Selective powering of modules

#1
In a project with multiple modules, powered by battery and communicating serially, is it viable to power down some modules when they are not in use to save battery power? For example, a main PICAXE controller and a OLED display, in which the main controller is only in use occasionally?
 

nick12ab

Senior Member
#2
Welcome to the PICAXE Forum.

PICAXEs have various low-power modes, including nap, sleep, doze, hibernate, the ability to change to a lower clock frequency, and other solutions, where the main controller can still "turn itself back on" or operate slowly whilst saving power.

A solution which requires another device to turn the device on and off involves using a transistor to control power to the device, but if this is used then no signals can be sent to the device as they will power the device via the clamping diodes and possibly cause damage.
 
#3
Thanks for the welcome, Nick. I'm currently trying to reactivate the part of my brain that did logic design in the 1970s ;-)

What I have in mind is two power rails, one of which is switched on only on demand by means of a timer. I should maybe have worded my enquiry more carefully. Putting things another way, will two modules connected serially suffer any damage if rail power is removed from the sending module?

I take your point about sleep states, which is a great alternative idea. But I need the unpowered part to respond to a switch; I was planning to do that by using the switch to power up the controller module by means of a timer and relay. The permanently powered part will be an AXE134. So the sequence would be: switch pressed, timer (not on PICAXE board) closes relay, supplies power to PICAXE controller, controller wakes up, does its thing, sends message to AXE134, timer times out, relay drops out, controller turned off. I'm trying, of course, to achieve maximum battery life. To clarify your point above, the AXE134 will be the only thing a serial signal is sent *to*, and it will be permanently powered.
 

hippy

Technical Support
Staff member
#4
The answer to the question is most likely "yes", but as nick12ab notes, there may be some issues in how things not powered should be interacted with.

Your having an AXE134 LCD/OLED display permanently powered, having something else power-up, send to the AXE134, then power-down should indeed work.

Depending on what it is you are waking up, you might even be able to use the AXE134 to monitor the wake-up button and handle the timer and powering up of the sending device.
 
#5
Thanks for that. It gives me the information I need to at least get the parts and start messing around with them.

I have to comment in passing that coming back to electronics after a long absence, I am dumbfounded at what is now possible at very modest cost.
 
#6
There is another issue that's only just occurred to me. As I understand it, the AXE134 and AXE033 display modules have a hardwired link that causes them to display the time from their RT clock chip. If (as in my case) one wants the AXE134 to display the time and date when it's not doing anything else, but also be able to display messages, presumably it is necessary to remove that link in some way before other messages can be sent to the display via the serial link. A switched rail operated a miniature reed relay would seem to be a way to do that.
 

AllyCat

Senior Member
#7
Hi,

I'm trying, of course, to achieve maximum battery life. .... the AXE134 will be .... permanently powered.
You do realise that the current drain of the AXE134Y can be around 100 times that of a humble PICaxe? There is of course one PICaxe on the PCB itself running (permanently) at elevated clock frequency.

What are your actual design targets (e.g. battery drain and ambient illumination), as IMHO the need to "switch off" the master PICaxe is probably irrelevant?

Cheers, Alan.

PS: And a transistor (bipolar or FET) will be much more efficient than a reed relay.
 

hippy

Technical Support
Staff member
#8
As I understand it, the AXE134 and AXE033 display modules have a hardwired link that causes them to display the time from their RT clock chip.
Only the AXE033 has an option for an RTC, the AXE134 and related serial modules do not.

The AXE134 would need an external RTC adding if you wanted it to display time data but that should be possible along with having the external PICAXE send to it. You would probably need to alter the 18M2 firmware to tailor it to what you need.
 
#9
No, I didn't. Aaargh! That (48 mA) seems like a crazy consumption for an OLED. For example, I've just bought a FitBit One, which has an OLED display. The whole thing is no bigger than a few sugar cubes, but contains a 3-axis accelerometer, an altimeter, a processor, a wireless comms chip and a battery, and it's looking as if it will run for at least a week on a charge.

There will be a fair bit of power available, from a LiPo cell, but it sounds very much as if I don't need to bother with switched powering.

Can anything be done to reduce the AXE134's power drain? Dimming, perhaps?

I followed your link - thanks for that. 48 mA. Hmmm. With my proposed battery system, which is chargeable via USB, it would give about 56 hours of operation. Not brilliant: need a re-think... maybe mains power...

On your reed relay point, yes, true, but I meant a SPNC relay, so no power in the 'off' state when the time is being displayed.
 
#10
Oh. I am tempted to swear a bit. I'd misread the AXE134 spec to read that it was 'the same' as an AXE033 but with a four-line display.

Does anyone know the AXE033's power consumption?
 
#11
This project is beginning to look like an AXE033 *and* an AXE134. That would actually simplify things, as the AXE134 with its power hunger would only be on occasionally, while the AXE033 could do nothing except display the time: no power switching needed. And I didn't mention the MP3 player ;-)

But I am still concerned about the AXE033's power needs. (Note to self: get one and measure it.)

Great forum - many thanks for all your replies and help.
 

hippy

Technical Support
Staff member
#12
while the AXE033 could do nothing except display the time
If it's doing nothing but displaying time you would probably be better off getting an actual clock module which will have been designed to use minimal power.

Perhaps some detail on exactly what it is you are thinking of creating will help people suggest the best ways to do that.
 
#13
Ok. This thing - a sort of executive toy as a present for a friend - is a desk clock that, when suitably prodded, displays a series of hopefully life-enhancing aphorisms and makes noises (to be determined). When not doing anything else, it's got to be useful (e.g. be a clock). Battery power and USB charge-ability (which I've got sorted) is preferable.

I had the idea ages ago, but I was delighted when I discovered the PICAXE modules, as they seems to offer a quick route to a solution.

I am sure that I came across a web link somewhere that showsed a four-line display from a DS1307 RT clock chip, but I've mislaid the link now. Maybe it wasn't a PICAXE system.
 

AllyCat

Senior Member
#14
Hi,

Can anything be done to reduce the AXE134's power drain? Dimming, perhaps?
Yes, I believe nick has described a method in his blog. But I doubt if it can compare with dimming a LCD backlight with PWM. I obtained a "useful" backlight at around 1% PWM (i.e. < 1mA) on a 20 x 4 LCD (actually driven by PICaxe software via one of the little I2C expander boards).

Cheers, Alan.
 

erco

Senior Member
#15
If it's doing nothing but displaying time you would probably be better off getting an actual clock module which will have been designed to use minimal power.
+1, I've made many executive/retirement/gag gifts using a clock or watch module with an hourly chime. I tap into the chime output to trigger a sound, action or both every hour. Actually very helpful in practice, it's nice to be reminded when you've wasted an hour on something and it's time to move on.

My favorites are the clock modules which have 2 chiming options, one which doesn't chime at night. Saves power and doesn't wake people up for home placement.
 
#17
No, I didn't. Aaargh! That (48 mA) seems like a crazy consumption for an OLED. For example, I've just bought a FitBit One, which has an OLED display. The whole thing is no bigger than a few sugar cubes, but contains a 3-axis accelerometer, an altimeter, a processor, a wireless comms chip and a battery, and it's looking as if it will run for at least a week on a charge.
The OLED on the FitBit One is much smaller than the AXE134, and the entire product will have been designed from the ground up using modern low-power microcontrollers, power supplies etc. whereas the PIC microcontrollers the PICAXEs are based on are rather old now.

If you're willing to dim the display or use a small OLED, substantial power savings can be made. The brightness of OLEDs using this chip (commonly found in small OLEDs) can be varied from completely off to fully on via the data interface, and current consumption is very low at the lower brightness settings. The Winstar OLEDs can be dimmed as well though it would still be higher consumption than using a small OLED.
 
#18
Thanks Nick - a lot to think about here, although at present starting at chip level is probably beyond my competence: I've got to teach my aged and trembling hands to do micro-soldering of SM chips first!
 
#19
Could anyone who actually has an AXE134 and/or an AXE033 tell me what their physical dimensions are. This information seems to be either missing or incorrect on the store site.

Thanks.
 
#20
Could anyone who actually has an AXE134 and/or an AXE033 tell me what their physical dimensions are. This information seems to be either missing or incorrect on the store site.

Thanks.
AXE133 PCB is 56.5mm x 19.0mm.
AXE033 PCB is 92mm x 43mm.
Both PCBs are 9mm thick including components but not solder joints on the bottom.
Dimensions of the Winstar OLEDs can be found in their respective datasheets, and overall thickness depends on what connectors you use.
 
#23
As I started this thread, I'm posting a follow-up on the power issue. I've also started a new thread on GPS time that relates to the same project, now reactivated as the hapless intended recipient is getting married this summer.

In the intervening couple of years Lithium Polymer batteries with colossal power densities have become more available, so this is the way I'm currently going. I say 'available', but in fact the simplest and cheapest way to get hold of, say, a 20,000 mAh LiPo battery seems to be to buy a 'powerbank' and rip it up. These are often offered - at present at least - at substantial discounts on everyone's favourite online bazaar, making the entire unit a lot cheaper than the value of the battery it contains. This will have the added benefit (?) of making the completed unit quite heavy, adding an aura of quality ;-)

I should of course add that tearing something that contains a large LiPo battery apart should be undertaken with caution and in the full knowledge of the risks involved. As for the battery itself, I intend to apply the same safety precautions as Boeing did to get the Dreamliner air-certified again after some of its batteries caught fire (put them in steel boxes). If it's good enough for the FAA, it's good enough for me.
 

PhilHornby

Senior Member
#25
but in fact the simplest and cheapest way to get hold of, say, a 20,000 mAh LiPo battery seems to be to buy a 'powerbank' and rip it up.
My other-half came back from one of her frequent trips to China, clutching such a device. I was astonished at the claim of 20A.h ... so I measured it. I can't remember the exact figure, but it was in the region of 3000mA.h. Not bad value for money, but nothing like the advertised capacity.
 
#26
Duly noted, and I share your cynicism. Mine are PowerAdd Pilot X7s, obtained 'on offer' from a well-known online retailer whose name rhymes with Brabazon. I have three. As soon as possible, I will measure the true capacity and report back.
 

oracacle

Senior Member
#27
My other-half came back from one of her frequent trips to China, clutching such a device. I was astonished at the claim of 20A.h ... so I measured it. I can't remember the exact figure, but it was in the region of 3000mA.h. Not bad value for money, but nothing like the advertised capacity.
that's pretty typical. there is a youtube channel called bigclive who has done teardowns of a few of the cheap battery units. some of them are truly terrifying. there are units available without cells, so you could in theory make on with as much juice as you like providing you run the cells parallel as most don't have balancer circuits.
 
#28
I'm not an expert, but I suspect that paralleling unmatched lithium cells would be asking for trouble.

[Scans 'Big Clive]... I think I like this guy ;-)
 
#30
Ok, consider me informed... clearly I need to do more reading. Trouble is, it's too easy to get diverted into watching YouTube videos of exciting battery fires.
 
Top