David J Taylor wrote: > "David Lord" <> wrote in message news:7jebcuf3303b...@mid.individual.net... > [] >> I never scoped mine but it did give a very poor ntp output by one >> of usb-serial I tried (nowhere near that without pps using serial >> port on server and much worse than you reported for yours using >> usb and Windows) and I had no output at all with other usb-serial >> I tried (that works ok connected to a modem) but was ok when >> connected back to serial port of server. I've just checked stats >> on server via serial port for reach !=377 and from 700+ it's < 20, >> which includes lack of sat's in view and INIT. My stats are at >> 30min so for radioclocks only catches about 25% when polled at 64s. >> >> David > > Well, it /could/ be the levels causing failure with a USB converter, but > it would surprise me. Much newer RS-232 kit responds, as you will know, > to TTL levels and may not have quite as much hysteresis built in as > older models did. When using a USB converter, I saw a jitter of around > 45 microseconds, compared to a jitter with a direct GPS/PPS connection > of 3 microseconds, but a jitter with a LAN connection of 110-140 > microseconds. I would suspect that figures much worse than that suggest > the PPS signal was not getting through on the DCD line. You need a > software breakout box, don't you, to see how the USB converter actually > presents such lines to the PC!
The pps via usb doesn't work with NetBSD. That is I believe lack of hardware interrupt capability via usb. Possibly latest development pps might work from usb on Linux. Problem I'm seeing is good ttl inputs (< +300mV) to usb converter not producing any output at pc whilst (1) ok using modem connected to serial input of converters. AT commands via cu work ok. (2) some erratic recogniable output with ttl input to converter (PL2303), no output with ttl input to converter (FTDI). (3) ttl outputs connected to serial inputs on server work ok from gps18x-lvc+pps (2us) (4) ttl outputs possibly a bit erratic for pps from Conrad MSF module via diy pps extractor as only 25% of polls with reach 377, but possibly this is to be expected. MSF via dsr or dcd almost 100% reach 377. So usb to pc is ok and it does appears my ttl levels at converter inputs aren't acceptable. I expect output +9v to -9V via a 1488 will be ok. I've not set that up yet as intended generating +/-10V using switched capacitor converters I ordered, but these didn't arrive and went from "in stock" to "no longer available" overnight between my order being accepted and delivery of remainder of order. I'll try this week with 1488 on +/-5V and higher from bench psu. David _______________________________________________ questions mailing list questions@lists.ntp.org https://lists.ntp.org/mailman/listinfo/questions
