Portable Power Supply

Status
Not open for further replies.
T

turtle9er

Well-known member
I am wondering if anybody has experience with using a portable powerpack to power the teensy 3. I recently purchased a Portable powerpack made by Verbatim, it works to turn the teensy on, however after about 10 minutes it stops sending power. I am not sure why it does this, maybe it is looking for some sort of signal from a device like a cell phone, that indicates it is actually charging. So, if anybody has another device that they have used in the past and works that would be great. I need something people can wear while walking around, so can't be too bulky, and only needs to last 8-10 hours at most. Thought this would be any easy solution since already has a usb port and not bulky. Thanks for your time and any suggestions are appreciated.
 
You should have gotten a lot longer on the Verbatim. Maybe you got a dud, or maybe it wasn't fully charged up. What is the rating on the pack?

I've used an EVO Powerbank 5000 maH with 1 amp max power USB charger on both the Arduino Uno and the Teensy 3.0. I haven't done an endurance test with the Teensy, but I did get several days with the Uno running the blink program. On the Teensy, I did run the Teensy with 2 neopixel rings for several hours and the rings are faily power hungry. However, it is a little larger than my 430 pin breadboard, or a 6 AA battery pack, so it isn't ultra small. http://www.amazon.com/EZOPower-Univ...d=1383627756&sr=8-1&keywords=ezopower+5000maH.

I recently got some 5v step-up converters to use with 2 AA battery pack, but I haven't done much testing with it. http://www.ebay.com/itm/DC-DC-5V-1-...616?pt=LH_DefaultDomain_0&hash=item3f116ad0e8
 
Last edited:
It is 2200maH, so lots of juice. It just turns off on its own. This pack has a button to turn it on, so maybe it is looking for something to indicate a device is charging. However, if the teensy is plugged in, it should be drawing power. I just assumed any pack would work. Guess I can try to buy another one and see if I get different results. If I need to bypass the usb plug and power it directly, do I need to cut those pads, or can I just make sure to not plug it in while having external power? Thanks for your help.
 
That reminds of a similar issue I saw over on the Adafruit forums. Adafruit sells a product called Power Bank, and one of the users noticed it wasn't giving any power for his Trinket (Adafruit ATtiny85) in a minimal configuration, but it lasted all Halloween evening with power to spare if he was running all of the lights. So maybe the Teensy 3.0 is too efficient for your battery. Here is the thread for reference: http://forums.adafruit.com/viewtopic.php?f=19&t=45295&p=227527&hilit=powerbank#p227527.
 
These USB power packs use a DC/DC converter to generate the 5 V USB from the ~ 4 V Li-Ion inside. However the DC/DC 'wastes' some power, even if there is no load, so the packs detect if there is no (significant) load and turn off the whole thing.

If your Teensy isn't consuming enough current to retrigger it, it will turn off. You could use a meter and try different resistive loads (be careful with the power in the resistors) to determine its threshold...

I haven't investigated in detail, so I don't know what threshold current they use (perhaps 10-100 mA ?) -- and it's time-consuming to experiment and discover. Perhaps you could have a Teensy drive a transistor every 5 mins which generates a 100 mA load for 10 ms to reset the pack's circuits ?
 
Thanks for all the info, I am going to assume I am not drawing enough current. Sorry for the noob question, but how would I set it up to draw 100mA for a short period, or at least tell me what to google so I can learn how to do it for my self. I have already made the circuit board for my device, so would have to jimmy rig something up for a quick fix. Thanks again.
 
Last edited:
Jp3141 said:
I haven't investigated in detail, so I don't know what threshold current they use (perhaps 10-100 mA ?) -- and it's time-consuming to experiment and discover. Perhaps you could have a Teensy drive a transistor every 5 mins which generates a 100 mA load for 10 ms to reset the pack's circuits ?
Click to expand...

That's a great idea!

I have one of those battery packs here too. I quickly discovered it automatically shuts because Teensy doesn't draw nearly enough power. Turning on a heavy load for only a brief time, if it works, seems like a great way to keep the thing going without actually draining its batteries too much.
 
Just did a quick test. Mine shuts off in 13 seconds with no load. It's 19 seconds with a 27 ohm load. A 22 ohm load seems to keep it on. So far, it's been running for over 3 minutes.
 
Hi Paul,

Can you let me know how to set up my teensy to draw a load. Searching on the web, but don't not sure exactly what I need.
 
Ok, I wired up a little NPN transistor circuit that puts a 230 mA load on the battery. I'm experimenting with a Teensy to pulse the transistor. With 8 ms pulses separated by 1 second, it ran for 45 seconds. Now I'm trying 9 ms and it's run for about 2 minutes so far... oh, looks like it just shut off about 2:20. Trying 10 ms pulses now....

Edit: 10 ms pulses ran only about 1:30. Trying 20 ms now...
 
Here's what I've got hookup up right now, just 3 parts, a red LED (approx 2V drop), a 2N2222 transistor, and a 10 ohm load.

When Teensy drives the pin high (to 5 volts), current flows through the LED (dropping about 2 volts) and base-emitter junction (dropping another 0.7V), for 2.3V on the 10 ohm resistor, for a load current of approx 230 mA. Of course, the transistor has high current gain, so if the load is providing 5V on the collector, most of that 230 mA flows from collector to emitter (the load on the battery). Only a tiny fraction flows from base to emitter. Pretty simple analog circuitry.

battery_load_pulser.jpg

Edit: 25 ms pulse seem to be working very well.
 
Last edited:
Along similar lines, if I have 4xAAA batteries and the measured voltage is 6.05 fully charged, can I connected these directly to the teensy, or is this too high of a voltage.
 
Also, note that alkalines only put out 1.5 volts when they are fully charged (and in fact may be higher depending on the battery construction). http://en.wikipedia.org/wiki/Alkaline_battery.

Over time the voltage will degrade (which is how most battery fuel gauges work on alkaline batteries, they measure the voltage). Nimh batteries tend to have a much flatter curve, and spend more of their time putting out 1.2 volts. So 4 rechargeable batteries would give you 4.8v most of the time. http://en.wikipedia.org/wiki/Nickel–metal_hydride_battery.

If you are wanting to feed your Teensy 5 volts through the USB socket, you might want to look at a step-up/step-down converter that will either boost or reduce the power to match the voltage level, such as this one that does 5v: http://www.pololu.com/product/2123. If you were bypassing the USB connector, I assume you would want a 3.3v vesion: http://www.pololu.com/product/2122. Finally, this one has a tiny potentiometer in it, so you can set the voltage you want: http://www.pololu.com/product/2118.

I believe if you know the voltage is always going to be higher or lower than you need, you can get the appropriate step-up/step-down converter, and can be more efficient. But the step-up/step-down converters will work with a mixed voltage level.

Note, I just bought the adjustable one (primarily to power an LCD that wanted a bit more than 5v), but I haven't used it other than hooking up a multimeter to see that it produced the voltage.
 
Last edited:
PaulStoffregen said:
I posted a quick blog article.

http://www.dorkbotpdx.org/blog/paul/battery_pack_load
Click to expand...
It may be a useful thing for PJRC to sell, if you combine it with a USB male pin. So you would plug your unit into the battery, and then plug the normal cable into your unit so you could power your Teensy. Digispark has this kit that has an on/off button so they don't have to keep inserting/removing the spark. I would imagine adding the components to keep the battery charged, perhaps adding a second switch to keep power, but don't do the auto discharge: http://digistump.com/products/42.
 
From looking at the circuit board, it looks like it is made to plug into the power pack to keep it on. But what if I need to also power the teensy, would I just connect my own usb cable to the 5v and ground on the circuit board? Also, thanks Michael for those links....I actually forgot I ordered some cheap converters from ebay a long time back, I will have to see how they will perform at stepping down the voltage, however I am not sure if they will step the voltage up to 5v when the batteries get low.
 
turtle9er said:
From looking at the circuit board, it looks like it is made to plug into the power pack to keep it on.
Click to expand...

Yes, exactly.

But what if I need to also power the teensy, would I just connect my own usb cable to the 5v and ground on the circuit board?
Click to expand...

My battery pack has 2 USB ports. That little board plugs into 1 to keep it on. I can plug a USB cable into the other to use the power.

If you had a pack with only 1 port, then yeah, you'd need to modify the PCB to give you a way to easily access the power. To be honest, I didn't think of that. I did this whole thing from start to finish in just a couple hours. It was only a quick hack, and it might only be compatible with the one and only battery pack I own. it's not not a well thought out product design. If someone were to make some more general purpose one (eg, to sell as a product), it's be good to test on other battery packs and design in features like access to the power.

I don't intend to do anything more with this, other than build one of the 3 PCBs when they arrive from OSH Park. I'll probably give the other 2 away.
 
Thanks for the info. I just need to pick up some PNP (whole drawer of NPN) transistors and I will see what I can come up with. This at least heads me in the right direction and had taught me some new tricks.
 
Hello, I made the circuit with the only difference that I used s9012 instead of 3906 transistor, but I experience high temperature to the 22ohm resistor. What do you think is the problem?
 
Ok, It was my mistake, the s9012 is not equivalent to 3906 transistor! I replaced the s9012 with the 3906 and is working great! Thank you!
 
I recently ordered a very cheap powerbank case from aliexpress: Safety 5V 1A USB DIY Power Bank 2X 18650
UT844UXXIBXXXagOFbXA.jpg
I run into the issue described in this topic of the device switching off after a delay with low current. The proposed workarrounds did not help. I opened the device and found the IC used is a IP5305. This is a fully integrated power bank IC with a fixed switchoff-time of about 25s when the current is lower than about 45 mA. With an already running timer I created a PWM with a puls width of 200ms and a pause of about 6s to switch a BC639 PNP with a base resistor of 1k and a load resistor of 22R. That apparently does the trick with this bank.I did not optimize a lot, but shorter pulses, lower current or longer pauses did not work stable.
datasheet of IP3505:
http://www.injoinic.com/doc/pdf/IP5305%20datasheet%20v1.0_en.pdf
 

Attachments

  • UT844UXXIBXXXagOFbXA.jpg
    UT844UXXIBXXXagOFbXA.jpg
    97.2 KB · Views: 306
Status
Not open for further replies.
Back
Top