Author Topic: Capture 1pps on a scope  (Read 6861 times)

0 Members and 1 Guest are viewing this topic.

Offline BryanTopic starter

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Capture 1pps on a scope
« on: January 30, 2016, 10:02:05 am »
Hello:

Trying to figure out how to capture the 1pps pulse from a GPSDO on my Rigol DS1052e. I just can't seem to see it, what settings should I be using on the scope.?
-=Bryan=-
 

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4694
  • Country: au
  • Question Everything... Except This Statement
Re: Capture 1pps on a scope
« Reply #1 on: January 30, 2016, 10:19:28 am »
first off, either rise edge trigger, or time base greater than 1s,
 

Offline BryanTopic starter

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Capture 1pps on a scope
« Reply #2 on: January 30, 2016, 10:30:32 am »
Yes, that is what I have been doing and it's just not seeing anything on the scope
-=Bryan=-
 

Offline nfmax

  • Super Contributor
  • ***
  • Posts: 1556
  • Country: gb
Re: Capture 1pps on a scope
« Reply #3 on: January 30, 2016, 11:04:59 am »
You probably already know this, but many (?) GPSDOs do not output 1pps if they are not seeing enough satellites to lock properly and get a position & time fix. Because they do not know when the GPS seconds occur without that information, and the 1pps is not just the 10MHz oscillator divided by ten million!
 

Offline awallin

  • Frequent Contributor
  • **
  • Posts: 694
Re: Capture 1pps on a scope
« Reply #4 on: January 30, 2016, 11:34:35 am »
first off, either rise edge trigger, or time base greater than 1s,

normal trigger mode usually works better than auto.
 

Offline BryanTopic starter

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Capture 1pps on a scope
« Reply #5 on: January 30, 2016, 11:58:03 am »
You probably already know this, but many (?) GPSDOs do not output 1pps if they are not seeing enough satellites to lock properly and get a position & time fix. Because they do not know when the GPS seconds occur without that information, and the 1pps is not just the 10MHz oscillator divided by ten million!

Thanks, but yes it is seeing about 4-6 satellites as I am testing. Wondering if there is software setting needs to be set as I am beginning to wonder if the 1pps output is dead or disabled   It's a Nortel  NTBW50AA
-=Bryan=-
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Capture 1pps on a scope
« Reply #6 on: January 30, 2016, 10:59:58 pm »

Thanks, but yes it is seeing about 4-6 satellites as I am testing. Wondering if there is software setting needs to be set as I am beginning to wonder if the 1pps output is dead or disabled   It's a Nortel  NTBW50AA

Hi

That one puts out a pulse ever other second ....

Make sure your trigger is DC coupled and manually set to a rational voltage level. The pulse should be around 10 us wide if I remember correctly.

Bob
 

Offline BryanTopic starter

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Capture 1pps on a scope
« Reply #7 on: January 31, 2016, 12:01:32 am »
Thanks, trigger is DC coupled and I have tried the trigger above and below the trace and still nothing. According to the manual
-=Bryan=-
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Capture 1pps on a scope
« Reply #8 on: January 31, 2016, 12:58:28 am »
Hi

If you  are on one of the ECL outputs (as opposed to the others) then it likely needs an ECL termination resistor. Even with the termination, you are looking at a few hundred mv. Setting the trigger will be tough.

Bob
 

Offline DimitriP

  • Super Contributor
  • ***
  • Posts: 1288
  • Country: us
  • "Best practices" are best not practiced.© Dimitri
Re: Capture 1pps on a scope
« Reply #9 on: January 31, 2016, 01:21:24 am »
On my analog scope I can see "something happening" (it looks like a jumping dot) to the trace every second if I set the timebase to sweep sloooooooowly (one sec/div works nicely).

For a "sanity check" it might make more sense to wire up a transistor turning on a  LED that you can see from across the room all the time, anytime.
   If three 100  Ohm resistors are connected in parallel, and in series with a 200 Ohm resistor, how many resistors do you have? 
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Capture 1pps on a scope
« Reply #10 on: January 31, 2016, 01:24:32 am »
On my analog scope I can see "something happening" (it looks like a jumping dot) to the trace every second if I set the timebase to sweep sloooooooowly (one sec/div works nicely).

For a "sanity check" it might make more sense to wire up a transistor turning on a  LED that you can see from across the room all the time, anytime.

Hi

If you see it jumping, then they already have some sort of termination on the ECL. Not the textbook way to do it, but also not uncommon. You still have the problem of it being a few hundred mv of change and likely biased up at some odd voltage.

Bob
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: Capture 1pps on a scope
« Reply #11 on: January 31, 2016, 01:52:45 am »
I was wondering if the DS1052 reduced its sample rate so much when looking at very low repetition rates,that it can't resolve the 10-50 usec pulse.

A bit of number crunching seems to say that isn't happening.

I note that it is specified as a "differential" output,----maybe you can look at it in the "pseudo-differential" mode,where you probe the two sides of the output,& set the 'scope to "A plus B".
 

Offline BryanTopic starter

  • Frequent Contributor
  • **
  • Posts: 618
  • Country: ca
Re: Capture 1pps on a scope
« Reply #12 on: February 08, 2016, 12:02:29 pm »
I think the output for 1 pps may be duff on the Nortel board. Perhaps a blown output buffer. No problems scoping a 1 pps output on a FE-5680B, so don't think it's the Rigol 1052E. There is some jumpers on the board that may need to be set, but information and schematics is somewhat sparse for the GPSDO. Not that big of a deal as the primary use is for the 10Mhz. I do note that Lady heather shows the 1pps, but it may be deriving that from software based on the 10Mhz vs hardware.

Cheers
-=Bryan=-
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Capture 1pps on a scope
« Reply #13 on: February 08, 2016, 12:17:18 pm »
Hi

LH really does not know much about the presence or absence of the signals. About the most it can access is if the signal is enabled.

Bob
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3743
  • Country: ca
  • Living the Dream
Re: Capture 1pps on a scope
« Reply #14 on: February 08, 2016, 08:22:37 pm »
I suspect it is there, but the pulse is ultra short on that unit. Anyway, PP2S is rarely useful - why not do the 1PPS mod on it. The output circuitry is easy to trace through as it is all very basic stuff.

https://www.eevblog.com/forum/testgear/economical-option-for-precision-frequency-reference/msg754222/#msg754222
VE7FM
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Capture 1pps on a scope
« Reply #15 on: February 08, 2016, 11:37:52 pm »
I suspect it is there, but the pulse is ultra short on that unit. Anyway, PP2S is rarely useful - why not do the 1PPS mod on it. The output circuitry is easy to trace through as it is all very basic stuff.

https://www.eevblog.com/forum/testgear/economical-option-for-precision-frequency-reference/msg754222/#msg754222

Hi

The basic problem there is that there is a disagreement about weather it's a pulse every other second or a pulse per second. If it's narrow, it would not be anything less than 100 ns. That might take some fiddling on the scope, but you could still catch it with a 1054Z.

Bob
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3743
  • Country: ca
  • Living the Dream
Re: Capture 1pps on a scope
« Reply #16 on: February 09, 2016, 02:49:58 am »
The NTBW50AA PPS is output is most certainly PP2S(common in the telecom industry).
VE7FM
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: Capture 1pps on a scope
« Reply #17 on: February 09, 2016, 06:36:30 am »
They are very short, intentionally.. I think.

might be easy to make a pulse stretcher with parts on hand..

Google "Fat PPS" ?

old but you can see how its commonly done for situations where the 1 pps is too short for an application\.
« Last Edit: February 09, 2016, 06:41:26 am by cdev »
"What the large print giveth, the small print taketh away."
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 16561
  • Country: 00
Re: Capture 1pps on a scope
« Reply #18 on: February 09, 2016, 06:41:34 am »

You can make a pulse stretcher.. 

Google "Fat PPS" perhaps..

If it's 10-50us wide then that's not needed.



 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: Capture 1pps on a scope
« Reply #19 on: February 09, 2016, 05:25:54 pm »
Fungus,

Your image is of an "inverted PPS".

On timing-oriented modules and boards there is often (always?) a jumper or switch which allows the user to select "inverted 1PPS" and/ or other kinds of time pulses. Then the pin may default to being usually in a high condition and will momentarily go low at the beginning of each second.  That is quite common.

But the norm is the opposite with my non-timing equipment.

The change, nomatter what direction it is in, is hard to see unless you trigger on the leading or falling edge as is appropriate. Because its so short. Also, it wont show up for some time when a GPS is started cold because it takes a minimum of around 13 minutes for the GPS to acquire ephemeris information, usually longer.

Assuming access to a good quality active GPS antenna with a good sky view, I would expect to see a fix within half an hour or less, but almost never less than around 13 minutes unless the ephemeris is already in memory (AGPS, etc).

The following is from a datasheet for one module I own and this is the non-timing GPS norm.

"One-pulse-per-second (1PPS) time mark output, 3V LVTTL. The rising edge synchronized to UTC second when getting 3D position fix. The pulse duration is about 4msec at rate of 1 Hz" (actually, that is short, compared to most- most are longer- 10 usec might be more typical, not 4)

Ive never owned a timing specific GPS and my experience with GPS in the time measurement context is relatively recent (the last decade, basically).

But the main thing I wanted to say here is - they all are picky about when they will deliver the pulses. They need to be.

There have been many times that I have seen GPSs  - especially indoors, and timing pulses, etc, suddenly spring to life after some long time.  Usually right about when I was going to give up on it.

You can look at the NMEA data stream and track the process by selecting the appropriate sentences. Some indicate acquisition of SVs, and signal strength and quality, some don't.
« Last Edit: February 09, 2016, 06:59:20 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Capture 1pps on a scope
« Reply #20 on: February 09, 2016, 05:46:45 pm »
Hi

For what ever reason, the old HP standard was a 10us wide pulse for the PPS. A CDMA cell site indexes the coding on an "every other second" basis. They started out with a pulse every second. The later gear went to a pulse every other second. The 10 us generally carried over when they did.

Part of the idea of sending a narrow pulse (rather than a 50/50 duty cycle was that figuring out inverted signals was much easier. It also was fast enough / short enough that you could couple it through a transformer. If that seems a bit odd, you lucked out by not doing this stuff in the 60's.

Bob
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf