Datum 4040A Problems

[bostonman]

Also, just to elaborate, older 5045 units had capacitor issues too, but not like the one I referred to.

The older capacitors would fail by leaking from the bottom, sometimes shorting the layers of the PCB. The issue I referenced was junk (Xicon I think) capacitors exploding from the top.

In older units, maybe one capacitor would leak from age and that could short the layers. In the newer units (within 3-4 years) the Xicon caps would explode from the top. Not just one or two, every single capacitor. If ONE capacitor was a different name brand, then that would be fine, just the Xicon would be blown.

[edpalmer42]

Also, just to elaborate, older 5045 units had capacitor issues too, but not like the one I referred to.

The older capacitors would fail by leaking from the bottom, sometimes shorting the layers of the PCB. The issue I referenced was junk (Xicon I think) capacitors exploding from the top.

In older units, maybe one capacitor would leak from age and that could short the layers. In the newer units (within 3-4 years) the Xicon caps would explode from the top. Not just one or two, every single capacitor. If ONE capacitor was a different name brand, then that would be fine, just the Xicon would be blown.

These ones seem to be silver Sprague 672D.  There are two blue ones with the 'shield' logo of Nippon Chemicon.  So far, leakage is minimal.  I haven't found any board damage because the capacitors hang upside down and they're spaced away from the board.  My unit dates from 5/25/00 according to the sticker.  The middle initial is a bit hard to make out.  Maybe GSH.

Ed

[bostonman]

Maybe GSH

Yes, GSH is correct. I know the person who worked on that unit - and he's still with the company - as a "manager".

Sounds like you have the thru-hole A/B power supply. You may also be referring to the date on the inspection sticker and not the manufacturing date. After a unit was repaired, we'd place a round tamper resistant sticker (that was easy to peel off) over a screw, and place our initials and date on it. The real way to tell when the unit was built is to look at the tube serial number (and maybe the serial number on the unit itself).

The thru-hole supply (07957 if memory serves me correctly) had many larger capacitors, silver and blue, and the caps had very little room so they were packed tightly. When those went, they'd have a fish odor, but again, they'd need to fail badly in order to cause locking issues with the cesium. If bad enough, I think they'd cause alarms because a voltage would go out of range, however, those were probably 'minor alarms'. As I mentioned, some of those capacitors got bad enough to burn the PCB (and I think it was a seven layer PCB). So layers would short to each other.

The surface mount version came around the 2000s or very early 2000s. It was nearly identical, only with surface mount components making much more room for the aluminum caps that became Xicon.

After a while, the approved vendor list included Nichicon, but, at that point, the company already purchased five trillion Xicon and insisted on running out stock even though they were known to explode; thus making room for many more failures years later.

[notfaded1]

Maybe GSH

After a while, the approved vendor list included Nichicon, but, at that point, the company already purchased five trillion Xicon and insisted on running out stock even though they were known to explode; thus making room for many more failures years later.

I bet you just loved this... I suppose it meant a LOT more rework and repairs for the repair shop.  Symmetricom probably liked this... I'm not sure many others did.  I'm going to take more pictures for you soon.  Time to get the beast back on the bench!

[ArthurDent]

Here's the Monitors and Indicators pages... it's a DB-15 connector so how does the second page make sense?

Here's what the inside of the FTS6006 monitor/control unit and the 37 conductor cable looks like.

[bostonman]

Here's what the inside of the FTS6006 monitor/control unit and the 37 conductor cable looks like.

I'm curious, how did you obtain one of these? I don't believe FTS sold these separately.

[ArthurDent]

Here's what the inside of the FTS6006 monitor/control unit and the 37 conductor cable looks like.

I'm curious, how did you obtain one of these? I don't believe FTS sold these separately.

I've got 4 of these FTS 6006 units over the years, only one cable, and no cesium beam standard. I have had S/N 04 (which I sold) and have S/N 01, S/N 023, and S/N 1007 which makes me wonder how many were made.

Other FTS units I have are a 4 input "multichannel phase comparator" FTS 6101/001 S/N 121 plus 2 prototype looking units that may be time interval units marked FTS 6102 S/N 023 and 037.

[bostonman]

I'm baffled how those boxes made it out of the facility. In any case, those phase comparators are not anything special. They convert the frequency to DC, and provide a voltage to a chart recorder. If in auto, they switch channels every so often by detecting which channels have a frequency on them.

I don't now what a 6102 is. Can you post a photo?

[ArthurDent]

I'm baffled how those boxes made it out of the facility. In any case, those phase comparators are not anything special. They convert the frequency to DC, and provide a voltage to a chart recorder. If in auto, they switch channels every so often by detecting which channels have a frequency on them.

I don't now what a 6102 is. Can you post a photo?

I had tried the phase comparator and it works as you described.

Here is what I found out about the 6102 units. I.C. date codes indicate they were made in 1987. I had quickly checked the circuit board to see if I could figure out what goes on and have a photo of the input section of the circuit board. There are 2 floating BNC inputs (A & B) that go through isolation transformers to buffers then to a phase detector. From the parts used it appears the circuit might be usable from near d.c. to about 100Mhz. The phase detector gives a d.c. voltage proportional to the phase difference and seems to be followed by RC low-pass filtering feeding the op-amp, which seems to make sense. There are 2 output BNC connectors marked ‘A-B’ and ‘analog’

There are also two LM211 voltage comparators that may be used to trigger the 8-digit counter chip that drives the front panel display. You can select 3 of the 8 digits with a switch and apparently the numeric value of that 3-digit number is converted with a 12-bit D-A converter to drive a recorder or whatever. Four decade averaging periods are also selectable. There is also a 25 pin ‘D’ connector on the rear marked ‘controller’ but I’m not sure what that does or if it is necessary to use the unit as a comparator.

One of the front panels was made from a plate that has holes drilled for 6-32 screws to be used as a back plate so I assume these were engineering samples and they weren't too concerned about parts and used whatever was on hand.

[bostonman]

I know what those are too..... I'll explain more later, but they are basically the 6101.

They were confusing to use, and the red triangles on top had to do with parts-per-billion and got selected by moving one of the two knobs.

I'll give thought to their function, but, to be honest, whenever I asked questions about them, it seemed others who used them couldn't answer the questions. They were exclusively (maybe used for other purposes before my time) for oscillators. So when I began my employment, I was exclusively repairing, testing, aligning, oscillators. The first oscillators I worked on were the ones in the 4060 in the cesium pictures above, then the same type of oscillator, only different model.

[notfaded1]

From the picture it looks like the cable may be straight through on the D-37 to D-37 cable and since you've got plugged into itself it must be male on one end and female on the other correct?

B

[notfaded1]

Actually if it's straight though I can buy a premade cable and not mess with the D-37 connectors.

[notfaded1]

Can you verify it's DB-37 M to F @ArthurDent?  I might as well order a cable now since you're sending me an FTS 6006.  Pretty neat that's it's serial number 001.  The first ones gotta work you'd think. 

[ArthurDent]

The photo I showed is a M-F extension cable.

If you make your own cable using connectors with solder pins, plug the two connectors together and make sure that each wire you solder to a connector pin on one side goes to the corresponding pin on the other connector.

If you use ribbon cable and IDC connectors make sure you don’t put one of the connectors on backwards.

Premade cables might be the best way out.

[notfaded1]

I'm making good headway.  I powered it up last night again and this morning took the covers off again.  I found both ends of the DB-37 inside...

[notfaded1]

the 5030 end of the cable...

[notfaded1]

pics...

[notfaded1]

The monitor DB-15 is basically useless except for telling you the same thing the LED's on the front tell you.  Major and Minor alarms.

[notfaded1]

Video of what the 10MHz is doing:
https://photos.app.goo.gl/2rVMtYbkXEzzCvBN9

It basically is snapping up and then counting down.

Bill

[edpalmer42]

Is your counter referenced to your GPSDO?  Assuming it is, when it's hunting as shown in your video, does the frequency ever drop below 10 MHz?  If not, there's your problem.  It's looking for the lock signal, but it can't find it because the oscillator is too far off frequency and can't be electrically tuned to 10 MHz so it hunts forever.  The same thing happens in a Rb standard.

The oscillator should have a frequency adjustment on the side with the connectors.  Give it a nudge and see if the frequency sweeps closer to 10 MHz.  If it happens to lock, keep adjusting in the same direction for a bit so that you're well into the adjustment range.

Ed

[ArthurDent]

here's the marketing info on the 5030.

[bostonman]

The oscillator should have a frequency adjustment on the side with the connectors.  Give it a nudge and see if the frequency sweeps closer to 10 MHz.  If it happens to lock, keep adjusting in the same direction for a bit so that you're well into the adjustment range.

NO! Don't do that now, the oscillator is too cold. Let it sit for at least 24hrs before touching the control voltage.

[edpalmer42]

I keep forgetting that we now have someone who's familiar with this equipment.  Usually we just have to try stuff and see what happens!   

Is there an official procedure to open the loop and allow the oscillator frequency to be set while in the chassis or would this normally be done with the oscillator on the bench?

Ed

[bostonman]

Is there an official procedure to open the loop and allow the oscillator frequency to be set while in the chassis or would this normally be done with the oscillator on the bench?

I didn't mean to blow up. I was trying to avoid too many cooks and have his Cesium unit go completely out of wack.

Yes, adjusting it is simple. It does involve opening the loop, but you either need a monitor box (which he should be getting) or short/open pins (I assume this is how it's done - I've never had to manually tweak logic lines). Also, you need a frequency reference so you know how far off the oscillator is.

Your advice wasn't far off because you could just turn at random, but ideally you don't want to do that. If you turn randomly and get the oscillator close enough, and the unit locks, then tweaking it closer to center is easier. So you're not wrong, but he's getting the monitor box, and the unit hasn't been powered long, so I'd prefer he approaches it methodically.

[edpalmer42]

We might both be premature.

If the oscillator frequency does cross 10 MHz and doesn't lock, then tuning isn't required and won't help.  We'll have to wait and see what Bill says.

But wouldn't it be nice if that's all that was wrong! 

Ed