about the measure rate and the nav rate

0 votes
Now, i use the M8T, and i want my receiver at a measure rate at 1Hz, and the nav rate at 1Hz. After i setted it,  i want to know the influence, maybe the positioning result is less accurate and etc, and i don't know how to verify it .
by liu168ad asked 5 days ago
by clive1 reshown 2 days ago
65 views
is this mean that the 1pps has no relation with the clockbias?。
They are handled in different ways, the drift infers the frequency of the local clock source, the 1PPS is placed against a higher MHz frequency clock, and the correlator integrates in some multiple of nominal milliseconds, for arguments sake say 20ms, when enough bias has accumulated a 19 or 21ms integration occurs to remove the bias. The tighter the clock accuracy the less frequently this occurs.
in my opinion, the adjustment of the 1PPS in every time unit is equal to the clockbias, because the definition of the clockbias in the protocol is the bias between the receiver and the timebase. May be i am wrong?
I think you're trying to apply concepts that the receiver is managing differently.

The placement of the 1PPS is reported with a quantization error based on the receiver's clocking speed and it's understanding of time.

It uses GPS to quantify the performance of the local clock (oscillator), how it deviates from the perfect frequency. The receiver calls this the drift. The measurement timebase in the receiver is nominally 1ms, for non-zero drift this will be slightly more or less than 1ms, over time this error will accumulate as a bias. If this bias accumulates to +1ms the receiver can shorten the integration time by 1ms to reduce the distance local time has walked off.

If the local clock is slow, by say 1ppm, that basically means each second in receiver time is actually 1s+1us in real time, after 1000s  the real elapsed time is 1000.001s and to remedy that 1ms is taken out of the receivers timeline, a slewing event that will occur every ~1000 seconds.

The UBX-NAV-CLOCK reports the clock bias and drift we're talking about, and defines the time the measurements in the receiver were made. This is the receiver's own estimate of when they occured, to the sub-nanosecond level, based on it's observation of the local oscillator vs timing of the GPS signals.

Based on it's same understanding of this timeline the receiver places the 1PPS on one of the edges of a multi-MHz clock derived from the same clock local source. The synchronous nature of this design means there's going to be a quantization error in this placement based on the granularity of the clock. The receiver reports how far off it is via UBX-TIM-TP
i am so sorry, maybe i don't describe the question clearly. because i don't know the meaning of the measure rate in the protocol, so i asked this question.
0 votes
I think you're trying to apply concepts that the receiver is managing differently.

The placement of the 1PPS is reported with a quantization error based on the receiver's clocking speed and it's understanding of time.

It uses GPS to quantify the performance of the local clock (oscillator), how it deviates from the perfect frequency. The receiver calls this the drift. The measurement timebase in the receiver is nominally 1ms, for non-zero drift this will be slightly more or less than 1ms, over time this error will accumulate as a bias. If this bias accumulates to +1ms the receiver can shorten the integration time by 1ms to reduce the distance local time has walked off.

If the local clock is slow, by say 1ppm, that basically means each second in receiver time is actually 1s+1us in real time, after 1000s  the real elapsed time is 1000.001s and to remedy that 1ms is taken out of the receivers timeline, a slewing event that will occur every ~1000 seconds.

The UBX-NAV-CLOCK reports the clock bias and drift we're talking about, and defines the time the measurements in the receiver were made. This is the receiver's own estimate of when they occured, to the sub-nanosecond level, based on it's observation of the local oscillator vs timing of the GPS signals.

Based on it's same understanding of this timeline the receiver places the 1PPS on one of the edges of a multi-MHz clock derived from the same clock local source. The synchronous nature of this design means there's going to be a quantization error in this placement based on the granularity of the clock. The receiver reports how far off it is via UBX-TIM-TP
by clive1 answered 4 days ago
by liu168ad selected 3 days ago
Flag
Best answer
And now, i need to do a application, i have to set the output  1PPS aligned with the GPS, but i need to know the bias between the output 1pps and the local clock reference,  how can i get the bias? thank you in advance.
Loop the signal back to EXTINT/TIMEMARK and have the receiver quantify the placement.
You can move the signal with the cable delay parameter.
The "bias" here would be in the nano-second realm. The milli-second bias reported in NAV-CLOCK is already known by the receiver and figured into the placement of the signal on the timeline. The quantization error is reported in TIM-TP.
Perhaps you should set up an experiment where you generate and measure the signals, and contrast them with the values reported by the receiver.

Perhaps use a GPS simulator and do an A-to-B comparison of the two 1PPS signals, again review the reported metrics.
Perhaps, i can't understand what you mean "Loop the signal back to EXTINT/TIMEMARK", when we do it , what is the definition of the 'bias' ?
I'm saying create a feed back loop and have the receiver report the time stamp for the signal. Better yet have some other reference and use this to measure the offset relative to that.

The receiver is running from the local clock source, it is not explicitly outputting that source.

You're going to need to do some experiments and determine if you can do what you need with the documentation and reporting the receiver provides.
maybe i need to do some experiments,perhaps i need to  think about your method for a while.
and i have another question, can the ublox report the synchronization error between the  multi-MHz clock derived from the same clock local source and the local clock reference. thank you in advance!
website banner