Datum 4040A Problems

[EMUD]

Hello,

now I am an owner of a Datum 4040A cesium frequency standard. But, the unit did not work, after starting it, it locked, but 2 hours later it unlocked and reported error 2 and 3. (I am using the terminal software monitor3.exe on a Win 7 computer).
After unlocking the unit did lock again after some minutes, then it unlockes again and so on. After switching it off an let it cool down the cycle above can be repeated.

Are there any ideas how to start the diagnosis?

Best regards and many thanks

EMUD

[edpalmer42]

Hello,

now I am an owner of a Datum 4040A cesium frequency standard. But, the unit did not work, after starting it, it locked, but 2 hours later it unlocked and reported error 2 and 3. (I am using the terminal software monitor3.exe on a Win 7 computer).
After unlocking the unit did lock again after some minutes, then it unlockes again and so on. After switching it off an let it cool down the cycle above can be repeated.

Are there any ideas how to start the diagnosis?

Best regards and many thanks

EMUD

There's almost no info available on the 4040A.  However, the 4040B lists errors 02 and 03 as:

02 Rabi-Ramsey error >40 mV
03 Rabi-Zeeman error >160 mV

By themselves, those don't tell me anything so .........

1.  Start reading:  https://www.eevblog.com/forum/metrology/no-luck-on-my-first-attempt-at-acquiring-a-cesium-frequency-standard/?all

2.  Post a copy of the Monitor3 Main Status Window showing the data while it's locked and a second copy showing the data when the lock fails.

Don't start tearing the thing apart yet.  Let's look over the Main Status Window data first.

What resources do you have available?  Can you confirm that the output frequency is at 10 MHz exactly or is it 'stuck' on one side or the other of 10 MHz.  Do you have a GPSDO or a calibrated Rubidium standard?

Ed

[texaspyro]

How is the ion pump current?   If it's been in storage for a while it can take weeks for it to pump down.   Worst case you may need an external high voltage supply for it to pump down... if it's even possible.

Also what about the beam current?  It could be out of cesium.   A new tube ($30K ?) may be needed.

[EMUD]

Hello,
ok, thanks for help.
I have the following instruments  :

- rubidium 10MHz systems (calibrated against a GPSDO)
- GPSDO 10 MHz and 1PPS
- 100MHz scope, 2 ch
- spetrum analyzer 8GHz
- rf generator up to 5.6 GHz
- hi res counter with ocxo

I have read the mentioned threat and the manual of the 4040B and the CSIII.
(I have tried to understand the theorie of operation in the CSIII manual,
but I have to read it several times to understand it)

The 4040A here seemed to be one of the last 4040A units, it was made 1998.

I have attached the pdfs from the working unit and the error condition.

About the ion pump : the unit was stored with pump active, the current is about 1uA.

Has anyone tried to check the tube by applying a rf generator instead of the 12,8MHz
modulation signal and test the response of it ? Could that be a way to check the health
(not how much CS is left...) of these tubes ...  The tube seemed to be somekind of a
resonator, so it could be possible to check it in this way. Yes the generator has to be a
good resolution.

Thanks

EMUD

[edpalmer42]

Your unit looks basically healthy.  Even when it loses lock, the Clock Signal (Cesium beam level) is basically constant and the level is good.  Even the Zeeman Signal stays good.  The only thing I'm questioning is the OCXO CV.  It should be between about 0.5V and 5.0V.  It's been a while since I looked at mine.  Is there a coarse tuning adjustment on your oscillator?  If so, try tweaking that to see if you can get that voltage to rise.  OCXOs drift, so it wouldn't surprise me if it needed a tweak to bring it back where it should be.  If there isn't a coarse tuning adjustment, I'd recommend removing the oscillator from the unit and powering it on the bench.  Leave it powered for a week, or a month, and see if it drifts back.  Don't try this while the oscillator is in the 4040A.  There's no way to have the oscillator powered without using up your precious supply of Cesium!  Actually, it would be useful to power the OCXO on the bench and run the EFC from one end to the other to see what the frequency range is.  However, if the OCXO needs a lot of time to drift back where you want it, you'll probably have to do something so that it's at a reasonable frequency after a short warmup so that you don't waste Cesium.

Two other things you should check are the power supply capacitors under the metal lid and the outputs of the STEL-1173.  You might consider adding a small heat sink to the 1173.  I don't know if it will help, but that chip is prone to failure and does run hot.  It can't hurt.  You should also check the outputs of the 1173 to see if it's started to fail.  You mentioned that yours seems to be a more recent unit so that might not be a problem for you - yet!

Regarding the tube testing, yes it's possible.  I discussed it in the other thread and posted the results for my tube.  You need a synthesized generator in the 12.8 MHz area that's referenced to your Rb standard, a DMM that has an input impedance of 100M or more, and a computer that can control the generator and read the value from the DMM.  If you're a real glutton for punishment you could take the readings manually.  Disconnect the tube's output and connect the DMM to the tube.  Configure things so that the tube sees a 100M ohm load.  Connect the generator in place of the 12.8 MHz source.  Sweep the generator from maybe 12.5 to 13.0 MHz and measure the result.  You should see something like the graph below.  I stepped the generator in 10 Hz steps.  Now, I wish I'd used 1 Hz steps.

This isn't a perfect test because while you're doing this, the processor is going crazy trying to change a bunch of parameters to try and find the signal that you disconnected.  But, my results seemed reasonable.  The level was significantly lower than you'd get from an HP 5061A Cesium, but the 5061A manual includes a calculation (paragraph 5-164) to check the beam signal quality.  According to that, my tube is great - if that calculation even makes sense on a different unit.

Ed

[EMUD]

Hello Ed,

thanks for your opinion about my unit. Yes, it makes sense to test the oscillator at first. I will use the following test setup on my bench:
 
24V Power supply for the oscillator and the oven
0 - 5V Simulator for the frequency control
TAIT Rubidium for the reference of my frequency counter
Frequency counter TR5212 Takeda (resolution 0.1Hz)

The oscillator is a DATUM 1111, a picture is added. It has a hole in its housing, so it may be recalibrated. I will check this.
If there are results I will post them this evening.


I have discovered, that the values displayed in monitor3 did not match the values in the real unit, the attenuator i. e. seemed to be a current controlled PIN diode type, its voltage is 0,7 to 0,4V. Do you have any idea about that?

The other thing is to check the 12,8MHz from the frequency generation (STEL ... ) with a spectrum analyzer. I think that this cound make sense to see the points where the tube is measured. (the osc frequency control should be blocked for this)

Thanks and best regard

EMUD

[EMUD]

Hello,
here are the results of my oscillator :

EFC   Frequency

+5V   9999989.9Hz
+4V   9999991.9Hz
+3V   9999993.9Hz
+2V   9999995.9Hz
+1V   9999997.8Hz
0V   9999999.8Hz
-1V   10000001.6Hz
-2V   10000003.7Hz
-3V   10000005.6Hz
-4V   10000007.6Hz
-5V   10000009.5Hz

I have checked the frequency at 0V for 4 hours, there was no drift (only the last digit jumps between 7 and 8 )
I think a recalibration is not needed.

Thanks for help

EMUD

[edpalmer42]

Your OCXO does seem to be centered around 0 volts doesn't it?  The funny thing is, I only know of a handful of units and all of them have tuning voltages that are positive. e.g. +1V3, +3V3, +2V8, +3V0.  It isn't clear to me that negative voltages are allowed.  Yes, I have seen documentation that says they are, but I've seen ranges of either +- 0V5 or +- 4V5 so I'm suspicious.  I'd still recommend that you tweak the frequency to move the EFC voltage higher.  Your symptom of lock then no lock with constant Cesium levels could easily be due to the OCXO tuning being at the end of it's range.  It's an easy test to do and can be just as easily reversed if necessary.

By the way, when the lock fails, what happens to the output frequency?

I have discovered, that the values displayed in monitor3 did not match the values in the real unit, the attenuator i. e. seemed to be a current controlled PIN diode type, its voltage is 0,7 to 0,4V. Do you have any idea about that?

Sorry, I don't understand the question.  Are you talking about the EFC voltage that controls the varactor diode in the oscillator to adjust the frequency?  If Monitor3 isn't reporting the actual EFC voltage then there's a hardware problem, but it might not affect the unit's operation.

The other thing is to check the 12,8MHz from the frequency generation (STEL ... ) with a spectrum analyzer. I think that this cound make sense to see the points where the tube is measured. (the osc frequency control should be blocked for this)

It could be tricky to look at that output on a spectrum analyzer, or a scope for that matter.  If you look at the circuit description, the processor bounces that frequency up and down many times a second to look at various characteristics of the signal to test the accuracy and health of the unit.  One time I tried to count all the frequencies.  I think I came up with something like 15 discrete values that are all tested each second (?).  I don't know of a way to stop that.  Based on these tests, the OCXO is nudged via a D/A converter to the correct frequency.  It isn't just a PLL.

You can go ahead and look at the 12.8 MHz signal.  Depending on the resolution of your spectrum analyzer, all the test frequencies could be so close together that they just look like one frequency.  Remember that since the frequencies are generated by a Numerically Controlled Oscillator and then filtered, they'll probably be rather dirty.  That won't have any effect on the unit's operation as long as the fundamental frequency is correct.

Ed

[EMUD]

Hi Ed,

the EFC seemed to be the same than the value in monitor3.  But the attenuator setting is just not the actual voltage at the attenuator (the unit between the multiplier and the tube ) ...

The osc control voltage seemed ok to me. EFC is just the same range as the often mentioned HP10811. In this unit the varicaps cathode is connected to a reference voltage of +6.4V  . So the range is from +5 to -5V specified (anode varycap).

OK, I will come back when the osc is corrected and the caps in the pwr supply are checked.

Best regards

EMUD

[edpalmer42]

I think your comparison to the HP10811 is quite correct.  Out of curiosity I pulled apart my Datum 1111B oscillator and thought I was looking at a Hewlett Packard 10811.  The aluminum block enclosing the crystal looks identical.  I didn't do a detailed reverse engineering on it, but I wondered if HP ever OEM'd the 10811 to other manufacturers.

My thought on the EFC voltage wasn't whether it would be accepted by the oscillator, but could the 4040A can actually produce a negative voltage.

I've never looked closely at the attenuator you mentioned so I can't help you there.

Ed

[EMUD]

Hi Ed,

yesterday I have disassembled the oscillator to look how to tweak it. There was a hole on the back, but no screw to adjust. I have added a picture of the oscillator pcb, but there is no variable cap and no pot to adjust. Only a pot to adjust the output voltage (the osc has a automatic level control !). The Datum people have replaced the variable cap with a normal cap ... so there is no possibility to tweak. I will examine the the path of the control voltage (I think from the BB 16bBit DAC) to the EFC in the next days.

I think the 1111B is a very similar to the HP osc, the pcbs have only minor differences.

best regards

EMUD

[edpalmer42]

Who was the idiot who said this would be an easy test??   

Okay, put everything back together.  Now, disconnect the output of the 1111B from the rest of the 4040A and use it as the external reference for your synthesized generator.  Connect the output of the generator to the point where you'd normally connect the output of the 1111B.  What you're doing is using the synthesized generator to tweak the frequency seen by the 4040A.  The 4040A will still control the frequency of the 1111B and the 1111B output will nudge the output of the generator.  The reference frequency for the generator has some wiggle room around 10 MHz.  Each type of generator has a different window that it will accept.  Often, that window isn't spec'ed.  My Stanford Research DS345 generator is spec'ed as +-2ppm (i.e. +-20Hz).

You've already determined that the 1111B shifts by 10 Hz for a 5V change in EFC.  (Note:  there appears to be a typo in the data you listed earlier:  +5V is only 1.1 Hz low - I'd expect to see more like 10.1 Hz).  Also, a higher EFC voltage results in a lower frequency.  So if you set the generator output for 10MHz + 5Hz, the 4040A will try to reduce the frequency by raising the EFC voltage.  Similarly, if you set the generator for 10 MHz - 5Hz, the 4040A will have to send a negative EFC voltage to drive the frequency higher.  This will prove whether or not the 4040A can generate negative EFC voltages or whether the -163mV shown on the Monitor3 screen is a problem.

Do you understand all that?  I'm not sure I do!    

If I got that all right, setting the generator output for 10 MHz + 5 Hz should drive the EFC to something like +2V5.  Depending on how clean the output of your generator is, you might or might not get a good lock.

Ed

[EMUD]

Hi Ed,
first I thought, what a strange idea to shift the frequency with a signal generator ... but, it works ! I have added the results the 4040a is able to generate a positive and negative EFT ! I have added the results.

I think it is fact that the osc is not guilty !  I will check the pwr supply next.

Best regards

EMUD

[edpalmer42]

Yes, that trick to shift the frequency is quite bizarre, isn't it?  I've used it once or twice for testing a system where the OCXO has drifted beyond its adjustment range.  I've also used it when I wanted an odd frequency that was tunable with an EFC voltage.  Just use a 10 MHz OCXO that has EFC as the external reference for the generator that generates the odd frequency.

The thing I find so bizarre about your OCXO is that it isn't adjustable, it's 20 years old, and the tuning range is almost exactly centered around 10 MHz.  Who needs Cesium, you've got a quartz oscillator that hasn't aged in 20 years!    But the measurements seem conclusive so yes, it's time to move on.  It's actually too bad that this wasn't the problem since the fix would be easy.

If you remove the metal plate that's inside the unit - the one that has the 5045A label on it - you'll see the capacitors that were bad in Hydrogen Maser's 4040A.  You can compare your voltage readings to the ones I posted in the other thread.  Also check for bad solder joints on the DC-DC converters.

When your unit loses lock, do you have it laying flat or is it in some other orientation?  I had mine standing on its side.  Everything was fine and then all hell broke loose!  Luckily, I was logging data at the time so I was able to capture the carnage.  The attached graph shows multiple incidents mashed together so the horizontal scale isn't significant.  The vertical axis shows the actual values of the four parameters.  Bottom line is - always keep the unit horizontal when it's operating!

Don't forget to check the outputs of the STEL-1173.  You don't have to try and analyze what they're doing, just make sure that they're toggling between 0 and 5V.  Apparently, when the chip starts to fail, the outputs randomly start failing one at a time.  Depending on which ones fail and whether they just stop or intermittently stop and start, you could see symptoms similar to yours.

Ed

[bostonman]

02 Rabi-Ramsey error >40 mV
03 Rabi-Zeeman error >160 mV

I didn't read the entire thread, but these errors indicate a poor cesium beam tube or a noisy oscillatore. What is the 9GHz microwave level? It should be around 2500 if I remember correctly.

[testpoint1]

Hello,

now I am an owner of a Datum 4040A cesium frequency standard. But, the unit did not work, after starting it, it locked, but 2 hours later it unlocked and reported error 2 and 3. (I am using the terminal software monitor3.exe on a Win 7 computer).
After unlocking the unit did lock again after some minutes, then it unlockes again and so on. After switching it off an let it cool down the cycle above can be repeated.

Are there any ideas how to start the diagnosis?

Best regards and many thanks


EMUD

please check the 10MHz output, if that is near 10.000000MHz ,  just several milli Hz offset, not stable, that means your OCXO is OK, nothing about it.

[notfaded1]

Rudi-

Boy you're lucky your 4040A really has the serial output on it!  That's wonderful news and that you're getting some lock.  I have a feeling you're in better shape than I am right now and I'm glad you started a thread too.  Any new info is good for me and my mine.

Bill

[bostonman]

Boy you're lucky your 4040A really has the serial output on it!  That's wonderful news and that you're getting some lock.  I have a feeling you're in better shape than I am right now and I'm glad you started a thread too.  Any new info is good for me and my mine.

4040A units do not have serial data out; 4040A/RS units do (the RS stands for RS232).

[notfaded1]

Boy you're lucky your 4040A really has the serial output on it!  That's wonderful news and that you're getting some lock.  I have a feeling you're in better shape than I am right now and I'm glad you started a thread too.  Any new info is good for me and my mine.

4040A units do not have serial data out; 4040A/RS units do (the RS stands for RS232).

I know that but I can show the front of mine it says 4040A/RS and believe me it doesn't have a working serial interface in it.  They used the same case for both types during the period mine was made.  Mine also does have the RS-232 port on it but it's totally empty of connection on the back inside.  When I bought it I thought it had a real RS-232 DB25 since it's on the back of it.

Bill

[bostonman]

They used the same case for both types during the period mine was made.

I know.... I worked there for over thirteen-years. It's possible that unit was worked on by me because I worked on every unit that was returned for repair.

[notfaded1]

They used the same case for both types during the period mine was made.

I know.... I worked there for over thirteen-years. It's possible that unit was worked on by me because I worked on every unit that was returned for repair.

I wish I knew how to get more info from mine.  It would be pretty wild if you did work on this one.  My next step would be to take apart the cage that holds the power supply in the back and test all the caps.  That's one thing I could fix.  The main issue I have is I don't have any documentation that tells what the correct voltages on the strip of power connections are supposed to be.  I don't have it front of me right now but there are a bunch of connection screws in a row next to the main power cage in the back.  The cage is a black screened metal box.

Bill

[bostonman]

I wish I knew how to get more info from mine.  It would be pretty wild if you did work on this one.

Does it have a round inspection sticker over one of the panel screws or a white rectangular tag on the front with the unit S/N and other info?

[notfaded1]

I wish I knew how to get more info from mine.  It would be pretty wild if you did work on this one.

Does it have a round inspection sticker over one of the panel screws or a white rectangular tag on the front with the unit S/N and other info?

It has a round inspection sticker.

[bostonman]

It should have initials then. If not my initials, it would be the initials of someone who worked on it, but I would have overseen the repair.

[notfaded1]

It should have initials then. If not my initials, it would be the initials of someone who worked on it, but I would have overseen the repair.

I'll check the markings tonight when I get back home!  That would be really wild.