Power consumption figures - including the internal regulator or not?

Genesis

Question that's not answered in the FAQ...· (unless I missed it!)

The various stamps have their power consumption figures given both running and in sleep mode.

Does this include the parasitic drain (e.g. quiescent current draw) of the on-chip regulator, or is this assuming you're providing regulated +5V to the module yourself?

If the latter, is the quiescent current draw of the on-chip regulator published?

Working up power budgets here to figure out expected battery life....

Tracy Allen

The on-chip regulator draws micropower quiescent current, around 15 or 20 microamps when the Stamp is SLEEPing. That applies both to the LM2936 on the BS2 and the LT1121 on the more recently introduced Stamps. That (small as it is) is included in the published figures.

The power budget calculation is complicated by the fact that the Stamp wakes up every two seconds or so briefly, even if it goes right back to SLEEP. The current it draws during that brief time needs to be accounted for. More data on the current accounting at these URLs:
www.emesys.com/BS2power.htm
www.emesys.com/BS2pe.htm#shortstart <--- The BS2pe has low SLEEP current

The on-chip regulators are low-dropout, micropower regulators that have the characteristic that you can leave their input disconnected and connect 5 volts from another source to their output, and current will not flow backwards into the regulator output. That lets a Stamp operate with an external 5 volt supply to Vdd, with Vin unconnected. The BOE uses the LM2940 in that manner to supply 5 volts externally to Vdd. The LM2940 is a hefty regulator capable of 1 amp of output current and lots of power to drive the Stamp and extra circuitry. However, it is not micropower. It draws about 12 milliamps of quiescent current for itself, plus the current for the LED indicator, even when the Stamp is in SLEEP mode.

In contrast, the on-chip regulators are both micro power and low dropout.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Genesis

Thanks - that's what I needed to know.

It makes for an interesting design consideration - potentially using a Power FET to literally "shut off" all external circuitry (e.g. have a second regulator) when the system goes to "sleep", thereby providing a hugely-extended "quiet" capability. If the "wakeup" is an input, then you can "nap" on a tight loop and check the wakeup input, then go right back to sleep. This should cut power consumption tremendously on battery.....

I know about the "interesting" characteristics of the wakeup code (the 10s of milliseconds that active outputs turn into inputs during that cycle) - I'll have to come up with something (e.g. a moderate-value cap across the gate and ground) to keep the gate on during times when it "wakes" during a NAP in the program - but it makes true micropower during "quiet" times possible way....

Tracy Allen

The wakeup code inside the BS2pe is different from all the other Stamps. It stays awake for about 1/100 the amount of time (0.18 milliseconds compared with 18 milliseconds). That both decreases the average quiescent current and also makes it much easier to carry external loads through the input glitch with a small capacitor.

Shutdown on micropower low dropout regulators (controlled by a Stamp pin) makes a nice way to control power to external circuitry.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Genesis

Yep.

The only "gotcha" is the LCD, which requires a delay before you can talk to it when it powers back up.

I'm going to have to look at the impact of in terms of what I'm doing, but a micropower mode is EXTREMELY attractive as it means I can have a "soft" power switch and yet provide some protection against someone failing to turn it on before being used (that is, it can "wake up" once in a while, check to see if the external world indicates its where it needs to be awake, if not, go back to sleep.)