Date: Fri, 22 Dec 2017 00:02:46 -0800 (PST) From: "Rodney W. Grimes" <freebsd-rwg@pdx.rh.CN85.dnsmgr.net> To: bob prohaska <fbsd@www.zefox.net> Cc: "Jukka A. Ukkonen" <jau789@gmail.com>, freebsd-arm@freebsd.org Subject: Re: Disapearing pl2303 usb serial adapter on rpi2 Message-ID: <201712220802.vBM82kZp035488@pdx.rh.CN85.dnsmgr.net> In-Reply-To: <20171222021909.GB20324@www.zefox.net>
next in thread | previous in thread | raw e-mail | index | archive | help
> On Thu, Dec 21, 2017 at 08:19:43AM -0800, Rodney W. Grimes wrote: > > > > What do you mean by unpowering both ends? > > If this is one of the serial adapters that has a power pin > > to plug in with the Gnd/Rx/Tx pins I would highly recommend > > you leave that power pin disconnected as connecting it can > > lead to power supplies in parallel and "that is a bad thing". > > -- > > Rod Grimes rgrimes@freebsd.org > > I should have written "disconnecting". The red wire wasn't in use. Good! That eliminates one source of potential "phase of moon" type problems. > In the past unplugging and replugging the pl2303's usb connector > by itself wouldn't allow FreeBSD to recognize the adapter, it was > necessary to also unplug the data pigtails. This is a very odd symptom, and leads me to a rare, but possible, cause of the problem, latch up. It is very strange that a device would not reset itself when its primary source of power is removed, but rather requires a ground path and possible low voltage input source to be removed before it properly resets. Cmos latchup could exibit that behavior, a next test would be to just remove the gnd pin from the data end next time the connection stops working and see if that clears the fault. You could also try removing and replacing the data end one pin at a time and see if the fault clears after any one of these, I am suspecting a latch condition in the FTDI TX pin output buffer caused by out of range voltage caused most likely by excess cable/capacitance induced ringing. > > The same appears true now; the adapter wasn't recognized by FreeBSD, > despite repeated disconnect/reconnect cycles of the usb connector, > but removed to an RPI3 (running Raspbian) it was recognized immediately. Perhaps damage to the I/O cell protection diodes over time has lead to an easy latch up condition. > When moved back to the RPI2 running -current, it was recognized and > at the moment seems to behave normally. For now I'll just leave it > alone to see how long it stays up. > > I've got four of these adapters, and they all seem to exhibit the > same behavior, but not to the same extent, among the four RPI2's on > my local network. Two FTDI usb/serial adapters were substituted for > the two most troublesome pl2303 devices, and all was well until the > update to r326951. > > Is there a way to determine where/what is getting stuck? There seem > to be at least three possibilities: usb, ucom and uplcom. Other usb > devices seem unaffected. > > Thanks for reading, and any suggestions. > > bob prohaska > > -- Rod Grimes rgrimes@freebsd.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201712220802.vBM82kZp035488>