Kitz Forum

Broadband Related => ADSL Issues => Topic started by: konrado5 on August 25, 2017, 04:37:21 PM

Title: higher power output-significantly more errored seconds
Post by: konrado5 on August 25, 2017, 04:37:21 PM

I have significantly more errors on synchronization rate about 18800 kbps with power output 18.6 dBm than on synchronization rate about 20400 kpbs with power output 18.3 dBm? On 18800 kbps I have about 200 CRC errors per hour. On 20400 kpbs I have about 30 CRC errors per hour.  How is it possible? Does DSLAM send less stable signal at higher power output?

Best regards
konrado5

Title: Re: higher power output-significantly more errored seconds
Post by: ejs on August 25, 2017, 05:49:11 PM

30 is less than 200, not more.

Title: Re: higher power output-significantly more errored seconds
Post by: konrado5 on August 25, 2017, 06:16:25 PM

ejs: it was mistake. I've edited the post.

Title: Re: higher power output-significantly more errored seconds
Post by: burakkucat on August 29, 2017, 12:03:45 AM

I have had to construct a table with the information you have provided so as to clearly see the effect --

Power (dBm)	Sync (kbps)	CRC per hr
18.3	20400	~30
18.6	18800	~200

I am unable to see any obvious correlation. Perhaps the decrease in synchronisation speed is a direct result of the increased number of CRCs.

As for your question about the DSLAM, I would think it was unlikely . . . or to put it another way, I would think it was not a designed feature. Surely the best people to ask would be the DSLAM's manufacturer?

Title: Re: higher power output-significantly more errored seconds
Post by: konrado5 on August 29, 2017, 12:09:37 AM

Unfortunately I missed SNR margin.

Power (dBm)	Sync (kbps)	CRC per hr	SNRM (dB)
18.3	20400	~30	~3
18.6	18800	~200	~6

[Moderator edited to put all the data into a table.]

Title: Re: higher power output-significantly more errored seconds
Post by: burakkucat on August 29, 2017, 12:40:48 AM

Now having the relevant SNRM values, we can see they align with the observed synchronisation speeds.

I have to ask the question: Is the 0.3 dBm change in power level particularly significant?

Let's do the calculations. The formula is --

P = 10^(X/10) mW

where P is the power (in mW) and X is the value in dB

For 18.3 dBm, P = 10^(18.3/10) mW ~ 67.6 mW

For 18.6 dBm, P = 10^(18.6/10) mW ~ 72.4 mW

Title: Re: higher power output-significantly more errored seconds
Post by: konrado5 on August 29, 2017, 12:48:06 AM

It is not significant for synchronization speed. Usually with power output 18.3 dBm I have merely 200 less synchronization rate but why I have significantly less CRC errors than on the significantly lower synchronization with significantly higher SNR margin? It is not only single observation. I always have this curious effect.

Title: Re: higher power output-significantly more errored seconds
Post by: burakkucat on August 29, 2017, 12:59:45 AM

I agree. It is a puzzle and I am unable to explain why it happens.  :(