I am trying to make an analog isolator of my own. The isolator needs to isolate analog signals (0-10 V) and should have relatively high bandwidth (up to 1 MHz) with unity gain. I am a beginner as far as electronics are concerned.

I found a design on the internet from Stanford research systems. The design looks great (input signal -10 to 10 V and up to 1 MHz bandwidth with unity to x100 gain) but I am puzzled by some of the components:

The schematics I found (https://xdevs.com/doc/Stanford_Research_Systems/SIM900/sim984_sch.pdf) contain three pages, the first page is the interfacing electronics, the second is the actual isolation electronics, the third is the power supply electronics. I am mainly interested in the second page below:

enter image description here

Again you can find the complete schematics here: https://xdevs.com/doc/Stanford_Research_Systems/SIM900/sim984_sch.pdf

First, the basics of the design (from the device's manual (https://www.thinksrs.com/downloads/pdfs/manuals/SIM984m.pdf)):
*The upper portion of Page 2 shows the floating input amplifier. Gain is controlled through latching relays U214 and U215, the control coils of which are earth-referenced.
*The (amplified) signal is optically coupled through U205, U206, U207, and U208.
*The (earth-referenced) output circuitry includes the output bandwidth control. This circuitry is on the lower portion of Page 2 of the schematics. The overall AC gain is trimmed using VR202 at the factory, and should not require user adjustment.
*DC offset can be adjusted with VR203, accessed from the front panel of the instrument. Note that the offset trim is referenced to the output, and comes after the gain is applied

My question:
What are the components in the purple frames, and in the bottom left light green frame doing ? These ICs don't even seem to be connected to any part of the circuit except the ground and the power supplies.

  • \$\begingroup\$ I suggest you instead use the application circuits here: docs.broadcom.com/doc/…. And breadboard the thing to see if you can't get anything remotely resembling coherent function. In the past I tried simulating these high linearity optocouplers but couldn't get it to work either. But my conclusion was that did not know enough about the model and so could not determine whether I could trust it one way or the other to know whether the problem was with me or the model. That's why I suggest breadboarding it. \$\endgroup\$
    – DKNguyen
    Jul 30 at 18:10
  • \$\begingroup\$ I will try breadboarding at some point, thanks for your comment. \$\endgroup\$ Jul 30 at 18:17
  • \$\begingroup\$ (As the scope of the question has been significantly reduced in response to it being closed as "needs focus", I'll reopen this to give the community a chance to respond to the improved version.) \$\endgroup\$
    – SamGibson
    Jul 30 at 22:17
  • \$\begingroup\$ This is a very complex circuit for a beginner. As DKNguyen said, you might have better luck looking at the example circuits in the HCNR200 datasheet. But unless you're more familiar with op amp circuits than you let on, it would be best to find someone who's familiar with analog electronics to help you with this. \$\endgroup\$
    – Adam Haun
    Jul 30 at 23:26
  • \$\begingroup\$ I think the components in the purple frames are unused sections of dual or quad op amps. But I would not leave the inputs of those unused portions floating. The inputs of unused amplifiers should be tied high & low to as to reduce power consumption and ensure that the outputs are in a known state and not oscillating. Those inputs may be tied that way, but I can't tel from the poor resolution of the schematic you provided. \$\endgroup\$
    – SteveSh
    Jul 31 at 0:20

1 Answer 1


The extra parts in purple appear to be supply connections. Often multi-part components (like dual/quad op-amps) are split up with a N+1th element representing the supply pins.

I can't possibly read the pin numbers on this very low resolution screenshot, but I assume you have the source document, and you can look up the part datasheets and verify the pinning is correct.

If a supply part were not used: consider how cluttered the schematic would be if the bypass capacitors and supply nets were mixed in with the signal flow. Schematics are a document that is to be read, as anything else is; and as the author of that document, consider your target audience. It's not just the EDA tool (a netlist and footprints to feed to the PCB side, and generate BOM data from). It's a human-readable format which should suggest to the reader where signal flow (generally left to right) and power flow (generally top to bottom) through the circuit, and how various sub-parts of the circuit interconnect.

For a simple circuit like this, connections should generally be wired; but it seems the whole system would not fit on a single sheet, so they opted to use off-sheet connectors or ports or net labels to make connections to the other sheets. Such connections can also be made within a sheet, which is annoying at times (you're stuck with a visual search that your eyeballs will inevitably skim over, or at best, a CTRL+F on a PDF if it contains searchable text at all). It is a compromise between reduced visual clutter (tangles of wires are impossible to trace), and reduced understanding (tangles of net names are impossible to follow). Which is to say, both extremes reduce readability, and somewhere in the middle lies best practices.

For further reading, search on electrical schematic best-practices, drawing tips, component symbols and what their elements mean, and just generally look at schematics and gain an appreciation for what looks good and what doesn't. (Lord knows, there's no shortage of newbie schematics on Stack..!)

The "what's this" appears to be manufacturing related elements, perhaps fiducial marks, mounting holes, logos, mechanical parts (added to the assembly BOM (bill of materials)), etc.

There is no bottom-left "green" section, but if you mean the khaki color, that looks like unused switches, and either a power converter, or maybe another supply block. Again, I can't tell at the low resolution. Cross-reference component designators and look up part datasheets and pinouts to confirm.


Your Answer

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

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