Tear down photos

Initial Impressions

This radio controlled clock uses the enhanced WWVB broadcast format. Even legacy WWVB receivers can lock onto the correct time within 2 or 3 minutes, but that is only under good reception conditions, and maybe facing the right direction. This new clock consistently acquires the correct time within just 3 or 4 minutes, at any time of day, and in any room in the house.
I'll guess the -SS in model 404-1235UA-SS means stainless steel. The clock is solidly constructed. It's clearly intended for wall mounted use and not as a portable clock. It requires 2 C batteries and allows 4 for double the run-time.
The fabrication quality of the clock is not overly impressive. Even radio clocks a fraction of the price use less tape and glue. However, from the outside and when mounted on a wall, it looks quite nice. No complaints.
One of the first two 1235UA clocks I bought had a second hand that was not aligned at 00 but at 59. That meant that every time it sync'd to the top of the hour it was actually a second late. If you've taken apart other pin-aligned, opto-sensing, self-setting clocks you will understand how this can happen. It's ironic that laboratory clocks are stable to a picosecond, our government can send a signal accurate to a nanosecond, a radio clock can receive the signal and know the time to a millisecond, but a defect in manufacturing / QA results in the time being off by a second.
Every time the clock sync's it agrees to within a tenth of a second with time shown by a dedicated iPhone NTP app.
The movement is made by U.T.S. who makes many of the RCC (Radio Controlled Clock) movements I've seen.
The main radio receiver decoder chipset is ES100 made by EverSet (Xtendwave). The orthogonal antenna pair mounted at 45 degrees is another clue that this chip uses the enhanced WWVB format. Earlier WWVB (AM code) receivers tend to use only a single loopstick antenna.
There are lots of test points on this PCB so this is a fun clock to experiment with. It's very easy to get a PPS (pulse per second) out of it.
There's a jumper wire on the PCB, near TP34. Perhaps I got an early prototype. We'll check next year to see if there's a new rev of the PCB.
The radio receiver / decoder chip uses a 16 MHz resonator. See near TP18.
The timekeeping / clock driver chip uses a 32 kHz resonator. See near TP25.

Detailed performance plots

The following is a 10-day plot of the error in a 1235UA UltrAtomic clock. For this test probes were attached to each of the stepper motor wires. Timing was obtained with a set of picPET timer chips which compared UTC ticks and UltrAtomic ticks. The plot shows that over 10 days the UltrAtomic clock stayed within 1/5th to 1/10th of a second. The plot also shows frequent corrections made to the clock's timing in order to keep it sync'ed.

50 days (mean -70 ms, stdev 50 ms):

ADEV+MDEV (Allan deviation, Modified Allan deviation):

TDEV (Time deviation):

Leap second video (31-Dec-2016)

How a LF radio clock or PC (NTP) or GPS receiver handles leap seconds is interesting to those who are picky about precise time. Most PC's as well as home or commercial radio "atomic" clocks ignore leap seconds. The good news is that this La Crosse 404-1235UA WWVB "atomic" clock correctly handled the positive leap second at the end of December 2016. In the video below you can see the second hand pause for an extra second before the top of the hour. Strictly speaking, a positive leap second is an extra second called 23:59:60 but for a clock with gears and analog hands pausing at the 59th second for 2 seconds before jumping to the 00th second is as good as it gets.

Pick your format: IMG_3126.MOV IMG_3126.m4v IMG_3126.mpg

Return to LeapSecond.com home page.
Comments/questions to tvb.