Author Topic: No luck on my first attempt at acquiring a cesium frequency standard (Read 55133 times)

hydrogen maser · « **Reply #100 on:** March 19, 2016, 03:49:43 pm »

The oddball ones on mine are:

CI - measures: 5.2 ohm across the terminals
CH - measures: 5.3 ohm across the terminals

CI + to CH + is 0 ohm
CI + to CH - is 5.5 ohm
CI - to CH + is 5.2 ohm
CI - to CH - is 0 ohm

It looks like CI and CH are in parallel

------

CG - measures: .2 ohm across the terminals
CJ - measures: .3 ohm across the terminals

CG + to CJ + is 0 ohm
CG + to CJ - is .2 ohm
CG - to CJ + is .3 ohm
CG - to CJ - is .1 ohm

It looks like CG and CJ are in parallel, but with so low resistance between all of those terminals, who knows

Scratching my head over this one.

edpalmer42 · « **Reply #101 on:** March 19, 2016, 04:18:03 pm »

There aren't many things that get power from that board. Have you unplugged whatever you can? That would at least show you if the 'shorts' are within the power supply or in the rest of the 5045A.

I think I understand the low voltage across and high offset voltage for CG/CJ. If you look here: http://www.ke5fx.com/cs.htm , John Miles says:

A 1-volt, 1-amp supply was applied to the hot-wire ionizer, elevated to 13.9V in accordance with the mass spectrometer specification marked on the tube.

He's working with a different brand of Cesium tube, but I bet the same applies to ours. CG/CJ provides ~1V for the hot-wire ionizer. My unit reports an ionizer voltage of 0V75. It also reports a spectrometer voltage of 17V5.

Ed

hydrogen maser · « **Reply #102 on:** March 19, 2016, 06:27:51 pm »

I unplugged everything and the shorts went away. As I plugged each connector back in I watched the resistance on those 4 sets of pads and tracked it down to one of the large multipin connectors. If you are looking at the bottom of the unit with the HV power supplies on the left then there are three rows of sets of connectors to the right of the HV supplies. In the first row there are two connectors, the top connector is a 5 position connector with only the top 3 pins populated, and below it is a 15 pin connector with 14 pins populated. This 15 pin is the connector that causes the "shorts". The top wire is yellow and the next connector down is unpopulated. Wire numbers are counting from the yellow wire above the unpopulated position as being position 1. The tan wires 3 and 4 are spliced together in the wiring harness and the tan 10 and 11 are spliced together - these two tan conductors end up going into the cesium tube on the left side just below the HV supplies. There is almost a dead short on those tan wires of about .2 of an ohm. There is continuity from the negative side of CJ/CG to wires 3/4 and there is continuity from the positive side of CJ/CG to wires 10/11.

The two brown wires in positions 9 (ground) and 12 go into the right side port in the cesium tube, these have a resistance between them of about 5.5 ohms - these pins have continuity up to CH and CI 12 on the positive side of the caps and 9 (ground) on the negative side - that is where the low impedance "short" of a little over 5 ohms is coming from for those caps.

So I guess the question is, it is normal for it to be that way? I am not going to try to open up the cesium tube box and find out. Those are very low impedance paths that could draw quite a lot of current, is that normal?

Vgkid · « **Reply #103 on:** March 19, 2016, 07:39:27 pm »

.2 ohms is not much resistance, could be a shorted tantalum inside.

hydrogen maser · « **Reply #104 on:** March 19, 2016, 08:00:50 pm »

I am ordering the capacitors now. Of course I don't have exactly the capacitors I want to use for this and I don't want to cobble together something just because I am in a hurry and want to see what happens next. Get the right stuff. Trying not to get my hopes up too much on this one, I know the odds are stacked against me in a big way in making this thing work, but hey, this is all about fun and learning for me.

edpalmer42 · « **Reply #105 on:** March 20, 2016, 02:37:18 am »

Good news!

I measured across CG/CJ and got 0.2 to 0.3 ohms. Across CH/CI measures 5.8 to 5.9 ohms. Those values do not include any lead resistance. So even though the readings are odd, it looks like they're good.

Ed

hydrogen maser · « **Reply #106 on:** March 20, 2016, 02:55:10 am »

Thanks! Caps should be here Tuesday. I will put them in and report back. Thanks Ed!

hydrogen maser · « **Reply #107 on:** March 24, 2016, 12:16:35 am »

IT IS ALIVE!

Kind of hard to see in the below pic, but the lock light is on

Thanks to everyone who has helped with this, but EXTRA SPECIAL BIG THANK YOU TO edpalmer42! Ed has put a large amount of time and effort into helping me and there is no way I could have gotten this far without his help. Thanks so much buddy! I know we don't even know if this thing really works yet or if it might have some unfixable fault or a bad tube but I consider this project to be a big success because I started out thinking that it was extremely unlikely that I was going to be able to get anywhere with this thing due to a lack of info about how it was built and how it works. I was just happy that it looked like the oscillator was working so it wasn't a complete waste of money. I learned a ton about how this thing works and I have confidence now that even if this one does not work correctly I can get more of these units until I get enough working parts to build one that does work. This thing is just way too interesting not to have one that works and I will get one that does work if this one does not, one way or another. Ed, what do you make of the output from the monitor program?

Vgkid · « **Reply #108 on:** March 24, 2016, 01:17:00 am »

Nice job. Fun following your progress.

edpalmer42 · « **Reply #109 on:** March 24, 2016, 03:56:17 am »

Sweet! Congratulations HM!!

Why is there no emoticon for a happy dance?

Looking at your numbers, I think everything looks good. Probably the most important numbers would be:
- Power supply voltages (for obvious reasons)
- OCXO cmd - I think this is the EFC to the oscillator. The range is +-4500 so yours is fine.
- Ramsey Level: 2500 +-500. Maybe on the edge. Watch this and see which way it goes.
- Zeeman level: 2000 +- 600. No problem there.
- Pump current: That's the ion pump. Lower is better. Yours will probably go to zero which is perfect.
- ShrtTerm StDv: Range is up to 225 so 158 is fine.

The only thing I'm puzzled by is that the temperature is only 29.1. On my unit, it's labelled as Cesium Module Temperature and tends to be about 20C above ambient. Is your room really cold?

Now the big question is where do you go from here? What was your reason for buying it? Remember that whenever you're running this, you're using up the Cesium, so you don't want to let it run 24/7 without a good reason. In particular, if your Ramsey Level stays low, it could mean that your tube is getting old so you want to be really careful with the runtime. I'd suggest that you get either a Rubidium standard or a GPSDO and keep the 4040A on the shelf for occasional checks against the others. A GPSDO would give you absolute confirmation that the 4040A is working properly, but I don't think it's really necessary. A good Rubidium standard will probably give you better short-term performance than the 4040A or most GPSDOs. A good OCXO will give you the best short-term performance of all of the above.

Congratulations again! It's great to see a wonderful piece of equipment like that working properly and saved from the trash!

Ed

P.S. If you'd like some more info on what all the parameters mean, check out this document:

http://www.ampere.com.mx/pdf/Manual_CsIII.pdf

It's the manual for a Cesium standard that's two or three generations newer than ours, but was made by the same people. They used a lot of the same terminology as ours does. It also talks about how their units work. It's different than many other Cesium standards.

hydrogen maser · « **Reply #110 on:** March 24, 2016, 04:47:12 am »

ShrtTerm StDv is up quite a bit now so perhaps there is more to repair? Ramsey level is going up after running for 4.5 hours and the ion pump is now at zero. That first pic was from when it had only been running for 24 minutes, the temperature has gone up now. Why did I buy it? Because it was relatively inexpensive and I like to repair broken equipment, I find it very satisfying. This thing looked really interesting and I could not really pass it up given the price. I knew it is was a very esoteric thing and I wanted to see how it worked. I figured that it would almost certainly not be fixable given the price but that it should have some interesting and unique parts in it and hey, who knows, maybe it would be fixable. Frankly I thought I would probably end up completely disassembling it and the oscillator would be useful at least if that part worked - I could use it to clock my frequency counters. This is the fourth piece of inexpensive non-working gear I have bought, the other three being my frequency counters and function generators. The function generator needed 8 new resistors at about 1 cent each, the frequency counters work perfectly, I did not have to fix them at all, I think the seller just did not know how they worked. The capacitors to fix the 5045A cost about $25, it could have been cheaper but I did not want to skimp on them on the off chance that it might actually work and fix this thing so I got really good capacitors. I have been really lucky in all my non-working for repair gear I have bought. There was one other reason I bought this thing, but that is for a different post I will make in the near future, I think it will be amusing to many.

edpalmer42 · « **Reply #111 on:** March 24, 2016, 05:27:41 am »

Yeah, that Shortterm StdDev isn't looking good. Something's wobbling around more than it should. Maybe it just needs to run a while to work the kinks out. Now's the time you want to start logging some data. For the time being, reset the alarms so that we can see if the unit thinks anything is really wrong and go into the 'graphics' mode so you can get a bit of logging. Start by displaying the 4 power supply levels. Make sure there's nothing intermittent with them. For the other two items, look at the Shortterm StdDev and the Ramsey Level.

Ed

hydrogen maser · « **Reply #112 on:** March 24, 2016, 02:17:42 pm »

I cleared the alarm and have not seen an alarm since.

Ramsey level was very different this time, it started at 2800 and has been trending down, the opposite of what happened last time where it started at 2000 and went up. Shortterm StdDev looks better for now.

edpalmer42 · « **Reply #113 on:** March 24, 2016, 05:30:02 pm »

Yes, that all looks fine again. It's nice to see those stable voltages! There's a bit of noise on the +5 line, but it looks okay. Maybe it was just working out the kinks. All you can do is watch it and see what happens.

Maybe whip up a program that queries the unit and dumps the data to a log file so you can see if there are any glitches or trends. Extra points if it parses the data and outputs it in some convenient format rather than just a raw data dump.

If you see the StdDev creeping (or jumping) up, look at the STEL-1173. It puts out parallel data to the D/A converter beside it. The failure mode is that the output drivers start to fail one by one. If you grab the data sheets for both chips, you might be able to probe the connections with your scope and see if they're failing. Yesterday, yours started to go strange after a few hours of operation. If it starts to go again, maybe cooling the 1173 would bring it back. I've noticed that it tends to run hot. Maybe try a fan and/or a small heatsink, or as a last resort, gentle application of freeze spray. Mine was totally dead when I got it so I wasn't able to see anything. I guess that the best plan would be to check now, when everything's working, to see which outputs/inputs are active. I don't think that all of them are used.

Ed

hydrogen maser · « **Reply #114 on:** March 24, 2016, 07:32:15 pm »

Yeah, that makes sense, measure what I can now while it is working. The serial port constantly spews data so no reason not to capture and save that data and to poke around in the 1173 and D/A and make some notes about what is what so that if/when I fails I will be able to tell if it is not working vs something else. Thanks so much again for all your help Ed, I could not have done it without you.

edpalmer42 · « **Reply #115 on:** March 24, 2016, 07:46:57 pm »

Glad I could help. Have fun with your new toy!

Ed

edpalmer42 · « **Reply #116 on:** March 24, 2016, 09:11:26 pm »

HM, now that you've got your 4040A working, maybe you can help me!

While I was exploring and repairing my 4065A, I changed the value of a pot. Now, I'd like to return the pot to it's original setting, but of course, I don't know what that was.

If you look at the attached picture, you'll see a scope probe attached to an unlabelled test point near one of the corners of the main board. This is the buffered signal that comes out of the Cesium tube. The blue pot that's nearby (R49, I guess) is the one that controls the level. It's a complex signal that sits at one level and jumps up and down to other levels. There's also a repetitive spike every 17 ms. Could you look at yours and tell me the values of the various levels it jumps to - in particular, what is the base level that it's usually at? They're in the 0 to -1 volt range.

Thanks,
Ed

hydrogen maser · « **Reply #117 on:** March 25, 2016, 01:02:05 pm »

Of course I will, leaving for work now - sometimes we call it early on Fridays, sometimes not, so I might be able to get to it this afternoon or I might not. If not today then tomorrow for sure. I assume you want measurements after it is locked? Assuming it will lock, I have not had any time to play with it since my last post and this will only be the third time it has been powered up post repair.

edpalmer42 · « **Reply #118 on:** March 25, 2016, 04:36:28 pm »

Yes, I should have specified that it's after lock. I have no idea what kind of mess you'd see before lock.

There's no rush. I have family obligations for the long weekend so it'll be early next week before I can get back to it.

Ed

hydrogen maser · « **Reply #119 on:** March 25, 2016, 06:56:11 pm »

Totally forgot it was Easter, just drank coffee for a few hours and called it quits.

If I understand what you want to know, the top of that waveform usually sits right at zero volts. I used DC coupling and the little yellow "1" carrot next to tVmin on the left shows the zero volt mark. It is 20ms horizontal and 16mv vertical per division. The waveform is centered right at -80mv most of the time. When it jumps, it jumps by approximately 40mv positive or negative. I am having trouble triggering on the lower excursions because of the downward periodic spikes. That doesn't bother the trigger for the upper excursions but it is causing issues on the lower ones. I think there is a way to tell the trigger not just the level but the min width to trigger on, I will play with it and see.

hydrogen maser · « **Reply #120 on:** March 25, 2016, 07:26:17 pm »

Better one that shows zero volts vs same relative position (approximately) for upper excursions.

edpalmer42 · « **Reply #121 on:** March 25, 2016, 08:08:59 pm »

Thanks very much for that. It'll be Monday or Tuesday before I can get back to mine.

Yours seems to be a lot noisier than mine. I'll try and post a picture. It'll be a bit of a challenge because I have an analog storage scope. Actually, I have a USB scope that might do the job. Got to think about that. First challenge is 'where did I put it?'.

Ed

hydrogen maser · « **Reply #122 on:** March 25, 2016, 08:16:14 pm »

Anything else you want me to look at while I have it running?

hydrogen maser · « **Reply #123 on:** March 25, 2016, 08:54:15 pm »

Looking pretty good on this run:

hydrogen maser · « **Reply #124 on:** March 25, 2016, 09:46:13 pm »

Ok, so the 4040A has at least one more problem or not? If you put it into storage mode and then later bring it out of storage mode you get a "Unit Restart Fault". It has always done this but I thought it was because the previous times I powered it down I removed the power cable to totally shut it down and it has complained about this before. In the manual it is Major Alarm 16 and it just says "Unit Restart" Perhaps this is not really an error and it is just letting you know that it was restarted? I assume these things are intended to be always running and never shut down? Just informational and not indicating that there was some problem?


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: No luck on my first attempt at acquiring a cesium frequency standard (Read 55133 times)

Share me