[Nut-upsdev] Driver for Cyberpower PR2200

[Doug Reynolds]

Arjen de Korte wrote:
>> One more update.  Moments after sending this email (approximately 3 or
>> 4 minutes after pulling the UPS plug to test the OB/OL status) the UPS
>> shutdown hard, crashing all attached computers!
> 
> As you've found out the hard way by now, it is not a good idea to do
> experiments on a UPS that is powering one (or more) system(s). Unless
> you're absolutely sure that the drivers work as expected, your test load
> should be something that isn't hurt by a sudden drop of input voltage.
> 
> Now back to checking out what goes wrong. Apparently, you have a working
> driver on a Windows (?) system that is showing the right indications. You
> can really help us out, by sniffing the communication between monitoring
> program and the UPS. Microsoft is our friend here, since they provide a
> free tool that will do just what we need:
> 
> http://www.microsoft.com/technet/sysinternals/utilities/portmon.mspx
> 
> Please fire this up *before* starting the monitoring program and post the
> first couple of seconds worth of output. We'll probably be able to decode
> the protocol from that.
> 
> Best regards, Arjen
> 
> PS  Note that the 'powerpanel' driver is for the text mode CyberPower
> devices, while the 'cyberpower' driver is for the binary version of this
> protocol.

>From the description of that shutdown (the exact same type of crash/shutdown the old cpsups driver did with my UPS), it sounds like we need to reset the RTS/DTR and clear the buffer like we did with the powerpanel driver when sending a command...

The newer upses must be a more sensitive with the DTR.  I don't believe it would break the driver.