[Nut-upsdev] Tripplite Smart Int 1000 - wrong numbers

[Charles Lepple]

[please use reply-all to keep the list CC'd, thanks!]

On Oct 6, 2017, at 1:29 PM, tuharsky at misbb.sk wrote:
> 
> Hello,
> 
> I've got Tripplite Smart Int 1000 UPS with lan4.1 serial interface. Quite sympathetic UPS, I must say from first glance. I have made a proper serial cable (btw i hate all those constructor's decisions to make distinct serial wiring for each and every UPS type from each and every manufacturer!)

Is the "lan4.1" cable different from this one? http://networkupstools.org/cables.html#_tripp_lite

> I use [tripplite] driver for NUT and I can use upsc to get report from UPS, but some of these numbers are nonsense. For example
> 
> battery.voltage    13.4
> (the UPS has 3x6V batteries in serial = 18V and it would never drop so low IMO when the UPS claims battery.charge is 99% that I also consider suspicious, after few minutes of charging from seriously depleted state, but the batteries are not new either)

This sounds very similar to the tripplite_usb driver: https://github.com/networkupstools/nut/blob/master/drivers/tripplite_usb.c#L436

> input frequency seems OK at 50,2Hz, but
> input.voltage (both with maximum and minimum) give 128-129V, where I have european grid with 230V.

Same thing, the voltages in tripplite_usb are scaled by the nominal voltage (more or less).

> Is there a problem with interpretation of these values in driver (version 2.7.4, internal 0.9.1), or have I UPS with faulty sensors? I don't deny this possibility either since I bought it second hand.

The driver hasn't been updated since 2009-2010, so it is probably that we did not know as much about the protocol then.

> 
> Could I diagnose it somehow and/or help improve the driver if necessary? E.q., send You some raw output if You tell me how?
> 
Are you comfortable with rebuilding the drivers?

I think the best thing might be to see if your UPS responds to the same protocol query command that is used in tripplite_usb:

https://github.com/networkupstools/nut/blob/master/drivers/tripplite_usb.c#L880-L896

If so, it is possible that we could reuse some of the code from that driver for the serial interface.

The protocol command has an embedded ASCII 0 character, so it might be tough to send this from the command line or a serial console program.

> I also miss values such as output.voltage, battery.runtime etc, but maybe this model simply dosen't have the numbers reported.

We have to calculate an estimate of battery.runtime in tripplite_usb, but we could try something similar.

> _______________________________________________
> Nut-upsdev mailing list
> Nut-upsdev at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsdev