It would be easy to serial shift through an SPI interface. At 1MHz, which is conservative for SPI, you would be done in a few uS, which would be more than adequate even for a mS clock.
Since my Thunderbolt Monitor already decodes the Trimble TSIP protocol, it would be easy to modify the software to drive an array of 7 segment LED displays through SPI instead of the LCD. A divider running from the 10MHz would generate a mS interrupt which would update the display. The offset caused by the slow serial data would have to be compensated, or maybe the 1PPS can be fed to the processor as well for synchronization. It seems quite feasible within the constraints of the small processor used by the TBolt Monitor. Now that I think of it, the VFD display I have been using with my TBolt Monitor is considerably faster than the LCD display and might be fast enough, at least for 100th of a second. Didier KO4BB Sent from my BlackBerry Wireless thingy while I do other things... -----Original Message----- From: "Tom Van Baak" <[email protected]> Sender: [email protected] Date: Thu, 3 Nov 2011 09:29:42 To: Discussion of precise time and frequency measurement<[email protected]> Reply-To: Tom Van Baak <[email protected]>, Discussion of precise time and frequency measurement <[email protected]> Subject: Re: [time-nuts] Precisione GPS based led clock I also have had need of a high resolution display clock for photographic time stamping. For those of you working on this make sure not to use a LCD or VFD display. The response time is too slow. Also you can rule out any sort of TV display. If using LED make sure not to multiplex the digits. This is a common trick, especially when using microprocessors, and works well for human eyes, but fails completely with high resolution photos. So having ruled out everything but direct drive LED the only other concern is to make sure the decade counters which drive the display are synchronous or at least that all the digits are latched at the same time. Otherwise you get false readouts due to ripple carry or sequential scan. Be careful using a microprocessor for this. For millisecond displays you need a total of 21 pins; for microseconds you need 42 pins. An external serial-parallel (e.g., shift register with latch) chip might be safer since neither a PIC nor an Arduino can update that many pins in one instruction. /tvb _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
