Yannick99

John, I don’t know for the transparency. Never try that before. Now, for the Paul’s gerber vs mine. It’s not the same. The schematic is the same but pcbs are not. So, t’s Normal if you are seeing differences. I did my own version of this with paul’s Schematic. That’s why

Ok i'll double checked. There no problem with r15. r15 is between 5v and transistor base T6. Note that surface mount pinout of 2n3904 isn't the same as a to92. http://www.ces-eshop.com/dir/image/cache/data/vn2/2N3904%20SMD-500x500.jpg

Your schematic is wrong. Look the carlson video at 11:40 and compare your schematic with his schematic. You will see that your schematic is wrong.

No problem. Have fun with this project. I'm using sot23 :)

Hi sea, ok, i got your point and you right... i maybe did a mistake here. I'll double check and i'll let you know. I just reformat my hd so i need to reinstall eagle first.

No i didn't do a component table. Sorry

I know john.... Try this: connect gpsdo1 to input and reset to default. Now the gpsdo will do at 10mhz with his load. Like that your counter will be good. After you will be able to check gpsdo2 with the counter.

It depend of your email i think. Ok, the gpsdo are drifting because of the load probably. The lcd 10,000,000.000 display have a refresh rate of 1000 seconds but it's high precision. if you read 99800000 (10mhz -20k) it's becasuse you have 99800000hz. The display doesn't lie. But the counter you are using show 10mhz because the reference it's the same in input and output.Now why the gpsdo are drifting ? Maybe the load is changing the frequency ? Maybe we can add a buffer at the output ? On my side as i remember, it wasn't drift at all...

Hi john. I dont thing to build 2 others gpsdo will be necessary. Try to register to eeblog. I create a thread. Will be easier for everyone and quicker. Comment section here is slow and hard to follow.http://www.eevblog.com/forum/projects/thread-to-he...

Hi John, have you turned off and on the gpsdo for a while ? If yes, it's probably the problem. I know the pwm value is in eeprom but at power on, after 15 minutes, all components are not exactly at same temperature or voltage as before. Best pratice is to always stay the gpsdo turn on. Sometime is't quicker to push button to reset default values and reconfig all the thing. Also, if you are in run mode at perfect 10,000,000.000 and frequency change for 9,999,995.948 without touching nothing, we have definitely a stability problem. Gps pulse not good, or ocxo is drifting. Hard to tell

All good. it's the normal operation. 9,999,999.999 is 10mhz +- 0,001 ppm

there are no problem John. It's a slow and very sensitive device. Your ocxo is just perfect. check here https://www.youtube.com/watch?v=zILwgQhjC_Q&t=88s Just moving the thing will change the frequency!!! :)

It can happen sometime one second is bouncing. It's why you have a number beginning by 8 instead of 9. You should read 999999909. Anyway the pwm read 09. So it will increase the voltage and the next count should be beginning by 9. If you never reach 10 000 000.000 it's maybe the pwm is at his maximun. This give me an idea... You are not the first person with this problem. Maybe i could do a 1.5 version and if the pwm is at maximun value the display could give a error or something like that.

https://www.instructables.com/id/Garage-Door-Opener-With-Feedback-Using-Esp8266/

Neo 6 is just fine. We only need the 1hz pulse, that's all.

Should be fine... 10mhz at 2.5v so the pwm will be able to find 10,000,000,000 Hz

Hi John, Yes neo 8 will be perfect. I tested this model myself. For OCXO it's up to you. Both will do. I also tried both. The important thing is the v ref. This one should be 0v to 5v. Some 12v model have a vref of 8v. In this case you must use an op-amp to match the uC 0-5v. So look the schematic before to be sure before to buy

The idea is good, i'll see what i can do if i release v1.5 a day :) Problem is, with assembler, to compare 2 large number isn't an easy thing.

Great job :) nice looking!!

Hi Igor, Very interesting, On my side, never happen. It'a probably a voltage related problem. dc-dc converter are cheap and handy but very noisy. So the zener diode is a good idea. thanks for letting me know. bye

Good work Mick!!! Very nice i'm impress :)

Hi, no it's not software. The problem is the gps pulse. This one come in late a bit. Look, you are only receiving 5 satellites. It can happen sometime. Look for a better antenna. Also, at 1000s scale. the pwm is only moving at 1/65535v. So it's no big deal.

Look at the bottom of this page: http://www.jrmiller.demon.co.uk/projects/ministd/f...His ocxo was 0-8v. So he did a amplifier. Good luck. You can click on i made it. And, if you want, picture :)

Hi, I saw the schematic of your ocxo and it’s a 0.3ppm 0 to 10v adjustment. This is the problem. You need an opinion amp buffer to match 0-5v to 0-10v or you can change your ocxo.

Maybe your ocxo doesn’t have a 0-5 v adjustment. You need probably an op-amp to match the pwm input of your ocxo. And by the way. Phase 2 is very easy to achieve. You can try to unsolder the pwm pin and install a potentiometer instead. You will control the frequency by this pot. If you achieve phase 3. Just take note of voltage. 10k pot should be ok.

Maybe noise on psu... it's a very sensitive device. what is the number of count achieved on display ? Is it moving ? Another possibilities is your ocxo variation is too large. So the pwm never hits the target. Always upper or lower. what is your ocxo model ?

Hi and thank you, what do you mean by sketch exactly? I think you are supposed to have all information you need ?

What is the voltage at pwm output ?

Done... available on instructable led board.rar

Sure, i'll do it when i have a bit of time. I'll let you know :)

On my board i have only one LED... but working fine. 23% of total memory to flash a LED lol

I'm using only one point of attachement. +5 et GND. Mr Carlson is doing the same thing. Look on the pcb, you will see GND. Solder the ground to this point. For the probe gnd, if you look well. You will see that the back of the board is a whole gnd. And this one is soldered to the inner copper tube as well. At the end, i suggest to put a shield on the nutshell if you want. On my side, had an accident, the probe did a short circuit because of the copper nutshell. keep going!!

No i dont sorry, you must use pdf file and perform toner transfer method or other. Or i could create gerber file for you. With those files you could sent this to pcb manufacturing and have professional result.

Sorry for delay, yYou can use regular pot and add a switch beside. So 2 regular pot, one on/off switch

Hi, capacitor are 0805. led resistor is 1k. I have only one pot with a switch. If you dont find that, no big deal, you can use a switch aside and use a regular pot :) For diode, package is sod-123. I'm using 1n914 on my side i just install one

Hi, Tantalum capacitor have polarity. Did you put the cap up side down. Normally the line is the +++

Do you see the nunber of locked satelittes ?

You must do this at hardware level. Just enter directly 1,5vpp directly. Change fusebit for cristal. And add .01cap, 10k,4,7k divider to raise the voltage above 0v. Here is the LU9GBR schematic with same code.

Very nice :) congrats! To answer to your question, you must do this at hardware level. Just enter 1,5vpp directly. Change fusebit for cristal. And add .01cap, 10k,4,7k divider to raise the voltage above 0v. Here is the LU9GBR schematic with same code. It's working

You can use the 1.4 version here from my newer instructable https://cdn.instructables.com/ORIG/FLS/7ABW/J9SVQV...

Very strange problem, this message appears when the watchdog timeout. The pps pulse reset the watchdog. which version of are you using ? Have you tried the 1.4 (atmega328 only) ? See my other gpsdo.Also, is your fuse bits are set correctly ? Be sure to uncheck CKDIV8. If not you will run 8 times slower and the watchdog will probably timeout.

Thank you :)

It's an sensitive noise amplifier probe. It can catch wire noise without touch the wire itself. Look at the video for more detail and exemple :)

All GPSDO made by LU9GBR :)

Here on instructable, just click 2 or 3 times on the schematic. You will able to download the (1657x1282) 26 KB size. Let me know if you have any problem. By the way, this one is obsolete. You can use my other instructable here https://www.instructables.com/id/GPSDO-YT-10-Mhz-Lcd-2x16-With-LED/

Here on instructable, just click 2 or 3 times on the schematic. You will able to download the (1657x1282) 26 KB size. Let me know if you have any problem.

Hi, NMEA. It read more specific GGA sentence at 9600 bps. But if, for exemple, you are using a rockwell jupirer, i can change the baudrate for 4800 instead.

Nice :) Bravo!

Hi, glad to hear that. For your question. Gps input is optional. Each second gps send information on the gps tx pin. If you put a wire from gps tx pin to gps input pin on atmega. You will see the number of satellite you are locked with.Your gps need to send data at 9600 bps. Ublox is already at 9600 by default. Just put the wire and see ? Let me know

Hi, glad to hear that. For your question. Gps input is optional. Each second gps send information on the gps tx pin. If you put a wire from gps tx pin to gps input pin on atmega. You will see the number of satellite you are locked with.Your gps need to send data at 9600 bps. Ublox is already at 9600 by default. Just put the wire and see ? Let me know

Hi, glad to hear that. For your question. Gps input is optional. Each second gps send information on the gps tx pin. If you put a wire from gps tx pin to gps input pin on atmega. You will see the number of satellite you are locked with.Your gps need to send data at 9600 bps. Ublox is already at 9600 by default. Just put the wire and see ? Let me know

Hi, glad to hear that. For your question. Gps input is optional. Each second gps send information on the gps tx pin. If you put a wire from gps tx pin to gps input pin on atmega. You will see the number of satellite you are locked with.Your gps need to send data at 9600 bps. Ublox is already at 9600 by default. Just put the wire and see ? Let me know

Cuita, after some other test i changed the schematic. I'm using the output of the 12V OCXO in the 5V in buffer before enter in uC. The modification is very easy to do... re-download schematic please. And excuse me for all of this. Thank you

Hi Cuita,Yes i told you :) 15 min to wait. It's a pre-heat waiting time to put Oscillator accurate before to start counting. Push button to pass.You will see this soon on your display:)Between you me, we are using oscillator. It's why i suggest to put fusebit in this mode. But now, thanks to you, i know that this can be a trouble. I'll add a note about this.When you do a hard reset. Uc reset... no choice, it reset. But if you was in run mode, uc will restart at same pwm value as before.Now for the LCD, i double checked my wire with my schematic. All fit but i saw something confusing. On the schematic, do you notice that pin 5 and 6 are reversed. Number and name are good. But position is reversed between 5 and 6. Check on your side it's maybe the trouble. And me i'll try to find another ...

Hi Cuita,Yes i told you :) 15 min to wait. It's a pre-heat waiting time to put Oscillator accurate before to start counting. Push button to pass.You will see this soon on your display:)Between you me, we are using oscillator. It's why i suggest to put fusebit in this mode. But now, thanks to you, i know that this can be a trouble. I'll add a note about this.When you do a hard reset. Uc reset... no choice, it reset. But if you was in run mode, uc will restart at same pwm value as before.Now for the LCD, i double checked my wire with my schematic. All fit but i saw something confusing. On the schematic, do you notice that pin 5 and 6 are reversed. Number and name are good. But position is reversed between 5 and 6. Check on your side it's maybe the trouble. And me i'll try to find another draw for this.Let me know, you can write me by email agaguk_99@hotmail.com will be more quick

So you are telling me it's working when lcd isn't connected but dont if it is ? Hummm... lcd is only a dummy in this circuit. Did you tried to add the TX cable from your gps module to RX pin 2 on atmega328 ? Probably not if you are seeing only Waiting.... Satellite locked is suppose to be at bottom of this line but it is optional naturally.I have double checked my code and i'm not understand why you don't come out of this waiting time. If you are able to load default setting, so the button works!!! Is your LCD needs a lot of power ?At waiting time, uC count 15 mins itself. No need of gps pulse. it's only a loop. So after 15 minutes it's done... So if you wait and nothing happen, uc is frozen! Why ? I dont know. Maybe a oscillator problem when lcd is connected ? Voltage ? But no you are ...

So you are telling me it's working when lcd isn't connected but dont if it is ? Hummm... lcd is only a dummy in this circuit. Did you tried to add the TX cable from your gps module to RX pin 2 on atmega328 ? Probably not if you are seeing only Waiting.... Satellite locked is suppose to be at bottom of this line but it is optional naturally.I have double checked my code and i'm not understand why you don't come out of this waiting time. If you are able to load default setting, so the button works!!! Is your LCD needs a lot of power ?At waiting time, uC count 15 mins itself. No need of gps pulse. it's only a loop. So after 15 minutes it's done... So if you wait and nothing happen, uc is frozen! Why ? I dont know. Maybe a oscillator problem when lcd is connected ? Voltage ? But no you are seeing something on the display!I don't know i'll think about that.

Great... i'll fix my schematic thank you :)On version 1.1 i changed something. Pwm value is saved only when run mode is acheved. So if you turn off the thing before that you must restart the whole process.But after, when pwm is saved, it start directly on run mode.I'm trying a new algorythm. Right now the frequency change very very sloly from 9,999,999.999 to 10,000,000.001. When frequency hit top or bottom, pwm is raised ou lowered. So most of the time we have the right frequency but we can have this more accurate.I'm logging 20 pwm last value = to 10 Mhz exactly. I do an average and i set the pwm with that and i'm restarting from there. And so on. Over time i think that will be more accurate.. I'll let you know if this is better.I also buy a metal box today. And i'm waithing for an b...

Great... i'll fix my schematic thank you :)On version 1.1 i changed something. Pwm value is saved only when run mode is acheved. So if you turn off the thing before that you must restart the whole process.But after, when pwm is saved, it start directly on run mode.I'm trying a new algorythm. Right now the frequency change very very sloly from 9,999,999.999 to 10,000,000.001. When frequency hit top or bottom, pwm is raised ou lowered. So most of the time we have the right frequency but we can have this more accurate.I'm logging 20 pwm last value = to 10 Mhz exactly. I do an average and i set the pwm with that and i'm restarting from there. And so on. Over time i think that will be more accurate.. I'll let you know if this is better.I also buy a metal box today. And i'm waithing for an brand new blue 2x16 lcd display. A day, i hope, i'll be able to finish this project :)I'll do the 2x16 version. 4x20 is too large for nothing.What you think, 2x16 is good enough ? I'd like to have one more space between HZ and frequency... 10,000,000.000HZ isn't pretty.Maybe i could use 10000000.000 HZ i dont know.

Done and tested... Hex is now available. Try it and please, tell me what you think :)

I almost finish. I did it. And i added number of locked satellite. Just add a wire between Tx from gps reciever and RX pin2 of atmega. .Hex soon

Hi Cuita, instead i'll do that for you. In a meantime i'll maybe add number of locked satellites. It's working with neo 6m but not with neo 8m... i must found why!! Probably 2x16 and 2x20 will be compatible. But code for 4x20 isn't the same memory adresses.After you could try the code and tell me what you thing about that. Change to make etc..,

i'm very glad to hear that :)Dont forget, at startup, you have a 15 min waiting time. PWM doesn't move at all. You can bypass this by pressing pushbutton. But it was not your ptoblem here.... Take care!

No need 1hz to program.Ok so you have programmed via arduino, and as per what i saw, it's working fine. It's running. 10Mhz/65536 = 152.59 Hz and that's what you have. This prove that you oscillator works well. Prove too that uC run with the OCXO clock. So far so good. The frequency will never change. Duty can change.Now... i'm seeing 2 possibilities.1- You do not have 1 Hertz pluse in uCOr2- The frequency of your OCXO is 10,000,000,000 at 2.467vOn start, pwm is at 50%... but if uC run already at 10 mhz, uC pass to next phase and pwm duty doesn't move.after phase 5, 6... duty varie very very slow. It a 16bits pwm. at phase 6, pwm change is 2... 2/65536. 5v / 65536 = 76,29uV76,29uV x 2 = 152,58 uV152,58 uV it's hard to measure with multimeter.Just put the proble on this pin, change the O...

No need 1hz to program.Ok so you have programmed via arduino, and as per what i saw, it's working fine. It's running. 10Mhz/65536 = 152.59 Hz and that's what you have. This prove that you oscillator works well. Prove too that uC run with the OCXO clock. So far so good. The frequency will never change. Duty can change.Now... i'm seeing 2 possibilities.1- You do not have 1 Hertz pluse in uCOr2- The frequency of your OCXO is 10,000,000,000 at 2.467vOn start, pwm is at 50%... but if uC run already at 10 mhz, uC pass to next phase and pwm duty doesn't move.after phase 5, 6... duty varie very very slow. It a 16bits pwm. at phase 6, pwm change is 2... 2/65536. 5v / 65536 = 76,29uV76,29uV x 2 = 152,58 uV152,58 uV it's hard to measure with multimeter.Just put the proble on this pin, change the OCXO speed. Very sensible. Your project is probably working fine.Maybe i could add a LED to monitor 1 HZ input. Red led = no hertz something like that.

Sure i can help you... How did you program the uC ?Also, check for good gps antenna. And be sure 1 Hz is working. If 1hz is missing. PWM doesn't change at all.

Hi, yes your right.... it's normal. Working fine. The duty cycle will change to set the right frequency.If you probe pin 15 after resistor. (between resistor and capacitor, pwm out) you are suppose to have a dc voltage.This one will correct the frequency of the OCXOLet me know

Thank you Jason, A contest :) Yes why not. I'll think about that thanks again