Hi,
I am using PTV5.1B-M3/84A for over voltage protection. Here is the circuit,
At, Vadc1 and 2, the voltage divider will generate 3.125-5VDC, depending on VIN1 and VIN2. But when D4 and D5 are populated, I see 3.05-3.995V. When D4 and D5 are removed, Vadc1 and 2 are correct. Vadc1 and 2 connections are correct, They are not shorted to anything else. Component installed is PTV5.1B-M3/84A (marked VE). Could anyone tell me what’s wrong?
Thanks,
Michael
Greetings,
The leakage current spec for the part is quoted as 20uA max at 1V reverse potential and 25°C temperature. That figure will increase with temperature and reverse voltage.
20uA across the 10K top resistance in your divider would make for a 0.2V drop, so the results you’re seeing would seem very reasonable at a glance.
It’s unclear what sort of over-voltages you’re trying to protect against, but the divider itself will go a long way toward limiting any offenses to a point where the '3217’s internal protections are adequate.
A person ought to consider whether those resistors themselves are up to the task however; the tiny SMD parts we use today are not nearly so ESD tolerant as the carbon comp behemoths our grandparents used.
Also, while the 10-bit ADC on that '3217 isn’t the most precise unit out there, it’s still going to draw some current during sampling that’s going to affect the result to a degree. It may or may not matter for your purposes, but it’s a thing to be aware of.
Thanks Rick. I am trying to protect against 10V accidentally applied to PB4 and 5. The spec says maximum voltage only up to 5.5V. Will a bi-direction diode help?
Michael
I am thinking about a TVS.
How about the – ESD145B1W01005E6327XTSA1CT-ND – TVS?
ESD145B1W01005E6327XTSA1 Infineon Technologies | Circuit Protection | DigiKey
A TVS might ad some insurance, but I would place it right at Vin1 and Vin2 rather than inside the ADC circuit to avoid negative effects on your voltage divider and resulting measurements.
Something like the SMAJ12A might be a starting point. It will stand off 12V and go into full breakdown by 14.7V. It’s maximum reverse leakage current would be no more than 1uA.
Do you mean via the test points?
In that case, a great deal depends on the source impedance associated with that 10V. If one’s talking about (for example) a careless technician accidentally connecting a 10V bench supply capable of sourcing substantial current, then there are few if any TVS diodes, zeners, etc that are going to be up to the task. These are all transient suppression devices designed to protect against short-term events, whereas the careless-tech scenario is for practical purposes a DC fault, which will turn TVS diodes and the like crispy and burny within the time frame required to recognize the error.
A crowbar-style protection device like the one Nathan mentions might potentially be better (note that that part would be more properly filed with the TVS thyristors) but still not really appropriate since the power dissipation issue remains, and at least in the case of that device it’ll do nothing until the fault voltage reaches at least 18.5v. So for that case of a 10V bench supply fault, it’d provide no protection whatsoever.
Thank you all for the feedback. A lot of good suggestions.
Michael
If you are planning on keeping it in the current position directly next to the analog input pins, then that seems to be a pretty good choice. The 10nA max leakage would cause a 0.1mV drop across the 10k resistor, so not too significant.
TSD05CDYFR works out beautifully. The drop is barely noticeable now. Thanks to all your help!
Michael
