EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: smoothVTer on November 14, 2017, 04:18:31 pm
-
I need to use relatively high-value capacitors to sustain the input voltage from a very weak battery in a low-duty cycle, pulsed power application. After running the numbers, I am using 2x 1206 size 47uF 16V tantalum capacitors for this buffering. There is zero room for any additional capacitors on the entire PCB. There is an input protection PFET, which acts as the diode in this case, and otherwise the 2 tantalum capacitors are across the battery contacts.
(https://photos.smugmug.com/photos/i-94Hx24R/0/737dd7e1/M/i-94Hx24R-M.png)
If the user inserts the battery in the wrong polarity, will tantalums be damaged despite the reverse-current blocking PFET?
Is the nA-level leakage from a reversed battery in this circuit enough to fry these tantalum caps?
Is the damage to tantalums specifically from reverse current, or is a reverse voltage enough to sustain damage?
-
You never ever connect tantalums directly across raw battery voltage.
They need current limiting. Typically used after a current-limited regulator.
If Q2 was replaced with a back-to-back configuration, it could be used to provide gentle precharge.
But just use standard aluminium electrolytics (and normal small-value ceramic bypassing where necessary).
-
Also respect the voltage derating, typically 50% below 70'C and more above. That means VBATT+ should be below 8 V with 16 V rated tantalum caps.
-
A much higher level of reverse current would be necessary to damage the tantalum capacitors.
Voltage derating the tantalum capacitors increases their surge current capability significantly.
-
You never ever connect tantalums directly across raw battery voltage.
They need current limiting. Typically used after a current-limited regulator.
You mean all those manufacturers who've been doing just that for the past forty years got it all wrong?
-
You never ever connect tantalums directly across raw battery voltage.
They need current limiting. Typically used after a current-limited regulator.
You mean all those manufacturers who've been doing just that for the past forty years got it all wrong?
Never seen that, personally; most of the time, the tants seem to be after a linear reg (which has current limiting). For a good reason, really, there's a lot of literature and appnotes available if you want to dive in.
AFAIK tantalum caps are the only type where even a small number of events of exceeding surge current ratings, even by modest factors, can be catastrophic. (There's no free lunch, however - using a large MLCC in the same place can be as dangerous, for different reasons though. It can also produce unwanted effects due to too low ESR.)
But of course it can be used directly in the battery line, if the designer ensures it's safe to do so. But an engineer capable of doing that doesn't necessarily come to forums to ask for a "rule of thumb", without giving even a part number :). I admit, giving rules of thumb is kind of lazy, and I apologize for it. But what else do we have to analyze right here?
For example, if the battery is installed only once, the caps can be precharged as a step in the assembly process (or, they just count on the fact that it's quite improbable that one surge event is disastrous, even though that does happen with tantalums).
Or, the tantalums are so heavily derated (in both voltage and surge current rating) that it's ok to do so.
Or maybe the battery you think about is an AA alkaline cell with ohms of DC resistance, while I'm thinking about a big lithium ion pack with 1mOhm of DCR. We need data!
Also, not all tantalums are the same! But when asked about a generic "tantalum", we have to assume worst case.
The difference is, some of us are engineers who can analyze these things, while some of us can only resort to blanket statements and meaningless appeal to authority. I guess all those "manufacturers" have either done the analysis, or did it by luck. Some products do end up in flames, so without looking at the failure rates of such products, appeal to the industry as a whole is not a good argument.
-
You never ever connect tantalums directly across raw battery voltage.
They need current limiting. Typically used after a current-limited regulator.
You mean all those manufacturers who've been doing just that for the past forty years got it all wrong?
Never seen that, personally; most of the time, the tants seem to be after a linear reg (which has current limiting). For a good reason, really, there's a lot of literature and appnotes available if you want to dive in.
Battery operated equipment does not always use linear regulators, as their efficiency is generally poor, & the designers are trying to get as long a life as possible from their batteries
But of course it can be used directly in the battery line, if the designer ensures it's safe to do so. But an engineer capable of doing that doesn't necessarily come to forums to ask for a "rule of thumb", without giving even a part number :). I admit, giving rules of thumb is kind of lazy, and I apologize for it. But what else do we have to analyze right
Or, the tantalums are so heavily derated (in both voltage and surge current rating) that it's ok to do so.
That is almost certainly what was done in the applications I saw.
Or maybe the battery you think about is an AA alkaline cell with ohms of DC resistance, while I'm thinking about a big lithium ion pack with 1mOhm of DCR. We need data!
But you jumped straight in -----my comment about 40 years might indicate that most such stuff I saw wouldn't have used lithium ion batteries.
Also, not all tantalums are the same! But when asked about a generic "tantalum", we have to assume worst case.
The difference is, some of us are engineers who can analyze these things, while some of us can only resort to blanket statements and meaningless appeal to authority. I guess all those "manufacturers" have either done the analysis, or did it by luck. Some products do end up in flames, so without looking at the failure rates of such products, appeal to the industry as a whole is not a good argument.
The only person to make a "blanket statement" was your goodself.
"Appeal to authority"?------I just pointed out what I had seen in the field over many years.
OK, you got your "Engineer feathers" all ruffled, but it doesn't take an EE degree to read a spec sheet.
That said, I'm not a great fan of tantalums, & have seen quite a few blown up ones, but none of these happened the first time the equipment was turned on, neither were those particular ones directly across batteries.