Datum PRS-50 cesium beam reference, no "power on hours"

[rockwell]

I recently got a Datum PRS-50 cesium beam reference source from eBay.
It's in a rather good shape and after replacing some leaking capactors from the power supply it's running.
The cesium lock state led comes on after only 10 minutes and the "critical alarm" led is off.
But now my question: why are in the Monitor3 program no "power on hours" shown ?
Also the "manufacture date" isn't shown.

[sahko123]

If there was a CMOS battery and internal RTC then it would explain why the data vanished.

If the CMOS battery died it would have cleared the data and maybe stopped the RTC from running again because of some strange system implementation where the battery is the only thing that powers the RTC.

[rockwell]

As far as I can see there is no battery on the processor board.
Also the very brief manual I have from this unit doesn't mention a battery.

But maybe someone else knows if this unit has a battery on board.

Anyway thans for your reply.

[rubidium]

I went through a rebuild of mine about 2 years ago (acquired for $150) - DOA, but recovered after replacing all of the leaking electrolytics. It's been running solidly ever since, although I don't leave it on 24/7. My Monitor3 output is similar. I've been through the PRS thoroughly, and there are no batteries, or provisions for batteries, whatsoever.

[rockwell]

Searched all over the Inet, found no info about this strange behaviour.
I would love to find more info about the values shown in Monitor3.
Also I tried to connect to the old DOS monitor program, but wasn't able to connect.

[rubidium]

Despite not having any "power on" indication, there is an indicator in your Monitor 3 output that does relate to tube life. According to discussions I've had with Robert Lutwak, who is the original author of Monitor3, both the DAC gain and Numeric Gain values reported in the Gain Servo section are indicators of tube health. As the cesium depletes in the tube and the beam current decreases, compensation is first made in the firmware by increasing a simple numeric multiplier, "Numeric Gain". When that value reaches 5.00, The hardware DAC Gain is increased by a factor of 2, and the Numeric Gain is reset to 1.00. Over the life of the cesium tube, the DAC Gain value will increase by factors of 2 from 2 to 128. The fact that my unit was experiencing a DAC Gain of only 8 was very gratifying; the fact that yours is showing a value of only 2 is grounds for celebration.

[rockwell]

Thanks for the explanation. So I have plenty of life left in this beauty.
Maybe there are more caps that need replacement in the future, I only changed 2 off them.

[rubidium]

In my case, I believe that the unit was taken out of telecom service due to premature power supply failure - thus sparing the tube from excessive cesium depletion. Sheer luck I would say. Before I made an offer for it, I asked the seller to capture a brief video of the front panel after applying power. The video showed very dim LEDs and also an LED pattern that repeated every 2-3 seconds - which I interpreted as a microcontroller repeatedly resetting due to brown-out conditions. So I gambled on the unit having one or more failing power supplies but hopefully still some decent tube life. Fortunately I was right. There were 13 leaking electrolytics inside, and after replacing them the unit powered up normally and achieved lock in about 20 minutes.

Perhaps your unit suffered the same fate - to your good fortune. Or, perhaps it was fitted with a replacement tube. In any event, it appears that you will get to enjoy it for years to come!

[rubidium]

Also, I reread one of your comments about assessing some of the values that appear in Monitor3. Table 2.8-3 in the March 2001 version of the User's Guide shows desirable ranges for many of the key parameters. I've enclosed it in case you don't already have it.

[rockwell]

Thanks much, for the nice read. explains a lot !

[bostonman]

I recently got a Datum PRS-50 cesium beam reference source from eBay.

Most likely that PRS50 has an old servo board. If memory is correct, that board doesn't store the same power on contents as models with new style servo boards.

That monitor screen shot you provided shows a perfectly good running unit (although the +/-15V is a bit more than I'd like to see). I don't remember the range, but it could trip an alarm if it continues to go out of range.

Also, the ION pump is 1uA, not high at all, but ideally it should be 0uA.

[rubidium]

I agree that the ion pump current should be zero, but we don't know how long the unit was powered on after some unknown period of being off in that Monitor-3 capture. Perhaps just waiting a few hours to a day longer will get that current down.

Another suggestion I would make to the OP is to replace the remaining electrolytics. These units are old, and it's not a question of if but of when the next failure will occur. I'll admit it's a pain removing the old caps, since they're mounted on a 4-layer PCB and one needs to be careful to not disrupt any internal vias. Plus they put a dab of epoxy on each cap where it butts up against the PCB - which raises the fun of desoldering to the next level.

Jim

[bostonman]

Another suggestion I would make to the OP is to replace the remaining electrolytics.

I would need to know whether the DC/DC power supply (the clam shell that's in the module internal to the PRS50) is through-hole or surface mount. Either way, the capacitors do leak, however, the surface mount power supplies had Xicon (if I remember correctly) capacitors (black color) that were notorious for leaking.

If this is the case, it will be obvious because all, or just about all, will be exploded with crud showing on the top of the cans.