Im curious why

[Altera_Forum]

Noone had any advice about running Cyclone II based boards off rechargeable batteries? Is there something I dont know where it's a bad idea?

[Altera_Forum]

Hi there. I saw your original post but had no particular advice to offer, so I didn't bother. However, I hate to see you neglected so here's my opinion. 

 I can see no reason why you can't run your Cyclone II system from rechargeable batteries. Depending on the device you use, I can imagine that the current drain could be pretty high. I am just about to get my first Cyclone II based prototype board back from the manufacturers and so I'm still feeling my way a bit with the power requirements.  In order to derive the necessary core, PLL and I/O supply voltages from your battery I'd look at using a multi-output switched mode controller. A rep. from Arrow showed me a little evaluation board the other day for a Linear Technology device which occupied a very small board area and needed only tiny external components. I'm sure National Semi etc. will do similar devices.  One other thought. If you haven't started your design yet, should you be looking at Cyclone III instead? Maybe others could comment.

[Altera_Forum]

Thanks Sharky, it's going to be a toy/game thing, so I dont know If my idea will work with the cyclone II if the battery life will be short, I'm planning a "laser tag" sort of device pair. I found that national semi stuff on google, I may have to figure out which option of theirs to use. 

 So sounds like you work with this stuff, thats great. I just do it as a hobby since I was educated as an Engineer and enjoy it. Defense plants around here just want people who have been on the job with this stuff for 10 years.

[Altera_Forum]

Hi Evil. I'm a bit surprised that nobody has come forward with a more helpful response on the power requirement question. I suppose it does depend a lot on the device you're thinking of using and what sort of external components you'll need. If the project is mostly for your own personal enjoyment and education then you should go ahead with it anyway. 

 My design uses an EP2C20F256 device with some external flash and SRAM for program memory (I'm using NIOSII). Apart from that I've got a few A/D converters, digital I/O etc. I tried to estimate the likely power consumption using the Powerplay feature of Quartus, but to be entirely honest it's going to be a bit of a 'finger in the air' approach. I'll let you know how I get on.  In the meantime, good luck in your endeavours.  Sharky

[Altera_Forum]

My needs for power are minimal (insert joke about me not really being evil here), the Cyclone II FPGA (144 pin type so I have a chance of soldering it). the EPCS chips, a few buttons, some small relays controling things from another bigger battery. It's just a toy/game prototype, so I'd rather carry extra batteries then build a 1.2v/3.3v warp drive.  

 This is my first PCB build and design, using PCB123 unless you know a better solution. I was a computer Engineer but never did much soldering, hoping the weller WES-51 I got may make it easier.

[Altera_Forum]

I haven't worked yet on battery powered FPGA system, but I would recommend that you try to keep the current consumption as low as possible in your design (being battery powered). When powering from a battery, you will also have to consider FPGA power up in-rush current requirements, which can affect more when the battery is partially drained.

[Altera_Forum]

Thanks, Im hoping these national seminconductor regulators will take care of that, I am beginning to suspect the external components for the regulators must be chosen carefully to do that, It needs a few capacitors, inductors and resistors that be soldered together several different ways. Hopefully I will have enough code written in a few weeks to start building designing a board, then building it.

[Altera_Forum]

National Semi regulators sounds like linear regulators. Switchers are much better for power limited applications. I have power limited Altera designs, and I use the LTC3415 for core voltages for its efficiency. (You have to pay a little more attention to power supply noise.) Linear Tech also has an easier to use LTM series. Also easy to use is a series from MuRata that specifically targets FPGA voltages, MPD family.

[Altera_Forum]

Just to prove you're not mad to try, look what turned up in my in-box today: 

 http://www.fpgajournal.com/articles_2007/20070904_proof.htm

[Altera_Forum]

Thanks, my research is pointing towards National Semiconductor LP3990s for the 1.2 and 3.3 volts I'd need to run a Cyclone II, not sure if the version of Quartus II I am running (free vers with cyclone II dev kit) will work with Cyclone III.  

Their reference designs show more complicated board designs with several capacitors. inductors and resistors, I may try to get by with just capacitors from input voltage to ground, and from output voltage to ground.  I am not sure if that linear technology one with the QFN package is something I can hand solder, with the contacts on the bottom. Maybe when I sell a few copies of my first game design, the second version can have that.

[Altera_Forum]

The Quartus II web-edition (free version) supports Cyclone III device family. See supported devices list at http://www.altera.com/products/software/products/quartus2web/features/sof-quarweb_features.html 

 I think that you could benefit from low power features of Cyclone III compared to Cyclone II if it suites your design requirements. May be a Cyclone III in a QFP package...

[Altera_Forum]

I've had a quick look at the LP3990 datasheet. Using a linear regulator is certainly a straightforward approach (which is why I used them on my prototype board) but it's not optimal as far as power consumption is concerned. Based on minimum input voltages recommended in the datasheet (2V for the 1.2V output regulator and 3.8V for the 3.3V output regulator) you're going to be wasting maybe 25% of your battery power. Depending on what your rechargeable cell voltage is, you might not be able to get these input voltages (e.g. you may have to use 2.4V and 4.8V in which case you'll be wasting even more power). 

 Anyway, as you say, once you've proved the concept and made some sales you can develop and optimise the design further.

[Altera_Forum]

[My opinion] - You should definately use a switching power supply instead of a linear regulator for this battery powered application, even intial builds. Using a linear regulator, you waste the power from the supply voltage down to the regulated voltage. 5V down to 1.2 V puts you in the 25% efficient range. A switcher puts you at the 90% range. Consider an easy to use one such as  

http://focus.ti.com/lit/ds/symlink/tps62243.pdf The effect on your battery life will be dramatic.

[Altera_Forum]

I would also recommend using a switching supply. The advantages are twofold. To begin with, you will get more efficiency compared to a linear one and you will be able to power the system with draining batteries as long as the current drain requirements are met. You won't have to bother about having the battery voltage above linear regulators' drop out limit.

[Altera_Forum]

If you want to learn a little more about cyclone devices for battery powered applications, you may want to try to look at this board that linear tech and arrow put together: 

 http://www.fpgajournal.com/articles_2007/20070904_proof.htm  this is a nice little write up from fpga journal.  good luck,  guano

[Altera_Forum]

Some more detail for the Arrow/Linear Tech's board can be found here: http://www.arrownac.com/mktg/resource_mag/may_2007/pdfs/resource-spr07.pdf (page 24-25 of document, 26-27 of pdf).

[Altera_Forum]

Then I will go with the switching ones that you recommend that make for the simplest PCB build.  

 In case you were wondering why I ask so many questions, I work for Staples now and I gpt my Engineering B.E. 12 years ago, I have no access to peoples or materials of this nature, so I really appreciate all your answers.  The national semiconductor page says about buck, buck switching with some complicated PCB setups and transformers they just label as 'custom'. Any suggestions if I can use those?

[Altera_Forum]

Looks like the LCD and GPS module are gonna need 5v, if I regulate the battery down to 5v to power those, then regulate that 5v down to 3.3 and 1.2, would that be more efficient that hooking up the regulators in paralell?

[Altera_Forum]

Hello again Twinzz 

 If you're going back to thinking about linear regulators, then the answer is no. All that would do is change where the power is being wasted. e.g. let's say your 1.2V regulator is supplying 100mA. That current is being drawn through the 3.3V regulator and the 5V regulator, adding to their individual dissipation. You've had to drop the same number of volts to get down to 1.2V. There's no getting away from it, linear regulators waste power.  If you're thinking of switching regulators, I think a similar argument would apply. I'd try looking for a multi-voltage output device or use separate regulators from a single supply.  As a matter of interest, my prototype board is now up and running with linear regulators and they're barely warm. However, my board goes in to a current injection set which could draw 10A from a 230V supply so low power consumption is not an issue.  Best regards

[Altera_Forum]

Thanks for info sharky, glad your proj is movin along. I don't see 'switching' in the National Semi page, but i see 'buck', so I will go with that. My concerns are designing and building a complicated PCB and not having it work right after the board house makes it. I really appreciate all the answers, I feel I owe you some Starbucks or something at this point.

[Altera_Forum]

Buck and boost regulation just refers to whether the output voltage is lower or higher than the input voltage. So a buck regulator is still a switching regulator. For example, you could have a situation where your battery voltage is 3.6V and you have a boost regulator up to 5V and two buck regulators down to 3.3V and 1.2V. 

 I understand perfectly why you are concerned about getting a more complicated power supply design right first time. I chose to go the linear regulator approach so that I could concentrate on the FPGA operation and not worry if my powers supplies were causing problems. On the production version I might well substitute in switching regulators.  You haven't mentioned whether you've looked at Linear Technology's switching regulators yet. I'm sure they are worth considering if you really only have one shot at the overall design.  Now, was that Starbucks or Starboosts?:rolleyes: