Slow Switching with Analog Switch

General Electronics
1

I’m working through Lab 12 of “Learning the Art of Electronics” and have encountered some odd behavior with the DG403 switch.

I got the first circuit (shown below and on page 490) working. This was my first time wiring a dual supply without using the AD3. I connected the positive terminal of one power supply to ground and the negative terminal of the other to ground, then connected each supply to the corresponding pins on the breadboard. While setting the voltage from the negative supply, it may have briefly spiked to around 30V due to the coarse adjustment, but this is within the maximum limits. I used the 5V output from the AD3 to apply 5V to the circuit.

first_switching_circuit
first_switching_circuit1920×2560 216 KB

With this setup, I measured the ‘ON’ resistance as the first task, but noticed strange behavior in the switching times. I decided to measure the turn-on and turn-off times. To turn off the switch, I connected pin 15 (the switch input) to ground. It took approximately 10 seconds to turn off and 7 seconds to turn on, though I just estimated these times.

Despite these odd readings, I moved on to the application circuits, starting with the sample-and-hold circuit on page 492 (screenshot below). The signal was generated by the AD3’s wave function generator.

Screenshot from 2024-12-30 15-32-17

In this circuit, the output slowly fell to the negative supply voltage, even when the switch was on! It would then randomly start ‘sampling’ (following the input waveform), but when I grounded the switch input, it wouldn’t ‘hold’ the voltage. Again, the switching time seemed very slow.

I read a bit about latch-up, which made me wonder if I had permanently damaged the switch. However, I don’t think so, as I’m fairly certain I didn’t exceed the maximum voltage ratings.

My plan is to take a break and then try rebuilding the circuit, in case I miswired something. Any help would be appreciated!

2

What was the maximum differential voltage that the switch saw at that time?
It can’t exceed 44V.

Also, if V- was at 30V relative to the GND pin, you’ve exceeded the absolute max:
image

1 Like
3

You’re right, thinking back, I might have exceeded the maximum ratings. I don’t recall the exact V+ setting, but it was likely around 0V. However, the voltage difference between GND and V- may have indeed gone beyond the absolute maximum. I focused on the V+ to V- differential, assuming that staying within that limit on the power supply’s display would be sufficient. I see now that I overlooked the absolute maximum rating for GND to V-.

I also checked the power supply’s range, and it’s possible it went higher than 30V, though not for an extended period. The switch did function afterward, but with the slow switching and other issues I described.

Is there a reliable way to test if it’s damaged? I was thinking of carefully wiring it up again in a known-good configuration and seeing if it behaves correctly.

Thanks for pointing this out! I appreciate the help.

4

Since the over-voltage was on the supply rails with low impedance connections, it’s almost a certainty it’s been damaged. The damage may be partial so that it might look like it’s sort of operating. I don’t have specific experience with EOS on this device, so I don’t have a specific answer.

I think you’re better off getting a replacement IC.

I have a “jar of sadness” where I toss failed devices and other detritus… While you are learning, that jar tends to fill up quickly…

1 Like