Looking for a bench multimeter

[Simon]

I'm looking to buy a bench multimeter. I'm particularly interested in taking accurate readings. For example it will be very useful to be able to measure a 10 million ohm resistor. Any suggestions?.

[TiN]

10 Meg resistor? Almost any bench DMM can do it.
What is your budget? Suggestions would be highly different based on that.
34465 seem to be nice daily units.

[ECEdesign]

Get a Keysight 3458A its pretty accurate 

[KJDS]

I've some Solartron 7150plus

Resistance accuracy spec at 1 year is 0.05% +/- 20 counts

Yours for £100 including UK shipping and VAT. £90 if you don't mind a scruffy one.

[uncle_bob]

Hi

If bench space is not an issue, the big old "rack mount sized" beasts are very hard to beat. The HP 3456A is another than has not yet been mentioned. They can (sometimes) be found for < $70 in good working shape.

If bench space *is* an issue, there are a bunch of things like the HP 34401A (looking at what's on my bench here). They are a lot smaller and still have some pretty impressive performance numbers. You *do* pay for the fact that they are a few decades newer ...

As nutty as it sounds, there are even hand held meters that do a fairly good job. I prefer a bench meter for use on the bench. Others like the ability to move things around easily.

Lots of choices.

Bob

[HKJ]

A good bench multimeter is the Keysight 34461A. If you need good precision I like then Fluke 8846A better and my best bench meter is the Keithley DMM7510 (That is a bit subjective). A 8846A that is a few years old matches a new DMM7510 in the measurements, my Keysight do not (The difference is fairly small).
The disadvantage with 8846A is the display, it will degrade over time, the other DMM's are not supposed to do that.

[GreyWoolfe]

KJDS has a pretty good offer.  I just peeked on eBay and saw one for 150 GBP and 35 GBP for shipping.

[John_ITIC]

If you need good precision I like then Fluke 8846A better ...

I have a 8846A that I would consider parting ways with. I'm a digital guy so don't need precision analog measurements. I received it a partial trade for my 6 GHz Agilent Scope. They seem to go for about $1K USD. See below for recent Ebay sale. Please PM me if interested.

http://www.ebay.com/itm/FLUKE-8846A-6-Digit-Precision-Multimeter-/321905776392?hash=item4af3145708%3Ag%3AtRUAAOSwhcJWMZ26&nma=true&si=wLmh6H1OHUr68P5Ib6hrsahbfc8%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557

[uncle_bob]

KJDS has a pretty good offer.  I just peeked on eBay and saw one for 150 GBP and 35 GBP for shipping.

+1

[nctnico]

I just spotted this but I have no idea how well it works:
https://www.eevblog.com/forum/buysellwanted/rigol-ds3064-for-sale/

[forrestc]

I'm looking to buy a bench multimeter. I'm particularly interested in taking accurate readings. For example it will be very useful to be able to measure a 10 million ohm resistor. Any suggestions?.

My $0.02:

10Mohms is pretty easy..   A lot easier than the other end of the spectrum where you start wondering if you're actually reading the resistor or the unseen corrosion on the component or test leads.

My daily bench meters are a Rigol DM3058E and a R&S HMC8012.   A lot of what I do doesn't rely on high accuracy, so 5 1/2 or 5 3/4 digits are plenty and then some.    The HMC8012 is primarily used where I am doing power supply work where simultaneous readings (DC+ AC ripple, or Voltage+Current) is important.  The DM3058E is used for pretty much everything else since it is a bit cheaper, and I wouldn't be as sad if I successfully destroyed it (which occasionally unfortunately happens).   If you're looking for a 'lower accuracy' dmm any of these would work.   

My 'reference' DMM is a Keithley 2000.   It actually is the only DMM that I send in for calibration every couple of years, and it doesn't get used other than as a 'quasi calibration standard'.  It's a 6 1/2 digit DMM, available on ebay for around $5-600 or so - some are even available from certified test labs which come with a NIST traceable cert for that price.     If I was buying this again, I'd go for a 2001 instead, which is a 7 1/2 digit for around $1000 on ebay.   (The 2002 is 8 1/2 digits but is a LOT more).   I really like the keithley units mainly because they are good solid DMM's, and they don't seem to be as popular to purchase as say a fluke or keysight unit - so the price seems to be lower on ebay.

For the keithley DMM's, if you are reading EXACTLY 10Mohm and want lots of digits, I'd try the 2001 instead of the 2000, since the 2000 only reads to the nearest 100 Ohms in the 100Mohm scale, and the 2001 actually has a 20Mohm scale which reads to the nearest ohm.   I'll let you interpret the accuracy spec since the 2001 is both better and worse than the 2000 depending on the scale.

[tooki]

Forrestc, sounds like you're confusing accuracy and precision. More digits gives you more precision, not more accuracy.

[Helix70]

I picked up a second hand Agilent 34401A, and love it. The new Keysight 34461A would be my pick for a new meter.

[teamSMITHusa]

The Rigol DM3058E is nice if you want to stay under $500USD.

[forrestc]

Forrestc, sounds like you're confusing accuracy and precision. More digits gives you more precision, not more accuracy.

I re-read my answer, and I don't think I was.  From my answer:

"For the keithley DMM's, if you are reading EXACTLY 10Mohm and want lots of digits, I'd try the 2001 instead of the 2000, since the 2000 only reads to the nearest 100 Ohms in the 100Mohm scale, and the 2001 actually has a 20Mohm scale which reads to the nearest ohm.   I'll let you interpret the accuracy spec since the 2001 is both better and worse than the 2000 depending on the scale."

The 2001 is a 7 1/2 digit meter, and the 2000 is a 6 1/2.  The 2001 is more *precise*, especially when reading a 10Mohm resistor since the 2000 only has a 10Mohm scale versus the 20Mohm on the 2001.  In many cases, you'd actually have to use the 100Mohm scale since there's a chance you were just over 10Mohms - ending up end up with a "100 Ohm" precision on the 2000 and a "1 Ohm Precision" on the 2001.   Yet the accuracy between the two is too close to call....   For instance, on the 10Mohm scale, on the 2000 is +-400+10ppm at 1 year, and the 2001 is 900+4.5ppm on the 20Mohm scale at 1 year, which means the 2001 is less accurate - until you go slightly over 10Mohms.   You then have to switch to the 100Mohm scale on the 2000 which is 1500+30 ppm which is much worse than the 900+4.5 of the 20Mohm scale.

So, no matter what the 2001 has better precision, yet the 2000 has better accuracy just below 10Mohms.   The 2001 has better accuracy from 10M-20Mohms, From 20-100Mohms, the 2000 wins for accuracy again and so on, both and above these ranges ( I didn't compare any further).  That's why I said " I'll let you interpret the accuracy spec since the 2001 is both better and worse than the 2000 depending on the scale"

The phrase "nearest ohm" might have been a bit inaccurate though....

And I guess I should have been more clear in the first sentence of my response since I really meant something like "I don't require high accuracy and as such the typical accuracy of a 5 1/2 digit DMM is enough and then some", since that is pretty close to what I meant.

[Performa01]

Unfortunately you haven’t been very specific on your requirements. Since even the cheapest handheld DMM will be able to measure 10Mohm, the big question is: what resolution, what accuracy?

In general, accurate resistance measurement isn’t exactly trivial, and it gets increasingly difficult for higher values. Very few modern bench meters offer better than 100ppm one year accuracy for the lower resistance values up to 1Mohm and 400ppm for 10Mohm. Exceptions at an affordable price can be found in some ancient high end meters, which are usually what I think is called ‘boat anchors’ in the English speaking world - I would probably call them fitted kitchen appliances

Some options for the used market:

Solartron/Schlumberger 7150(plus) is a very nice looking compact meter which can be had for cheap due to its age, and its AC accuracy was outstanding back then so it can still compete today, but resistance measurement isn’t particularly good at 500ppm up to 20Mohm. It has some other shortcomings too, so I’d consider it a good choice for the occasional need for high precision, but not as an everyday bench meter.

Fluke 8840A and 8842A are very nice meters with excellent long term stability, often found at reasonable prices <400€, but with only 200000 counts and one has to be aware that they will only do true RMS AC measurements when option 09 is fitted. Display and UI are most convenient, but there are no bells and whistles like continuity check or even diode test.

Beyond that, the usual suspects come into play: Agilent 34401A, Keithley 2000 and 2015THD can sometimes be had at reasonable prices, and they all offer 1200000 counts together with state of the art accuracy specs.

If you want to buy new, I would like to mention:

Rigol DM3068, one of the very few modern DMM that maintains high input impedance up to 20V. Specs are pretty much the same as the before mentioned Agilent/Keithley devices, even though I would expect inferior long term stability. Functionality is a bit of a mixed bag, as it has some shiny features like histogram, but lacks others that might be more important at times.

My personal favourite is the Picotest M3500 (aka Keithley 2100), which is about the same price as a Rigol DM3068, but is pretty similar to an Agilent 34401 spec. and feature wise. It comes with a ready to go Excel plugin that makes external data logging through USB really easy, a feature that requires more effort with most older meters that only come with GPIB (if at all). Long term stability is quite good, but still expect it to drift out of spec some 2-3 years after initial purchase for the V and A ranges, whereas of all things the ohms measurement might go out of spec after only half a year. This is the downside of buying a new precision bench meter – you will need to re-calibrate the unit for at least the first three years…

[uncle_bob]

Hi

Calibration brings up a whole other world of issues. IF a calibrated meter is what you are after, it is well worth looking into the cost of a cal trip. A newer meter from the big guys will plug into a tester and virtually calibrate all on its own. No tearing into things, no tweaking pots. It is all numbers stored in memory somewhere. The real cost of calibration is the couple of minutes indirect rental of that fancy box it connects to.

A big old beast (like my 3456) is a bit more labor intensive to calibrate. Yes a *lot* of people know how to do it. It does hang around on the bench getting poked for a bit more time than a new meter. You also need to have (to my mind) a bit more skill / training to calibrate one.

Check with your "cal guy" on anything you are thinking about. My 3456 gets calibrated as a favor under a bulk calibration contract. I suspect that if it was priced out individually (or calibrated more often than every 5 years) it could be a bit expensive compared to a newer device.

Bob

[Performa01]

Two more hints:

On the subject of calibration, all the meters I’ve mentioned in my previous post have the ‘closed cabinet calibration’ – even the ancient Solartron/Schlumberger 7150. This is also true for the HP 3457A mentioned below.

I forgot to mention the HP 3457A, which is still a reasonable size, i.e. despite being full rack width it is only 300mm deep, thus fitting on regular benches and shelves. It offers a wide choice of measurement ranges and an exceptional accuracy on resistance measurement. But it is hard to find a reasonably priced unit and it has a few downsides, like the ugly LCD display without backlight, high input impedance only up to 3 volts and the very poor AC accuracy, as is typical for HP gear from that era.

[pxl]

The DM3058E is used for pretty much everything else since it is a bit cheaper, and I wouldn't be as sad if I successfully destroyed it (which occasionally unfortunately happens). 

But how do you destroy these DMMs? Is it possible to destroy it at all? The current input(s) are protected by fuses, the voltage input should tolerate anything below 1000V, I guess. What do I miss here?

[forrestc]

The DM3058E is used for pretty much everything else since it is a bit cheaper, and I wouldn't be as sad if I successfully destroyed it (which occasionally unfortunately happens). 

But how do you destroy these DMMs? Is it possible to destroy it at all? The current input(s) are protected by fuses, the voltage input should tolerate anything below 1000V, I guess. What do I miss here?

Honestly I don't expect to be able to destroy either of these....  but in the past I've destroyed my fair share of dmms..

The culprits were usually either a dmm which wasn't all that well protected in all the ranges (high voltages on ohms or current scales is a common culprit), or susceptible to esd.  Some DMMs also don't particularly care for high levels of RF.  I also know that I've fried at least one DMM as a result of that scary warning about one of the current ranges being unfused.

If these are both truly 1000V protected in all the ranges and properly fused,  then they will live a good long life.   I'd just rather tempt fate on a lower cost dmm.

Sent from my SM-N900V using Tapatalk

[EEVblog]

It's all about the budget, how much you willing to spend?

[nctnico]

Also very accurate bench DMMs may not tolerate high voltages that well. For example: my Vici VC8145 is rated CAT II 1000V while my Keysight 34461 is rated CAT II 300V even though both are capable of handling 1000V DC / 750V AC.

[mongo]

Note if you are looking at the Keithley units the Model 2100 may be nicer than the 2000.  You do lose the ability to use the relay card but you gain built in USB and a BNC trigger input.  I personally would prefer to have built in USB but I am not looking to run the meter in a fixed install, programed test environment.  Being blind the LCD display in the Rigol's tend to be a lot less useful for me in a higher precision meter compared to the VFD.  Note that I picked up my unit with a recent cal cert used from a rental company and it was in like new condition for ~$150 less than the DM3068.  They are not multi-voltage so that may not be an option for you but if the specs hit your need I would think the extra $50 in street price would be worth it if the specs meet your need.  While the Rigol is probably a good unit you will make far more than that if you ever choose to sell it too.