Glossary entry

Japanese term or phrase:

クラスタ毎に1[LSB],最大28段で28[LSB]ずれて

English translation:

time difference of 1 LSB per cluster, maximum time difference of 28 LSB at 28 stops

Added to glossary by Lekhika
Jul 30, 2017 07:04
6 yrs ago
Japanese term

クラスタ毎に1[LSB],最大28段で28[LSB]ずれて

Japanese to English Tech/Engineering Electronics / Elect Eng
The following paragraph is a part of a paper on ADC for Backside Illuminated Global Shutter CMOS Image Sensors.

What does the phrase "クラスタ毎に1[LSB],最大28段で28[LSB]ずれて" in the last line mean?

リピータは,ADCの書き込みと読み出し動作を時分割駆動(Time interleave)させる128画素をひと単位としてクラスタが構成され,縦方向に28クラスタが縦続接続されている.
ADCへのコードは,CK入力とは反対側で生成したグレイコードを,フリップフロップを介してデジタルバケツリレー式に順次転送することで供給する.
グレイコードは,1CKで遷移する信号が1bitのためタイミングエラーに強く,動作時の電源変動を一定にする効果を持つ.
本構成ではGCから近い側から出力までクラスタ毎に1[LSB],最大28段で28[LSB]ずれてFixed Pattern Noiseとなるが,CDSにてリセットノイズともにキャンセルされる.

Discussion

Port City Aug 3, 2017:
段 in photography refers to something like a unit of exposure. So, I thought ずれ was about time. But I'm totally unfamiliar with what is written in the passage. I'm not quite sure if 段 was used in the same sense as "stops" in photography. But 時分割駆動, whatever it means, seems to indicate that the passage has something to do with "time".
Marc Brunet Aug 3, 2017:
Alert to Asker... Lekhika, see Wikipedia re: Gray Codes and how they increment compared with vs natural binary series. This forces us to reconsider whether this (ずれ) of your passage refers to a 'Time-shift', as we thought, or simply to a 'binary value shift' affecting only 1 bit -- the Least Significant one of the signal's binary string. Gray's safer single bit value change per transition is not aiming for time saving nor time lag offset/correction, but for reliability in signal change execution even if it takes longer to jump from one state to one more distant in the serial set. So my post does not take this LSB element into account and should be ignored as off the track. We can exchange thoughts on new findings but I am starting on this one from scratch again. Sorry for everything. Cheers,
Marc Brunet Aug 3, 2017:
time shift cancellation story (#2: prerequisites) GCから近い側から出力までoffers more than 1 parsing solution.
Settling for this one, and as a paraphrase, rather than a tight translation:

"By the time outputs from both ends (close end and GC end) are transmitted, a 1 [LSB] time shift per cluster and, for a stack of 28 clusters, a maximum time shift of 28 [LSB] will have occurred, ..."

(Makes sense in that no time-shift offset can occur without the issue (in matching numbers) of both N CK codes that cause such time shifts, and N Grey Codes to cancel those.)

For your comments...
Marc Brunet Aug 3, 2017:
this time shift cancellation story .... By the time an output from the end close to the GC reaches destination, a 1 [LSB] time shift per cluster and, for a stack of 28 clusters, a maximum time shift of 28 [LSB] will have occurred, generating a Fixed Pattern Noise; but with this CDS arrangement, both Fixed Pattern Noise and Reset Noise are cancelled out.

What is GC? = Group Control? No, of course not, = Gray Code! so that calls for some changes.... still working on it, then...

Proposed translations

1 hr
Selected

time difference of 1 LSB per cluster, maximum time difference of 28 LSB at 28 stops

I assume there are maximum 28 clusters... I also assume from the context that ずれて refers to time difference concerning transfer of signals.
Note from asker:
Thank you. I had not interpreted ずれて as time difference.
Something went wrong...
2 KudoZ points awarded for this answer. Comment: "Thank you very much"
+2
4 days

....a time shift of 1 [LSB] per cluster and, for a stack of 28 clusters, a max.time shift of 28 will

....a time shift of 1 [LSB] per cluster and, for a stack of 28 clusters, a maximum time shift of 28 [LSB] will occur, ...

Assuming LSB = least significant bit; CDS= coding sequence
(This passage intrigued me, and I itched to nail it for complete sense.
Did I managed to? You be judge of that.. (see discussion.) )


--------------------------------------------------
Note added at 4 days (2017-08-03 07:12:20 GMT)
--------------------------------------------------

Oh, No! Apologies for "managed
Note from asker:
Your discussion entry really helped understanding. Thank you very much.
Peer comment(s):

agree Chrisso (X)
15 hrs
agree DPurohit (X)
3 days 5 hrs
Something went wrong...
Term search
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search