Control signal unexpectedly drops from 5v to 1v

[jjzn]

I am debugging my PCB where a control signal unexpectedly drops from 5v to 1v, hence toggles the logic.

After power up, the signal is at 5.0v.

The interesting thing is when a tweezer or any metal probe touches that pin, the signal drops to 1.0v right away, not all the way to 0v though.
I wear gloves for insulation and hold the tweezer.

A more interesting one is when I put the tweezer aside, still wear gloves, power cycle the whole, and touch that pin with my finger, the signal stays at 5v and never changes.

I guess there might be short-circuit somewhere on the PCB, but two experiments confused me a lot.

The series resistance of the a tweezer, a glove, and my body are the sum of them, but the tweezer is nearly nothing.

Could EE veterans help me to brainstorm? Thank you

 

[dick_freebird]

May be rectifying ambient AC hum and causing some
internal misbehavior. Tweezer is a conductive link between
your body-as-antenna and the test article and reduces the
"source impedance" by the increased capacitance and
reduced resistance you are able to poke in, with (finger
won't be as good especially with dry skin).; both R and C
epend on the contact area from you to the work at any
"choke point" (here, skin to tool).

Large AC hum on +5V might be triggering chip ESD
protection central clamp, or something.

 

[D.A.(Tony)Stewart]

This is a classic positive feedback instability causing oscillation where the signal may be distorted and you are reading average voltage.

Since the feedback is null when DC, but exists while injecting sufficient hum with your hand on low ESR tweezer-hand interface.
The coupling pF and injected line E-field hum is much larger while holding the tweezers than the finger tip on tiny pins due to area ratio.


Your design has poor phase margin and needs analysis, and correction. (whatever it is)

 

[BradtheRad]

My reason may already have been implied in above replies. Your problem could be that high-freq oscillations spring up after any (conductive, capacitive, static charge?) touch.

Your circuit powers up okay and works okay, however once they begin such oscillations tend to continue. Volt readings might be disrupted. Average readings might be disrupted.

 

[KlausST]

Hi,

I´d say: without a schematic .. all we can do is guess.

It may be mains frequency introduced into a high ohmic node .... or it may be oscillation ... maybe something different.


Klaus

 

[D.A.(Tony)Stewart]

A good Engineer will share all details on schematic and layout to reduce your guessing.