[This topic continued from
/g/TekScopes/message/158778 ]
I know the discussion was about the later CSA803(A/B/C) and 11801(B/C). I have an 803 and an 803A. The U500/U511 batteries in my 803 were dead, so factory calibration constants would have been lost and I could freely experiment with the CCALCORRECTION command. To my surprise this command was unknown to the 803. This is also confirmed by the commands lists in U810/U910 EPROM images. The CSA803 Service Reference is now available at Tekwiki (thanks H?kan!). There is no reference at all in that manual to factory set calibration constants. The corresponding texts in the C version was obviously added later on.
So I am convinced that in the CSA803 nothing is lost when an A5 time base board Dallas chip dies.
[[BTW in the pdf manuals provided by Tek you won't find words in the added text with an ordinary search. Likely Tek used a custom character coding in the added text. If you copy-paste for instance T1331 from the added text into the search field then you will find the other instances of T1331 in the added text.]
Albert
On Thu, Aug 15, 2019 at 04:21 PM, freemail8877 wrote:
Hi, first post to this forum. I have very much enjoyed browsing, and thought I
could contribute with this information. Thanks guys for keeping this forum so
interesting. The T1331 (loss of battery backed up information in the Time
Base/Controller board NVRAMs) seems to be an error to avoid, by backing
up/replacing the NVRAMs on that board. See e.g. the topic "Saving A5 Timing
Board NVRAM calibration constants CSA803 (A/C) and 11801(B/C)" for how to. To
answer some question raised there, here is some information from the Tektronix
documentation.
In the 11801C User Manual (070-9971-01), on p. 3-112 (p.262/440 of the pdf),
I found this:
"Initializing and Erasing Nonvolatile RAM:
...
The following information is not lost:
...
- Factory calibration constants, which are established at the factory and
cannot be changed by the user or by the 11801C
- Enhanced Accuracy calibration constants, which are periodically calculated
and updated
..."
And in the 11801C Service Manual (070-9972-02) on p. 6-108 (p.244/286) it is
stated:
"Time Base Calibration Errors:
...
Calibration data to enhance the time interpolator linearity is stored in NVRAM
on the Time Base/Controller board. If this NVRAM fails, a T1331 error occurs
during power-up diagnostics. The values are written at the factory with the
GPIB command CALCORRECTION.
CALCORRECTION<ui>:N
where <ui> = 0..32 and N = -128..+127
CAUTION: The calibration enhancement values are set at time of manufacture.
There is no query form of this command. Do NOT reset these values. If a T1331
failure occurs during extended power-up diagnostics or you suspect a time
interpolator error, contact Tekronix factory service.
A T1331 failure causes all N values in the 33 ui locations to be reset to
zero. The oscilloscope will still run, but with decreased accuracy in the time
interpolator linearity. Resetting NVRAM with the Teksecure feature does not
affect the Time Base/Controller board NVRAM."
So now the big question seems to be:
Does anybody in the group know, or know who might know, how the factory
calibration procedure is, how the 33 N-values are determined?
Does anybody know how much the accuracy is affected?
Monica A.
PS:
I have for many years been a Tektronix scopes addict, It started with the
fascination in late 60's and early 70's, when I saw and used a 535 with Type
CA plugin at my father's workplace. Being interested in electronics, this was
heaven. I now have a 7904a with sampling and non-sampling plugins, a 7704a
which rudely enough let the smoke out of the power supply, a newly acquired
7603, and a 468. The design and layout of the fronts are just so great. An
11801C is on the way in, so I have used search a lot lately. It is said to
pass POST, and a picture shows the graticule on the screen, so hopefully it's
OK.
