?2- I see that you have included ublox.h in your program, is there a write up on the functionality of that library or do you know what protocol is used to communicate between the Arduino and the ublox chip.In a personal note, i'm an old fart, I loved electronics when I was young but life took me away from that path. I collect an repair antique radios, scopes and TV as a hobby, but this is the kind of things that I really like. Experimenting, blowing thing up because you put too much current and having a led blink at your will across the room. I'm always looking at your channel to see if you have posted something new.Thanks, I really appreciated.

[@scullcomhobbyelectronics1702]

+dan3460 Thanks for your comments. I am still working on cleaning up the 10MHz signal. The ublox.h library software was written specifically for this project (it did not exist before) so that we can configure/program the ublox Neo-7 module directly from the Arduino. Normally you would have to use ublox's UBX proprietary protocol to program and configure the module with a normal PC which then limits its scope in projects like this one. This new library is included (ublox.cpp and ublox.h) in the software zip file I provide in the download. I am currently working on putting together a manual for this project which will include some details. I was helped on this software by Andre Eisenbach who worked out the code so we can program the ublox module directly.

[@alangilman8015]

Most informative.

[@valordk]

Thanks for sharing! Really neat project. I learnt a new use for mosfets today :)

[@scullcomhobbyelectronics1702]

+valordk Thanks.

[@VoeViking]

New subscriber, very nice videos.

[@cmuller1441]

The problem of clock jitter propably comes from how it's generated in the gps module. It seems that signal changes (low to high and back) are synchronized with the 48MHz clock. Is it cleaner when outputting 480 000Hz for example ? (try any 48MHz / n with n an Integer >= 4)

[@cmuller1441]

The 1MHz and 2MHz tests in the video are clean... If the goal is only the number of cycles per second, the current design is ok but if you want a clean output you can just round to the closed "n" in 48MHz/n. If you want really any value with clean output you'll need a PLL I think.

[@cmuller1441]

+cyril muller not "closed" but "closest" ...

[@daque1960]

+cyril muller I was thinking something similar myself. The way the last 2 digits bounced around I think it was trying to average 10 MHz. It would be interesting to see what the counter shows with a much longer gate time of 100 seconds for example

[@scullcomhobbyelectronics1702]

+cyril muller Thanks for the comments.

[@PapasDino]

Nice project Louis. Does the Racal counter not have a reference input to feed your Rb standard into? I'm guessing you feed your reference capable test equipment from the standard (mine is a real nest with three distribution amps feeding all the boxes and radios). Always enjoy your videos and learn something new every time!

[@scullcomhobbyelectronics1702]

+Dino Papas Thanks Dino. Yes I could have feed the 10MHz Rubidium reference as an external clock to the frequency counter which would have ensured accuracy. I normally use it the re-calibrate the internal clock reference.

[@PapasDino]

+Scullcom Hobby Electronics Ok, just curious. I leave all the gear that can accept a external reference connected to my GPSDO and have the Rb as a backup (might say I'm a little OCD about frequency accuracy across the bench!). 73 - Dino

[@OCybress]

great video, a suggestion would be to use the star symbol as the back button and the hash symbol as the enter button, that to me seems more logical.

[@RobB_VK6ES]

+Weistek Engineering video blog and make the battery voltage display show with a delay and revert, saving a key

[@FlyingShotsman]

Pressing A shouldn't be necessary; just make the code watch for any numeral key. Use the star as a decimal and the A, B and C as Hz, kHz and MHz, respectively, saving a lot of zero key presses. Use D for battery check with auto reversion, or better yet, a short battery voltage display at each power up, leaving D and Hash for future expansion. Great project, Louis. Very clever logic level shifter.

[@scullcomhobbyelectronics1702]

+Weistek Engineering video blog Thanks for your comments. The choice of buttons can easily be changed in software to suit individual preferences.

[@scullcomhobbyelectronics1702]

+FlyingShotsman Thanks for the comments. I think the software will develop over time and so improvements can be made later. I am putting an instruction manual together for this project which should help those that build this project to modify it to suit their needs. When I have finished the manual, I will make it available as a PDF download file.

[@pirateman1966]

I like to build one. I think it's a real useful tool to have around. I have everything except the GPS module and the PCB. That's a lot of professional and detailed work you have put into this project. Thank you for sharing it with us. Where may I find the module for less than $59?

[@robertcalkjr.8325]

Thanks Louis.

[@scullcomhobbyelectronics1702]

+Robert Calk Jr. Thanks Robert.

Wow, what a nice project :) Now I can just think to try to "fuse" projects, ie having some larger (in terms of displayable characters, maybe that graphical OLED based one) display to show your geographical position (maybe also calculating the speed - if you are moving with the unit - from coordinate changes), using as a "regular clock", and also for a frequency reference by the way :) However I am not sure how hard would be to allow internal and external antennas as well as an option, it's not so easy to electronically switch between them (a switch would introduce just too much noise?). Anyway, I guess, this is more or less the question of the software then. Surely it won't be a small box then as with the other projects, which was kinda neat! Thanks again for the great videos, and ideas :)

[@scullcomhobbyelectronics1702]

Thanks. Good luck with your own version.

[@G0WZB]

Keeps getting better, Im glad a tube was broken on on my UV lightbox and i didnt manage to do the pcb's, I can now do the new one. Will you be adding the number of satalites found screen and gps lock led as the previous version? Good use for the * and # keys. Many Thanks again for a fantastic project. Brian

[@scullcomhobbyelectronics1702]

+Brian Burdis Thanks Brian. I am looking at possible software upgrade with regards number of satellites. See the reply I gave to Eric Gee for details.

[@djsbriscoe]

Great project. I can make this from a lot of the bits I have lying around at work. I see your schematic is made using Kicad. Do you have the PCB file available? I use Kicad for all my work and I'm willing to contribute back to you the schematic and PCB files in the latest format (V5.02) if that's OK.

[@Chris-di1xp]

Just come across this project, exactly what i am looking for as a 10mhz reference for the shack but i only need 10mhz and probably 5 or 6 outputs of the same. How easy is that to do ? Cheers and an excellent channel you have me hooked 🙂

[@hankus253]

Did I miss it or will the final PCB include the LED indicating GPS lock.

[@scullcomhobbyelectronics1702]

+hankus253 The new PCB does not have a LED indicating GPS lock. With regards to seeing number of locked satellites on the display I am still looking at that as a possible software upgrade. As the project stands at the moment, with the new PCB, if there is no locked satellite signal the output frequency is set at a fixed 3.33KHz, which makes it quite obvious (although this frequency can be changed in software) and can not be changed by the keypad. As a result you will only be able to change frequency with the keypad entry when the unit is locked on to a satellite. So in essence that tells you you have a locked frequency.

[@SandeepKumar-jj7zi]

Does this code and ublox.h also work with NEO-M8N gps module ?

[@G0WZB]

Just changed the 2 resisitors...imbarrased i didnt spot that quicker lol I tried the membraine keypad both ways and it was still mixed up keys, I've also got another membraine keypad and that did same. Im guessing that the ones i have are different to yours in some way, They look same unless mine are both faulty. Do you have a link to where you got yours from? As i say im happy with normal keypad but otheres may want to use membraine type, and have same issues as myself, Again Thanks for the info. Brian

[@scullcomhobbyelectronics1702]

+Brian Burdis Hi Brian, I am just wondering if you soldered the membrane keypad plug pins on the PCB and also the 3 wire links under the normal keypad. You can not use the same keypad socket which is used for the rigid keypad as that will give you the effect you are explaining (if you look at the print side you will see I had to reconfigure the wiring for the membrane keypad pins on the PCB). I have bought several of these 4x4 membrame keypads from a number of places and there all the same and work fine. Below are two links for them: hobbycomponents.com/keypads/166-4x4-16-key-matrix-membrane-keypad www.ebay.co.uk/itm/272059892793?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[@DJCaverna]

beautiful board :)

[@scullcomhobbyelectronics1702]

+DJ Caverna Haroldo Bradaschia Neto Thanks :)

[@dh1nbe_90]

Very nice videos and projects! Is it possible to get the gerber files for your projects? Thanks!

[@williamconkling8308]

I may have missed it but do you have a parts list for this project? I am interested in building it and I don't see capacitor voltages or resistor wattage (assuming 1/4 watt and 25V for electrolytics).bill conkling

[@scullcomhobbyelectronics1702]

Hi Bill, I did add a parts list originally to part 1 of this project but it did change a little as the project progressed. Below is a summary parts list which may help (all resistor were 0.25W 1% Metal Film and the electrolytics can either be 16V or 25V): Parts required for Part 1 of this project below: 1 x GPS NEO-7M Module 1 x Arduino Nano module 1 x 78L05 - 5 volt regulator 1 x 16x2 LCD display 1 x 74AC14 (or 74HC14) 1 x 14 pin DIL IC socket 2 x 2N7000 (N Channel Mosfet Transistor) 1 x IN4148 diode 1 x 470uF capacitor 1 x 100uF capacitor 5 x 100nF capacitors 1 x 10nF capacitor 2 x 100 ohm resistors 3 x 150 ohm resitor 1 x 220 ohm resistor 1 x 120 ohm resistor 6 x 10k resistor 1 x 47K resistor 1 x 100K resistor 1 x 4700uH inductor 1 x 4x4 Matrix Keypad 1 x Battery Holder/Snap On Connector for 6 AA 1.5 volt batteries 6 x 1.5 volt batteries 1 x CR2032 (3 volt lithium battery) 1 x 16 pin SIL socket for LCD display 1 x 8 pin SIL header for Keypad If I get a chance later I will try an put together a more detailed parts list. Busy at the moment on Part 4 of my latest Electronic DC Load Project. Regards, Louis

[@allanpierron2296]

Did you ever add the artwork for the circuit board for the the GPS Freq Stand /if so where can i find it/?

[@scullcomhobbyelectronics1702]

Hi Allan, You can download the PCB artwork from the link below (unzip the file and you will have the PCB artwork in PDF format which you can then print a transparency from on an inkjet printer to make your own PCB): www.scullcom.com/GPS_Freq_Ref_PCB_artwork.zip Checkout Part 4 of this project and look for the comments below the video for all the downloads: kzbin.info/www/bejne/h4bMh418oKebbs0 Regards, Louis

[@G0WZB]

Yes I soldered the membraine pins in and added the links but still not working. Its got to be the membraine pads that im useing as yours are working, i must have found a strange batch...i did get them cheap from china lol As i say the normal kepad is working great. I dont really want the membraine keypad but as i had them thought i would try them out of curiosity. Thanks for the advice its much appriciated

[@scullcomhobbyelectronics1702]

+Brian Burdis For the purpose of other followers. I note from your other comments below that you may have had a faulty membrane keypad as having tested a membrane keypad from your mate all now works OK.

[@G0WZB]

+Scullcom Hobby Electronics I actually had 2 keypads that did not work, Can't decide if they are faulty or maybe wired differenty inside. I cant find any info on them so im going to pull one apart when i get a chance. Just to confirm My friends keyboard worked 100% and my keyboards DID NOT work in his device....thought i was loosing the plot lol MANY thanks for the project its been great fun

[@audiotecnologia4084]

Is very good project !!!

[@scullcomhobbyelectronics1702]

Thanks.

[@YashKudale]

Awesome !

[@scullcomhobbyelectronics1702]

+Yash Kudale Thanks.

[@YashKudale]

I need some advice regradind my electronics career can you share your email or you can email me [e-mail]

[@scullcomhobbyelectronics1702]

+Yash Kudale If you wish you can ask your question in the private messaging section of You Tube.

[@G0WZB]

Last question I promise What is C7? It does not say on Schematic, I used 100nf and all seems ok. Thanks

[@scullcomhobbyelectronics1702]

+Brian Burdis Yes C7 is a 100nF. Also the inductor is a 4700μH.

[@G0WZB]

Just finnished building this project and it works great, One question. My voltage check says 13.8volts even though i'm only feeding 9 volts in.....what have i done wrong or is this a software glitch. Checked board and voltages are all as they should be so guessing its a software glitch. Many Thanks Cant wait till next installment Brian

[@scullcomhobbyelectronics1702]

+Brian Burdis Hi Brian, Nice to hear the project went well. The reason your voltage check is reading 13.8 volts is because I made an error in the schematic diagram with regards the resistor values for the resistor divider feeding the battery voltage to the Arduino A6 pin. R11 should be a 100K and R12 should be a 47K (this then effectively divides the battery voltage by 3). Let me know if that fixed your problem. I will try and change the schematic in the download I gave for this project.

[@G0WZB]

Thanks I realised late last night what the problem was. haven't got round to changing them yet but sure all will be well. One other issue I found is that normal keypad works great but the membrane pad keys are a bit mixed up. I'm assuming it's a different kepad library file for that pad? I'm not bothered about the membrane pad as I like proper key lol. Thanks for the reply and a great project I've really enjoyed this

[@scullcomhobbyelectronics1702]

+Brian Burdis I have just updated the schematic download with the correct resistor values. But if you click to download it your PC cashe may still have the old one so you may need to press key F5 on you PC to refresh that page. I have tested the membrane key pad again and mine works fine - I think you may have the membrane keypad plug in reverse. If you look at the video from point 18:06 you will see how I plugged it in then folded it back over the board. Let me know if that solves your issue.

[@daque1960]

Awesome.

[@scullcomhobbyelectronics1702]

+David Queen Thanks.

[@surplusdriller1]

for some reason i can only upload a maximum of 24MHz as T5 and its what i get out, is it any way to get around this?

[@dvhammer]

I am wanting to use this unit as a 10 MHz reference for my frequency counter but am not sure what the accuracy would be. I looked through all of the videos and comments but probably missed it. Does anyone have a ball park estimate? Thanks. Dale K9NN

[@scullcomhobbyelectronics1702]

+dvhammer As you will see on the video the frequency of the 10MHz signal is accurate (see 30:07 of video) but the wave shape has some phase distortion due to the fact that it is derived from the 48MHz oscillator. However, this timing signal can be used as a reference for a frequency counter. It will be more accurate than the frequency counters own internal clock reference. The 10MHz output from the GPS module is locked in to the GPS carrier signals which originate from an on-board atomic clock (oscillator) on the satellite which is also monitored and controlled by ground stations in the United States to agree with the Coordinated Universal Time (UTC). UTC is the primary time standard by which the world regulates clocks and time. Atomic clocks are based on the natural atomic oscillations of gases in resonant cavities. When isolated from magnetic fields, rubidium and cesium gases will resonate at specific frequencies under controlled conditions. In the next part of this project I will look at some way of cleaning up the phase distortion of the 10MHz signal. If you check the output at lower frequencies they do not have any phase distortion.

[@andreweastland9634]

What happened to the filter/buffer follow up touted in part 2?

[@scullcomhobbyelectronics1702]

Sorry Andrew I got side tracked on to other projects as a result of requests. In Part 3 and 4 of this project I moved to an updated design. I may cover filtering again at a later stage in general.

[@sauerdrops484]

I like this project and find it very interesting. But I am not sure how useful this solution is if the waveform on the scope at 5 MHz and 10 MHz shows jitter. The datasheet shows not the best jitter performance for the Neo-6 and Neo-7 modules. If so there is nothing you can do about it. But anyhow keep up the good work and lets hope there is a solution for this. I will stay tuned :-) It would be interesting to see a comparison of the modules sinewave output to your Rb standard.

[@scullcomhobbyelectronics1702]

+Sven Dibbert (LA2STA) Thanks Sven for your comments.