Looking for Benchtop Multimeter - Cheap but good enough

[Martin72]

Hi,

At work I want to create a dedicated place only for testing pcbs.
Most of the components are "fix", for example:

- 4ch scope will be the siglent sds2104xp
- 2ch function generator will be the siglent sdg1032x
- several 4ch supplies from peaktech

But I also want to install some benchtop multimeter, appx 4 pieces.
For our purposes, we don´t need 5,6 or even 7 digits ones.
But what it measure, you should count on it.
So I´m looking for a relative cheap thing, cheap to buy several ones.
But good enough to trust its measures and to calibrate it every year.
Favorite is this siglent :

https://www.welectron.com/Siglent-SDM3045X-Benchtop-Multimeter

Question : Are there other meters nearly as this one, but cheaper ?

[pope]

Also interested 

[tautech]

Martin, before you make any decision I'd urge you to carefully study this thread:
https://www.eevblog.com/forum/testgear/siglent-sdm3045x-enough-is-enough/

[mark432]

I recently bought an Owon XDM3041.
It works reasonably well, but I've found some issues:
-USB interface does not work with the latest version of Windows
-Ethernet interface does not work on my network.  If I isolate to just the lab PC and the meter it works.  Some bit of ethernet traffic causes the meter to lock up and require power cycle.
RS232 control works fine.
-Thermocouple measurement needs to warm up for 1/2 hr to be accurate.

As a bench meter it seems to work fine, but if I was going to do a let of test automation, I'd watch out for the issues with USB and ethernet.

I wrote a driver file for Testcontroller, and that works well with serial interface, and an isolated Ethernet interface.

I do like the ability to log to a USB stick, and the graphing and stats.

Hopefully Owon get the USB and Ethernet issues fixed.  I sent them traffic logs of the interface data in both cases, so they should have enough information to fix it.

As far as accuracy, the next best thing I have is a Keithley 179A, and it agrees with that within the expected error.  I have not sent it for calibration.

[Martin72]

Martin, before you make any decision I'd urge you to carefully study this thread:
https://www.eevblog.com/forum/testgear/siglent-sdm3045x-enough-is-enough/

After the first post, it turns into a hack thread...

[nightfire]

Which are the requirements for that multimeter? Do you simply need some reliable values put out, or do you need some communications stuff for logging purposes?

Is maybe buying pre-owned gear from a commercial seller with warranty (and fresh calibration) an option?
In this case, you could get some reasonable Fluke or Agilent options in that range...

[Kleinstein]

The SDM3045 seems to have quite some software problems. It looks odd to have hardare with ranges of (200 mV ?), 2 V , 20 V, 200 V, ... and software that makes 600 mV / 6 V / 60 V from this. The hack is partially a way to circumvent buggy firmware.

Price wise the difference between the SDM3045 and SDM3055 is relatively small - possibly an option if this would allow for a longer calibration interval.

On the long term anual calibration can be a significant part of the cost of ownership, especially for lower cost meters. The cal costs can vary locally and between meters.

[armandine2]

"Cheap but good enough" seems to be where BK Precision were heading with theirs - BK2831E may fit your budget.

[nctnico]

On the long term anual calibration can be a significant part of the cost of ownership, especially for lower cost meters. The cal costs can vary locally and between meters.

The calibration requirement is going to drive up costs considerably. Probably even dwarf the purchase cost. It may be hard to find a company that will calibrate (check) low costs meters. Doing a cursory check in-house with a reference can be an option as well. For service work the measurement results are likely within certain parameters so when a lot of boards suddenly are marked as 'failed' it is time to check the DMM but with 4 identical setups, you can also re-measure at a difference bench.

IMHO a good option to consider are the bench meters from GW Instek if you also want to have some service level. GWInstek has been around a lot longer than Siglent so their service network had more time to evolve. Farnell offers calibration service for example: https://nl.farnell.com/gw-instek/gdm-8341/digital-multimeter-bench-4-3-4/dp/2460248

Depending on cost versus TCO the lower end (Owon, Vici, etc) may be an option but is also depends on how well the technicians care for their equipment. Getting nice, good quality test leads is probably a good idea. Even the highest end DMMs come with crappy leads.

[rsjsouza]

On the long term anual calibration can be a significant part of the cost of ownership, especially for lower cost meters. The cal costs can vary locally and between meters.

The calibration requirement is going to drive up costs considerably. Probably even dwarf the purchase cost. It may be hard to find a company that will calibrate (check) low costs meters. Doing a cursory check in-house with a reference can be an option as well. For service work the measurement results are likely within certain parameters so when a lot of boards suddenly are marked as 'failed' it is time to check the DMM but with 4 identical setups, you can also re-measure at a difference bench.

IMHO a good option to consider are the bench meters from GW Instek if you also want to have some service level. GWInstek has been around a lot longer than Siglent so their service network had more time to evolve. Farnell offers calibration service for example: https://nl.farnell.com/gw-instek/gdm-8341/digital-multimeter-bench-4-3-4/dp/2460248

Depending on cost versus TCO the lower end (Owon, Vici, etc) may be an option but is also depends on how well the technicians care for their equipment. Getting nice, good quality test leads is probably a good idea. Even the highest end DMMs come with crappy leads.

Indeed. If the company requires calibration, your lowest cost options would be limited to a few choices such as the Siglent, Instek and Rigol (SDM3058E being their cheapest) - I couldn't find service and calibration procedures for any other brand.

Out of that requirement (or if you can surreptitiously slap NCR tags on them )  the options increase by a large margin. Heck, even Uni-T and its 8000E series looks quite cool. Otherwise, I have seen plenty of UT803, VC8145 and similarly looking meters on various repair benches.

As for test leads, it all depends on how abused they will be by the techs. Get the sturdiest ones. 

[Fungus]

Is the requirement for "good enough":

a) "Meter for basic measurements that can be left on all day"
or
b) "Meter with traceable calibration at a reasonable price"

If it's (a) then there's loads of cheap options. If it's (b) then the calibrations can easily swap the cost of the meter over time.

[Martin72]

Hi,

Thanks for all the answers so far.
For the daily purposes we don´t need a 5.5 digits or above meter (we got 2 keysight 5.5).
When adjustments of the circuits are needed, we didn´t this over 3 digits after the comma, mostly 2.
The meter must be calibrated every year as we are under ISO9001.
It should have mV range.
Low cost china "we don´t know it, we only sell it" brands are out of the race.

Martin

[tautech]

Martin, before you make any decision I'd urge you to carefully study this thread:
https://www.eevblog.com/forum/testgear/siglent-sdm3045x-enough-is-enough/

After the first post, it turns into a hack thread...

Look past that to see the different spec between the west and east models that are now proven to be the same HW.
3045X is the oddball in the range, some of its measurement ranges have been SW changed whereas the Chinese 3055X-E measurement ranges match the 3055 and 3065X we see in the west.

Look, SDM3045X have been more popular here than I expected and buyers seem to be satisfied with them and AFAIK use them as you intend, precise measurements are not required but only good enough...service bench type use.

[nightfire]

Does the calibration needs to be done with official certificates (=external lab), or, given a sufficient number of devices could be done inhouse to check against a known good 5.5 multimeter?

In this case, I would make the decision dependant upon the recurring calibration costs from the calibration lab they should quote- maybe a Fluke or Agilent/Keysight device will come in the long run (as they can operated for 10 years with no big problems) cheaper.
In that case, ask your cal lab for some pricing and then choose the one with the cheapest cost that will follow...
Maybe get some used stuff like HP 34401 or old Fluke 45, if they are sufficiently cheap to calibrate...

[Fungus]

Does the calibration needs to be done with official certificates (=external lab), or, given a sufficient number of devices could be done inhouse to check against a known good 5.5 multimeter?

If you're "ISO9001" then there has to be a full paper trail in case of audit.

[2N3055]

So to recap:

- 4.5 digit is enough
- yearly calibration service
- benchtop (mains powered device)

Questions:

- does it have to have connectivity, is there interface preference ?
- does it have to be 4W Ohm
- Is there any specific measurement that it is needed (frequency, capacity, duty cycle, sensors, temp, continuity with adjustable threshold)?

[Martin72]

Hi,

- benchtop (mains powered device)

Yes and for a simple reason:

As described, I want to create a place in the testfield only and explicit for testing pcbs.
And I want to have instruments there, which only will moved away for calibration.
Everytime a colleague will test a pcb, he should find all the measure tools on this place, not wasting time for searching around.
We got "tons" of handheld meters, but only two benchtops.
And they´re always there, nobody likes to work with them because of the mains needing, you´ll always find them on their dedicated places...
So I´m quiet sure, when I order 4 pcs, they will be always on the place.
What features and accuracy concerns, we got lots of fluke 87 here - their specs would fulfill nearly all of our needs in the testfield.
Therefore the "little" siglent benchtop came into my mind.
As we are a manufacturer of power supply solutions, strongly customer tailored, we don´t have mass production.
So automatic test procedures are not important.

Martin

[bdunham7]

The previously mentioned BK 2831E is actually a very decent basic bench meter w/ dual display and USB connectivity.  It is very easy and quick to use.  PC software and calibration procedures are available from the company and these are more widely used than you might think, so annual calibration should not be a problem.

[andr1985]

Thank you for your feedback on OWON XDM 3041.

1. Concerning Ethernet. We purchased a few XDM3041 for the automated measurements setup in 2020 and they work fine (Windows 10, software written in Visual Studio 2019, communication via Ethernet).
We purchased more recently (June 2022) in order to replicate the setup, but there are troubles now (Windows 10 with recent updates, multimeters connected to the computer via Ethernet). We have checked the cables and Ethernet switches, but it does not help. It looks like XDM3041 looses connection from time to time.

Have you solved this issue or maybe received any reply from OWON?

2. We tested a thermocouple (K-type). It seems that XDM gives a constant error +2 deg.C in the range 0 - 100 C.

[MarkF]

You might also look at the VICTOR 8145B.
   http://www.victor-multimeter.com/products/digital-multimeter/bench-type-multimeter/victor-8145b-dual-display-multimeter-638.html

I personally don't know anything about them, but have seen several people use them and a teardown somewhere.
 
One person did several videos where he developed software the control it remotely.


Edit:  Unboxing and teardown video found (He also has two videos on a software interface.)

 

Better teardown

 

[nctnico]

Vici is not Viktor! The VC8145 and Viktor 8145B are different devices.