Prema 5000 6.5 digit bench DMM repair project

[Traceless]

Hey everyone,

I have a nice Prema 5000 6.5-digit bench DMM here. While the device is faulty its condition is pretty good overall, especially considering its age. Since Prema is a german brand and the meter does not seem to be too common abroad I thought I'd share some pictures. manual, specs and schematics are available online, albeit in German. While the meter might seem very yellowed on the first glance the chassis is actually painted metal so I believe what you see is the original color.



A view under the hood reveals that the meter must probably have been made somewhere around the mid 1980s.



The meter actually powers up but once you put it in auto ranging mode it goes bonkers . I also get "Error 1" in certain modes which according to the manual indicates "overrange". I think a good place to start would probably be to swap the battery (bottom left). I'm also not super happy about the open glass fuse on the right side. I might drop a ceramic substitute in. Here is a short demo of the running meter:

Faulty Prema 5000 6.5 digit bench DMM demo.

As you can see, in the V, I, R measurement modes the last to digits do not seem to be displaying anything, that might be related to the ranging issues, the 7-segment displays seem to be all in working order though, because in the beginning of the video the meter shows "Contr. 1".

Let me know what you think and merry Christmas to everyone!

[Kleinstein]

The bevavior looks normal: with on open input the voltage dirfts as the inpedance is very high and the input bias is enough to make it drift up. Once it swiches to the 20 V range the 10 M divider is engaged and discharges the capacitance. A better test would work with a short at the input or a capacitor to slow down the drift, so that one can better see it.

[Traceless]

Hello Kleinstein,

thank you for the response I took a quick take with shorted inputs as requested. Some drift with open input seems okay, however what makes me suspicious is that the relays keep switching as well (Starting at 0:13 in the first video I posted). Are you sure that is normal? It seems that this would unnecessarily wear the relays (and annoy the user )


Prema 5000 test run with shorted inputs.

[Kleinstein]

The prema meters use quite some relays for range switching. Especially the connection of the high voltage divider is usually done with a relay also with many other meters. One gets a similar relay clicking with an HP34401 or HP3478/3457 - how fast can vary between the units and usually the relays are not as loud.

There could still be a bias current that is on the high side or a relatively small input capacitance.  For a test of the input current one could connect a capacitor (PS or PP type) of some 10 nF with low leakage / loss. The input bias current can than be estimated via  I = C * dU/dt.

AFAIR (it is quite some time since I used such a meter) the meter is not super low noise and one often sees a little noise and drift even with a sharted input. Overall it is an OK, low end 6 digit meter.

[Traceless]

Hi Kleinstein,

I read through the Prema manual today. You were absolutely right the meter seems to be fine. The ranging problems occurred because the zero point calibration was off. I performed calibration for all modes and so far everything seems to be in order, even the internal lithium battery still measures 3.4V, what are the odds?

However what completely blew my mind was how ridiculously good that meter is at AC voltage measurements even at very high frequency and low amplitudes. Here is how the Prema 5000 performs against my Brymen BM869s. Consider my mind blown .

Prema 5000 high frequency, low amplitude AC voltage measurements (vs. Brymen B869s)

[Kleinstein]

It is not so unusual that the bench top meters perform better with AC than the handheld meters. The handheld ones are battery operated and thus need to conserve power. A fast AC part needs more power (e.g. for a fast amplifier) and a higher supply also helps. So the mains powered meters have it easier and at the price point they can effort the better and faster of the RMS converters.  The prema meters actually don't have an especially good AC performance for a 6 digit meter. AFAIR it was more a low cost 6 digit meter.  From the video it is more the performance of the Brymen meter that is surprisingly good.


The Battery backuped memory does not need much current. Especially if storred cool and dry the battery can last quite long - if hot or humid the battery is consumed quite a bit faster.
It is also possible that essentially not current is needed from the battery while the meter is powered on and some such meter were run 24/7 for quite some time.
It depends on the battery type if the voltage measurement alone is a good indicator: some battery types show a resonable voltage even if essentially dead. It is more the internal resistance that goes
up. Maybe check if a 1 M resistor as additional load makes a difference.

[Traceless]

It is not so unusual that the bench top meters perform better with AC than the handheld meters. The handheld ones are battery operated and thus need to conserve power. A fast AC part needs more power (e.g. for a fast amplifier) and a higher supply also helps. So the mains powered meters have it easier and at the price point they can effort the better and faster of the RMS converters.  The prema meters actually don't have an especially good AC performance for a 6 digit meter. AFAIR it was more a low cost 6 digit meter.  From the video it is more the performance of the Brymen meter that is surprisingly good.

Fair point, the Brymen is a pretty good handheld, and while it needs to run from battery it still benefits from ~40 years of technological advancements compared to the Prema.  Maybe I just had low expectiations but it seems to do pretty well for a meter that probably hasn't seen calibration for ages. What would you consider the bar for a bench DMM to be a "good/high end" 6.5 digit DMM.  BTW according to the manual I linked in my original post above the Prema 5000 is a 6.5 digit DMM. Why do you consider it a "low end 6 digit DMM"?

The Battery backuped memory does not need much current. Especially if storred cool and dry the battery can last quite long - if hot or humid the battery is consumed quite a bit faster.
It is also possible that essentially not current is needed from the battery while the meter is powered on and some such meter were run 24/7 for quite some time.
It depends on the battery type if the voltage measurement alone is a good indicator: some battery types show a resonable voltage even if essentially dead. It is more the internal resistance that goes
up. Maybe check if a 1 M resistor as additional load makes a difference.

I'll give that a try tomorrow. Still considering to change the battery even if it reads good. Apparently the lithium batteries are not as bad as the alkalines but I'd rather avoid the cell spilling its guts inside the meter, which given the age should only be a matter of time. I've done a quick google search for through-hole lithium buffer batteries, but haven't found similar ones - what would you recommend as a suitable replacement?

[Kleinstein]

I would consider the HP34401 or Keithley2000 about the mid range 6 digit meters. They still are more cost saving with only a few current ranges.

The Prema5000 is limited with only a 2 V high Z range and also only a few current ranges. The AC ranges are also limited to DC coupled.
From the design with the same amplifier for AC and DC is a bit limiting for the bandwidth and also more a low cost solution. It still comes with the positve side of having high impedance for 2 V AC.

I have used a Prema5000 quite a bit for low voltages and the last digit is not really stable and it looks like the ADC is also limited getting a somewhat uneven histogram (visible DNL errors - maybe limited resolution before rounding). I faintly remember that the meter was also relatively afordable, more competing with a HP3478 and Keithley 197 price wise.

[Traceless]

Okay, but aren't both those meters (HP34401, Keithley 2000) significantly newer than the Prema (which was probably designed in the late 70s/early 80s, my unit seems to be from 1984)?  The HP34401 and Keithley 2000 look like early 90s designs to me. Sure would be nicer if high Z wasn't limited to the 2V range. For a more fair comparison instead of using the Brymen I took some measurements of the Perma 5000 vs my Keysight 34465A (AC Volts at 60 Hz, 1 kHz, 10 kHz, 100 kHz and Resistance 10 Ohm to 1 MOhm with .1% SMD resistors). I also took some VDC measurements. Since my bench PSU is not exactly a high precision DC source I unfortunately could not really perform sensible measurements below the 100 mV level.

As for the results (see below), the Prema seems to hold up pretty well but apparently could use some calibration in the > 2V ranges.
[ Specified attachment is not available ]

[coromonadalix]

the 34401a is about 30years old now if not older

there is some meters with 20v  high z  inputs in the giga ohms


what strike me on the 5000   is why 2x or the custom made adc's  in it  someking of compare between the 2 ??  null out deviation, aging  etc ... oversampling ?

[Kleinstein]

There is only 1 custom ADC chip, but 2 different custom chips that togehter (and with more parts) make up the ADC.

There is a prema DMM with 2 ADC chips: the 6048  8 digit version uses a 2nd ADC chip to measure the temperature. The ADC chip itself is not really expensive - it's more the resistors around it.

The HP34401 should be from around 1989. So it is newer and even the HP3457 is from around 1985 or 86 and a little later than 1984.  I have used both a Prema5000 and HP3457 for quite a while and on most aspects (except the display) I would prefer the HP meter. It was also likely a lot more expensive.

[voltsandjolts]

Here is a short demo of the running meter:
Faulty Prema 5000 6.5 digit bench DMM demo.

My OCD very much approves of you using made in germany hand inside made in germany glove to control the made in germany DMM

[Traceless]

@voltsandjolts:  didn't notice the "made in germany" on the gloves until you mentioned it.

[Traceless]

@Kleinstein: Unfortunately I haven't managed to get my hands on HP gear. It seems just hard to get here in Germany for a reasonable price. Most of the stuff you find is completely wrecked and grossly overpriced at the same time (compared to US prices). So I don't have a comparison in terms of usability. After reading the manual I think the Prema is very easy to use - cosmetically if I had to change one thing it would be the mustard color. I think I'd rather go with the classic test instrument grey like they used on the bigger brother (see picture below). It's a bit like the old Keithley brown, not exactly my taste, still I do like the old Keithley gear.

[Kleinstein]

The usage is reasonable easy with the Prema - just the old style 1 letter commands, like the old Keithley 19x series and many other meters of the time. However the commands are only similar and not exactly the same or compatible. What I liked was that the scanner function also includes the current ranges (though with some limitations on the contact rating). This worked well with a AD590 temperature sensor that gives current out (1 µA/K).  This in part seems to come with the way they use the same terminals for current and 4 wire ohms.

Yes, used HP/Keithley  gear is often expensive and not that common in Europe.

[MadTux]

With a bit of luck, that might end up quite cheap
https://www.ricardo.ch/de/a/tisch-multimeter-keithley-2001m-75-digit-1248998294/

[Traceless]

Thanks MadTux, the Keithley 2001 is a really nice meter. Assuming the calibrator shown on the pictures is precise, the meter's DC readings seem to be a bit off (-3.54 mA and +500µV).

[Traceless]

During further testing I noticed that there seems to be some interference with AC measurements at mains frequency (50 Hz) above the 2 VAC range. The interference declines the further the measured voltage source frequency deviates from mains frequency. The < 2VAC range seems (mostly) unaffected. I guess the source would be the power supply, maybe bad filter caps, then again wouldn't that also affect the < 2 VAC range in a similar fashion?

Short demo is available here.

[Kleinstein]

The meter uses the same 10 M input divider for AC as for DC (just like most handheld meters) and is thus a bit sensitive capacitive coupled hum with the divider. Combined with a lack of shielding, chances are this is just capacitive coupled hum comming in at the divider (still some 900 K impedance for the 20 V range). With the generator set to 50 Hz and the mains frequency not exactly at the same frequency one can get a beat frequency where the 2 signal add and possibly also subtract.

Chances are that adding some electrostatic shilding (e.g. to the top cover of the case and maybe the bottom could improve things quite a bit. Looking at the output of U3 with a scope should give a hint on how much hum and what frequencies (maybe 100 Hz if it is bad filter caps).

Large hum should also show up with a shorted input and fixed 20 V AC range.


Either the generator or the meter is also quite a bit off at 2 V from the generator.

[Traceless]

The meter uses the same 10 M input divider for AC as for DC (just like most handheld meters) and is thus a bit sensitive capacitive coupled hum with the divider. Combined with a lack of shielding, chances are this is just capacitive coupled hum comming in at the divider (still some 900 K impedance for the 20 V range). With the generator set to 50 Hz and the mains frequency not exactly at the same frequency one can get a beat frequency where the 2 signal add and possibly also subtract.

Chances are that adding some electrostatic shilding (e.g. to the top cover of the case and maybe the bottom could improve things quite a bit.

Thank you very much Kleinstein. Do you have any tips on how to improve shielding, simply adding sheet metal seems a bit redundant since both top and bottom of the prema case are metal.

Looking at the output of U3 with a scope should give a hint on how much hum and what frequencies (maybe 100 Hz if it is bad filter caps).

Large hum should also show up with a shorted input and fixed 20 V AC range.

I'll take a look with the scope and report back. I don't recall seeing any issues with shorted inputs but I'll double check. Maybe I just didn't notice.

Either the generator or the meter is also quite a bit off at 2 V from the generator.

The meter is definitely off in the higher ranges, I already noticed that during the test-run against my 34465A above. So some calibration is needed. Unfortuntely I don't have the necessary equipment and realistically I won't get my hands on a Datron 4700, Fluke 5700A or the like. I think I'll create a thread over in the Metrology section to see if anyone has suggestions on how to set up a "Poor man's calibration lab" I'm thinking in the direction of PDVS2-Mini + Homebew voltage multiplier board for 10x and 100x VDC and something like this for resistance. No idea how to tackle current and VAC yet.

[Kleinstein]

If the case parts are metal, they would have to be connected to PE (or maybe floating), but they should not be connected to the ADC part circuit ground. A good shield would need to be connected to the ADC ground = low side of the voltage measurement and well insulated from the case and high side of the input.

With the typical cheap generators with SMPS supply and no PE connection the signal has a good chance to have quite some common mode voltage to ground (e.g. about half of mains via the class y capacitors at the supply). With just a short at the input one would not notice that. Chances are the hum comes in with common mode voltage.
For a test one could check what happens of the DMM low side / generator is also connected to PE (e.g. meter case).

[Traceless]

Hi Kleinstein,

If the case parts are metal, they would have to be connected to PE (or maybe floating), but they should not be connected to the ADC part circuit ground. A good shield would need to be connected to the ADC ground = low side of the voltage measurement and well insulated from the case and high side of the input.

you are absolutely correct. Chassis is connected to PE, and the ADC part is isolated except for the ribbon connection to the PSU board.

With the typical cheap generators with SMPS supply and no PE connection the signal has a good chance to have quite some common mode voltage to ground (e.g. about half of mains via the class y capacitors at the supply). With just a short at the input one would not notice that. Chances are the hum comes in with common mode voltage.
For a test one could check what happens of the DMM low side / generator is also connected to PE (e.g. meter case).

I did a test run connecting one function generator's BNC connector's shield to the Prema's metal back-plate via alligator clip and that helped a lot. The meter is pretty stable now in the 20 VAC range and only reads about 40-50 mV low (assuming the function generator is precise). So I guess additional shielding is not needed the noise came in via the Uni-T's power supply.

Low amplitude AC measurements - Prema 5000 / UTG962E common earth

I'll use a power bank to run the Uni-T from battery and avoid PSU related noise issues. I'll probably also make a banana-to barrel cable so I can hook it up to my linear bench PSU which should be a lot better than the 5V adapter that came with the function generator.

I also noticed an interesting issue where I'm getting a declining reading upon range switch between high Z and non-high Z ranges. Looks almost like some capacitance related artifact. A short demo is available here.

While the meter was open I also measured the big 4700µF 16V cap at the bottom left of the PSU board, just because it was easy to access and does not require removal of the board. It measures a bit high ~5500 µF but ESR is >0.1 Ohm (in circuit). I heard that older caps tended to rather read high then low - so I guess its still good.

Also the lithium cell must still be good (enough) because according to the manual the Prema runs 3 self tests (Contr. 1-3) at startup. Contr 2 "calculates a check sum over the calibration factor data stored in the buffered RAM and compares those with a checksum". If the battery was to weak I think I'd just get an "Err. 8: Error during self-test 2: Calculated checksum and control checksum do not match (data-loss in battery buffered RAM).".

I guess now I need to come up with a plan to get the meter calibrated. Resistance range is good and pretty close to the Keysight. The <=2VAC range looks also good.

Also there is an interesting safety remark in the manual chapter. 8.1. translation:

"It must be considered that the maximum permissible values of 50 VDC or peak voltage between the "LO" input and guard (see section shielding) and 200 VDC or peak-to-peak AC between guard and chassis are not exceeded. For high-voltage devices without mains isolation this must be considered during polarity selection."

So if I get that right they are saying: If you measure high voltage DC, don't connect DC+ to the black input and also if you measure high voltage AC make sure you don't connect any AC signal (or anything floating 200V above chassis/PE) to guard, correct?

[Kleinstein]

Since connecting the input side the PE makes a difference an added shield and maybe added isolation material would absolutely make sense.
A good meter should still measure OK with such a common mode signal as from the cheap generator, at least with the guard connected to the Lo input. Grouding the voltage to measure is not always an option.

There seems to be a guard / shield terminal, but somehow the shield missing inside  :'(.  There may be something on the underside, but it looks like nothing on the top.

It is just just that there may not be much space for it. At least on the top, there should be space, reserved for the scanner PCB.

[Wollvieh]

The prema meters actually don't have an especially good AC performance for a 6 digit meter. AFAIR it was more a low cost 6 digit meter. 
(...)

The Battery backuped memory does not need much current. Especially if storred cool and dry the battery can last quite long - if hot or humid the battery is consumed quite a bit faster.
It is also possible that essentially not current is needed from the battery while the meter is powered on and some such meter were run 24/7 for quite some time.
It depends on the battery type if the voltage measurement alone is a good indicator: some battery types show a resonable voltage even if essentially dead. It is more the internal resistance that goes
up. Maybe check if a 1 M resistor as additional load makes a difference.

In the manual they boast a little about their patented "double slope"-AD, which allows a "cheap but still high end" device. So maybe that comes at a price, as the AC V and A error of the 6 1/2 digits Prema 5000 is the same as of a BBC MA5D with ony 30000 counts, making it 4 2/3 digit (if I calculate correctly). In all other ranges The MA5D is a decade worse than the Prema.

The battery and +5V-CPU-Line are each connected to VCC-RAM by a diode. So as long as the meter is powered, the battery is not discharged, maybe even slightly charged by reverse current through the diode.

I wouldn't test the old battery with a load, as that (w|c)ould be a destructive test with regard to the RAM's contents.
Old Li-batteries with no power draw often get Ri in the hundreds of ohms which makes them "empty" for many uses (for example in 10 year smoke detectors which die after two or three years regularly). But in the HP 3478 there is a 470R in line with the battery, so any inner resistance of the battery below 100k will most likely be ok.
 
By the way the user's manual linked in the first posting is for a more modern version containing a 48z02 "Zeropower RAM" with integrated battery. I attached the (apparently) correct version for this (and by the way my) meter here.

My meter has the Scanner option, so 125V max allowed and no 1000V Range. 
But maybe better shielding because of the additional PCB above the analog part. 

I wonder if the 125V restriction is technical (I think the relay contacts are speficied to 250V) or for safety reasons only, as the D-SUB-50 on the back though female can be accessed freely.

[Traceless]

My meter has the Scanner option, so 125V max allowed and no 1000V Range. 
But maybe better shielding because of the additional PCB above the analog part. 

Hi Wollvieh, thanks for sharing and pointing out the differences in the manual. If you have the time I'd be super curious to know if you can replicate the drift I observed when measuring low amplitude AC voltage at mains frequency (see post and video above). If your meter does not misbehave in a similar way that could either point at the scanner option providing better shielding or my meter having some kind of problem.Who knows maybe the meter originally had a top shielding and the previous owner threw it out after servicing the meter or temporarily installing the scanner mod.