"Absolute Maximum Ratings" vs "DC Characteristics" in a datasheet

homosapien

Specifically pertaining to the SX micro. I have some older chips that I want to use up on a battery powered application, was hoping to use 4 AA cells for power, producing ~6 volts. The data sheet for the SX states:



17.0 ELECTRICAL CHARACTERISTICS:

17.1 Absolute Maximum Ratings - Voltage on Vdd with respect to Vss - 0v to +7.0v

17.2 DC Characteristics - Supply Voltage (Fosc 32 MHz) - min 2.7V, max 5.5V




So I'm wondering if the 6V as a supply is acceptable. Does anyone know specifically what is the difference between "Absolute Maximum Ratings" and the "maximum" in the DC Characteristics?

I could use 3 cells for 4.5V, but part of the circuit is a RF transmitter that works considerably better at the higher 6V. Since this is a battery powered application, I would like to avoid using a regulator and just run everything off the 6V.


Thanks,
Nate

Phil Pilgrim (PhiPi)

homosapien wrote:

So I'm wondering if the 6V as a supply is acceptable.

No, it is not. The operating voltage range extends only to 5.5V. The Absolute Maximum Ratings tell you only what can be tolerated outside of an operating mode without physical damage.

-Phil

homosapien

Hi Phil, thanks for the reply.

So operating mode would be a transistors changing states, while the "outside of operation mode" would be Vdd applied but no transistors changing states? Trying to imagine where the electrical breakdown would be and why one mode (operating/non-operating) would be more sensitive (to, I presume, voltage/current jumping some dielectric) than the other.


Nate

Phil Pilgrim (PhiPi)

Outside of the operating voltage range, but inside of the absolute maximum ratings, the device might not suffer physical damage, but it is not guaranteed to work correctly, and its operating lifetime could be shortened.

-Phil

Chris Savage

Depending on what operating voltage you run the SX at you could use a small LDO regulator to power your project from the batteries. This would work especially well if you're running the SX at 3.3V.

LoopyByteloose

the 5.5v limit is for reliable operation. From 5.5v to 7.0v, the chip is getting hotter and may be unreliable.. but recover from the overvoltage. This leeway is mostly for transient conditions. Just consider it in terms of heat buildup. After all, when things get too hot, the semiconductors change composition and teenie-tiny wiring begins to fuse.

One chip might do fine with 6.5 volts and the next one do terrible. There is a certain amount of variation in all production. The maker provides the 5.5v as a good operating design. And the 7.0 volt figure so you know when to throw out the garbage. Between is the leeway in testing, repair, and construction.

homosapien

Thanks for the replies guys.

I am using the SX 'SLEEP' function to save battery power and 'wake up' about every 10 minutes to check and transmit a parameter, so I don't think there is a way to use a LDO without it running all the time (even when the SX is in the 'SLEEP' mode) and using precious power. This is a device that hopefully will only need a battery change once a year or so (ie batteries will self-discharge faster than the device actually uses them up).

Might rig something up so the SX uses 4.5V, and when needed a transistor switches the 6V to the RF module...


Nate

piguy101

You might be able to rig a few diodes or schottkys in series to crudely reduce the voltage.

Peter Jakacki

homosapien wrote: »

Thanks for the replies guys.

I am using the SX 'SLEEP' function to save battery power and 'wake up' about every 10 minutes to check and transmit a parameter, so I don't think there is a way to use a LDO without it running all the time (even when the SX is in the 'SLEEP' mode) and using precious power. This is a device that hopefully will only need a battery change once a year or so (ie batteries will self-discharge faster than the device actually uses them up).

Might rig something up so the SX uses 4.5V, and when needed a transistor switches the 6V to the RF module...


Nate

I would favor the LDO approach as there are many modern LDOs that have very low quiescent current. The other approach is to run the SX from 4.5V or even 3V if it works and use a boost regulator to give your transmitter as much voltage as it might need. Since the boost regulator would only be turned on as required you have the best of both worlds.

BTW, the diode dropping approach is very wasteful.
@Loopy: In short, it might get hotter but if you stopped the clock it wouldn't, but it would still stress the chip. So it's not about getting hotter.

LoopyByteloose

@Peter J
Are you saying that heat is not the destructive force in a marginal over-voltage setting. From what I can understand, VxI= Watts. And the watts are either passed down the line in useful work, or kept within the chip to cause heat.

In the Absolute Maximum Ratings is usually a 'watt' limit for the chip. The smaller SMDs seem to have lower limits which implies they can't dump heat as easily due to size.

Peter Jakacki

Loopy Byteloose wrote: »

@Peter J
Are you saying that heat is not the destructive force in a marginal over-voltage setting. From what I can understand, VxI= Watts. And the watts are either passed down the line in useful work, or kept within the chip to cause heat.

In the Absolute Maximum Ratings is usually a 'watt' limit for the chip. The smaller SMDs seem to have lower limits which implies they can't dump heat as easily due to size.

Heat can be a destructive force at normal operating voltage but even if you didn't clock the chip so that it would draw hardly any current you could still "fry" it from overvoltage. So to "fry" just refers to the fact that it's been done, it's had it. Take for example a 5.1V zener diode which as we know "breakdowns" or conducts when too much voltage is applied but it doesn't fry it because normally the current is limited. All devices have a breakdown voltage, even a normal diode, it's just that zeners are doped until they behave like a very low voltage breakdown diode. Overvoltage on a Prop pin doesn't fry it either as long as it is suitably current limit too whereas you can't really current limit the supply voltage and make it useful at the same time.

There is a watt limit to chips too but this is a different parameter altogether and happens when you are clocking it too fast and/or driving relatively heavy loads etc. Bear in mind too that while the package has this watt rating you can still develop hotspots in the device leading to destruction without exceeded this watt rating.

evanh

One can take the attitude of experiment and find out.