New DMM for Chistmas / embedded development

[2N3055]

It is possible to not get what is 'in spec' from a not-broken meter.  Or, you might get an unexpected result that technically is within specs, once you read them carefully, but is not what you want. (This happened to me and although it can be dismissed as 'RTFM', I still think it is interesting).

It is entirely acceptable in some cases to characterize an instrument to use it in an application that requires a tighter spec than that instruments listed uncertainties.  And protocols are only necessary where you need paperwork at that stage of the process. 

I think the OP has made his decision, so when get time I'll start a new thread with my particular example.  Then you can decide whether it is worth discussing or should be dismissed as an inconsequential 'RTFM' case.

I'm not disputing what you wrote here.
But as you said, "if you want simple answer", which is the way people usually want it. Of course I want simple answer, I want to use my instruments, not playing with them for the sake of it, fun as it might be.

You cannot characterize instrument to some super tight specs without having advanced equipment and knowledge and time to do it. So I would need 5000-10000 € of equipment to be able to characterize 300 € meter to perform the role of the 500 € meter....    Of course you can characterize to tighter spec than original, but in practice it is cheaper, easier and faster to simply buy Keysight 34460 than to characterize  BM869S to 0.01% accuracy, although it usually keeps that without problem. 34460 will have that guaranteed by design. You don't have to do anything.

Most users will be in two categories.

One group will buy meter and hope it will be OK, maybe occasionally check it with DMMcheck or one of simple resistor / voltage boxes like on Joe has, as a functional verification...
Others may want to know exactly how accurate their instruments are, so they send them for calibration..
Or you're Voltnut, and pretty much playing with instruments is what you do. Which is fun, but most people are not interested in that.

Getting suboptimal results from user error and not reading manuals is always a possibility, and any time you encounter some interesting scenario that is not obvious, please share it with us..  Knowledge is always welcome.
We just shouldn't confuse metrology practice with instrument capabilities/maintenance/verification.

[bdunham7]

Others may want to know exactly how accurate their instruments are, so they send them for calibration..

And that may be a serious mistake in some instances. 

 

[2N3055]

Others may want to know exactly how accurate their instruments are, so they send them for calibration..

And that may be a serious mistake in some instances.

For a handheld?  Topic here is not Keysight 3458 and calibration history..

[AJ3G]

Calibration in my opinion is a bit of a scam. There are reputable CAL houses out there, but most of the ones I have seen turn the instrument on, do nothing with it, apply calibration sticker and then charge you an arm and a leg. Seen this with my own two eyes, so I know it happens. For most people, having another known good device to see if you are close should be sufficient. We do have contracts at our company where all the instruments used must be NIST traceable. We ultimately charge the customer for this.

Rich 

[tautech]

Again ignorance of what calibration actually encompasses.
2 parts:
Accuracy verification, which in most cases is all that is required.
Adjustment to meet manufacturers spec. Rarer and more costly.

[AJ3G]

Maybe so, but when a known failed instrument comes back with a Calibration Sticker, not sure who is ignorant in that case.

[2N3055]

Maybe so, but when a known failed instrument comes back with a Calibration Sticker, not sure who is ignorant in that case.

You. I would ask money back and take my business elsewhere..

[AJ3G]

Precisely what has happened, and at more than one Calibration Facility in the local area. Where I live, its become quite a racket. Funny, NIST is just down the road 45 Minutes! 

[bdunham7]

For a handheld?  Topic here is not Keysight 3458 and calibration history..

The mistake I was thinking of was sending to the wrong calibration lab.  But you're right, I'm not just referring to handheld issues which is why it is time for a new thread...

But, as a final thought on handhelds, I don't like the fact that Brymen has not made the calibration manual available.  See my new thread for thoughts on that.

[bdunham7]

Precisely what has happened, and at more than one Calibration Facility in the local area. Where I live, its become quite a racket. Funny, NIST is just down the road 45 Minutes! 

Not naming names, but I've seen some stuff.....

Part of the problem is the practice of charging a lot more money for as-found and as-left data.  Most of the lower end obligatory (they have to have the paper) customers want the cheapest option, so they offer them the no-data or else the OOT data (just the ones that were out of spec).  And try asking the sales rep a technical question at some of those outfits.  So why charge so much more for data that should be generated automatically with MET/CAL or whatever they are using? 

I feel bad for any reputable, conscientious non-OEM cal lab.  Their 'competitors' are dragging them all down.  I would likely only use an OEM lab at this point.

[bdunham7]

Again ignorance of what calibration actually encompasses.
2 parts:
Accuracy verification, which in most cases is all that is required.
Adjustment to meet manufacturers spec. Rarer and more costly.

Be careful not to generalize.  Many times the price is the same, and some OEMs actually always adjust as a default.  And the basic service at many cal labs is you get it back with stickers on every external screw and fitting along with a paper that says "It's OK, trust us".

[AJ3G]

Yes, we use a company we trust now. But it's been more than once I have had something like I describe happen, and from three different calibration labs. I too will not name names, but I observed another Calibration Lab in the Hampton Roads area, in person, and I can tell you what they were doing had nothing to do with Calibration. It had to do with selling stickers, and paperwork. It's kind of like home contractors, if you find a good one, pay them what they are worth, because the cut rate outfits here are horrible.

[bdunham7]

Yes, we use a company we trust now. But it's been more than once I have had something like I describe happen, and from three different calibration labs. I too will not name names, but I observed another Calibration Lab in the Hampton Roads area, in person, and I can tell you what they were doing had nothing to do with Calibration. It had to do with selling stickers, and paperwork. It's kind of like home contractors, if you find a good one, pay them what they are worth, because the cut rate outfits here are horrible.

Are they making bank screwing the government?  I'm in SoCal, so a different set of labs, but I'm wondering if it might be the same large company that rhymes with Kimco.

[joeqsmith]

Again ignorance of what calibration actually encompasses.
2 parts:
Accuracy verification, which in most cases is all that is required.
Adjustment to meet manufacturers spec. Rarer and more costly.

https://www.transcat.com/calibration-services/what-is-calibration

A calibration is not an adjustment, correction, or repair though. Those are separate actions that could be taken as a result of the calibration action, which is simply to report how much error is in the UUT. If the error exceeds a pre-determined limit (usually the Original Equipment Manufacturer’s warranted performance specifications) then an adjustment or repair may be needed in order to minimize the UUT error.

https://fieldcal.com/calibration/

FCI does not charge for adjusting your equipment. That’s correct, no extra fee to adjust your equipment, it is part of our instrument calibration services.

Make sure you understand what's being offered beforehand. 

[AJ3G]

I will not even get that close to disclosing names. They know it's tax payer money.

Yes, we use a company we trust now. But it's been more than once I have had something like I describe happen, and from three different calibration labs. I too will not name names, but I observed another Calibration Lab in the Hampton Roads area, in person, and I can tell you what they were doing had nothing to do with Calibration. It had to do with selling stickers, and paperwork. It's kind of like home contractors, if you find a good one, pay them what they are worth, because the cut rate outfits here are horrible.

Are they making bank screwing the government?  I'm in SoCal, so a different set of labs, but I'm wondering if it might be the same large company that rhymes with Kimco.

[tautech]

Again ignorance of what calibration actually encompasses.
2 parts:
Accuracy verification, which in most cases is all that is required.
Adjustment to meet manufacturers spec. Rarer and more costly.

https://www.transcat.com/calibration-services/what-is-calibration

A calibration is not an adjustment, correction, or repair though. Those are separate actions that could be taken as a result of the calibration action, which is simply to report how much error is in the UUT. If the error exceeds a pre-determined limit (usually the Original Equipment Manufacturer’s warranted performance specifications) then an adjustment or repair may be needed in order to minimize the UUT error.

https://fieldcal.com/calibration/

FCI does not charge for adjusting your equipment. That’s correct, no extra fee to adjust your equipment, it is part of our instrument calibration services.

Make sure you understand what's being offered beforehand.

Quite.

Yet with the instruments I deal with, their service manuals have a large chapter on performance verification and as most modern equipment is closed case calibration accuracy verification is the first step in the calibration process.
Service manuals and datasheets are free to anyone and also list specialist equipment requirements for adjustments which in this day and age is a PC with specialist SW, a automated calibration script and a universal calibrator.
Therefore if a Cal lab hasn't got a universal calibrator and your instrument is out way of whack then run a mile ! 

Hopefully they will have sufficient other traceable equipment to at least do performance verification and if so can provide a legit Cal cert if your instrument still meets OEM spec.

Luckily in 8 years in this game not one customer AFAIK has required the Cal services of the 2 NZ companies I would recommend.

[joeqsmith]

...

Most users will be in two categories.

One group will buy meter and hope it will be OK, maybe occasionally check it with DMMcheck or one of simple resistor / voltage boxes like on Joe has, as a functional verification...
Others may want to know exactly how accurate their instruments are, so they send them for calibration..
Or you're Voltnut, and pretty much playing with instruments is what you do. Which is fun, but most people are not interested in that.

...

I certainly do fall into the first category for the most part.  Testing the meters, I really only want to know if I have damaged them.  The box makes for a quick sanity check.  Some of the parts in that box are fairly decent but it's not for calibration.    I have aligned a few handheld meters myself after repairs.  I have a box of resistors and capacitors that I will use to get me in the ballpark.  Most resistors are 10 ppm and lower.   The box has an Apex VRE305AS 5V reference but to set the meters, I use my old HP bench meter.   I attempted to build a source meter that I use to measure the battery life of the meters and for other tests.  I used it recently to align a meter.    The source isn't very good.  Well, the readout isn't all that good.  Two plots were taken at the same time, one from the source meter, the other from my hold HP.   Still for basic alignment of these handheld meters, it's good enough for me.   

The bench meters and one of my scopes have baby papers but nothing current.  The endless story of a hobbyist on a budget.     

[wizard69]

...
You aren't going to see a lot of 'innovation' in handheld general-purpose DMMs because they are a mature product.   People will hang flashlights on them and put in bluetooth comms or whatever else the fad of the day is, but I don't see the general functions changing or improving in any significant way.  Some might even say they've gone too far already.

In many ways I do believe that we are seeing a bunch of useless features on mutimeters, at least in the form currently being delivered.   One good example is the capacitance function on many multimeters that for the most part is not of much use.   This especially considering that very good handheld LCR meters exist that do a much better job.

As for room for innovation I do believe there is plenty of possibilities available.   The trick would be t make new features worthwhile and not close to useless lick many of the capacitance functions out there.   Bluetooth is a good example of an innovation that would be very useful if manufactures would learn to leverage it.   An app that runs on a Bluetooth enable watch should exist for any meter claiming to offer Bluetooth support.   Along with that open source software should exist to make it easy to use those Bluetooth capable meters on the various PC operating systems.   I never understood the need manufactures have to make using their hardware difficult in conjunction with general purpose computing devices.   It should be easy for me to write an app that connects to a Bluetooth capable meter.

Other things like large displays that deliver multiple parameters to the user should be high on any meter manufactures list.   One usage here is ACV measurements that also display frequency.   Back in the day I hardly ever saw a VFD, these days they are as common as grass.  Now the real trick is getting such a simultaneous display to work well.   Oh and my pet peeve, give me a battery system that doesn't LEAK.

In general I think you are right, the hand held meter these days pretty much is well defined.   There is room for innovation but the basic functionality has been zeroed in on long ago.   

[2N3055]

In many ways I do believe that we are seeing a bunch of useless features on mutimeters, at least in the form currently being delivered.   One good example is the capacitance function on many multimeters that for the most part is not of much use.   This especially considering that very good handheld LCR meters exist that do a much better job.

For one, that measurement was, i believe, driven mostly by people repairing household appliances with single phase induction motors (that was huge market), and many single phase motors in industry (pumps and such). Built in capacitance function in meters was more than adequate to check various blocking and starting capacitors, and meters are also safe to use on charged ones... 
LCR meters are notoriously underprotected for that job.
Multimeters are not only for electronics at the desk in controlled environment.Most of them, are, in fact not used for fine electronics at desk. Other markets are much larger..

Many years ago (more than 20..) there was already Wavetek DM27XT that had full LCR and frequency measurement.... Not a bad instrument either.. But nobody followed the trend, it seems market didn't think it was necessary...

I agree about battery system. That definitely can be made more robust.

One thing manufacturers could do is separate AC and DC measurement path with separate ranging (so DC is in 600 V, and AC is in 400mV), that would yield much faster AC+DC measurements too.
One other thing they could do is to add additional current measurement path, isolated one, to make real VA measurements.. There are benchtop meter with that..
But those would add lots of electronics, and would be battery guzzlers..  And also not everybody would use that, meaning limited market..
They could add math or this or that, and there would always be some people who would find it useful and many that wouldn't.

I personally would like standardized PC/phone connectivity (both cable and bluetooth) and protocol, and few advanced log/analysis softwares.

[joeqsmith]

I would like to see an event counter.  Say the voltage or resistance changes by some threshold.   Even if it could only work to 1Hz and required 500ms, it would be helpful in many cases.   

A sealed battery pack would be nice although, I have not had a leakage problem after changing to rechargeable batteries.   

[Wytnucls]

My old Gossen 26S has an event counter for voltage. 0.5Hz max. Set low and high thresholds first.

[joeqsmith]

I wonder why it didn't catch on. 

Dave made a video some time ago where he was trying to count laps of a toy train going around a track.  It seemed he was struggling to get it to work. With so many features in the 121GW, surprised it wasn't included.

[rsjsouza]

My Keysights 1273A and U1282A count events just fine via a voltage threshold - one of the trigger options of the datalogging feature.

[joeqsmith]

Looks like it's limited to 10,000 counts.  Or does it continue to count after the memory is full?

[rsjsouza]

Looks like it's limited to 10,000 counts.  Or does it continue to count after the memory is full?

I don't remember if it wraps around or is limited - after all, it is a data logging feature and therefore it uses its internal memory to store the voltage as well. 

However, with the proper USB cable and software I vaguely recall it can count indefinitely.