EEVblog #544 - Fluke 5450A Resistance Calibrator Teardown

hi welcome to tear down Tuesday we got another vintage bit of gear today and it's from a cowb something you wouldn't ordinarily get your hands on it's a resistance transfer standard it's the fluke 5450 a straight from the mid 80s I think 85 86 vintage something like that this would be the transfer um standard used for calibrating uh multimeters back then it's OB now but uh you know it's still not a bad bit of Kit and look at the size of it it is absolutely enormous and uh uh designed to either go into a rack or be uh bench mounted like this and basically it is nothing more than a bunch of precision resistors in a box that's pretty much all it is yes it's got a big lead display on here which shows you the value but it doesn't actually uh measure anything all it is is basically a bit of automated gear for automated testing of uh multimeters in a typical calibration lab and the way it works is um it's a four terminal uh capability two or four terminal measurement it's got uh really Shmo Precision resistors inside and when you calibrate this thing um itself then you actually program in the exact value of the calibration resistors in here so this doesn't actually measure its own internal resistors it can't actually do that but what it can do is actually display the value um what use is that I hear you ask well in automated measurement with the upside down all the electrons are going to fall out gpib on the bottom you can automate the testing this and it can read the value out of this to know exactly what value the calibration resistor is in here and then it canare it with the device out test and so on so it's a really just a very fancy automated Reay switching box and it could be interesting inside not only in the uh type and construction of the Precision uh resistors calibration reference standard resistors in this thing but uh yeah in the way physical construction and stuff like that keep out noise leakage all that sort of stuff in the analog section of it could be really interesting so you always say on E blog turn it on take it apart and if we take a look at the specs here for it I mean it's you know compared to uh sort of modern multimeters um it really wouldn't be uh suitable for say calibrating a modern well I don't probably not anyway calibrating say a modern agilant 3461 A6 and2 digit multimeter or an equivalent digital multimeter I mean we're looking at you know nominal tolerances here of 5050 PPM uh something like that that absolute uncertainty is actually quite good you know down in the order of uh 5 PPM or under for like 24 hours or 1 year you know 1230 uh PPM and uh they're the various uh test currents but you know compared with say my uh Precision resistor uh box I've got here which has just got a couple of $20 resistors in a modern uh modern vet resistor there's the uh part number let's get the best one here which is the 10K one you know we're talking 05% or 50 PPM accuracy so it's sort of you know in the same order as this and if you compare the stability like for example a 90day s you're only talking you know 3 PPM or something like that really quite good and temperature coefficient here 1.5 PPM per degree C and if you compare that to here this is like you know 2 PPM per degree C so you know it it's not a huge amount better than just a modern uh you know prec prision uh resistor from the likes of vishe that you can buy for like $20 odd but you know in practice I'd bet my money on the resistors in this box any day of the week so this is basically a single-purpose resistance uh calibrator only designed for calibrating the resistance or calibration checking the resistance ranges of multimeters that's it doesn't do anything else unlike you know modern multi-function calibrators that can do everything you know ACDC uh voltage resistance current all sorts of stuff like a fluke 5700 or a 50a multifunction calibrator for example so they they're pretty much uh made something like this uh obsolete I'm not sure of a cow lab that would still be running uh one of these things a 54508 but if you are aware of any CBS still using them for a specific uh purpose please chime in but uh yeah basically here is the uh control panel for it it's designed for two and four wire measurement it's got a two wire compensation function which basically just doesn't do anything fancy it just adds a uh a pre-entered pre-measured and pre-entered value on top of the actual internal calibration uh value in there so it just adds a stored number uh basically to the value and of course in four wire mode you would just hook this directly onto your multimeter the sense wires go to the sense terminals your multimeter the output goes to the input terminals of your multimeter that's it it's also got a ground Shield which you can turn off and on external guard for when you're actually calibrating this thing and I won't go into the whole uh way you would actually calibrate this in terms of uh shielding and everything else but can be really really important that external uh guard Shield that's just not just Ms Earth willy-nilly it's you know has very specific uh requirements in terms of noise reduction for calibrating this thing and then all you do is you just choose your range and that's it um it's pretty basic it's got a uh multiplier so it can only uh select values of 1 ohm 10 ohms 100 ohms in each decade up to 100 megga ohms but you can actually multiply that by 1.9 so if you wanted a 1.9k standard resistor press that button with that one and you would get 1.9k it' switch the relays internally clunk clunk and you'd get 1.9k reference resistor across there and that's pretty much all ARS to it it also allows you to display the PPM error but once again it's not actually measuring anything it is just displaying a pre-recorded value in there and that's all it does on the uh 2 4 6 8 digigit uh display here it can show the um output value uh or the percentage error for there and that's all there is to it as I said it's dumb this display is not reading anything just showing you a pre-stored value on the back nothing fancy of course Mains input um I e48 uh G gpib standard plus a set of uh rear terminals on the back for when it's uh mounted into a rack or something like that once again these are really going to these are low EMF uh connectors in there you can see they're probably some sort of uh a torium copper or something like that really Schmo I don't know what are they worth like a hundred bucks each to get a replacement set of those and it's made in the United States of America still has the calibration uh void seal intact um but the thing has been uh opened as you can see it uh was calibrated by the fluke Metrology lab but uh it has been opened so inside this thing um is going to be I think when we open it up we're not going to find anything fancy at all uh we're just going to find some uh a whole lot of digital stuff which of course drives the um gpib and the Seven segment displays on the front front um nothing doing there that has you know nothing to do with the uh actual real operation of this unit what we're interested in is all the uh the uh Precision reference resistors in here and the relay switching and all that that will be probably triple Shield or something ridiculous like that and uh it will probably be laid out quite well because this thing you know if you bre if you open this thing and you breae on it um you know you're probably going to have to recalibrate the thing I mean especially when you're talking or if you touch it you know all the grease in your fingers and stuff like that probably not if you breathe on it but you know the oils from your fingers if they get on and contaminate the analog board in here oh here we go yeah contaminate that analog aha Tada look at that and there it is check that out it's actually quite big I've had to uh prop the thing up vertically on my bench here to take a look and uh as I was saying um the uh oils and stuff from your fingers if they contaminate the analog circuitry in here as I said You know here it is behind all the um shielding in here leakage sensitive printed circuit assembly use special handling there you go because when you're talking about 100 Meg you know a Precision reference resistor which is you know 05% accurate or something it doesn't take much in terms of leakage for that value to change all right let's just pluck a number out of our ass to see what sort of scale magnitude we need to get in parallel with our 100 megga reference resistor for you know it to start being a problem I chose my Casio FX 61f because it has a parallel button the only commercial calculator in existence that ever did I think oh beautiful anyway um 100 Meg in parallel with let's say 100 Gig 100 Gig ohms okay let's have a look what we get look at that 99.9 Meg we're still outside of our tolerance and if we go divided by 100 Meg like that there we go subtract one and uh let's shift that and times 100 what do we get in percentage there we're talking you know 0.1% okay so that's is you know outside of our 05% or better just just forget about it so even 100 Gig ohms in parallel with uh that 100 megga ohm reference resistor gives us that .1% error way out we need something in the order of you know even like Terra ohms even a one Terra ohm can actually affect this thing and of course that's aide from noise issues I mean we're at 100 Meg at 1 Vol you're only talking 10 nano amps I mean you know it's go it's na all half a bees dick so really any noise that gets into this thing can be a real pain in the ass that's why that um switchable Earth terminal on the front panel is such a big deal so this is really interesting they've actually got a relay table here which shows you which relays are energized for each of the particular ranges a Reay maap look at that that's just gorgeous engineering right there so it's 34 re inside this thing and it even gives you the fluke part numbers for the different uh Reay types there we've got ourselves what looks like the uh Transformer box in here and uh the internal voltage selections are in there and look at this thoughtfully provided two spare fuses in little holders on the side accessible ah it's a thing of beauty and really there's nothing uh exciting in there at all um we've just got some uh range selection switches for the mains range and there's the mains Transformer down there but you know the main point is it is inside its own shielder box they're not taking any chances and as far as the digital board goes well our care factor is around about 0 plus 50 PPM I mean it's not that exciting at all we've got a z80 CPU uh down here and we've got some rums and some memory and well whoopy do I mean all it's doing um there's some jumpers down there that's some programmable jumpers but really you know we don't care ribbon cable going off to the uh front um uh lead uh Matrix uh display and really we don't care you don't need much horsepower at all to drive a gpib and uh then you know drive a display and store a value so check out that monster Mains filter cap though that's absolutely insane manufactured by uh sprag computics um jeez sprag are they still around I don't know but that's a 22,000 microfarad 15 volt cap plus 85° C rated jeez look at the size of it what the hell do they need that for it's absolutely enormous uh for the Relay switching I guess klunk and we have ourselves a date code well it's actually 1970s look copyright 1970 9 John fluke manufacturing Co so this model dates at least back to then I mean the user manual I got says I double checked out 1984 it doesn't say anything uh less than that but uh we've got ourselves a date code we're looking at you know 1989 uh vintage that could have been changed but look you know we've got 88 Vintage Parts bit of dust on this sucker but yeah so this thing was manufactured in late 80s um you know 89 maybe 1990 all right let's lift the skirt up this is what we want to see I better not breathe on this thing better not sneeze because uh we could be in trouble if I do that I definitely won't be uh touching anything inside here I have a poke around with the uh plastic stick I was wondering why there's a hole in there well the hole is a finger lift look at that there we go and I'll try not to spit when I talk to I try not to get too excited and salivate over this thing here we go Tada holy doly look at that oh look at those reays look at the resistors if we take a look at the reays here company I've never heard of Potter and Brumfield who to Google sure enough Potter and Brumfield still exist but well the brand uh does but it's now owned by TE Connectivity and yes they still make reays go figure and as we've saw before on the front panel there are various types of reays in here this one is a 4.5v dc1 um yeah double pole double throw is it and on the analog board we've got ourselves copyright 1983 and down here we got ourselves a bunch of um unusual looking Coto Reed relays of course Cotto one of the uh Best Brands in the business in terms of uh you know low leakage high performance uh relays and these would spe specifically be low leakage types had a quick look of a data sheet couldn't find the exact one but I found a 1240 series 1 uh contact leakage in you know up in the order of uh Power 10 the power of 14 and I think I've done a video Way Way Back on uh relay matrices for measuring uh high value resistors I will try and find that and Link it in and now Precision resistors in this thing check them out they are manufactured well I assume they're manufactured by uh fluke they've got um individual uh part numbers well fluke part numbers I presume and look they've got two in this case uh 40.5k uh 05% um you know I technology I don't know some sort of wire wound and the 05% does seem to be the best that they've got in here we can see but some are you know 0.01 the ones that you know aren't as critical cool there's another couple up there 0.01 uh 05 for 4K 1es 1.8k once you get down in those values 0.1 there's another 162 ohms for example and here we go these low value are 40 ohm ones 05% each but take a look at the configuration what we've got here is uh four paralleled uh 40 ohm resistors so that gives us a 10 ohm better than 05 5 uh% and then that 10 ohms in series with another uh 10 ohm there and then we've got a then that is in series with 262 ohms in parallel and check out the configuration it's uh you know it's all over the shop they got some complex Arrangement here that's for sure and then we've got some 280 ohms in parallel and then we've got uh 1.62 K here and then that's branching off there and no and then 1.8k so this really long string of precision resistors look at that isn't that pornographic special low leakage codo relays well of course on the uh very high side resistor values this one is uh oh you probably can't see that but that's um 81 Meg there it is and it looks like that is you can check the arrangement down in here of this uh string with the relay switching in there and they're 4.5 Mega pop and these ones here are yeah these are all 4 1/2 Meg getting down to 450k here so this part of the circuit is where our leakage really matters and you can see that they've conveniently marked the uh tap points I mean there's the 19 Meg tap there so when you include that 1.9 multip when you switch that button on the front for the 1.9 multiplier there we go that tab for the 10 Meg range that's the tap it's going to use and then there's the 10m uh tap there and so on down the string there's the 1.9 Meg tap there's the 1 Meg tap and you can bet your bottom dollar that they know the leakage of this board down in here and the uh uh solder it's got that crinkly uh solder resist coating on it fluke really would have done their homework on that you can bet your bottom dollar but of course when you get to the really top end of the range the 100 Meg the PCB itself is not good enough you get too much leakage what do you have to do you have to have uh special low leakage insulators like that with uh the uh copper removed from around there and then you have to run insulated pointto Point wires like that right into special output relays like this they couldn't those Cotto ones that we looked at weren't even good enough these are electrol brand ones W look at those how much do they cost each or did cost each there we go electrol brand another one I've never heard of um R 8538 couldn't find any uh info on that on a first pass except that uh Electro were bought out by Hamlin so you know I'm not sure if you could still buy those so check it out we've got the input coming over here switching through here once again on these isolated standoffs here all point to point wiring they cannot use the PCB this wire down there is sneaking off through a hole in the board not only air is good enough for this and for those who love their connectors look at that low EMF tum copper no doubt and in all the excitement over the pornographic reays I missed this little cludge look at this a 7805 linear regulator That was supposed to just probably be freestanding in the board there or maybe um you know attached to the uh back panel there they've M look it looks like they've done a staggered offset maybe they didn't get the PCB right to line up with the shazzy or vice versa so they've I don't know they've done a cludge there mounting it up off the board and uh there's a there's a seal pad in there to insulate it uh from the back but jeez look at that and they've done got to the trouble to add a flat Flex there unbelievable I mean it's not a huge amount of current obviously but you know your guess is as good as mine who goofed up there so you can really see the arrangement of the resistors here I mean here's our 10 ohm tap 9 ohm tap 100 ohm tap 190 ohm tap uh and so forth up there blah blah blah goes around like that comes down here like this but what I don't see is well there's our com that that says com down there there's a common terminal but I don't see the 1 ohm tab it's not there and if you're curious uh 10 ohms the spec for that or the nominal tolerance is uh 500 PPM or uh 05% um in terms of absolute uncertainty we're down to about you know 4050 PPM something like that so that's basically to get that 05% we're looking at 5 milliohms uh maximum error there so in terms of you know relay switching going all the way through the wire into the front panel and all that sort of stuff um that stuff well it doesn't have to be better than that but in terms of the uh you know the stability I mean you can calibrate the uncertainty out of all that but um you know still it's got to be very very low these relays have to be ultra reliable in terms of uh contact resistance so I don't have the uh schematic for this one to hand if anyone does if anyone actually has the service menu which has the front end schematic in it then that would be fantastic the uh actual you know so we can see the actual relay Arrangement wouldn't surprise me if they you know paral paralleled up some relay contacts or something that like that for these really low order ranges I can't imagine where the 1 ohm range is actually haven't found it yet H shielded can wonder what's under there silly me I did find the schematic in the um in the operational manual for this thing was right at the back I swore I checked it and it wasn't in there but duh it is and Tada here's the schematic for the um analog section here with all the relay switching and the wiring and everything else fantastic all beautifully handrawn as you got back in those old days so let's take a look at the string here here we go here's our 1 ohm resistor and our 1.9 Ohm resistor up here which we haven't found yet I bet they're under that can for sure and there you go there's the four paralleled uh 40 ohm resistors that we got before and then all the various uh points there's our common terminal so that's sort of on the so these lower value resistors are on the flip side of that and then we went up the chain went up the chain and we can switch various configurations of these resistors through to a high and low bus very similar to what I've done in a um a previous relay switching Matrix which I said I'll link in and I'll try and do that um and going all the way over to our final output and check it out there we go sure enough they've identified on the schematic look they've put a circle around that which indicates to the PCB designer you know that is to be marked off the PCB I'm not sure over here number one so we're obviously got some detail we'll have to look at but uh yeah that indicates that that's raised off the PCB and this is all just wiring and there we go they've got 4401 cable they're specifying the type used over there so all these high impedance uh parts are all mounted off the board and it looks like I might be right in that they do use multiple uh relay contacts in parallel for the 1 and 1.9 ohm ranges check it out there we go three in parallel on each line just to get that contact resistance down and to within a manageable uh you know tolerance variation over the switching life of the Rel there you go unless noted otherwise there's our note for our points down there here we go Teflon standoffs there we go seven places resistors are specially match set there we go R5 to R39 so R5 yep all of these R5 through to R well R5 R6 yes all of these are specially matched so yeah they don't just roll off the assembly line they're probably individually hand tested hand uh you know hand measured and sorted by you know someone rubbing their Gray beard as they do it sorting them into the individual bins and then you know putting a bit of spit and polish on each one oh beautiful and you can see the beautiful soldered ends of the cans of those resistors there ah just absolutely beautiful goldplated pins oh each manufactured and hand assembled by nude virgins in Utopia land just beautiful if anyone knows if fluke actually uh made these resistors themselves or still do roll their own specialty resistors please let us know so I reckon the uh one what is it the 1 ohm and 1.9 ohm resistors are under this can down in here does it have an extra Reay down in there I don't know I haven't looked at the schematic yet dare I open this sucker it's almost sacriligious isn't it it's just oh oh couple of big standoffs on there and uh yeah I mean as I said you know even at the 10 ohm range only talking what was it 5 milliohms or something to get within that 05% band and uh really it's even going to be well it's an order of magnitude actually less for oh look at those yes wire W resistors on card look at that now I'm sorry that it's not easy to get in here but look at the wire W resistor that looks like one big wide strap on a clear you can actually see through it you can see that's like a some sort of uh you know a perspect uh you know acrylic sort of uh uh sheet that they've wound that on but you can see that that's tape wrapped over there and look individually serial numbered down in there oh yeah these things are definitely manufactured by nude virgins in Utopia land that's for sure and individually tested by gray beards no doubt look at that beautiful they would have been manufactured inous by fluke I'm guessing and check out the Dual terminals coming off that thing both there and there on that side they've just tapped these off individually right at the point there beautiful yeah so that's probably some form of uh nichrome resistor uh well you know wire as in like a flat um strip wire and they've just wound that around that uh plastic seethrough former there and then they've just solded sorry it's very difficult to get in here but then there we go you can see the strap coming over and then solded directly onto that uh dual point contact there one of the interesting things to note is that they have socketed all of these um low value relays by low value I mean used on the low side of the resistant low value resistance side of the reference uh string in there I I would have thought that wouldn't have been the go you would have sold of these directly onto the board but um that certainly makes them replaceable and I'm sure those sockets are worth a fortune and goldplated to the hill and you might be wondering why it was shielded like this well this sucker is a coil in there you don't want anything getting in there because as we saw before these are Precision wire wound resistors these are already uh shielded on one end of the can there so once again it's interesting to see the progression in the contact and resistor technology as you go up right down at the low end here special shielded wire wound you know hand manufactured wire wound resistors down in here and uh multic contct relays down in here to ensure the uh tolerance of that and then we've got some fairly beefy ones for the low value resistance on the Precision string along here so once again you really need good top quality low contact resistance reays possibly multiple contacts along there as you go up you know in the tens of ohms the hundreds of ohms sort of range and then as you get up into the kiler oh range you can go for these uh smaller contact resistance uh by smaller I mean physically uh smaller relays so you know probably not as greater or as lower contact resistance as the other relays along here not nearly as critical on these lower ranges down here and then you get up to these Cotto re relays on the sort of the Meg ohm range up here where contact resistance doesn't matter at all it's down in the noise you can't even measure it what matters right up at this end is the insulation resistance of the relay which wouldn't have mattered for all these ones down here who gives a rat's ass what the insulation resistance of these relays is when it's in parallel with you know 1 ohm 10 ohms hundreds of ohms but when you're in parallel with you know 10 Meg 100 Meg things like that as I said you know ter ohms uh gig ohms you hundreds of gig ohms up into the Tera ohms region actually matters so that's why you need those very high um insulation resistance values up here from contact to contact you know in terms of uh up to up to the point where it's not necessarily the relay that dominates it's becomes the PCB that dominates itself the surface of the PCB as I said you you know oils from hands contaminants from the air other stuff there's no fans in this thing so um you know nothing's actually uh blowing through this thing but dust has accumulated uh to the point where down in this section with the really high values we needed more special relays read relays again and then that Teflon standoffs on the board so we can't rely on the PCB anymore right up at the high end and that as you progress all the way through you got different challenges at each design stage fantastic so I hope you found that uh rather interesting I certainly did and you know there's more technology which goes into actually designing an analog section like this for one of these Precision uh Precision calibrators not only just designing but manufacturing the Precision resistors themselves and uh as I said if anyone has any info on manufacturing fluke manufacturing their own Precision resistors in here uh please let us just know so there you go I love looking in high-end gear like this you always find nice little uh you know design tweaks here and there to get these things um you know to get the operational performance of these units and this isn't even bleeding edge these days I mean this is late '70s you know early to mid 80s technology designed to calibrate multimeters of that era you know it it's probably maybe not even good enough to uh you know calibrate a modern 6 and 1/2 digigit multimeter like uh the agilant one I've got there so there you go unbelievably fascinating I love this sort of stuff and if you want to discuss it best place to do it is the eev blog Forum link below as always I'll link in any available uh data sheets and the manual for this thing which has some beautifully drawn uh circuit diagrams for this thing uh in terms of like all the digital stuff all hand drawn just beautiful really is I love it and here it is working connected to my agilant 3461 a 6 and 1/2 digit bench meter and it hasn't been warmed up for long but it doesn't seem to drift uh well neither of them seem to uh drift much at all really so um this is on the 1K so as you can see it's uh got uh programmed in as the calibration value is 1.66 K and we're reading 1.86 not quite uh bang on but you wouldn't expect it based on the uh tolerance and I can uh change the range there and you can see it seems to be on the high side for uh various uh I can go Auto Range here but often it doesn't quite get the auto range quite right and as you can see this always seems to read on the uh high side of things and point there you know there it is 10 Meg oh that one's reading actually lower than what we're getting on here but as you can see a bit of noise on there probably have to do some extra uh Power Line Cycles and at 100 Meg it's reading low but on most of the uh low ranges it uh certainly seems to be measuring um slightly on the high side I think that one's almost bang on if you see the Auto Range doesn't quite work when it's on this thing when it's quite when it's right near to the full scale there so we can actually change that there we go that one's reading slightly but most of the mid-range stuff seems to be reading slightly high now we're getting into tricky real tricky business down in the uh 10 ohm range down in there but uh as you can see it's um still within spec let's take uh 1K well you know it's it's within margins of this entire uh uh test set up let me give you an example here let's um do this one and see how far we're out in terms of uh percentage so we're looking at 1. 85 for example divided by 1.66 and we're looking at that subtract one change the register there and we're looking at uh times change that to percentage and we're looking at um you know round about o uh let's say Point round it to 0.02% or 20 PPM and that is with that is better the difference there roughly 20 PPM or .2% is within the 24hour accuracy spec of my agilant bench meter here so as you can see um you know which one's right which one's wrong uh you know where we just don't know we would have to uh take this to a standards cow lab that has an order of magnitude or better better than this agilant unit and certainly um at least maybe four times better say than this unit itself to actually have it recalibrated I it doesn't have a cow sticker and I have no idea when it was last uh actually calibrated and these values actually programmed into the thing have no idea so who knows it may not it might have been 10 15 years ago I don't know it might have been a couple of years ago got no idea anyway it does uh work but hey um because the resistors are Ultra stable in this thing um we could just have this re recalibrated we just hook it up take it to standards C lab programming the new values and bang I'm sure it's not going to drift a sausage so as always if you enjoyed it give it a thumbs up catch you next time

Channel: EEVblog

Views: 96,640

Keywords: fluke, fluke 5450a, fluke calibrator, teardown, resistance standard, calibration lab, reference resistor, reference standard, precision resistor, vishay, diy, review, how to, Fluke Corporation (Organization), reed relay, coto, calibration, multifunction calibrator

Id: QFOH8n43kY4

Length: 36min 50sec (2210 seconds)

Published: Tue Nov 05 2013