0
\$\begingroup\$

I am using the ‘Allion USB HS Device Electrical Test Fixture’ with a Lecroy oscilloscope (4 GHz, 40 GS/s) in order to make a USB upstream Near-End Signal Eye measurement (for High Speed Measurement).

The fixture connects via two 50 ohm identical coaxial cables, to the oscilloscope terminals.

The problem I am having is the following:

Even when testing with USB-Compliant devices, the voltage level of the device in the signal eye is too high or too distorted. This is not normal, since it should show without any points touching the red area.

An example of the signal eye I am getting is below:

Eye Diagram measurement of USB Certified Device

Any help on the matter?

\$\endgroup\$
3
  • \$\begingroup\$ Help on what? The circuit or the scope? \$\endgroup\$
    – Ron Beyer
    Feb 13, 2020 at 16:30
  • \$\begingroup\$ the circuit is a test fixture connected to the scope via 2 cables. allion.com/test-fixture/usb-a-b-compliance-pre-test i do not really know which of the 2 is causing the problem, so i would appreciate help on both \$\endgroup\$
    – anxiousPI
    Feb 13, 2020 at 16:31
  • \$\begingroup\$ Before trying HS eye, did you try "Test_J" and "Test_K"? Are the DC levels at 400 mV plus-minus? \$\endgroup\$ Feb 14, 2020 at 5:49

3 Answers 3

1
\$\begingroup\$

That eye looks like its levels are designed for HS operation. LS and FS modes use a higher differential voltage. This waveform looks like it's clamped by diodes between the lines.

\$\endgroup\$
0
\$\begingroup\$

The quoted Allion set of test fixtures is not designed to test HS eye with naked 50-Ohm cables. The test fixture for device measurements is enter image description here and high-speed differential probes (or two active singe-ended probes) must be used with this fixture. The fixture with two SMA connectors is used for INPUT from signal generator to test the receiver sensitivity. If you are using the signal sensitivity fixture as direct 50-ohm connection to a scope, you will have improper termination, and signals will be severely distorted.

\$\endgroup\$
2
  • \$\begingroup\$ Yes, that is true. I have used this test fixture with 2 active single-ended probes with SMA to BNC Connectors. what is wrong about it ? \$\endgroup\$
    – anxiousPI
    Feb 14, 2020 at 9:28
  • \$\begingroup\$ @anxiousPI, if you could provide a picture of your setup, I will tell what is wrong about it. So far the statement about "single-ended probes with SMA to BNC connectors" doesn't make much sense. \$\endgroup\$ Feb 15, 2020 at 0:50
0
\$\begingroup\$

The solution: I contacted the support from the oscilloscope and they mentioned the firmware version was too old and had to upgrade it since USB IF changed the impedance on the new fixtures to 100 ohms, the oscilloscope was expecting 90 ohms.

\$\endgroup\$
2
  • \$\begingroup\$ It looks like the scope support people bs you. Change from 90 to 100 is just 10%, while your signals differ by (650-400)/400 = 60%. Did you get the right eye? \$\endgroup\$ Feb 15, 2020 at 0:57
  • \$\begingroup\$ I did. It could be that some other settings that I am not aware of changed. \$\endgroup\$
    – anxiousPI
    Feb 17, 2020 at 9:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.