I am working on a circuit reading the TCD1103 image sensor. In the datasheet, the shown 2 peripheral components are advised.

Datasheet connection reference

I wonder if the BJT connected to the OS signal would be neccessary, if an operational amplifier in voltage following mode was connected instead.

Since the output is an analogue voltage signal, I got the impression, that the BJT might have a role in linearizing the signal.

My (simplified) circuit would connect the TCD to an ADC as in 3. In the real circuit there would be additional analogue amplifier stages as per the ADC datasheet, but notably it is lacking the BJT.

enter image description here

Update 1

As suggested in the comments, I will keep the BJT. Since the datasheet does not specifiy any particular model, I'd like to extend the question to which BJT model should be used. My choice wold be 2N3906.

  • 1
    \$\begingroup\$ I'd suggest building it as instructed by the datasheet first. Simplification can then be tested against a correctly operating circuit. \$\endgroup\$ Jun 23, 2019 at 13:06
  • \$\begingroup\$ The BJT is functioning as an emitter-follower, which both introduces a voltage offset (which is probably desirable) and reduces the output impedance of the sensor. \$\endgroup\$
    – Dave Tweed
    Jun 23, 2019 at 13:45
  • \$\begingroup\$ I've updated the question; If you don't mind, formulate a short answer so I can accept it! \$\endgroup\$ Jun 23, 2019 at 15:12

1 Answer 1


Processing the TCD1103 output signal

You can certainly buffer the OS output signal with an opamp, but for the LMH6658 you can't power the opamp from the 3.3V supply used for the TCD1103.

The OS signal dark level is very close to 3.3V so you need to be able to handle this as the CM signal for the opamp.
The LMH6658 CM minimum is not characterized for a 3.3V supply but at a 5V supply it is limited to 2.8V worst case. The ouptut votlage swing is also a problem for this opamp, but the input restrictions dominate.

enter image description here

A better choice may be the OPA837/OPA2837 which has a CM minimum of 3.7V for a 5V supply.

enter image description here

For either opamp choice you would still need to use higher supply voltage to allow the OS signal to be correctly buffered.
For the LMH6658 I'd suggest a minimum 6V supply.
For the OPA837 you could use a 5V supply (which you may already have).

From your datasheet you can see the OS output signal is high (close to 3.3V) for Dark: enter image description here

SS = Ground. The Dark voltage level output is specified as an absolute value (but it's from an inverted perspective).

The datasheet also shows that the OS outpur is a negative going signal (much like you might expect from a video signal) from whatever the OS high level is.

enter image description here

  • \$\begingroup\$ Thank you for the advice! I am not sure how to understand the "OS signal dark level" being close to 3.3V - There is a table in the datasheet which gives the dark level voltage (VMDK) at 3 mV and a total DC output signal voltage (VOS) of 1-2 Volts. Otherwise the OPA837 seems very good advice. I've spent significant time browsing and comparing the available parts and the LMH6658 was a more or less arbitrary choice. \$\endgroup\$ Jun 23, 2019 at 18:22
  • \$\begingroup\$ @antipatternAdded to the answer ....you can clearly see the OS output signal is at it's highest level (closest to 3.3V) for Dark signals. \$\endgroup\$ Jun 23, 2019 at 19:09
  • \$\begingroup\$ Okay, I got the part about VMDK being inverted wrong. But isn't V_OS in the table the same as the OS value in the figure? Then I expected it to be SS+1.5V typical in the dark case? \$\endgroup\$ Jun 23, 2019 at 20:27
  • \$\begingroup\$ @antipattern Certainly not the way I read the datasheet. \$\endgroup\$ Jun 23, 2019 at 20:59
  • \$\begingroup\$ I am absolutely certain that you are right. Just trying to understand. How did you conclude that "OS output signal is high (close to 3.3V)" ? Because the figure does not give any values, and the numeric I can find is the one I cited, V_OS in the table ("Optical/Electrical Characteristics"). Obviously this has implications in the following stages, so I really would like to understand this comprehensivly. \$\endgroup\$ Jun 23, 2019 at 21:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.