DMM

[LunarMist]

I have a Fluke 8060A/8062 or some DMM about 25 years old. Is it worth calibrating or even possible?
What today has comparable accuracy and quality? Most of the stuff on the Amazon looks cheap and unreliable. Thanks.

 

[Stereodude]

You can probably get it calibrated. What are you trying to do with it? The Uni-T UT61E is a popular not Fluke money choice.

 

[LunarMist]

You can probably get it calibrated. What are you trying to do with it? The Uni-T UT61E is a popular not Fluke money choice.

Thanks, but that looks like some funky Chinese with no US service presence. How do you know if they are accurate or get them calibrated?

 

[jtr1962]

A few questions:

1) What do you intend to use it for?
2) What kind of resolution/accuracy do you need?
3) Why is calibration important here?

The answers to these questions will at least give you a starting point. I have a Fluke 189. I really needed the accuracy and 50,000 counts of resolution but for most people it would be gross overkill. I bought it on eBay and the seller had it calibrated. However, calibrations in good DMMs rarely drift enough to matter. Unlike bench multimeters which might be powered on 12 hours a day, 7 days a weeks, battery-powered DMMs rarely see enough hours to cause any significant drift in the calibration. That's doubly true when the calibration is done electronically, hence no trimpots which might degrade over time. So unless you're running a secondary standards lab, or have some real need to check voltages or currents to 0.1% or less, you don't have to worry much about calibration. Just check that the accuracy specs of the DMM you want are good enough for your needs. The factory calibration should be good for decades unless the meter sees rough use.

The DMM you already have may well still be good enough for you. If you're concerned about the accuracy, you can get it calibrated, or just buy a voltage standard like this one on eBay and check it yourself.

For really basic stuff like checking battery voltages even $3 Harbor Freight multimeters are fine. They're accurate to better than 0.5% and most have an adjustment pot which can get you to 0.1%, or +-1 count on the 10.00V scale.

 

[LunarMist]

A few questions:

1) What do you intend to use it for?
2) What kind of resolution/accuracy do you need?
3) Why is calibration important here?

The answers to these questions will at least give you a starting point. I have a Fluke 189. I really needed the accuracy and 50,000 counts of resolution but for most people it would be gross overkill. I bought it on eBay and the seller had it calibrated. However, calibrations in good DMMs rarely drift enough to matter. Unlike bench multimeters which might be powered on 12 hours a day, 7 days a weeks, battery-powered DMMs rarely see enough hours to cause any significant drift in the calibration. That's doubly true when the calibration is done electronically, hence no trimpots which might degrade over time. So unless you're running a secondary standards lab, or have some real need to check voltages or currents to 0.1% or less, you don't have to worry much about calibration. Just check that the accuracy specs of the DMM you want are good enough for your needs. The factory calibration should be good for decades unless the meter sees rough use.

The DMM you already have may well still be good enough for you. If you're concerned about the accuracy, you can get it calibrated, or just buy a voltage standard like this one on eBay and check it yourself.

For really basic stuff like checking battery voltages even $3 Harbor Freight multimeters are fine. They're accurate to better than 0.5% and most have an adjustment pot which can get you to 0.1%, or +-1 count on the 10.00V scale.

I have a fundamental issue with measurement error. NMT 1 ppt (0.1%) is just how my brain functions, and even that seems high. I'd probably get a bench meter if more accuracy and resolution were needed. My Fluke was fine, but I have no idea after 25 years. I will find one of the precision voltage sources and check just out of curiosity.

From what I am reading 0.025% is the target basic DC accuracy for a good handheld in the 40,000-60,000 count range. I suppose that would be the 289, though it looks to be a rather aging model.
What do you think about Agilent meters or whatever they are now?

 

[jtr1962]

I have a fundamental issue with measurement error. NMT 1 ppt (0.1%) is just how my brain functions, and even that seems high. I'd probably get a bench meter if more accuracy and resolution were needed. My Fluke was fine, but I have no idea after 25 years. I will find one of the precision voltage sources and check just out of curiosity.

From what I am reading 0.025% is the target basic DC accuracy for a good handheld in the 40,000-60,000 count range. I suppose that would be the 289, though it looks to be a rather aging model.
What do you think about Agilent meters or whatever they are now?

Agilent is fine. I was seriously considering getting the U1253B for a while. Love the OLED display, although because of it battery life sucks. For a bench DMM something like the 34461A is great if you want to spend that kind of money. I couldn't justify it personally. Or you can opt for its predecessor the now discontinued 34401A which can often be had on eBay for under $400. For example, here's a nice used calibrated one. If I was still getting a lot of hours on my consulting job I would probably get one for myself but at this point I can't justify it. I already have the Fluke and an HP3478A Bench multimeter.

 

[Stereodude]

A few questions:

1) What do you intend to use it for?
2) What kind of resolution/accuracy do you need?
3) Why is calibration important here?

None of which were subsequently answered...

 

[LunarMist]

Well I answered #2 and #3 is related. #1 is mostly VDC ADC and Ω. I rArely measure caps and diode or continuity can be tested with a £20 meter.

 

[Stereodude]

I beg to differ, but it's basically irrelevant to me.

 

[LunarMist]

If an instrument is not calibrated properly the data obtained is not valid.
Nobody cares about 17025 anymore?

 

[jtr1962]

If an instrument is not calibrated properly the data obtained is not valid.
Nobody cares about 17025 anymore?

That's one reason I have a few voltage and resistance standards. Occasionally I check my instruments to see if they're still calibrated. Other than the Harbor Freight multimeters, which might drift a few tenths of a percent, everything else holds its calibration nicely. I seriously question if DMMs need to be calibrated as often as the calibration places say. Maybe in heavy, rough use they do, but not if they sit mostly unpowered on a work bench in climate-controlled conditions.

Frequent calibration is probably more relevant to very high resolution instruments with very high specs. References can easily drift enough in a year or two to throw the readings in a 7.5 digit DMM off by a few counts. Not so when the resolution is only 50,000 counts.

 

[Stereodude]

If an instrument is not calibrated properly the data obtained is not valid.
Nobody cares about 17025 anymore?

So what? Do you think I care if my DMM is 1% off when I'm measuring a battery for a flashlight or an RC vehicle?

Since you won't tell us what you're measuring and why you need a NIST traceable calibration for the data taken with it why should we take your concerns seriously? It sound like you're taking hobby related unimportant measurement of nothing of importance but are stressing over things that just don't matter.

 

[LunarMist]

It matters to me. OTOH the color temperature of a flashlight not so much.
I'm sure DD would say it matters if a vehicle takes 4 vs. 6 or 8 seconds to 60, but in the city it wouldn't matter much to me either.

 

[Stereodude]

Then get out your wallet and buy a high precision DMM with NIST traceability.

 

[LunarMist]

Then get out your wallet and buy a high precision DMM with NIST traceability.

I will try the AD584JH reference next week and see where my Fluke stands. I'd be surprised if it is close to twice the natural tolerance.

 

[jtr1962]

I will try the AD584JH reference next week and see where my Fluke stands. I'd be surprised if it is close to twice the natural tolerance.

You might be pleasantly surprised. My friend had a Fluke of similar vintage which he asked me to check. Most of the ranges were either dead-on-nuts or a few counts off.

 

[LunarMist]

You might be pleasantly surprised. My friend had a Fluke of similar vintage which he asked me to check. Most of the ranges were either dead-on-nuts or a few counts off.

From what I can decipher the AD584 chip was made in 1997 and the Chinese allegedly measured the voltages in 2017 at 25°C with a 34401A.
I don't like that temperature, so I used 21°C. It is amazing that after nearly 25 years the measured values are within ±1 in the LSD compared to the reference.
I doubt it is just a coincidence. The Flukers sure knew what they were doing back then.

 

[jtr1962]

From what I can decipher the AD584 chip was made in 1997 and the Chinese allegedly measured the voltages in 2017 at 25°C with a 34401A.
I don't like that temperature, so I used 21°C. It is amazing that after nearly 25 years the measured values are within ±1 in the LSD compared to the reference.
I doubt it is just a coincidence. The Flukers sure knew what they were doing back then.

The AD584s used in those cheap eBay references are salvaged from equipment. That's actually a plus since this means they're nicely aged, and hence already did all their drifting (most of which occurs in the first few months of powered operation). The only open question is whether or not the calibration values the seller gives are fake or not. I trust the ones that are hand-written on the boards themselves. I used to do something similar back when I was making LED drivers. I wrote the measured drive current directly on the board. I don't trust anyone who gives a printed slip of paper. That's mostly the China sellers.

Anyway, if your Fluke checked out at ±1 in the least significant digit that's great news. It just supports what I said. If you buy decent DMMs, they won't drift enough to matter. The Harbor Freight stuff of course is different. I never use them for anything critical. Mostly I use them for stuff like checking batteries where it doesn't matter if the meter is a few tenths of a percent off.

 

[LunarMist]

I like a good dessert wine, after dinner or in lieu of a cocktail. A fine Sauternes can be treasured - yes, sweet, but with a fine balance of acidity.

A few months ago, I had a "Pedro Ximénez" sherry, and it was lovely, dark and deep.

Then a couple of weeks ago, at some friends' Italian restaurant, we tried a Malvasia, a red, "frizzante" (semi-sparkling), dessert wine, and that was a very nice introduction to the genre. It was a Malvasia Costelnuovo Don Bosco "Annalisa"; I tried to find a Malvasia at a local shop, and got some "Ronaldo ReDream", but it didn't have the same complexity as the Annalisa.

Years ago, we used to get Essensia from the Quady winery in Madera, California ("have some Madeira, my dear"). This fabulous wine is made from the Orange Muscat variety of grape, slightly fortified to about 15%, but it is hard to get on the east coast of the U.S. It has a distinct orange-ish flavor, rich and lush.

I don't like that sweet stuff. After dinner there should be coffee.

Lately there is a trend for the rosé wines, but I don't really like them. We have decent wines in CA, so I don't need to buy any of the foreign ones anymore.