[Nut-upsuser] SU2200RTXLCD2U and protocol 4006

Charles Lepple clepple at gmail.com
Mon Jul 3 20:16:09 UTC 2017

On Jul 3, 2017, at 12:48 PM, Ambrogio Coletti wrote:
> I wouldn't use the NUT at this point because I don't think I will be able to write a software update to adapt the NUT to my needs.
> My idea was first to test the protocol via a serial terminal client directly connected to the UPS, and then write some simple code from scratch to implement the protocol and use it for my basic needs.


If you are considering writing serial code from scratch, I don't think it is that much more complicated to modify the existing NUT drivers (if that is even needed). In general, protocol revisions from Tripp Lite have not been very different from the previous versions.

First, which OS and distribution are you using?

Rudimentary support for the 4006 protocol over USB has been in the usbhid-ups driver since NUT 2.6.0, but since that information was apparently added from another project, we don't have an entry in the HCL for that protocol. So I would try the USB connection first.

Bear in mind that some UPS firmware cannot switch between USB and serial without first powering down completely (including disconnecting AC). We don't have good data on which UPSes require this low-level reset, but it is something to try.

I have an older Tripp-Lite UPS that uses a different driver, so when I run "lsusb -d 09ae:", I see the following:

$ lsusb -d 09ae:
Bus 006 Device 002: ID 09ae:0001 Tripp Lite 

Yours will probably say "09ae:4006" instead.

A note: drivers/tripplite_usb.c was written for older models which seem to have a rudimentary USB-to-serial converter built in. Some of the commands and responses match those in drivers/tripplite.c (which talks to a serial port). It would be interesting to see where your UPS fails during detection with the drivers/tripplite.c code. I seem to remember it does not have much in the way of debugging (but strace should work).

You mentioned "set the communication parameters to those indicated in the protocol spec" - is this published somewhere?

Thanks for the kind words!

- Charles

More information about the Nut-upsuser mailing list