1
\$\begingroup\$

I have a programming setup where I plug in a GoldX USB cable plugged into my programming laptop (in this case a MacBook Pro). Cable picture: GoldX Cable

This is plugged into an eBay sourced RS232 TTL converter: RS232 TTL Converter

Which I plug into my testing board. Everything works fine, most of the time... Except I've noticed that occasionally, when I plug in just the cable to the converter board, without connecting to my circuitry, the IC is powered. The LED turns on, and the VCC pin reads ~ 2.1 V. Note, the leads on the chip pins are connected straight to the meter; that isn't clear from the photo.

Test Setup

But, this doesn't happen every time I plug in the cable. Occasionally it works as expected, where the LED is not powered until I connect VCC and GND to the testing board.

I tried with a couple different programming computers, and the problem seems to be worse when powering from my MacBook; about 50% of the time when I plug in. I tried 10x on another laptop, and an ATX desktop and didn't see the issue with 10 attempts on each. Not a ton of data, but relevant.

My gut instinct is buying the RS232 to TTL converter from eBay left me with a knockoff MAX3232 chip that is misbehaving and leaking one of the data lines back through VCC. I'd be fine leaving it there, but then why would the behavior be so different when connecting the GOLDX cable from my macbook versus a different programming computer? Or, maybe thats just the issue -- the behavior will be completely inconsistent, and I shouldn't try to find an expected output from a given input.

TLDR: Should I just assume the MAX3232 is counterfeit and move on?

\$\endgroup\$
  • 1
    \$\begingroup\$ Usually these things are taking the power from RTS/DTR lines. You might try to put some serial extension cable between the usb adapter and the board without these lines. \$\endgroup\$ – Eugene Sh. Feb 13 '18 at 16:36
  • \$\begingroup\$ Is there a problem here you want to solve, or just a curiosity you want resolved? I agree that you have the data lines leaking to VCC. The Max3232 says "When MAX3232 is unpowered, it can be safely connected to an active remote RS232 device." But im not sure if "Safely" guarantees no body diode leakage, or just that it wont break. \$\endgroup\$ – Selvek Feb 13 '18 at 16:46
  • \$\begingroup\$ @Selvek good point. Added the TLDR; should I assume the MAX3232 is counterrfeit and move on? \$\endgroup\$ – Jim Feb 13 '18 at 16:52
2
\$\begingroup\$

I'm not surprised, inputs usually have some form of protection diode up to the rail which, when the power is off, will conduct. This will be from the Rx and DTR lines and will depend on the state (voltage) of those lines and how much current whatever it is attached to can source.

If it bothers you, then adding some series resistance to the input lines will reduce this effect.

\$\endgroup\$
  • \$\begingroup\$ Thanks Trevor. Going to do some testing using this info as a starting point. \$\endgroup\$ – Jim Feb 13 '18 at 17:33

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.