TSP #72 - Teardown, Repair & Calibration of a Fluke PM6680B 225MHz High Resolution Frequency Counter
ฝัง
- เผยแพร่เมื่อ 18 ม.ค. 2016
- In this episode Shahriar repairs a Fluke PM6680B 225MHz High Resolution Frequency Counter. This unit which is equipped with a soft power button, does not power on. The standby LED is dimly lit and a high-frequency noise is emitted from the unit. This points to a switching power supply failure. The power supply is a simple fly-back converter with three outputs (5V, 18V, -8V). The 5V output is regulated while the other two outputs are single diode rectified.
The fault is traced to a bad precision reference IC which is replaced. The performance of the power supply is verified and the instrument powers on. An OCXO is used to calibrate the 10MHz reference crystal of the unit after the repair is completed.
The Signal Path Blog
www.TheSignalPath.com - วิทยาศาสตร์และเทคโนโลยี
Thanks for making and sharing these repair videos.
Once again a very nice repair video, thanks and please continue to upload content like this!
Really like ALL your videos, especially the repair ones you walk us through everything. I learn a great deal each time. Thank you, please keep it up
Pooch is smart......."oh thats a power supply, Better Back off some", and yes just keep filming and working, we love it!
wow, your analysis is so good to learn from. most people would jump and assume its the capacitors, but a simple opto gone bad would be hard to diagnose.
thanks great repair!
In the U.S., residents are only billed for KW, not KVA. When you measured the the standby current of the counter, and multiplied by the RMS voltage you calculated KVA. I would anticipate that the KW of the unit is significantly less, because the unit probably has a PF (power factor) of around 0.5. Remember the difference between real, reactive and apparent power. Btw I like your channel.
These are great!. I heard your Dec 14 interview on The AmpHour and decide to check you out. I'm new to the field of electronics, but I'm already picking up valuable techniques and troubleshooting procedures as well as general knowledge in the electronics eco-system. Thanks!
Very nice of Pooch to allow you to use his lab for these videos.
Thanks for another interesting repair video.
the sarcastic sticker thing was hilarious I hate those stickers ruin everything thanks for the video learn so much from you apreciate you taking the time to share your knowledge with us
I agree with Rufuak 100% Thank you very much, big thumbs up :)
Nice repair video.
Very interesting as usual..
Good Video Thank You. I loved it.
Pooch really loves the spotlight :)
I had one of these units and also had to repair the power supply. I tested the power consumption after the repair and it was drawing around 12 watts on standby but as it has the oven controlled xtal I did not think much of it but I did notice that the mains input filter was getting warm. I should have changed it but it is an odd size so I decided to change it 'later'. What a mistake. I came into my lab a few days later and wow what a stink. The block was oozing brown goo and the smell took days to clear.With the oven option the accuracy is fantastic and I actually use the 10MHz output as my lab reference.
You make everyone else look bad. Nobody can match your level of preparation and clear explanations.
Nice video and a lovely friendly cat too.
I was a little surprised that you didn't isolate circuit areas that are ribbon cabled (GPIB, LCD front panel) just to see if they were loading the PSU. As it happened though they were not part of the problem.
Nice score with that unit but WOW, the reference oscillator has virtually no compensation and looked very old school and this from Fluke !
Great video.
Once again the Signal Path delivers quality content.
Hope your cat gets at least 50% of your revenue.
What an underwhelming frequency reference! Actually impressive they where able to build a counter around that thing, my HP counter has a massive oven and doubles as a space heater!
Great video, as aways! :)
I think I would upgrade that to a decent ovenized oscillator.
Good Video...
Nice repair video. However, the Rb clock standard does not operate based on radioactive decay of isotope, but rather on super-fine structure in its absorbance - it is basically a crystal oscillator phaselocked to very narrowband "resonator", which has its resonance defined by physical properties of rubidium atom. (The resonator actually being cavity filled with trace amounts of Rb, ovenized, driven by output clock that is multiplied up to the frequency of Rb fine transition around 6,4GHz, and changes in optical power of the light transmitted* trough rubidium cell being the feedback vale for the PLL.)
*fixed wrong explanation - the Rb cell itself is not excited to produce any visible radiation, it's driven externally by secondary rubidium RF discharge lamp, and its absorbance is monitored (much like with fine stabilisation of lasers with iodine).
+Martin Konečný Yes, you are correct. Thank you. I have updated the video. I had taken this fact for granted. This is exactly why I like to read the comments. I learned something new today which I thought I knew already!
+The Signal Path Blog Not even my original explanation was correct, I needed to give it second read and think a bit more about it.
However it made me wonder about effect of external magnetic field due to Zeeman effect - I have no idea how well is the cell shielded inside the frequency reference, but perhaps you can try if there can be any frequency shift measured against second Rb reference, when strong external field is applied to one of them (neodymium magnet)?
+Martin Konečný Actually it came to me surprisingly similar to way how lasers are stabilised to very precise wavelentgh with iodine cell, and aksed my chemist firend if he can at least approximately calculate zeeman energy levels splitting due to external magnetic field. He found me nice article instead (DOI 10.1007/BF00325411), interestingly enough if I get that correctly, I should be able to make to detune the cell out of absorbance of stabilised laser with reasonably achieveable magnetic fields with solenoid.
I have literally no clue if I can extrapolate the frequency shift ratio to rubidiun to get at least order of magnitude approximation, but still'd like to hear if there will be any shift detectable in output freqency - I actully haven't found out yet how the phasenoise spectra of rubidium clock looks alike at low frequencies, nor am I sure that the freqency shift at 10MHz output is at order of several miliHertz. Perhaps a suggestion for tutorial video on Allan variance for next time? :)
+Jack Hudler Oh, so it can be used for precision distance measurement reference - nice. How is the recieving side done then on that thing? I actually wondered if that requires digital lock-in amplifier or not.
I laughed unreasonably hard at 1:05.
+Godfrey Poon me 2 I always love how ebay resellers think the lcd is a great place for a sticker!
nice video
Always love your videos. Great repair. Wish my videos could be this good.
+The Radio Shop Your videos already are!
+su pyrow Why thank you very much for that nice compliment. Much appreciated.
Nice video and fix. There is a "Ext Ref" button on the front panel and I didn't catch if you showed a pix of the rear panel - is that a typical 10MHz reference input? I turn on my GPSDO in the winter time to add a little warmth to the workbench area and the Rb in the summer during air conditioning season since you only have to turn it on and wait a few minutes for it to lock and be useable and no heat added to the environment!
I've got a PM6666 :) funny name, great module.
Wow, a bad 50 cent voltage reference IC caused all of that heartache. You might call it a...fluke. :)
To this day, Shahriar is still scrubbing the LCD of the Fluke PM6680B, trying to remove that dang sticker!
your cat is so sweet
Another nice repair, thank you! On this occasion, I would like to suggest the principles of switch-mode power supplies as a topic for your tutorials series (should you ever find the time to conitnue it...). They are such an omnipresent component in almost any modern device, and at the same time a common cause of failure. You seem to assume that everyone is familiar with its functionality, so maybe it would be too basic of a topic?
You can't use full-rectifier on output of flyback converter (if you put full-rectifier on output it become one transistor forward converter, but then you need to provide way to reset transformer).
To be accurate all outputs on that power supply are regulated, but only one is tightly regulated (using voltage reference) and other are less accurately regulated via transformer action (which is commonly used technique for flyback with multiple outputs and was used even in old PC power supplies).
Since you aren't using the metal pointer any longer, how much do you want for it? I always like that metal pointer.
I case you care... I really dont mind if you skip a few reviews and do more repair videos! ;)
I didn't know you were going to do a CAT scan on the board ;)
👍👍
The Agilent 53181A is also 225Mhz for the top of its spec (it will measure higher). Why have multiple vendors use this frequency as the target for the spec?
Great repair as always, little more lengthy this time. When you first turned it on I thought transformer whinning. I have seen it before in older TV's. The whining always indicated some kind of feedback or oscillation problem. Are you intentions to use that instrument, because it seems you have more precise test geer then that. I sometimes myself go to flea markets and pick up dead equipment to fix for the hell of it. I find it quite fun and challenging and then usually give those things away. Anyways look forward to seeing the next video, Cheers {:>)
Hi - quick correction to the guru! You mentioned varactors a couple of times to tune the crystal - but surely it was just a trimmer cap you were tweaking - or was there a potentiometer and a varactors?
Putch found the problem with the psu :)
the items you are showing to recalibrate the frequency counter what companies sell the items, at 39:18 seconds on video or where to buy them,
Can you characterize IV curve of the TL431 with the Keithley source meter? Will heat induce the failure mode?
+James Damewood Yes, you can definitely characterize it with the SMUs. I am not sure what caused the failure, but this is an old power supply which is always running then the unit is plugged in.
+The Signal Path Blog I was thinking what may have caused of the source of component's failure. It might have been damaged by a lightning strike, causing EOS (Electrically Over Stressed) damage. Bad wire bond from the terminal bridge to IC die. Thermo-cycling from reflowing the soldering sometimes self-heals the component from cracked die, IC delamination from die's paddle or poor wire bond. Leaky transistor, causing a latching effect from particle defect created during the constructions of IC's die etching process.
I didn't think about this at the time.... but that 2460 is WAY noisier than it needs to be. It's cranking the fan up to full speed while sinking 500mA at 5 volts ... that's only 2.5W on a unit capable of sinking 100W. The fans should be able to be idle at that power. Why are they spooling the fans up so fast? It makes no sense. It must sound like a hurricane at 100W. Shahriar, I would REALLY love to see you investigate this.
I thought that there was a temp sensor on the Power supply, so as he drew current it detected an increase temp and fan came on, which didn't help as power supply was out of the case, open loop, fans max out.
Also maybe the 5V rail isn't the one that delivers most of the power, I wouldn't imagine it is, my 2c :)
Cats are from Venus, and Dogs are from Mars. You have got to laugh. Nice.
KITTY CAT!!! :)
Funny you never heard about efratom since it is the first I think of when hearing rubidium. Might be because they started their stuff in the 70s in germany and through several steps ended up being part of microsemis symmetricom, so whenever you have an efratom, it is old. I wonder where your 10MHz distribution network and the hydrogen maser is ;) anyways since you have multiple of them it might be a good opportunity to compare them and talk a bit about how their performance is measure and allan deviation and stuff like that.
I think that the power consumption in standby will have really misserable power factor, so most of that current might be just reactive and not real power.
45:40 120V * 0.072 does not have to be 10W @ AC...
Maybe next time you could use a cats whisker as a voltage probe. :-)
I would personally had expected an OCXO or TCXO in there instead of that simple reference oscillator circuit.
Link to the Agilent video: th-cam.com/video/21_9Dspzu_w/w-d-xo.html
Could you not add an external reference BNC and tap the 10mhz crystal and switch it? Possibly use a GPSDO as an external reference which is probably as good as a rubidium reference in this case. Once buffered and filtered this would become a formidable counter. Failing that add a separate 12vdc supply and use a double oven OcsilloQuartz XS8663 10mhz reference internally?.
I believe the US government use this counter to track the Caesium clocks onboard GPS satellites (how do they do it “season 5 episode 20”, on Discovery channel). And now OscilloQuarrz has moved away from Rubidium to Caesium clocks too!!
Yes, that would be a nice addition.
I'ts possible to swap that crystal with one of yours? (that thing doesn't really give it justice)
+marco56702 There is an option for a better frequency standard just to the back of the instrument, and 2 jumpers to select either the crystal or the option close to the crystal. I guess the original owner was just not into more precision...
I have hardwired my oscilloscope to be floating just because I am to scared to blow it up.
If you have so many OCXOs, why you don't plant one into this Fluke ? I mean they cost less than $US20-30 each. Those cheap single XOs with trimmer cap are crap for a counter like this.
lol, the witty comments in this video are priceless
21:00
Kitty!
+Justin Bell yeah pooch likes to poke around while hes working,
Exactly, I was thinking from first second, "what idiot put sticker on display..."
am sorry, but by any means radiation decay is predictable. it is the best RNG generator in fact.
Has your cat ever gotten shocked? I keep worrying his nose is going to short a cap... A cat cap...
I would also like to send you gifts, but you wont reply to mu emails 😢
+A. R. Jasso Sory A.R., I honestly don't have that much time. I get emails from a lot of people with detailed technical questions. It would be a full time job if I were to answer them. :(
The Signal Path Blog I understand. However i would like to collaborate with your excellent channel with a couple of goods that i think you can find useful.... ill just wait patiently for you to go trough your emails ☺
The Signal Path Blog I understand. However i would like to collaborate with your excellent channel with a couple of goods that i think you can find useful.... ill just wait patiently for you to go trough your emails ☺