Quick Test: Advantest R6581T 8.5 Digit Multimeter

this video has been supported by jlcpcb forest major has cancelled everything this year including the electronica exhibition in munich so jlcpcb is inviting everybody to an online alternative and wishes the following message to be broadcasted the event will start november 10th and will last for 15 days in that time four layer pcbs will be available for only two dollars if you also register and visit their virtual showrooms you can win up to three prizes for example an iphone 3d printer or gift cards after visiting all three showrooms you can play spin to win for an additional price with a 100 win chance jlcpcb is looking forward to your visit end of broadcast let's check out the advantage first of all i've got another piece of advantage test equipment here which has been waiting for its time to shine for over two years now it suffers from the same lack of freely available documentation as the device we are looking at today additionally for someone like me who is hardly smart enough for the dc world it's quite an intimidating task to make a detailed video about such a high frequency spectrum analyzer there is no excuse sufficient for a two-year delay just wanted to say this thing has not been forgotten and it's coming soon today though something a little bit easier this is an advantaged r6581t if you keep an eye out you can often get these for less than 1 000 us dollars making them one of the cheapest eight and a half digit multimeters out there personally i like mine and i'll put a link to the ebay seller where i found it in the video description but i wouldn't spend more than a thousand bucks on one because there are some quirks which i'll get to in a moment this model the t version doesn't have an rms converter it can only measure dcv dci and resistances a model r6581 without the t exists with ac functions but otherwise no differences these were and in one form or another still are made in japan so full of prejudice as i am i'm expecting supreme build quality and tentacles of the only documentation i was able to find on the internet is 600 pages worth of manual full of illegible hieroglyphs luckily both front panel and gpib operation are wonderfully intuitive no manual necessary you just press the configuration key followed by whatever function it is you want to configure and then it gives you these easily understandable menus in english what is very much missing however are specifications nobody knows what kind of performance the manufacturer promised in terms of stability accuracy linearity or influence of ambient temperatures smart move that way no promises can be broken but i'm kind of curious what i just spent my hard-earned jlc sponsorship money on so here are a few tests nice okay next we are looking at some of oh okay sorry at first i configured the calibrator for 10 volt dc output and wrote down the multimeter readings for four days straight something must have spooked solatron massively i think i must have disabled its internal drift correction in the first half that's good for low noise performance on that meter but obviously not for long term drift keithley 2002 in green remained well within 1 ppm of its readings peak to peak its 24 hour specification is 1.2 ppm by the way so a very good result here advantist r6581t the meter under test in this case is also not bad at all almost exactly one pppm ppp peak to peak over four days and with four degrees c ambient temperature swings over the course of the days for the low price i paid for that meter that is fantastic just like its internal temperature sensor which can be read via gpib i'm really liking that too next i made a linearity test by sweeping the calibrator output voltage red from -10 to 10 volt over the course of one night i've since been told that that might be too long because there might be a risk of capturing long-term thermal effects but unfortunately the results i got look believable and similar to those of others i've seen as it turns out r6581t in blue is the least linear of the three meters i've tested and to really rub it in it's also sporting that ugly distortion at the zero crossing the same test twice as fast confirms r6581t's poor linearity but there is a nice surprise in here can you see it for a short time solartron 7081 in pink is actually flatter than keithley 2002 scandalous quick to the next point before anybody sees this here i compared two 10 kilo ohm resistance standards a commercial one and a diy they are getting a video of their own soon i noticed a pretty significant temperature coefficient on the one that was connected to the r6581t yellow one that i wouldn't have expected from either model so i swapped them including the test leads and unfortunately found another shortcoming of the advantage it has shown this kind of undamped sensitivity to the room temperature in the voltage measurements already but that was over multiple days in this graph one can really appreciate how promptly and how how dutifully almost to the japanese candidate responds to me opening a window at eight o'clock in the morning in conclusion i would say stability and noise are great linearity and temperature coefficient or not it has some fantastic convenience features though like artifact calibration which lets you calibrate the entire meter with only an external 10 volt and a 10 kilo ohm reference to fully calibrate solatron 7081 for example which doesn't have this feature you need an entire arsenal of standards or a fluke 5700a the front rear inputs are selectable digitally which is unusual i don't know yet if the relays that they are using for that are ppm safe but it certainly is a useful feature in many situations sort of a miniature 2 input scanner card also cool it can read resistance thermometers like pt-100s directly and with a lot of digits i have no way of testing how trustworthy those micro kelvin are but hey more digits more better i also like how fast it is one can get reasonably low noise measurements with only 10 powerline cycles integration time okay let's take it apart and see if we can find out where these results are coming from well there's the expected japanese construction quality already it's beautiful its overall layout is very similar to a 3458a with an outguard digital section and an option slot on the right the big mains transformer has some separate windings which feed a small sub transformer that one lives in and supplies power to the guarded analog section on the left is that an effective technique to reduce common mode voltages because i've only just started to look at special transformers with low interwinding capacitance and a bit of space to be able to introduce a shield between primary and secondary you know recessed shrouded banana jacks are already a pet peeve of mine these look nice and coppery at least but then they lead to internal tin plated crimp contacts [Music] this is where the ltz 1000 voltage reference lives it gives the meter its good stability which we've observed in that first 10 volt for 4 days test it has a pretty normal entourage of resistors except for r2 and r3 the ones that are usually around 70k it's true those are not the most important configuration resistors for the ltz but i don't like the looks of those normal 1 axial types that they are using here when i saw this hybrid module i immediately assumed that this would be advantageous answer to the u180 device in 3458as that is essentially the magic heart of one of the best adcs in existence containing some logic some switches and a resistor network this advantaged part is not that too many of its pins go directly to the input relays its adc is over here controlled by that nec device right next to the input amplifier it says so right there on the board i think we are looking at a multi-slope implementation again with all these metal can jfets acting as switches for the different rundown slopes just above them an integration capacitor and its three best friends so a similar principle to that of the u180 hybrid but with discrete components on a circuit board okay then where the problems at well for example there are only three hermetic precision resistors in here whereas solartron 7081 for example is full of those grey barrels remember i know nothing about these epoxy dipped sip networks that advantaged seems to love so much but if i had to improve that temperature coefficient these parts would be the first i would look at the input linearity thing is difficult it has been discussed by eev block forum heavyweights for years but little to no solutions were found my solution is going to be to put everything back together to be aware of this slight disadvantage and to enjoy this multimeter where input linearity is not the most important thing for that discipline in particular i've got a new champion coming up all the way from cern stay tuned for that it's going to be epic thank you for watching and see you then

Channel: Marco Reps

Views: 43,326

Rating: 4.9532671 out of 5

Keywords: fluke, keithley, calibrator, 2002, multimeter, voltmeter, repair, schematic, nanovolt, adc, slope, mtultislope, nplc, filter, average, digit, count, solartron, 7081, INL, ltz1000, linearity, ambient, temperature, pt100, rtd, grafana, influxdb, raspberry, pi, lxi, python, vxi, gpib, datalogging, stability, noise, guard, common mode, cable, copper, low, thermal, emf, ethernet, logging, inl, japan, advantest, r6581, teardown, electronic, engineering, 3458a

Id: d4LbBeUiUWM

Length: 10min 21sec (621 seconds)

Published: Wed Nov 11 2020