TSP #106 - Teardown & Repair of a Fluke PM6685R 4.5GHz Rubidium Frequency Counter

[Music] hi welcome to the signal path in this episode I'm going to do another repair with you guys this is a fluke PM six six eight five R this is a universal frequency counter now what makes this a special unit is that this particular one has a rubidium atomic clock built right into it this makes it very attractive because it means that it has a very stable very precise internal reference making this not just a frequency counter but also a calibrator and it means that you can calibrate other frequency counters or other synthesizer to this reference that's built into it making it a very nice unit for repair definitely worthwhile now it has channel a it doesn't have a channel B option but I can feel the indent behind it where this could have potentially been populated by channel bina I'm not sure if there's a version of this with two channels but it has channel C which goes from 0.15 to four and a half giggles meaning that you can then of course compare frequencies as high as four and a half liters to the internal rubidium standard now I have done a repair on a fluke universal counter before and in fact it's right up there and this particular one fluke PM six six eight zero B was a two channel one it doesn't have option C installed on that but I've done two repair videos on this where on the second one I believe i retrofitted it with an oven control crystal oscillator which one of you guys sent in as a little module board which was fantastic making this a much more usable unit because the internal crystal was just absolutely horrible so that means that if we can fix this one we're going to have a very nice frequency counter here on our hand be able to calibrate everything around the lab in fact and other synthesizers now I do actually have a separate rubidium standard which I came across very recently also so we could even compare these two against each other the rubidium against each other if we can fix it so what is going on with that I'm gonna turn it on and show you now when I do turn it on and even when I just plug it in because it does have a standby power if you remember it makes this a horrible screeching noise making me seem like that maybe power supply is bad now I know you guys don't necessarily like power supply repairs but you know I don't choose what is broken in an instrument in varies widely as you've seen with all the repair videos that I do and again I want to thank my patreon supporters they are the reason why I can do so much repair videos and this is a very important because I know use of your favorite episodes and I definitely enjoy doing them as well and I don't charge every episode to patreon only episodes that are more than 30 40 minutes I actually charge the patron so if you want to look at that you'll see that only a fraction of my videos show up so I saw that I don't charge my patreon supporters you know too often maybe once or twice a month at most but that anyway so this is going to be an interesting one most likely a probably a short one but we'll see how it goes should be cool to take it apart so thank you again patreon supporters you're making this all possible all right let's give it a try I also have an assistant here making sure everything is going well and he seems to be quite happy about it that's a good pooch all right here we go so let's take a look and see what happens when I just enabled the power directly from my variable isolation transformer here so let's go ahead and turn it on and there this and this is just in standby now initially of course it's going to have to turn on the rubidium standard so that it can be warm and ready to go so it is definitely what is attempting to do but burning 46 watts seems like a lot and if I go ahead and take a look at the unit itself you can see if I can ever achieve focus here you can see that the unlock light of the rubidium is on because it's not warmed up yet that's not in necessarily a problem but we can go ahead and try on turning it on and see what happens now this I don't know if you can hear it or not but it sounds really really bad and then it's going just nuts and I know that there are some threshold that just you can do on this and any it might stop that but it's obviously false triggering quite a lot and I left that on for a while I'm gonna leave it on for a while a little bit longer I'm just a little worried about the noise the power supply makes it shouldn't be doing that and I couldn't get this unlock light to go off so maybe the voltage is unstable and he just can hit the point where he needs to and it sounds absolutely terrible so let's go ahead and open it take a look inside of it take a look at the power supply see if we see any you know necessarily damage and we can measure it and lucky enough I think schematic of this is actually available we should make it a lot simpler and we can try that on and see if we can figure out what's going on with it all right here is inside of the unit and we find everything that we expect to find except that there is another power supply in this and that makes sense the original power supply which is the one that I repaired for the or old frequency counter is probably not strong enough and doesn't have the correct voltage to supply rubidium standard as well as everything else and instead of changing the design of this power supply they just have a standard one and a standard PCB and the mechanicals all done for that but they add a separate power supply it's the 24th power supplying can give two and a half amps and that would then power only the rubidium standard which is sitting over here and we will take a look at that and if you look at the power cable directly coming from here it only goes to this these two cables that you see coming from the rubidium is a 10 megahertz output and it goes right where the ocx ol would normally go if you remember when I repaired the old Phillips one at the fluke one at that one when I install the oven control crystal Arthur I put it right here and then the original one that was there is just absolutely horrible and there are a lot of Phillips branded ICS here not to be again it's expected because this is a philips technology and it's flu quad on my belief and yeah so and then everything else is pretty straightforward this one has the added channel C and we will take a look at that for sure because I didn't have one in the old one so I want to take the can off and see what technology it uses likely it's a divider based one like the one that I showed for the agilent before other than that it has a very straightforward design i think i've talked about this before so i won't do all on it too much but yeah let's go and take a look now but i'm going to do is i'm going to separate the two power supplies so we can see which one is the one making noise i could be either one and then we can go from the army beats both of them and then we'll go and figure out what to do next i'm quite eager and it's most likely this one because this is the one that's trying to deliver crime when you power it on but it would be a good test so let me plug this in and make sure my cat doesn't walk on it when it's plugged in and we'll take a look and here we go unplugged and i'm just going to make sure it doesn't touch anything else because it is going to be live when i enable the power there we go and let me just hold it like this and let's go ahead and turn it on oh no it still makes that terrible terrible noise but I just noticed something interesting the fan is not and this fan is powered from this board go ahead and turn this on so if the fan comes on yep the fan does come on so indeed it's working interesting it's kind of weird that this fan doesn't turn on even though this power supply is always on even when this is not powered so I guess it's supposed to cool the whole thing but doesn't seem to have a very good thermal profile as a result of that a little bit unusual design from that point of view so yeah this is still making noise so let me go ahead and turn this off and I wasn't gonna try and see you I don't know if you can hear it or not but it's kind of home winding down now we can go ahead and open it a little bit more and they're close to look and see what you find so I took the cover off and it's a measurement and in verified that the rectifier is working now if you want to get more information about this particular power supply watch my other video where I do a full repair on this so I don't want to waste too much of your time so I went ahead and I took this out again just like last time so we can take a look at it and I suspect that this capacitor might be bad and this is one of the very first things I'm going to change and see if the behavior of the changes I'm also going to change that the capacitors because in these are pretty old and then we can try it now as you know because the rectifier and the filter capacitor are on this board the output of this over here is DC and this is just a DC DC converter so this just has to accept the rectified DC voltage and it gives you a couple of DC voltages at the output and there some of them are just diode rectified along the 5 volt is fully voltage regulated so we can go ahead and try that separately so we don't need this main unit anymore because this is I just want to make sure this runs quietly and it runs smoothly so that it's reliable and then we can turn our attention to some of the other components that are there that's why I don't want to spend too much time on it so let me show you how I plan to test this on its own and here is our module and in order to test this as I said because it needs a DC voltage I made this now I definitely do not recommend that you make this this is a an incredibly dangerous thing because of the exposed wires because of the we can hold such a large amount of charge and it will rectify it it's just something I put together the reason I'm using it again I'm familiar with high voltages and I'm careful around them and I am using an isolation transformer and I can bring it up slowly so I can have a multiple layers of protection here but be very careful don't necessarily do this in my other video I use the unit itself to power this and you know that this time I'm just doing something different but just be please be very very careful if you're going to attempt anything like this so before I even turn it on I'm just going to go and change the capacitors again just want to quickly get this no matter what I want to change those capacitors anyway let's go ahead and change these and see if their noise changes and then if it works we can put it back in the unit and and turn our attention to something else well here are the components I took out so we can quickly give him a test and see what happens for instance let's start with this largest capacitor here so I can go ahead and try and measure them yes all of this particular one and 0.19 ohm ESR is not that great but it's also not terrible and the capacitance is there you go a thousand my 10,000 micro farad which is correct this was supposed to be so this is not a terrible capacitor I does have a bit leaky door let's try this one and this particular one what do we see here analysing open circuit or low capacitance oh this is a bad capacitance I can try it one more time and what does it say yep open circuit doesn't like it at all can't even detect that is it is there here's another one it's the other capacitor there and analyzing oh yes about a 40 ohms that one is definitely dead and the last one here what do we get with this one take a quick measurement on it and nope doesn't even get detected so this is an open and this explains a bit of the behavior because this is the capacitor that's sitting on the PWM I see so this would be bad if it is open that's a quite a quite a problem so I have replacement capacitors right here we can go ahead and put those in these are all good these are good Japanese capacitors and they should last a long time unfortunately they are 85 degrees Celsius rated but I'd only cared that much so we should be able to get this thing going so I'm eager to try it let's put it on see what happens all right here we go I did all the replacements I didn't replace that big capacitor because I don't have a ten thousand microfarad version it's good enough for now so we can go ahead and turn it on see if you get it anything called the output and I have it connected directly to the file the output and let's turn in the Sun and here we go I don't know if you can hear it or not but it's completely silent and the output voltage can be read over there it's 5.4 volt DC there is an adjustment partner I'm not gonna adjusted yet I'm gonna adjust it when it's in circuit so we can see the exact load that's on it so this certainly seems to have solve the problem of this power supply now we can take it a step further put it back powered on and start investigating other issues with it all right all good and ready to test I did actually end up finding a capacitor and I plugged it in I turned on and it is very very silent so it's making no noise and we just have to turn it on and to see what happens on the front panel I'm just gonna clean up and get to that alright let's give it a try now so it sounds pretty good and here we go and there this nice there's no signal which is what it should say it doesn't do that gibberish ting and what what more is that I now have the unlock sign of the rubidium turned off so it looks like the rubidium may also be warmed up and actually functional so we should be able to put a signal in it and see what we get as just a default a situation and see what without any calibration if you're getting our 10 mega and whether this even works and then I want to take apart and take a quick look at the RF input and then test that as well and then we should be done okay let's give it a try so I'm going to use this AT&T rubidium standard that you can buy from eBay they're not that expensive actually and then we're going to hook it up I cross and see what we get so let's set this guy to 50 ohm input and there you go on channel a and connect this over here and connect this to the 10 mega output and let's see what happens and oh my god that's that's crazy look at that that's that's us because it's gonna get so this rubidium standard compared to the Ruby M standards in here are basically a perfect match and this is to be expected that obviously because these are both atomic references so we have one two three one two three so these are in the melee hurt so this is a melee hurts at the end of that so pretty pretty close to each other I'm not even sure if I'm going to try adjusting and I'm just curious what would happen if you try and adjust it and I to be to be honest I don't even know which of these two would be out so it's difficult to say but I will wiggle the frequency adjust a little bit just to see if you'll do anything out of curiosity so I went ahead and I change the integration time on the fluke and now we see a slight difference between the two of them so we can go ahead and and try and change it and then change just a little bit if we can get some difference let's turn with a quarter of a turn and I think I'm going in the wrong direction I think I am so let's give it a full turn and a half now we're at nine oh there we go nine nine very very close now nine eight I think I can make it a little bit more another quarter of a turn there we go ten ten my guards it's really really close now I'm gonna make it a little bit tiny more and that should be good enough now interesting this last digit I don't see a changing at always always seems to be six I'm not sure if that's because my integration time is not long enough but either way I mean I'm pretty happy with this it's gonna be good enough I mean I don't do this side I have not a calibration how so it doesn't matter but this is a good sign I mean they're very very close to each other to begin with so now we have a perfect 10 megahertz rubidium driven frequency counter that's a calibrator also this is a full full calibrator that now you can use you don't even need to have this with you and here is the module the four and a half gigahertz front end so I can take the cage off of this and what we see is unfortunately a single hybrid module and we cannot really see what's inside of it but we can make some judgment on what's going on so the input comes over here you can see some matching in some network over here some filters over there then we will a couple to some clamping diodes and then we go into this module now this is going to have amplifiers prescaler even maybe an automatic gain control and of course the divider which is a pre scalar and the divider signal then comes out and goes into this and I couldn't find a part number for this this is a CA 3 1 9 9 e but it looks like it itself is a pre scalar perhaps a lower frequency one to give you a last maybe a factor of 4 divisions or something these two are just voltage regulators so it's really straightforward your architecture of it is very similar to the one that I showed before for my keysight agilent unit but i accept that this that one uses discrete dividers and amplifiers and this one does and it seems to all be one hybrid block probably made by philips and that's really it that's all there is to it and unfortunately this is one of those situations but if this guy breaks you're gonna basically be stuck having to redesign this entirely because there's no way you're gonna find this and anyway it doesn't have even have any numbers on it so it'd be very hard to find this proprietary component but either way if you have some information on it you can share with me but I wanted to show you what's inside of it so that we can go and hook it up to my synthesizer and to see if it's actually working all right let's go ahead and measure the Tektronix TSG 4106 a now the reason I'm choosing that particular unit for measuring is because if you remember when I did a review and teardown of it it has a very good oven controlled crystal oscillator for very precise reference built into it this is based on Stanford Research Laboratories system there so it looks up pretty good and right now I have a set to 250 megahertz with 0 DBM and we can go ahead and see what the instrument says and check it out we are getting a measurement of 250 and this is kilohertz so we are very very close in the in the Hertz range to the rubidium standard which is pretty reasonable and as you can see it doesn't tell me the reference value coming in that the power coming in because it doesn't have a detector so it doesn't know it can only give me the power value coming from here and I'm doing a two-second gating so 250 megahertz works reasonably well and let's go ahead and try one giggles so let me just go to the frequency here and enter one gigahertz never this and it's going to settle down again there this one gigahertz lots of lots of zeros and that's pretty good I am pretty happy with that and oh I get it okay never mind okay I was wrong before this this number is not the last digit this is ten to the something that's a bit confusing so it's one times ten to the nine Hertz that's why before that six was there it was you know ten times ten to the six megahertz so anyway that my apologies for that this is that for that reason so there we go so one gigahertz not too bad and we can go ahead to two all the way to this upper range four and a half gigahertz and it should work hopefully let's see and yep four and a half gigahertz not a problem very stable and a nice and clean and it should probably go above the rated value you can try five gigahertz let's see if you can pick that up okay feel good yeah it's gonna pick it up no problem the question is how small can I make the amplitude and let me see if it even works up to four and a half here I should say five and a half yeah yeah it's beginning to not be able to trigger correctly so it seems to work perfectly fine at five gigahertz so let me go and change the amplitude to that same minus six and it still detect five vigorous at minus six DBM yeah it has no issues - 12 DBM and yeah it's gonna start having problems there stuff but it works really well I have to say so without for its rated output and even above the rated output I have no problem being able to measure so I'm curious if I'm at four and a half gigahertz which is exactly where it's supposed to be what is this smallest amplitude I can detect reliably so here's four and a half kickers in the upper range of it's a specification let's go to minus 10 it's minus 10 DB M it has no problem with minus 10 DBM minus 15 DB m and minus 15 DBM it's still happy with that - 20 and yeah even - 20 DBM it works really well so it looks like within its own specification it's quite stable so here's - 25 DBM and yeah it's now beginning to break in about minus 20 to minus 25 DBM is a typical range for no external amplifier counters to be able to detect reliably so it looks like - 20 DBM is still not an issue and it can pick it up perfectly fine at the upper range of his frequency so and it's very good and you can see it's very accurate again we have quite a few digits there that match perfectly and this is a very good synthesizer for its absolute frequency accuracy so I'm not surprised that we're getting a good value out of it so really happy with that now I'm curious on going around the lab and measuring everything because it is a calibrator so we could actually go around and calibrate a whole lot of different things and then out of curiosity I'm going to just measure one other little thing on one of my Agilent synthesizers one of the old ones and see how good those are so let's quickly try the agilent 33522a from generator i'm just curious on how well it behaves so let's set it to 10 megahertz here and the amplitude instead of to one volt peak-to-peak and that should be good enough for purposes and right now it's probably set to a high impedance output but doesn't matter so let's go ahead and see what we get when I enable it and output here we go let's see and let's settle down for a second you know what oh no nevermind I was gonna say it's pretty good but I didn't notice there's actually five here so you can see how different this is from another rubidium standard or even from what I was measuring the Tektronix TSG 4106 it's the difference between having a really stable loving controlled crystal oscillator sitting in in your instrument and what a difference it makes in terms of your frequency accuracy so yeah I'm quite happy with that and we can go ahead and measure a whole bunch of things around the lab but this is a really nice unit I'm quite impressed with it and it's going to be great for calibrating and keeping everything on the right time base and this includes oscilloscopes and spectrum analyzers so yeah quite nice I'm going to hope you liked this video you going to be a short one none of the ones that are is not going to be one of the ones that gets charged to patreon of course because it's a quick one but I am working on a tutorial y'all hopefully I'll finish today and that was pretty exciting so hopefully you will like that I'll see you soon in the comment section [Music]

Channel: The Signal Path

Views: 24,292

Keywords: Frequency Counter, Atomic Reference, Calibration, Power Supply, Capacitor Repair, DC-DC Converter, Repair, Teardown, Fluke, GHz Counter, Rectifier, Filter

Id: _Rclt5ptc1U

Length: 22min 35sec (1355 seconds)

Published: Sat Sep 02 2017