Cost of opening a calibration lab?

[opa627bm]

Hi all,
My boss is asking about the possibility of opening a internal calibration lab for mostly scopes (Tek 2GHZ DPO7000), Fluke 4.5 digi meters (handheld) TEK diff probes (TDP1000..).
I think it is going to be a lot of work besides equipment and personnel.
I think nowadays, each company have their own scrips to run the calibration procedure, hand tuning is no longer the case , am I correct?
Regards,
Li

[meeder]

The question is how much work do you have for the lab. How high are the current calibration costs?
It is not only the cost of getting all the equipment but also the calibration of your working standards and so on.

[opa627bm]

The question is how much work do you have for the lab. How high are the current calibration costs?
It is not only the cost of getting all the equipment but also the calibration of your working standards and so on.

Pretty much constantly, We have 100+ scopes an 200+ diff probes and 400+ meters,
The cost of contract calibration is sky rocketing ....

But buying the equipment an hire a Metrologist is very costly too..

[JS]

  Let's do some economics here...

  The actual cost of having a running cal lab is the same as using an external cal lab, excluding the shipping costs. This is assuming you are using it at ~100%, not much idle time, but if you have much idle time you can sell that time calibrating for a 3rd party. This is basic economics, if you have any doubts get back to the books.

  The decision then is about if the interest (plus reduced shipping costs) of having it in house are high enough to justify the investment. Plus maybe the extra benefits as fastest time response, custom jobs, etc. Without forgetting the headaches it would bring, as the need of special trained staff, higher standard calibrations, etc.

  Then, economically is all about the duty cycle of the fully working cal lab, the shipping costs and the interests of the initial investment. You are here asking, particularly, for that last point. Which does depend on the specific equipment. To give an example 4.5 digit DMM isn't specific enough, most 4.5 digit DMM could be calibrated with an 34401A or equivalent, but not all of them as some may have an specially low range or wired function that even a 6.5 digit DMM wouldn't be able to deal with. Of that will depend if you could get away with a reasonable prized 6.5 digit or you have to go even higher. With that the external supplies and transfer standards. I don't know particular cases of scopes but I guess similar things could happen. Other factor to consider is the amount of work, to know the level of automation needed, I guess that with the info you gave plus the timing of the cals, an experienced person in the field should give a first approach of what you need.

  You would need to do in one day, 2 DMM, 1 Diff probe and 1/2 scope to have the lab working 200 days for once a year cal for everybody. That leaves space to send out the gear to an external cal lab twice a year. As you would be using the lab a lot you can't work directly on your standards, but you need transfers to be calibrated in house maybe after 90 days if not shorter. Then you are probably good with one 6.5 digit DMM and one capable of calibrating it which would be the reference one. If it happens that there is a few 4.5 digit DMM as the one I told you do have the better one to deal with it. This is just a guess and only for the DMM with my limited understanding of the topic.

JS

[AlfBaz]

Most companies I know keep a master set of instruments which are kept calibrated to nationally traceable calibration standards and are used as transfer standards for the rest.
This still requires coming up with procedures and personnel to carry out the tests but I imagine it's much cheaper than what your boss is suggesting.

[e61_phil]

Even with a set of calibrated Scopes and DMMs you will need some very stable sources for URI (DC and AC).

[AlfBaz]

Even with a set of calibrated Scopes and DMMs you will need some very stable sources for URI (DC and AC).

Very stable?
Reference sources can be probed simultaneously by master and DUT. Comparing instruments readings side by side on some arbitrary source should reveal any major discrepancies don't you think?

[e61_phil]

Even with a set of calibrated Scopes and DMMs you will need some very stable sources for URI (DC and AC).

Very stable?
Reference sources can be probed simultaneously by master and DUT. Comparing instruments readings side by side on some arbitrary source should reveal any major discrepancies don't you think?

A normal bench power supply isn't stable enough to compare two 6,5 digit meters. You will need up to 1000V stable to at least 10ppm over a few minutes. Similar for AC. And you can't measure one resistor with two instruments at the same time.

[AlfBaz]

A normal bench power supply isn't stable enough to compare two 6,5 digit meters. You will need up to 1000V stable to at least 10ppm over a few minutes. Similar for AC.

Why does it have to be that stable, if at all? If there is difficulty in comparing 2 readouts at the same time and there is no coms access to the meters, would not a series of photographs be sufficient or does a fluctuating voltage cause some measuring artefact? P.S. the op mentioned 4.5 digit meters

And you can't measure one resistor with two instruments at the same time.

No but surely some nulled switching mechanism could be used to take several alternate samples

[manganin]

My boss is asking about the possibility of opening a internal calibration lab for mostly scopes (Tek 2GHZ DPO7000), Fluke 4.5 digi meters (handheld) TEK diff probes (TDP1000..). I think it is going to be a lot of work besides equipment and personnel.  I think nowadays, each company have their own scrips to run the calibration procedure, hand tuning is no longer the case , am I correct?

I happen to know a company which has quite similar situation: hundreds of 3.5/4.5 multimeters and oscilloscopes.

They had an in house cal lab for decades. Old calibration equipment and an old metrologist. When he retired the company changed to external calibration services. It didn't cut down the costs as expected.

Instead of the traditional "white coat" solution, the company bought a Fluke 5520A Multiproduct calibrator, which they calibrate once a year in an external lab.

They also hired a (very good!) consultant who wrote the instructions and prepared the software routines for the specific multimeters and oscilloscopes. Two company workers participated the training held by the consultant. Neither of them had any metrological background, only basic knowledge in electronics. The whole process took less than a week.

Both new "metrologists" have their day job in quality control, calibration being just an avocation. One day per week average. If there is need for manual adjustment or repair they simply buy a new meter. Or if the instrument is expensive send it to the external lab.

The new arrangement works surprisingly well and costs much less than the previous external services. (For a volt-nut the old lab was a nice place to visit, but nowadays there is nothing to see anymore...)

[ManateeMafia]

Do the calibrations have to be accredited or is the company subject to audits of calibrations for compliance purposes?

If so, then the cost of accreditation is another charge that the cal labs pass on to their customers. I also understand that Fluke charges a fortune for Met/Cal if you were looking to use an automatic procedure.