In February, I sent my HP 3457A in for cal. In late March, I sent my EDC VS330 DC Voltage Reference in for cal. Today, I just got the VS330 back and thought I'd compare the two instruments. Both instruments have been on for a couple of hours. It is about 25 C in the room. Setting 10.00000 on the VS330, I measure 9.99891 VDC on the 3457A. According to the manual, on the 30 VDC range, the 90 day spec is 0.0035% of reading, plus 19 counts. The VS330 lists the "Limit of error accuracy" as 0.003% of setting plus 0.001% of range (30 V range). The "Calibration accuracy" is 0.002% of setting plus 0.0005% of range. Now, if we take the worst case for both instruments, what I am reading is marginally inside the total error for both instruments. However, considering that both instruments were recently calibrated and adjusted, I would expect much better. The report on the 3457A, dated 16 Feb says the 30 VDC range read 29.9991 VDC. The report on the VS330, dated 31 Mar says 10 VDC read 10.0001 VDC. Both instruments required adjustment. These numbers would lead me to believe that I should be getting a lot closer agreement to 10 VDC for both of these instruments. Am I expecting too much? Joe Gray W5JG

You really won't know what is going on until you do the next calibration cycle and examine the as received data for both units. Or find someone with a better meter or reference and compare them. Who did the calibration? You might leave both units powered up for a few days/weeks and see how they behave together. Also, you are two degrees C above you lab calibration temperature. That could add to the uncertainty. ----- Original Message ----- From: "Joseph Gray" <jgray@zianet.com> To: "Discussion of precise voltage measurement" <volt-nuts@febo.com> Sent: Thursday, April 07, 2016 11:20 PM Subject: [volt-nuts] Just back from cal > In February, I sent my HP 3457A in for cal. In late March, I sent my > EDC VS330 DC Voltage Reference in for cal. Today, I just got the VS330 > back and thought I'd compare the two instruments. > > Both instruments have been on for a couple of hours. It is about 25 C > in the room. > > Setting 10.00000 on the VS330, I measure 9.99891 VDC on the 3457A. > > According to the manual, on the 30 VDC range, the 90 day spec is > 0.0035% of reading, plus 19 counts. > > The VS330 lists the "Limit of error accuracy" as 0.003% of setting > plus 0.001% of range (30 V range). The "Calibration accuracy" is > 0.002% of setting plus 0.0005% of range. > > Now, if we take the worst case for both instruments, what I am reading > is marginally inside the total error for both instruments. However, > considering that both instruments were recently calibrated and > adjusted, I would expect much better. > > The report on the 3457A, dated 16 Feb says the 30 VDC range read > 29.9991 VDC. The report on the VS330, dated 31 Mar says 10 VDC read > 10.0001 VDC. Both instruments required adjustment. > > These numbers would lead me to believe that I should be getting a lot > closer agreement to 10 VDC for both of these instruments. Am I > expecting too much? > > Joe Gray > W5JG > _______________________________________________ > volt-nuts mailing list -- volt-nuts@febo.com > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts > and follow the instructions there.

On 8 Apr 2016 04:21, "Joseph Gray" <jgray@zianet.com> wrote: > > In February, I sent my HP 3457A in for cal. In late March, I sent my > EDC VS330 DC Voltage Reference in for cal. Today, I just got the VS330 > back and thought I'd compare the two instruments. Did you send them to the manufacturers or a third party lab? I am personally VERY skeptical of the competence of a number of so-called calibration labs. A few months back there wss a 40 GHz vector network analyser on eBay for auction which included a cal certificate. The certificate listed the items used for the calibration. The calibration kit used was a 6 GHz economy type N kit made by HP. The N connector will not work properly at 40 GHz, but even if it did, the use of a calibration kit is insufficient to calibrate a VNA. I have seen countless examples of this sort of thing. I would not personally consider getting any of my HP kit calibrated by anyone other than Keysight. > Both instruments have been on for a couple of hours. It is about 25 C > in the room. That is quite warm. Are both specified at that temperature? > Setting 10.00000 on the VS330, I measure 9.99891 VDC on the 3457A. I would try to find a local volt-nut with a 3458A. I would also keep a log of the differences between your voltage standard and multimeter. If the difference is changing significantly over time, it would tend to indicate one or both are unstable. If however the difference remains very similar then I would be more suspicious that one is incorrectly calibrated. > The report on the 3457A, dated 16 Feb says the 30 VDC range read > 29.9991 VDC. The report on the VS330, dated 31 Mar says 10 VDC read > 10.0001 VDC. Both instruments required adjustment. I have heard several reports of 3457As being in spec many years after they were last calibrated. The fact your unit required adjustment seems a little unusual. If the VS330 was out by 1 on the last digit, that suggests to me it didn't require adjustment. > These numbers would lead me to believe that I should be getting a lot > closer agreement to 10 VDC for both of these instruments. Am I > expecting too much? I have not checked your numbers, but would agree if both units have been recently calibrated, the agreement between the two should not be just inside the limits if you take the worst case of each. I would expect much closer agreement. If the two appear to be drifting with respect to each other the the purchase of another 3457A, even if uncalibrated, would probably allow you to determine what one of your > Joe Gray > W5JG Dave, G8WRB.