3458A Cal Results

[Tony_G]

Thanks - The statement that:

Knowing what I know, I'd rather send my 3458A for Standards Lab cal on a 3 year cal interval, instead of a one year cycle for STE9000 cal.

makes sense - I might do that as saving the one year cost for 3 years is roughly the Standard Lab Cost - I wasn't planning on having the unit calibrated every year but if I need to have the unit calibrated for the 3 year warranty then it might make sense to just save up each year and do the Standards Lab calibration and warranty extension at the same time.

[MisterDiodes]

Quick price update for anyone interested on Keysight 3458a service pricing - I just signed the invoice:

The extended warranty is $192 per year, and if the unit already has a current cal sticker, that will still qualify to start the warranty. If the meter is not under a valid cal period it will have to get cal'd first.

For me in Pacific Northwest USA, extended warranty contract also includes 1 free trip per year via courier service from either Seattle or Portland metro area to Rocklin, CA or Loveland, CO facilities for calibration.  Cal cost runs around $668 for standard, $1625 for high accuracy @ Loveland.  Otherwise you have to ship or drive the unit in yourself.

The warranty coverage is a very good deal, and if your business depends on a 3458a, it is highly recommended.  It helps you avoid those $2800 surprise repair bills.

If you have 732a/b's, I'd skip the expensive cal on the 3458a and keep the 732's on their cal schedule since those will get you in the much better 2ppm uncertainty range anyway for DCV.  732's start ~$575 per cal and that's a very good deal also.  Once your freshly cal'd 732's come in then you can keep your 3458a's checked against those.

[Tony_G]

Thanks for the info - Really appreciate it.

For me in Pacific Northwest USA, extended warranty contract also includes 1 free trip per year via courier service from either Seattle or Portland metro area to Rocklin, CA or Loveland, CO facilities for calibration.  Cal cost runs around $668 for standard, $1625 for high accuracy @ Loveland.  Otherwise you have to ship or drive the unit in yourself.

Where did you find this information? I have my unit, coincidently also in the PNW, under the 3 year warranty but I missed the bit about free shipping apparently 

Also do you know if they'll extend the warranty while under warranty but without a subsequent calibration? I ask because they want the cal to confirm the unit works but if it is already under warranty then any failures would be repaired anyway so why not extend it.

TonyG

[orin]

Thanks for the info - Really appreciate it.

For me in Pacific Northwest USA, extended warranty contract also includes 1 free trip per year via courier service from either Seattle or Portland metro area to Rocklin, CA or Loveland, CO facilities for calibration.  Cal cost runs around $668 for standard, $1625 for high accuracy @ Loveland.  Otherwise you have to ship or drive the unit in yourself.

Where did you find this information? I have my unit, coincidently also in the PNW, under the 3 year warranty but I missed the bit about free shipping apparently 

Also do you know if they'll extend the warranty while under warranty but without a subsequent calibration? I ask because they want the cal to confirm the unit works but if it is already under warranty then any failures would be repaired anyway so why not extend it.

TonyG

I don't see why not.

If it's anything like the 34461A, when I extended its warranty, they asked how many years I wanted quoted - up to 5 years.  I chose 3 and they gave a tiny discount for the multi-year contract.  They mustn't break very often as it was $36 a year.

However, if they have to repair a 3458A, they would likely have to recalibrate it as part of the repair.  I wonder how they handle that if you send it in for repair out of cal.

[ManateeMafia]

Mine failed last year under warranty. I haven't opened it to confirm but they said they replaced the A1 board. The 5 year contract was bought and paid for at that point.

Of note, the instrument returned with no detailed repair information. It just came back with the usual cal cert.

I have always paid for shipping to Keysight but they paid for return shipping. The return delivery gets marked hold for pickup (at my request) so it doesn't bounce around in the truck all day in the Florida heat.
If you call before shipping, they may tell you to ship to Rocklin. When I ask about Loveland, they tell me to ship it there instead.

[MisterDiodes]

Tony:

RE: Shipping - that's how it works for us, ask your sales rep.  Sometimes you do get conflicting info depending on who you call, some of the reps know, some don't.   Maybe you have to be near a Silicon Forest tech area where they are running courier service anyway.  I'm near I-5 corridor so maybe that why it works out.  I don't think they are going to drive 400 miles out in the sticks for one meter, but I'm not sure.  I'll ask next time I have occasion to talk to them.  But that is something they definitely offer us for the $192 yearly charge - you call to setup a cal date, and then they tell you what is a nearby date a courier could be scheduled with other errands in your area, and you go for that.  It might not be on the exact date you wanted, maybe +- a week or two, something like that.

Warranty - Actually our rep told us as long as it's working, you could even send it to a partner lab for cal (but she was a bit confused).  We just use KS.  I don't -think- there is a requirement to send it to KS for yearly cal, as long as it -stays- in cal with at least one of their cal partners.  I don't know that for sure but one of our clients does that, somehow:  They have the ext-warranty with KS but Transcat(?) does the yearly cal (just a basic cal of course, not the fancy-pants Loveland cal) - just because they have a contract with a local TransCat for a bunch of equipment anyway.  Somehow that all works out for everybody.

[Tony_G]

Thanks - they actually didn't mention it needing to stay in cal at all to me. Just that it had to be in cal to start the warranty. I'll have to contact them again.

TonyG

[MisterDiodes]

That sounds like KS sales...every time I call I get some slightly different story.  I guess everyone has to "compare notes" to get the whole picture.

[ap]

With regard to the KS cal report attached above, it is a little surprising that the 5720 used is specified for calibration on a yearly basis. That means a pretty bad test uncertainty ratio. I think I saw a similar one with a KS asia report. Anyone has some further insights?

[orin]

With regard to the KS cal report attached above, it is a little surprising that the 5720 used is specified for calibration on a yearly basis. That means a pretty bad test uncertainty ratio. I think I saw a similar one with a KS asia report. Anyone has some further insights?

Given these notes on the cal report:

"The Input value under TEST CONDITIONS column is the nominal value applied.
The MINIMUM and MAXIMUM test limits are based on the actual value applied
that is expected to be the indicated (MEASURED) value.
This can be calculated from:
Expected Value=(Minimum+Maximum)/2"

I'd expect that the 3458A in the calibration standards used section is used to measure the actual 5720A output and generate the limits.  This 3458A would no doubt be very well characterized and on a very short cal cycle.

[ap]

Thats probably how they do it, so that the 5720 is used as a source only. I actually overlooked this. Still not that great in some cases, but anyway. Would also be in line with the standard cal procedure.

[alm]

If you read the link I posted earlier with information from the engineer who designed the procedure, that is exactly what they do. The 3458a is on a 90 day cycle and is well characterized. The calibrator acts as a stable source and only its 15 min stability counts towards the measurement uncertainty. But TUR is indeed borderline.

[MisterDiodes]

If your interested mainly in lowest uncertainty on DCV and resistance, we don't consider the Loveland cal service the best value.  Over long term, money is better spent keeping some cal'd 732's (and KVD's + 752) right next to the stack of 3458a's, and right next to 732's you keep an SR-1010 for resistance checks.  The SR-1010 is easily much more stable over time than any 3458a and can be counted on as an excellent Rref.

Now you can keep your 3458a fleet tuned up and locked on to ~2ppm uncertainty DCV and check them any time for drift - and just send them out for standard cal as required if you need ACV, amps etc. or if you're doing an ISO900x audit and need outside verification.  That's probably as good or better than Loveland will get you for DCV anyway.

If you count the risk of bouncing that meter along in a truck on the trip to and from Loveland, keeping the 732's cal'd generally will get you a better result for lower uncertainty at the same or less cost over time, and at much less risk.