CYIH1SM1000AA-HHCS Cypress Semiconductor Corp, CYIH1SM1000AA-HHCS Datasheet - Page 66

• Manufacturer Part Number: CYIH1SM1000AA-HHCS
• Description: IC SPACE IMAGE SENSOR 84-JLCC
• Manufacturer: Cypress Semiconductor Corp

Specifications of CYIH1SM1000AA-HHCS

• Lead Free Status / RoHS Status: Lead free / RoHS Compliant
• Lead Free Status / RoHS Status: Lead free / RoHS Compliant, Lead free / RoHS Compliant

Question:
In your datasheet, ADC High/Low bias voltages are recommended to be set with a resistive divider. But the datasheet doesn't mention
anything about temperature stability. For the STAR-1000, there was an internal resistor between ADC_HIGH and ADC_LOW that
had temperature dependence. Because of this, for STAR-1000 designs, I used to set my ADC bias voltages with buffers that would
keep the bias levels constant over temperature. Do I need to repeat the same principle for the HAS2? Or does the HAS2 remove
any temperature dependence for the ADC bias voltages?
Answer:
For good temperature stability, it is better the same principle as the STAR-1000. So use external buffers to keep ADC_HIGH and
ADC_LOW to a fixed voltage level
Question:
In your datasheet, in the
a troublesome output delay variability of 20 - 60 ns, some parts had even 70 ns! Have the digital output drivers been significantly
improved for the HAS2 ADC? What are typical rise/fall times for the outputs?
Answer:
The output delay and stability has been improved compared to STAR-1000
Question:
Could you please discuss the differences between BLANK, CAL, and PRECHARGE? The STAR-1000 only had a CAL signal.
Answer:
The extra BLANK signal is used to reset the internal CLKX divider. PRECHARGE is used to pre-charge the column lines and column
caps to ground
Question:
I liked the flexibility of the STAR-1000. The HAS2 seems more restrictive. For example, your application note says, "…repeated use
of pixel re-addressing (register X1) potentially injects offset-noise into any windows that overlap in Y-coordinates." If I understand
correctly, this means I cannot address each pixel along a line individually? I cannot readout every other pixel, or every 2nd, or 5th,
or 10th? I have to readout all the pixels in a line? Can you think of any options?
Answer:
You still can start reading at any X or Y position. You have only keep in mind that there is an analog pipeline on the pixel data. So if
you individual read 2 pixels of the same line closer together then the analog pipe, the second pixel will be addressed when you are
only interested in the first pixel. So when you want to read that second pixel by a new SyncX, it will be the second time you address it.
As a result, there is a risk of a deviated value. Probably some deviated offset on the pixel value. You have probably the same problem
with STAR-1000 but maybe the analog pipe is there smaller.
Question:
For NDR/CDS mode, there is parasitic exposure given your suggested algorithm. Can I do this algorithm instead?
Answer:
I don't see a problem with your algorithm
Document Number: 001-54123 Rev. *A
a. Reset Row X
b. Readout Row X
c. Reset Row X+1
d. Readout Row X+1
e. Reset Row X+2
f. Readout Row X+2
g. (repeat to region of interest)
h. (wait for integration timer completion)
i. Readout Row X
j. (wait for time to reset a row)
k. Readout Row X+1
l. (wait for time to reset a row)
m.Readout Row X+2
n. (wait for time to reset a row)
o. (repeat to region of interest)
i. Start integration timer
“ADC Timing Diagram”
on page 55, the table lists t5, output delay, as typically 10 ns. The STAR-1000 had
CYIH1SM1000AA-HHCS
Page 66 of 71
[+] Feedback