Powering Arduino from a 60-100V battery pack

[dmwahl]

Is anybody powering an Arduino from their main battery pack? I'm trying to think of ways to do this with minimal power dissipation, while ideally still keeping the drain on each cell equal.

It would be easy to do with a linear regulator across the whole pack, but as the Arduino (Uno R3 specifically) can easily draw 20-30mA or more the power dissipated by the linear regulator would add up quickly. On a 100V pack, 20-30mA translates into 2-3W. This would be a fair amount to dissipate from a small board.

If anybody is doing it by only drawing from a few cells I'd be interested to hear how that is working as well.

 

[dnmun]

did you follow otmar's discussion of using the depletion mode n channel mosfet to use as part of the voltage regulation? it is in richard's cellog hacking thread. it was a very educational read.

http://www.ixys.com/Documents/AppNotes/IXAN0063.pdf

you can buy the TO-220 type and attach a heatsink to help the regulator.

 

[fechter]

30mA would make a lot of heat at 100V input.

One approach would be to use a 'wall wart' switching power supply that can run off the pack. Minimal heat and they can be pretty efficient, but I don't think you'd want to leave it always on. I've found a few that can run off 36v up to rectified line voltage (>300v) dc. You just have to do trial and error on those to find one that works.

Another approach would be a 'hybrid' linear/switcher like I found in a chinese watt meter. This used a Simple Switcher chip, which is limited to about 60V input. Ahead of this, they used a linear regulator. The switcher is in buck mode, so the input current is way less than the output current. This drastically reduces the heat dissipated on the linear part, along with the fact that it doesn't need to drop the entire voltage difference.

 

[dnmun]

i was thinking more like 20mA and 85v so i thought 3.5W would be ok on the big type of mosfets that use the TO-220. the little one he used on the cellog was only a few watts as i recall. the smaller TO-89 package of the Dmode mosfet i found was only 1.6W. but it is so small it is half the size.

this is what the balancing chargers use for the DC converters that make the circuit current to power the charger electronics, but it is limited to 40V too.

http://www.ti.com/lit/ds/slls636m/slls636m.pdf

http://www.ebay.com/itm/400284671411?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649

 

[etriker]

I am using one of these packs. 9.9v bms included .

I did not take them all apart.

http://endless-sphere.com/forums/viewtopic.php?f=31&t=36589

 

[dmwahl]

I'll have to try the wall wart idea, I have a box full of them that I've collected assuming I'll find a use for them some day. Thanks for providing some further justification Did you find any correlation between what worked for different input/output voltages?

I didn't follow otmar's thread but I'll look through it. Thanks for the suggestion.

 

[fechter]

I'll have to try the wall wart idea, I have a box full of them that I've collected assuming I'll find a use for them some day. Thanks for providing some further justification Did you find any correlation between what worked for different input/output voltages?

About the only thing is the ones rated for a wide input range, like 90-240 VAC, seemed to be the ones that work at the lower voltages. It's good to test at a voltage slightly lower than the lowest expected pack voltage to make sure there is some head room. Some supplies will start up at lower voltage but may not deliver the full rated current until a higher voltage.

I've done several two stage linear regulators that work fine, but will still have the heat issue. First stage is a FET or Darlington based zener pre-regulator that drops the input to around 24v, which then feeds a 78xx linear regulator. These can work well in excess of 100V with an adequate heat sink and very easy to DIY.

Otmar's circuit was a N channel depletion mode FET feeding a linear regulator. This has fewer parts and very low dropout voltage. Dropout won't be an issue in your case.

 

[etriker]

Or just use a dc to dc converter ?

http://www.ebay.com/itm/1pcs-DC-DC-Buck-Converter-Step-Down-Module-Power-Supply-25V-80V-to-5V-12V-15V-/400349994535?pt=LH_DefaultDomain_0&hash=item5d36b81e27#ht_3079wt_813

 

[DVDRW]

I have two of these. Adjustable? but there is no pot
80V is the highest dc-dc input voltage stepdown converter from eBay you can find

 

[etriker]

I have two of these. Adjustable? but there is no pot
80V is the highest dc-dc input voltage stepdown converter from eBay you can find

Hi,

What is the output voltage ?

Thanks, Steve

 

[dmwahl]

I have two of these. Adjustable? but there is no pot
80V is the highest dc-dc input voltage stepdown converter from eBay you can find

Hi,

What is the output voltage ?

Thanks, Steve

Looks to be 1.5V on that one, at least if I'm reading the datasheet (TPS61001) correctly.

 

[etriker]

The TI chip has 10 pins and the IC in the picture on ebay has 8 pins ?

Can't find a data sheet for that IC with 8 pins ? I don't know what it is ?

I emailed the seller and asked him how to adjust it.

 

[etriker]

You have to change resistors.

http://www.icstation.com/ebay/IMAGE_DING/1859.1.jpg

 

[sixty_cycle_hum]

Per the advice of this thread, I was able to find a USB power supply that I already owned that worked great all the way down past 48VDC.

Works great for charging the iPhone, but I have not tried it with the Arduino yet, since I don't have plans for a bike-duino. Without the Ethernet shield it would probably work great. In my experience, the Ethernet shield takes more current than the actual Arduino.

Here's the charger. http://www.belkin.com/us/F8Z630-Belkin/p/P-F8Z630