DMM Product Cycle

[Rypht]

What seems to be the average length of time between new versions of DMMs in this industry?

I come from a computer hardware background that has distorted my perception of many different product cycles,  because of the rapid pace of computer advancements.  So when I'm going through some of Dave's back catalog,  I see him reviewing something six or seven years ago, and it sort of throws me to find it might still be that company's flagship DMM.

I guess I'm used to there always something just over the horizon to factor in when it comes to purchase decisions.  So how long after a DMM comes to market, should one start wondering when the next model will be showing up?

[EEVblog]

The big names like Fluke resist releasing upgrading meter designs as much as humanly possible, and hence have a very long time between models.
Let's take the Model 87 as an extreme example.
First released in 1989
Then the 87 Series III (there was no model II AFAIK) in 2001 (12 year period)
Then the 87 Series V in 2004 (rare short period of 3 years)

The 87-V is still current, so that's coming on 17 years now!

There are reasons for this.
Allowing the customer to build up "measurement confidence" in the product, and also supply confidence. This is a deal in defense where they will write procedures around a particular model meter and they still have to be able to buy that meter and use the procedures in 10-15 years time.

[Vtile]

Dave said it already.

It is safety driven mature equipment for specific task, like a hammer. Not much have happened for a regular hammer for a few hundrets of year now, some material changes and some mainly marketing driven cosmetics.

My multimeters from 70's early 80's are still as good if not even better (case sensitive) than new DMMs, the usage haven't changed too much. What are the latest big feature updates? What comes to my mind is mostly the the data logging capabilities (and trend displays) and some have wider bandwidth, but there were really high bandwidth models (multimeters) already many moons ago.
TrueRMS vs. RMS is the last really big update?

Personally I'm satisfied by the fact they stay pretty constant, nothing is worse than consumer style product cycle in industry settings.

PS. AVO8 were in production roughly 50 years and were the Triplett 630 still in production? But both are/were old needle movement analogs

[rstofer]

Some variant of the Simpson 260 VOM has been around since 1939.  There have been some changes over time but it wouldn't matter much which meter you picked up, they would all work the same.

Interesting tidbit re: Simpson:

In 1985 the company and plant was purchased by the Chippewa Indians of Wisconsin and remains in operation today.

http://www.electronicproducts.com/Test_and_Measurement/Benchtop_Rack_Mountable/What_s_It_Worth_Simpson_Electric_Company_Testers.aspx

The Triplett 310 VOM has been around since 1955 and the 630 was around before that.

[Rypht]

Hey, thanks for the replies!  You guys make some excellent points, especially in terms of long term contract support and safety.  That hammer analogy was perfect too. 

I suppose it's also largely down to material science too, improving the components in some way, and what those improvements do to the price of production and unit sale prices.

[TiN]

On benchtop/system applications long life usual too, as many of these meters are used for production lines QC/lab experiments and testing. And people who use them prefer not to change code/software written to older model without absolute need too, as their previous data may be incomparable with new code, etc.

Precision metrology-level DMMs, standards, calibrators often also have 30+ year design life, as their stability and issues are known and tested over very long periods with many different labs/people. Often calibration labs have magnitudes more money spent on calibration and testing their DMMs/gear than actual unit cost itself. That's exactly why meter like 3458A, designed in 1989 is still top notch model and available from Keysight even today with very minor design changes due to parts obsolescence.

Another reason, while electronics today is evolved a lot since 80's, most of that advancement happen on digital side. And DMM is instrument mostly depending on analog input front end quality. That part did not improved as much, not even close to justify need of scrapping old meters and getting new ones. Yea, modern DMMs may have fancy touchscreens, LAN/WiFi interfaces and LEDs everywhere, but actual analog performance and measurement accuracy of units designed back in 90's still often superior and PROVEN by decades. 

[james_s]

It's all about product maturity. At some point a piece of gear does everything you need it to do and does it well, innovation slows down and there is little reason to upgrade. We are finally seeing a similar trend in computers too, after years of hectic development and exponential gains in performance things are really slowing down. A 10 year old computer is still going to do pretty much everything the average person needs to do with it, that notion would have been absurd 20 years ago when even a 2 year old computer was often hopelessly obsolete.

People who buy expensive test equipment expect to use it for decades. I have no plans to replace my Fluke 87 unless it fails and I cannot think of any feature that would entice me to move up to a newer model.