1
\$\begingroup\$

Circuit 1:

enter image description here

Circuit 2:

enter image description here

The only difference between Circuit 1 and Circuit 2 is`that in Circuit 2 a parallel 2.2k resistor R8 is added as a load next to R6.

Why adding this R8 is halving the output voltage here?

I tested this in real with a scope, and the voltage remains at 12V pulse when I add R8. So unlike in simulation R8 didn't change the output voltage. In real implementation the grounds of the input side and output side are isolated.

edit:

enter image description here

\$\endgroup\$
2
  • \$\begingroup\$ @replete Can you explain your suggestion? I don't see how this would help. \$\endgroup\$ – Elliot Alderson Jun 27 '18 at 12:13
  • \$\begingroup\$ @ElliotAlderson Thanks for the catch, I've removed the suggestion. I misread what was happening and assumed the OP needed to isolate the two circuits while satisfying SPICE's ground requirement. The suggestion I made is useful in that situation. \$\endgroup\$ – user133493 Jun 27 '18 at 12:17
4
\$\begingroup\$

I tested this in real with a scope, and the voltage remains at 12V pulse when I add R8. So unlike in simulation R8 didn't change the output voltage.

The simulation model will likely have a typical current transfer ratio of based on the data sheet and your real test may be using an opto isolator with a very high CTR. With a very high CTR the output opto-transistor remains saturated and you largely get the same output peak-to-peak voltage as you got when you didn't load with the extra 2k2 resistor.

The minimum CTR for the 4N26 is 20% but typically this can be 50% and if you had a really good one it could be a lot higher such as over 100%. The model parameter can be changed and you should be able to resimulate and obtain a result that is closer to what you measured.

\$\endgroup\$
7
  • \$\begingroup\$ I use R6 as a pull down resistor. The load R8 is 2.2k as shown. Do you think increasing R6 to 10k or 100k can be a remedy in general? In simulation when R6 around 22k and above the output voltage approaches to 12V. On the other hand Im afraid if increasing R6 too much might have side effects. (?) \$\endgroup\$ – GNZ Jun 27 '18 at 12:34
  • \$\begingroup\$ Making R6 too big in value in conjunction with a much lighter load can increase rise / fall times considerably. If you are not switching at particularly high speeds then making R6 10 kohm (or greater) should not be a problem. Alternatively you can buffer the opto output thus you get the best of both worlds. \$\endgroup\$ – Andy aka Jun 27 '18 at 12:38
  • \$\begingroup\$ Do you mean something like that? i.stack.imgur.com/whMx7.png I can live with 11V as well, this works in simulation. \$\endgroup\$ – GNZ Jun 27 '18 at 13:15
  • \$\begingroup\$ This worked in sim but in scope I see like 4V offset :( \$\endgroup\$ – GNZ Jun 27 '18 at 13:37
  • \$\begingroup\$ oh sorry I swapped collector and emitter wrong on breadboard. \$\endgroup\$ – GNZ Jun 27 '18 at 13:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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