HomeWork Board - does it Sleep at 50uA or at 50uA plus 15mA?

LoopyByteloose

Okay, I have a Version C HomeWork Board that is suppose to Sleep at 50uA according to the documentation and I have no doubt that the PIC16c57 chip really does that.

But there is a LM2940 voltage regulator on the board as well and it states that its quiescent current it 15ma.

I take that to mean that I should at 15ma plus 50ua together to get a total of 15.050ma when in a Sleep condition.

Am I correct, or is there an 'off condition' in the voltage regulator that uses less current?

Mike Green

The Stamp circuitry on the HomeWork Board may draw only 50uA in sleep mode, but there are other pieces that draw current. The regulator does have a quiescent current and the USB interface will leak a little bit of current even though it's designed to be powered by the USB connection. According to the graphs on the LM2940 datasheet, the quiescent current should be more like 10mA. There is no "off condition" for the LM2940.

If you need a lower sleep current, you'll need to use a serial HomeWork Board so it doesn't have the USB interface built-in and you'll need to replace the regulator with one with a lower quiescent current.

LoopyByteloose

Hi Mike, I have the Serial HomeWork Board - no USB. I really don't need to have a lower quiescent current. It seems the documentation is a bit odd as I have the LM2940 regulator on board (which I like and trust) and provides 1amp of power, but it thinks that it is an LM2936 on board (which will never provide 500ma, as it is rated for 50ma).

The LM2936 is a low-quiescent device (would add some uA to the PIC16c57) , but it would never go on this board as a replacement.

The main thing is that the LM2940 is low-drop out, so I use much more of the 9V battery AND it is internally protected from a backwards battery insertion (which can happen quite regularly).

I am just looking at teaching with these and want to make sure I am teaching reality. Some bright student might insist that one should leave the battery in and not suffer an eventually dead cell.

Mike Green

It's a useful discussion point to talk about the issues behind the choice of a component like a voltage regulator. There are all kinds of tradeoffs. I'm sure the original choice of an LM2936 was based on its very low quiescent current. There were customer complaints about the low power rating that trumped the low quiescent current advantage. Since these devices were introduced, there have been newer devices developed with both features to one extent or another. For most users, a 10mA quiescent current rating is more than adequate. When do you switch? What impacts does that have on costs, manufacturing, testing, multiple sourcing, etc.?

It might be a useful "special credit" project to set up an LM2936 on the breadboard area to bypass the LM2940 and supply regulated +5V to the Stamp, then measure the 9V current drain through the LM2936 to the Stamp under various conditions. The Stamp could be programmed to flash one or two or three or four LEDs using the built-in resistors. In addition to measuring the current drain, the student could look for the point where the average current exceeds the 50mA limit of the LM2936 and it shuts down. Each LED would pull about 15mA with the built-in resistors and the Stamp would pull a few mA.

Tracy Allen

The Stamp documentation mentions the LM2936 prominently because that is the regulator that is mounted on the Stamp module itself. It provides the 5 V power when you apply power to the stamp module at the Vin pin. In that case the Vdd pin on the module is a 5V output. On the homework board, the LM2940 provides power instead, directly to the Vdd pin, and nothing is connected to the module's Vin pin. The 5V from the LM2940 is connected to the output of the LM2936, but that is okay--it does not conduct in reverse.

For any of these regulators it is the graphs in the data sheet that tell the full story, for teaching.

quiescent current vs load current: positive slope, more load current, more quiescent current, due to PNP output stage.

quiescent current vs input voltage: funny things happen when the input voltage approaches dropout, depending on load current. Also may show how the regulator handles reverse input voltage.

dropout voltage vs load current: lower output current requirement allows lower input battery voltage.

There are regulators that have both low dropout and also low quiescent current, but one always has to look at the whole story.

LoopyByteloose

Hi all,
Hard to sort out 'the whole story' with at least 5 revisions of the HomeWork Board existing, not including the variations of USB and RS232. But I think I've gotten it. The board has evolved into a better device for beginner classes.

The simple fact is a have 10 WAM kit with Revision D Homework Boards and I see documentation with photos of Revision E that claim the LM2936 regulator is still in use, but providing an impossible 500ma (somebody goofed).

I AM getting 9VDC on the Vin pins, contrary to claims that they have been disconnected. Not sure what that is about.

I do indeed comprehend the problems that might occur with being only limited 50ma of regulated +5VDC in an educational setting. Besides, it seems that r/c Servos really begin to suffer with +9VDC. Even the r/c airplane industry is now providing +5VDC regulated for
r/c servos on their ESC devices.

There is NO WAY to by-pass the LM2940 without removing it from the board - a rather significant bit of board surgery. And I wouldn't want to. THIS BOARD is PERFECT FOR TEACHING FIRST TIMERS. The traditional BS2 in 24 pin is ideal for someone that wants micro-power and I have at least one.

The LM2940 5VDC published documents state 15ma quiescent current in text, but then provided the curves shown below.

With REVISIONS A, B, C, D, E and RS232 versus USB version, it might be wisest to no longer try to claim 50uA in SLEEP mode and just point out that the board has EVOLVED into being the most rugged board for novice users at home and in the classroom.

After all, it is delightfully rugged and doesn't require the extra tangle of a wall wart. Keeping kids away from 115VAC mains is very wise in a beginners class. Many parents would never let their children use even a wall wart when starting to explore electricity.

How about.....
"An early revision, Rev B of the HomeWork Board was limited to 50ma due to being faithful to the micro-power abilities of the BS1 and BS2, but experience in actual teaching environments has shown that more can be taught by providing higher amounts of regulated power, so even in a SLEEP mode, there will be a drain of 10-15ma on the 9V battery."

Of course, my assertion of 15ma is too general. Tracy Allen pointed the way. Please see the attached curve chart.

Tracy Allen

Understanding those graphs and the summary parameters is a big part of analog design. A lot hinges on choice of parts that have the right combination of characteristics. Another important graph is the one that shows quiescent current vs output load current.

The shape of the graph is due mainly to the base current required to drive the PNP output transistor harder at higher load currents. In a conventional, the output stage is an NPN transistor in a follower configuration. All of the base bias current is delivered to the load rather than being lost as quiescent current, so the shape of the graph is more or less a horizontal straight line. Micro-power is a different aspect, simply very low quiescent current. These days it is possible to get micro-power regulators with quiescent current measured in microamps.

The propellor demo board is nice because it has a plug jumper that allows you to disconnect the prop and its associated circuitry from the regulator. You can put in a meter to measure the current consumed by the Prop itself, or use a battery or a different regulator if you want. That is a nice feature.