PTP 700 w/ NIDU- T1 line has errors

Have a PTP-700 link in our lab with dual NIDUs (performance of the link is solid with RX signal strength at -41 dBm).

Using a T1 test set to generate QRSS test patterns with the far side T1 interface in "wireless" loopback. T1 test set is indicating mutliple errors and error types on the RX; ony stable for short periods. System Status on both microwaves indicates that T1 TDM Channel Status is Up and Interface status is "ok."

Line code is set to : "HDB3 or B8ZS" on microwaves. T1 test set is configured for ESF, B8ZS, 1.544Mbps, and is generating a clock (set to internal clock). 

Thanks for the information.

Correct selection of the clock sources is critical, but you've selected the internal clock in the test set, which should be okay.

We're happy to look at the diagnostic files if you can provide them. In the interim, here are a few basic suggestions:

  • Does the test set report normal error-free operation if you loop back the T1 traffic without connecting to the NIDU?
  • What about if you use the Copper loopback at the end nearest the testset?
  • Can you confirm that the Wireless loopback is at the ODU furthest from the test set?

Thanks, Mark

Mark,

Which diagnostics reports would you like to see?

When I do a copper loopback on the local NIDU (test set is connected to the local/slave side), I see no errors on the T1 test set. I let it run for >1 hour on copper loopback with no errors reported on the test set. The test set appears to be working fine.

When I do a wireless loopback on the T1 on the far side (master) it works, but I see errors. For example, over a 5 minute period I see 2 LOF, 5 FE, and 10 CRC errors. A small amount, but significant since my conditions are ideal. I tried changing LBO on the test set to vary signal level (worried that I might be overloading the NIDU) but I saw no change in results over the range of values in the test set.

I do have Link Mode Optimization set to TDM Traffic. System Status indicates TDM latency of 5042 microseconds.

My number is 757 686 4284 if you would like to discuss this problem. 

Steve Farthing

USCG

Mark,

I did find one discrepancy: When we entered the MAC address for one of the two radios now under test, I entered the MAC from another box rather than the box the unit came in. Is this a problem?

Steve 

Hi Steve,

Let's take the MAC address first: As far as I know, the only place you need to enter a MAC address is in the Target MAC Address attribute. This is used when Access Method = Link Access. Access Method provides protection against accidentally connecting to the wrong remote ODU. If you entered the wriong address here, the link would not establish.

The ODU reports the MAC address of the local and remote NIDUs in the TDM Local/Remote MAC Address on the TDM Configuration page, but these are read-only controls that simply report status.

You can create the field diagnostics files by pointing the web browser to: X.X.X.X/field_diags.cgi, and click on the button. You should see a dialog like this:

Please do this for both ODUs.

The field diags file allows us to see the status, configuration and recent performance of your unit. We don't get to see passwords, encryption keys or any other secrets.

You mention Link Mode Optimization; this should be set automatically to TDM when you configure the units for the TDM service.

Mark

Hi Steve,

Thanks for the field diagnostics files.

It looks like you have a lab test with two Connectorized + Integrated (Connectorgrated) ODUs, Integrated antennas selected, arranged to provide a healthy receive signal at about -50 dBm.

It's actually really difficult to do this lab test with Integrated antennas, because the propagation between units consists of reflections off the building structure, furniture and people. It is often better to create a sandwich of the two units with some RF absorbing foam in between, but not everybody has suitable foam. A better option is to use the connectors, with N-type cables and 60 dB of inline attenuators. But not everybody has the cables and attenuators.

Here's the Vector Error plot from the Diagnostics plotter:

The Receive Data Rate plot shows that modulation mode is changing continuously

The V/H Ratio is hitting extemes of -10 dB to +10 dB, indicating the presence of significant reflections.

Looking at the Spectrum Management page, we see the presence of large interfering signal:

Steve Farthing Spectrum.PNG

We don't know how big this signal is, but it's clearly bigger than the -40 dBm at the top of the scale.

Data transferred over this link will have a significant number of errors. This affects the T1 data and the end-to-end Ethernet data in the same way.

I think we can assume that your T1 tester is making an accurate measurement of errors in the TDM service.

You might want to do one of the following:

- Use the sandwich approach with some RF absorbing foam

- Get some N-type cables and attenuators

- Find a local site where you can install the ODUs a few hundred yards appart, with the antennas roughly aligned, and transmitter power configured quite low.

Also, watch out for the interference at 4430 MHz to 4500 MHz.

It will be interesting to find out if you get the lab test to work in a convenient way. Please keep us updated.

Mark.

Steve,

Have you managed to get this lab link to behave? If so, do you have any lesson learned that you can share with the community?

Thanks, Mark

Mark,

We had to ship the equipment to AK before I could make any of the changes you suggested. We will have adequate separation on Kodiak to set up the system and test it again before we do the installation.

Can you suggest a source of supply for the attenuators? We do want to establish a baseline in the lab to support the system on a more permanent basis.

Steve

Hi Steve,

For everyday lab testing, we generally use four 30 dB attenuators to provide a total insertion loss of 60 dB in the H path and 60 dB in the V path

Look for 30 dB co-axial N-type Male-Female attenuators with frequency range of at least 6 GHz. We have attenuators of various vintages in our lab, but we seem to have been buying Inmet 6N-30 most recently. They cost about $30 each, so $120 for four. Similar parts are available from many different suppliers.

Fit one attenuator directly on each of the connectors on the connectorized ODUs, and then connect between attenuators using two short N-type Male-Male cables.

Mark