I have a LITE-ON LTV-816S optoisolator I want to use to isolate a USART communication channel.

The inversion of case 2 is not a problem as I can handle this in software.

Which case is better in terms of signal integrity, the pull-down or the pull-up?

What effects will the pull-up have compared to the pull-down in terms of the slew rates of the signal?


simulate this circuit – Schematic created using CircuitLab

  • 2
    \$\begingroup\$ I think the difference is elsewhere. Do you want the LED to be on or off when UART is idle, do you want to detect a BREAK condition if cable is disconnected? Also, the rise and fall times ahould be equal so there should be no difference, the rise and fall times are in the data sheet. \$\endgroup\$
    – Justme
    Feb 17 at 9:39

2 Answers 2


No difference to signal quality. The isolation capacitance is far smaller than the phototransistor output capacitance (or the internal B-E junction that actually receives photocurrent), and common-mode immunity is about the same either way.

330R is pretty low for a logic-level output; that's ~10mA, which will need ~20mA IF to ensure saturation.

If you need higher baudrate (above 10kbps or so), high bias doesn't help too much; you need a different type of isolator. There are a handful of tricks that can be employed to speed up a two-terminal phototransistor receiver, but you're always limited by the couple microseconds it takes for the internal VBE to charge up/down. (Typical strategy is to cascode it, keeping VCE steady and eliminating Miller effect.) (Three-terminal phototransistor types, like 4N35, can be improved to the low 100s kbps by loading the B-E junction with a resistor, or a control circuit to use the phototransistor as a transresistance amplifier.)

The next better type is photodiode into (non-photo) transistor, such as 6N136, SFH6345. That, and proper logic types (6N137, various HCPL-xxxx), are good for a few Mbps up to 10 or so.

Digital isolators are cheap and plentiful; these use monolithic capacitors or transformers to couple signals, and are available in quite high bandwidth (100s Mbps). Current consumption is likely better than opto types too, especially for lower bandwidths (single Mbps or below).


For signal integrity or slew rate, it makes no difference whatsoever.

So you can choose based on the polarity of your signals. The typical UART connection is case 1 because it saves power when idle, but you might want to choose case 2 to be able to quickly detect a broken connection.


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.