EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Gyorgy Albert on June 01, 2023, 01:20:40 pm
-
Hello forum members,
I have two HP3120A waveform generators, and it came into my mind to calibrate them under computer control, over GPIB. I have the necessary instruments, and prior to enter into the calibration process, I proposed to save the actual calibration values. According to the manual, there are a total of 87 calibration setups (noted as SETUP 00 to SETUP86, some are alternatives). After the instrument is unsecured, the calibration setup must be set with the command "CAL:SET nn", where nn is in the range 00..86. At this stage, with the query "CAL:VAL?" the actual value stored in the instrument can be read out, without performing the new calibration. With the command "CAL:VAL vv", where vv is the new value, the instrument will enter the new calibration value (some conditions must be met to save the new value).
Since I want just to read out and save the actual values, I simply looped "CAL:SET nn" and "CAL:VAL?" for the range 00 - 86 and saved the values I got. For my great surprise, for some of the SETUPs the 33120A displayed error messages on the VFD and beep-ed (and in these steps the returned value was 0 in most of the cases). So I repeated the test, and this time I read out in each loop cycle the error with the "SYST:ERR?" query. These were the errors (listed after their error numbers) for certain SETUPs:
851 Negative offset gain cal required (CAL:SETup 50)
852 Flatness DAC gain cal required (CAL:SETup 64)
853 AM cal 1 required (CAL:SETup 30)
854 AM cal 2 required (CAL:SETup 31)
855 Cal load resistance not specified (CAL:SETup 33)
856 Square wave positive offset cal required (CAL:SETup 60)
857 Square wave 50% duty cycle cal required (CAL:SETup 62)
858 AM cal 3 required (CAL:SETup 32)
At this point I am a bit surprised. I am just reading out the values, why the instrument is returning an error code? And for reading out the actual values, why it is complaining that some other calibration steps are required (prior). I have not done any attempt to perform a calibration. I performed the tests on both instruments, the results were the same. It seems that I am doing something wrong, but I could not configure out what.
Following what I experienced till here I looked in the manual of the 33220, which is a "bigger" brother. The description of the procedures for calibration are similar, same syntax, just different number of SETUPs. So I did not get further. The reading out of the calibration coefficient seems to work in the end, just for certain values an error is reported. For SETUP numbers greater than the specified 86, the error is different : -222,"Data out of range". So obviously, these are not valid SETUPs (above 86).
The firmware of both instruments is rev 7, I know about one single firmware dump here :
https://www.eevblog.com/forum/repair/help-anyone-have-hp-33120a-function-gen-rom-firmware/msg1119713/#msg1119713 (https://www.eevblog.com/forum/repair/help-anyone-have-hp-33120a-function-gen-rom-firmware/msg1119713/#msg1119713)
which seems to be of fw revision 8. A newer one.
In the end, I read all the calibration coefficient from both of my instruments. And at this moment I noticed something strange. The serial numbers of my instruments and the corresponding calibration strings are the following:
Instrument 1: US36015082, " 6 NOV 1997 :CALIBRATED"
Instrument 2: US36005606, "30 JAN 1997 :CALIBRATED"
But the calibration coefficients are identical (up to bit level) for both instruments. For me, this seems incredible. The calibration coefficients are floating point numbers, so if anyone ever performed a calibration on two (let we say) identical instruments, he could not get identical coefficients. This is for sure.
What I can imagine, that these are the default factory values, the calibration strings denotes the date of manufacturing (this seems to match, one manufactured in January 1997, the other with a ~ 10000 higher serial number, manufactured in November 1997). And practically in the factory they loaded a default (identical) calibration coefficient table in every instrument. Probably they made in the factory the performance tests, and as long as the instrument satisfied the specification, they do not touch the default calibration coefficient table.
I can suppose, that while I am reading the calibration coefficients, the meaning of the error would be: no calibration table is defined, the existing one is the default. I am thinking to perform the calibration myself, as I planned initially, and I will read out after that the coefficients. If this time I will not get errors, my assumption might be correct.
For those interested, I attached the logs of the reading over GPIB of the coefficients. The ESR (result of the query *ESR? after a GPIB Command/Query) value of 8 means "Device Error" (but the format of the command is correct, and it is interpreted correctly), while the ESR value of 16 means "Execution Error" (for the setups above 86, which means that it cannot be executed, since it is invalid).
Is there anyone who tried the calibration over GPB and succeeded? Any help/comment/ideas are welcomed.
Thanks,
George/Gyorgy