Rigol DM3068 Calibration

[slugrustle]

Has anyone here had their Rigol DM3068 bench DMM calibrated?  If so, did you send it back to Rigol or have it calibrated at a 3rd party lab?  How much did it cost?

I was surprised to find the following statements in the service guide:

To ensure accuracy, the meter should be completely recalibrated when the interval is reached whatever interval is chosen. Even though the meter has passed the performance test, the specifications (see “Specifications”) cannot be guaranteed unless recalibration is done.

As the DM3068 does not support user calibration, please send your meter back to us for professional calibration when necessary.

[Fungus]

I was surprised to find the following statements in the service guide:

To ensure accuracy, the meter should be completely recalibrated when the interval is reached whatever interval is chosen. Even though the meter has passed the performance test, the specifications (see “Specifications”) cannot be guaranteed unless recalibration is done.

That's easy: Choose a longer interval! 

Do you want (a) "calibration", ie. a report on it's accuracy at the moment, or (b) "adjustment", ie. tweak it as needed to improve accuracy?

[slugrustle]

The service guide also says

An interval higher than 1 year is not recommended under all circumstances.

Based on the earlier quoted statements, what the manual refers to as recalibration, you seem to be referring to as adjustment.

Edit: What I'm really after is other's experiences with having this unit calibrated, whether by sending it to Rigol or to a 3rd party lab.  For example, perhaps the 3rd party lab read the service manual and said, "we can't calibrate this device."

[Fungus]

The service guide also says

An interval higher than 1 year is not recommended under all circumstances.

Based on the earlier quoted statements, what the manual refers to as recalibration, you seem to be referring to as adjustment.

"Adjustment" is the correct term.

Pricing and who can do it will depend on whether you want calibration or adjustment.

Almost any lab can do calibration. Not all of them will be able to do adjustment of every single multimeter on the market.

If you want adjustment you might end up sending it to Rigol.

[tautech]

Please get the terminology correct.
Performance verification is the first part of calibration and if the instrument passes, that is it is found to be within manufacturers spec, it is then declared calibrated and documentation proving such is made that also lists the equipment used, their SN#'s and their calibration expiry dates.

OTHO if it fails, then adjustment may be required and much modern equipment requires a universal calibrator controlled by a PC script to automatically check accuracy and make the SW adjustments.
Some Cal labs have such equipment and the manufacturers calibration script however those that don't have access to scripts can write their own providing they have adequately trained and capable staff.

[AVGresponding]

If it were to fail to be within spec, I'd be quite concerned there was a fault somewhere. As has been pointed out in numerous iterations of this type of thread, adjusting out of spec readings on an instrument of this calibre might well mask an ongoing problem.

[2N3055]

If it were to fail to be within spec, I'd be quite concerned there was a fault somewhere. As has been pointed out in numerous iterations of this type of thread, adjusting out of spec readings on an instrument of this calibre might well mask an ongoing problem.

That is why many places don't even quote "adjustment" but "repair". They will open it, verify critical points and only then adjust calibration coefficients.

[slugrustle]

I'm asking this question because of the two statements below from the service guide that also appear in the original post:

To ensure accuracy, the meter should be completely recalibrated when the interval is reached whatever interval is chosen. Even though the meter has passed the performance test, the specifications (see “Specifications”) cannot be guaranteed unless recalibration is done.

As the DM3068 does not support user calibration, please send your meter back to us for professional calibration when necessary.

Even though the service guide does not use the correct terminology, it does seem to be saying (really, read both quotes) that passing the performance verification is insufficient to ensure the meter is performing to its specifications, which is kind of nuts. I have an email out to Rigol on the matter.

I did a better job searching the forum and found this post by smgvbest, which does answer my question. They emailed Rigol on this matter and received the following reply:

We do have a PV document that allows users to test and verify their DMMs which would be the typical first step that a cal lab would do. Any cal lab can do that.
If for some reason it were out of specification, then our current process would be to ship it to Beaverton, OR where we would investigate, adjust, or repair depending on the nature of the calibration results.
We are also working on ways to improve how 3rd party labs can work with these and other of our units.
When customers asks us for where to go for calibration we have several options currently available:
- Excalibur Engineering in Southern California
- Norway Labs in Beaverton, Oregon
- Broadview Instrumentation Services in Ohio
Each are familiar with our instruments and have done a number of calibrations for us. But any lab is able to do a verification on most any of our equipment.

So it looks like the service guide is just poorly worded.

If anyone does have an experience of sending out their DM3068 for calibration to Rigol or a 3rd party lab, I'd still be interested in the cost and process of that.

[alm]

OTHO if it fails, then adjustment may be required and much modern equipment requires a universal calibrator controlled by a PC script to automatically check accuracy and make the SW adjustments.
Some Cal labs have such equipment and the manufacturers calibration script however those that don't have access to scripts can write their own providing they have adequately trained and capable staff.

And assuming the manufacturer publishes documentation about the adjustment procedure. Rigol is the only major bench DMM manufacturer that doesn't publish this information. I find this ridiculous.

Adjustment are not necessarily only happening if the unit fails its performance verification. That's something the customer can specify. Some customers might require adjustment to as close as possible to nominal, or to within its 24 hour specs. Good luck sending it to Rigol every time

[tautech]

slugrustle
All equipment accuracy drifts with age hopefully slowing and stabilizing with age.
Its annual drift is important to quantify and document and successive PV's can build up a picture that should indicate a predictable drift.

It would be a reasonable position to NOT adjust the instrument for the first few years until accuracy drift slows if it remains in spec and then have it adjusted a little down the track if required.

Only you can make these decisions ultimately based on the recorded accuracy drift of successive PV's and the inherent accuracy you ultimately require.

[Kleinstein]

OTHO if it fails, then adjustment may be required and much modern equipment requires a universal calibrator controlled by a PC script to automatically check accuracy and make the SW adjustments.
Some Cal labs have such equipment and the manufacturers calibration script however those that don't have access to scripts can write their own providing they have adequately trained and capable staff.

And assuming the manufacturer publishes documentation about the adjustment procedure. Rigol is the only major bench DMM manufacturer that doesn't publish this information. I find this ridiculous.

Adjustment are not necessarily only happening if the unit fails its performance verification. That's something the customer can specify. Some customers might require adjustment to as close as possible to nominal, or to within its 24 hour specs. Good luck sending it to Rigol every time

If you need accuracy near the 24 hour or even 3 month specs, than the rigol meter is the wrong instrument. Getter a high grade meter and do calibration less frequent.
There could still be cases when a rather frequent calibration is needed. However this would not be because of high accuracy, but more because of high costs from failure. This may still be a reason for a better meter, or maybe run 2 meters in parallel.

There is no absolute need to have the adjustment published. As long as the certified cal labs get the procedures or scrips, all would be fine.
Frequent adjustment may make it hard to follow the drift / cal history.  It depends on the way the adjustment is done:
Ideally one would get the exact adjustments made. If done as a direct adjustment to measured values (external reference point), one may not get accurate factors to reconstruct a cal history without extra errors (e.g. noise of measurement and standad).
In many cases there is still the option to not adjust the meter, but apply correction factors to the date one the PC side after the measurement. This even works backwards for data just before a calibration.

[Kean]

I'm asking this question because of the two statements below from the service guide that also appear in the original post:

To ensure accuracy, the meter should be completely recalibrated when the interval is reached whatever interval is chosen. Even though the meter has passed the performance test, the specifications (see “Specifications”) cannot be guaranteed unless recalibration is done.

The mention of "performance test" in the above quote appears to be referring to the end-user checks listed in section 2 of the service guide (self-test, fast test, & routine test), and it is saying that even if those end-user tests pass you should have professional calibration (verification) done at regular intervals.  The mention of "recalibration" here can be misleading, and doesn't specifically mean adjustment (as others have indicated).

[slugrustle]

Here's what Rigol came back with, quite similar to what people in this thread have been saying, so thanks for that.

3rd-party calibration is possible and preferred - our recommended partner is Transcat. They would be able to provide the certificate of calibration that would be good for a year or more depending on the standard you are conforming to. They can provide you a calibration price if you request a quote.

The only time it would need to be sent to us directly is if it has failed calibration and requires repair/adjustment.

The manual also says, under the section "Routine Test"

The meter passed the routine test must be tested again when the test time interval is exceeded.

The meter failed to pass the routine test must be calibrated or repaired before it is put back into use.

This part matches very well with routine test = verification, what people in this thread are saying, and how Rigol replied via email.   The quote below, which really confused me, might simply be referring to the 24-hour and 90-day specifications based on what alm and Kleinstein said.  If that's the case, then it all makes sense.  I'd be happy if this thing performed half as well as the 1-year spec over the long term.

To ensure accuracy, the meter should be completely recalibrated when the interval is reached whatever interval is chosen. Even though the meter has passed the performance test, the specifications (see “Specifications”) cannot be guaranteed unless recalibration is done.

[slugrustle]

Unfortunately, I have a DM3068 purchased in July of last year that was just out of spec on a few ranges last time it was calibrated in December or January.

I contacted Rigol, and it took four emails to get them to help me out; the first three responses only said they don't do calibration, their outside partners do.  I said up front that the meter failed calibration and I wanted it adjusted back into specification.

The meter was under warranty, so they gave me an RMA and I shipped it in.  They replaced the mainboard.  I just received the unit back, and they did not include a calibration certificate of any kind.  I'm guessing they just replaced the mainboard and ran self test.  I'm following up with them on this because it's unacceptable.  It's like they don't even understand what these meters are for. The whole point in buying one is to get a device that measures within its specifications.  That was also the whole point of me sending it back.  Mindboggling.

As a Rigol owner, I highly recommend either Keysight or Tektronix for bench DMMs.

[J-R]

I'm curious, was the company doing the calibration not able to adjust it back into spec?  Or they could, but it was more money so you wanted Rigol to cover it since it was so new and under warranty?

[slugrustle]

Rigol seems to say in the service manual that outside parties cannot adjust the DM3068, and it must be sent back to Rigol for repair.

I say "seems to" because the language is a little unclear.  But Rigol has clarified this in email.  Look at reply #7 in this thread.

I'm still working with Rigol on this, but all they did was replace the main board.  It turns out that the main boards are calibrated at the factory in China, but calibration certificates are not sent along with the replacement main boards to the service center in the USA.  And the service center doesn't have a calibrator.

[J-R]

OK, still trying to figure this out, cal lab says which:
- "we can't adjust this, it's broken"
- "we can't adjust this because we don't know how and Rigol won't tell us"
- "we can adjust this but we need more money"
- "we didn't adjust anything because you didn't ask us to"

[bdunham7]

Rigol seems to say in the service manual that outside parties cannot adjust the DM3068, and it must be sent back to Rigol for repair.

That might be outdated information and they may be disorganized enough that you will not find anyone there who can actually answer that question.  People get a little hung up on metrological semantics over what is "calibration", but in my experience is that for an instrument like a modern 6.5-digit DMM a calibration service includes adjustment if needed (sometimes at extra cost)  and "calibration failure" means that the adjustment process did not complete properly.  Most reasonable calibration providers will not take in an instrument that they can't adjust without having a conversation about it first. 

I'd recommend you get a quote from Transcat and not bother with anyone else.  Transcat actually sells these units and I doubt they would do so unless they had the in-house capability of calibrating them for their customers.  They list an accredited calibration service for this instrument at $364, but that is for a certificate with a new meter.  You can get a quote and ask them specifically if they can do the adjustments on this form.  The thing is I think you really want the calibration data in cases where you suspect the meter is misbehaving.

https://www.transcat.com/quote-request

Obviously the lesson here is that if you need a calibrated meter--and why one would need an uncalibrated 6.5-digit DMM I don't know--it's cheaper to buy an A-brand meter that is stable for decades and just have your local yokel cal lab whack a no-data cal sticker on it once a year for $80 or whatever they get these days. 

It all seems a bit Kafkaesque.

[DaneLaw]

.. I just received the unit back, and they did not include a calibration certificate of any kind.  I'm guessing they just replaced the mainboard and ran self test.  I'm following up with them on this because it's unacceptable.  It's like they don't even understand what these meters are for. The whole point in buying one is to get a device that measures within its specifications.  That was also the whole point of me sending it back.  Mindboggling.

As a Rigol owner, I highly recommend either Keysight or Tektronix for bench DMMs.

Then you should purchase a new 6½ digit meter from either Keysight or Tektronix / Keithley..
Pursuing the lowest common nominator from CN, and being annoyed that you don't get lab-grade calibration certification..then don't pursue the lowest bang for the buck.
Aint a Rigol DM3068 like 500 to 600 bucks for 6½ digit DMM, if you pay 300% to 400% more, you will likely get a better service with segmented cal reports and a western service that actually gives a damn about their customers.

There is often a reason why they cost a fraction of the bigger brand names.
Same with Hanteks 5½ & 6½ DMMs.. they will ditch a few hurdles to obtain these impressive specs/tolerances, and some customers are okay with that if the price is 1/3 but it goes without saying the experience, software & build-quality, will be of a lower standard.
A bottom-of-the-barrel 6½ digit meter to 500 to 600 bucks, aint gonna be the same as a +1500 bench DMM.. and even in this higher price-ballpark with the likes of Keithley & Keysight, you can also get a lackluster experience.

[bdunham7]

Then you should purchase a new 6½ digit meter from either Keysight or Tektronix..
Pursuing the lowest common nominator from CN, and being annoyed that you don't get lab-grade calibration certification..then don't pursue the lowest bang for the buck.
Aint a Rigol DM3068 like 500 bucks for 6½ digit DMM, if you pay 300% to 500% more, you will likely get a better service.

I agree that he should purchase Keithley (now that Fluke has abandoned that particular market point for now) but Rigol is marketed as an up-and-coming competitor to A-brands, not low-bux junk on Aliexpress.  That meter costs $882 today.  The Keithley DMM6500, which has way more functionality, two-year calibration cycle, better specs, is $1670--less than 2X--and it used to be quite a bit cheaper when they had competition.  In addition, the Keithley 2100 is about $1300 and Keysight has two models similarly priced.  So it's not like the Rigol is dirt cheap.

Rigol wouldn't have to do much--just answer the phone, respond to reasonable questions and maybe make arrangements for routine calibration services which are a pretty standard thing for that product line.

[J-R]

I'm surprised this question hasn't come up before.  It looks like this DMM has been out for well over a decade: https://www.eevblog.com/forum/testgear/video-review-rigol-dm3068-6-5-digit-digital-multimeter/

[coromonadalix]

34401a 34410a 34411a  would have been better ....
Keithley 2xxx series ...

some Keysight to avoid because of the failing spear mcu

[slugrustle]

I'm in a bit of a unique situation in that I have a lab setup with my own equipment, but I use it for work as well as hobby stuff.  Work agreed to calibrate the critical pieces of gear that I use for testing.  So the failed cal report is just what came back this calibration cycle from my employer and whoever they used didn't say a thing about adjustment.

I naively assumed that Rigol, who makes 6.5 digit meters and told me they could repair the unit, would at least do a cal on their end and send a certificate equivalent to the one that originally came with the meter (no numbers, just a pass/fail result).

I have two other of these units that have always passed cal at work and that are older.  I got the first one when starting out because of the perceived value for money.  The next two were in order to keep the same SCPI command language across meters for consistency.  In hindsight, this was a mistake.  Live and learn.

[bdunham7]

I'm in a bit of a unique situation in that I have a lab setup with my own equipment, but I use it for work as well as hobby stuff.  Work agreed to calibrate the critical pieces of gear that I use for testing.  So the failed cal report is just what came back this calibration cycle from my employer and whoever they used didn't say a thing about adjustment.

I naively assumed that Rigol, who makes 6.5 digit meters and told me they could repair the unit, would at least do a cal on their end and send a certificate equivalent to the one that originally came with the meter (no numbers, just a pass/fail result).

OK, now I understand your predicament.  The first time I sent something to Fluke for repair (under the lifetime warranty) I was surprised that they offered not only a free calibration certificate, but your choice of standard, Z540 w/data or ISO17025--at no cost.  That seemed generous for a 10+ year old meter that didn't come with a cal certficate in the first place.  Then I realized--as you have here--that issuing whatever cal cert that the customer needs or requests is just a necessary part of the warranty service in order to avoid disrupting the customer's business.  They just assume that the majority of their customers have their instruments on a calibration program and any repair or failure interrupts that.  I'd have to guess that the majority of Rigol's customers are NOT on a calibration program. 

[slugrustle]

The story gets better.

The after-sale technical support engineer at Rigol is a really nice guy and helpful.  After I admittedly gave him the riot act about the lack of a calibration following a main board replacement on the meter, he went above and beyond and said he could get me a calibration certificate for the meter, but I had to wait because of the spring festival holiday going on in China at the time.

I just got the calibration certificate, and the PDF is not worth the paper that it isn't printed on.
1. It is dated 2024-01-29, one day before I was given notice that the repaired meter was being shipped back to me.  I highly doubt that Rigol overnight shipped, from China, a mainboard that had just been calibrated the day before.
2. The calibration certificate bears the serial number of the out-of-calibration meter that I sent in for repair and not the serial number that the meter now displays via its utility menu.

You can't make this stuff up (the story, I mean, not the certificate).  I'm honestly too amused at this point to be mad.

Rigol is not professional equipment, and I should have known better.

[bdunham7]

You can't make this stuff up (the story, I mean, not the certificate).  I'm honestly too amused at this point to be mad.

Shocking...but not surprising.   

After all, the meter is (probably) accurate within spec and you have your certificate in hand.  They're probably genuinely perplexed as to what the problem might be. 

Is the serial number physically on the back of the case as well as shown on the display?  And now they don't match?   

[slugrustle]

Shocking...but not surprising.   

I think I need a better dictionary

After all, the meter is (probably) accurate within spec and you have your certificate in hand.  They're probably genuinely perplexed as to what the problem might be. 

This attitude would be more understandable if it wasn't coming from a company that makes a 6.5 digit bench DMM.

Is the serial number physically on the back of the case as well as shown on the display?  And now they don't match?   

Correct, but it's a sticker on the back of the case, and I have a label maker...