Hi all,
I built a circuit to generate 2 manually adjustable voltages (~0.7V) and selectable by a 3.3V digital signal. I used 2N7000 MOSFET (with internal source-drain protection diode) to implement NOT gate and Switches. The circuit ran ok for some days. But today I found all MOSFET were damaged (may not at the same time). I wonder if they were destroyed by power switching pike or ESD. It is also said that MOSFET is likely be damaged in soldering. Could you give some suggestions to use them?
Thanks a lot!
I think Q1 and Q3 damaged because of over voltage on Gate-source pins.
you can use one zener diode , say 5.6V parallel with gate-source pins (input-gnd).
by the way,
you could use IC 4066 (analoge swith) insted of 3transistors.
How strong is your 3.3V drive (I mean how much current can it provide)?
It seems that you might need some resistance on the gates of Q1 and Q3 (between th e gates and the SEL node) to slow down the voltage from gate to source. Of course you should not slow it down too much so that it effects circuit operation. the gate resistors will make the gate voltage rise exponentially toward 3.3V -- instead of slamming them immediately to 3.3V. If you SEL voltage has a very fast rise time then you could see large current spikes for the gate current and you could be damaging the gate.
Does anybody agree (disagree) with this?
Davood -- since the drive is 3.3V do you think the 5.6V zener will protect the gate-source? I guess what I am thinking is that the problem might be the initial impulsive current rather than the voltage. I guess it would not hurt to put the zener there. I think this might be the classic chicken-egg problem where the chicken is the voltage and egg is the current (or the other way around).
OK -- that's good! As long as the resistance looking back at the SEL drive is high enough your current drawn from SEL will be limited. If you have a stronger drive it is not hard to damage the MOSFETs. I have seen this type of damage with power MOSFETs for example -- this is not what you have of course.
yeah, i think that is a esd problem.
you know, the actual line you used on your board has ESL, and if
your switch frequency is high and rising edge is sharp enough, your conponent
will suffer from a high votage.