PowerSource

Technical Editor Margery Conner's PowerSource streams the latest developments in electronic power design and related technologies.

Profile

About Me

RSS Feed

Add this blog to your RSS newsreader!

By Category

Power Management Articles

Blog

PowerSource

« Smart Grid Scorecard suggestio... | Main | Tire pressure monitoring syste... »

Thursday, August 28, 2008

Camera’s clunky, fat capacitor no match for a supercap

Here are the guts of an old Canon PowerShot A520 that gave up the ghost recently <sniff>. I was impressed at the amount of space dedicated to the capacitor that powers the xenon flash, and for comparison’s sake pulled out the supercap that Pierre Mars of CAP-XX gave me when we met last spring. Here’s picture of the camera and its 80uF, 330V capacitor next to the 0.3 F, 5.5V supercap. (I did a double-take at the 3/10 of a Farad supercap – I don’t think I’ve ever seen an “F” without a “p” or “u” in front of it before.) The supercap is intended for use with LED flash, hence the 5.5V rating, versus the 330V apparently required by a xenon flash.

Supercapacitor This is a graphic example of the advantage a supercap has in cameras and camera-phones: Not only does it take up about half the space – I measured the supercap as only about 2mm thick – but its rectangular, flat package is much easier to tuck into a camera’s or cell phone’s tight real estate. The Canon’s capacitor was held down with double-sided sticky tape that had begun to decompose, and must be a packaging nightmare for devices that undergo temperature extremes as well as shock and vibration. The supercap has tiny surface mount tabs that mount it snugly to a pcb. Supercaps look to dominate the camera-flash market going forward.

Related entries in: Capacitor | Passive components |

Reader Comments

(Add your comments)

at 8/29/2008 11:04:19 AM, Tracy Hall said:
Note that the *Energy Storage* (the real measure for this application) of the two devices is nearly identical. (1/2)*(80e-06)*(330)^2 == 4.3 joules vs (1/2)*(0.3)*(5.5)^2 == 4.5 joules [1/2*C*V^2]. the 330V of the former is needed for the Xenon, although getting that from the supercap (i.e. transform it) wouldn''t be that hard.

at 8/29/2008 11:07:48 AM, Bauerml said:
J = 1/2 CV^2 80uF/330V -> 4.3 J 0.3F/5.5V -> 4.5 J The difference is less impressive now. Yes, the flat pack does package better in today's thin electronics. Martin

at 8/29/2008 11:10:33 AM, Dave F said:
It would be great if there were high capacity, high voltage caps that are small, but physics dictates against it. And the xenon flash lamp does require the high voltage. Plus the purpose of the cap is to supply the energy to the flash lamp directly as it discharges. I don't think many LEDs are capable of putting out nearly the same energy as a xenon flash lamp, even with a high current, short pulse.

at 8/29/2008 11:12:09 AM, RobS said:
Of course, if you store the charge for the flash at 5V, you then need to convert all that energy to HV during the brief flash. I haven't designed such a circuit, but I would imagine it would be more difficult (read: expensive) than charging a high voltage cap at a fraction of the discharge rate.

at 8/29/2008 11:17:29 AM, Brian D said:
Also note that to freeze action, the Xenon flash duration is about 1 ms. To match that with the supercap and LED (CV coulombs in 1 ms) would imply 1500 amps. What was the series resistance of the supercap again?

at 9/2/2008 8:53:05 AM, James Perkins said:
The real measurement for the application of the two devices is almost identical. (1/2)*(80e-06)*(330)^2 ==4.356 if rounded off = 4.4 joules. And for the super cap is ==4.5375 or 4.5 joules. The real advantage is the packaging.

at 9/2/2008 1:59:37 PM, ted said:
I did the same calculation before I looked at the letters. It appears that the tech editior does not have engineering degree.

at 9/2/2008 2:09:30 PM, Stiggle said:
It might be hard to match the short duration of intense light the Xenon tube puts out. Also the Xenon lamp can discharge the 330V capacitor down to about 80V or less before dropping out (used about 4J of energy for the flash) while a white LED will drop out at about 3V and used only about 3.5J out of 4.5J of the stored energy.... I don't see much of a problem making HV capacitors in the prism shape like the super capacitor. There just has not been much of a need to do this. The Xenon circuit can be quite inexpensive and charges from a phone's 3V power source.

at 9/2/2008 2:21:46 PM, EE Geek said:
Geez Louise...supercaps have been available for a good three decades, and in common usage for at least half that period. With a comment like "I don’t think I’ve ever seen an “F” without a “p” or “u” in front of it before", I have to wonder about the editor's credentials.

at 9/2/2008 2:39:06 PM, Steve H said:
Boy - aren't you folks glad you don't get "dissed" on the web like you are dishing out the "Diss.."! Read Margery's other columns - you'll find out that she is a plenty smart engineer. I tried to build a flash from LED's - it just doesn't have the 'snap' of that good 'ol Xenon tube. Not saying that they won't eventually - but they aren't there yet. The Xenon is also pretty energy efficient - but like Margery points out - not so space efficient.

at 9/2/2008 2:44:36 PM, ted said:
I goggled her and it says she has a BSEE and 10 years of experience. But most of it talked about her jobs in pubs. I would think the design experience is very little and a long time ago.

at 9/2/2008 4:20:04 PM, bboyes said:
We've done several commercial LED strobe designs which replace Xenon tubes. It takes 5-6 3W LEDs to match one Xenon tube, but the result is a much more rugged (tested to 200 G's of shock - not a typo) and long-lived device. Total 12V energy draw is about the same, actually a bit better for the LEDs. LEDs do require a fresnel lens to meet DOT beam requirements. The LED pulse width and pattern is infinitely controllable where the Xenon flash is not. One model uses the multiple LEDs to make a "rotating" beacon with no moving parts for example.

at 9/2/2008 4:25:34 PM, bboyes said:
Another issue of LED vs Xenon is the quality and color of the light. "White" LEDs actually use indirect phosphor emission which seems a bit harder to control from device to device. Also the phosphor fades over time. White LED light is more blue than Xenon flash. But recently "warm" LEDs are appearing to improve this. And, you could mix RGB LEDs to control the light color. Some products which do this are now appearing.

at 9/2/2008 8:11:47 PM, Mito said:
The point of these articles is to keep us informed of technology trends. When we see something interesting, we can learn the design details in any number of places. This is not the forum for personal attacks. You can do that on many blogs and newsgroups!

at 9/2/2008 11:31:31 PM, MrSkruggs said:
Hmm. Hard to believe that a Flash LED can generate the necessary 4.5J/1ms burst to measure up to Xenon. Assuming LED VF=4V, and E=VF*A*time... A=4.5/(4*.001) = 1125A !! This means a much longer time is required - and may not be practical.

at 9/3/2008 4:30:36 AM, Shane said:
Unless you are photographing fast moving subjects, why do you NEED such a short flash ? If you really do need it, why not string several led's in series .. Taking it to the extreme ... a 300V string ... and compare performance.

at 9/3/2008 10:19:22 AM, MrSkruggs said:
5-6 series LEDs seems a reasonable alternative to a single Xenon bulb (co$t!?). 300V string would require quite a few devices at 4V/each. Still, a much longer illumination time would be needed to match Xenon energy.

at 9/3/2008 11:15:01 AM, Chris PE said:
New technologies are very exciting and supercaps have awesome uses in LED flashlights and emergency storage systems for safety.I am not a professional photographer, but LED based flash would only be good for basic , inexpensive cameras.Even high power currently used flash tubes are often not enough ,once we exceed 8-10ft.I have a few "old fashioned" flash units as well as newest once with microcontrollers in them and we would need a huge pack of LEDs to TRY matching them.For point-and shoot cameras they may be fine.As far as longevity LEDs die slowly if you drive them to extremes.Xenon flashes live almolst forever.Just my personal opinion.

at 9/4/2008 12:34:49 AM, Dave said:
The advantages of the high current LED flash solution in the smaller cameras are its reduced volume, ease of implementation, it is complementary to existing LED torch/movie light and is able to sustain light output for the long capture period demanded by the CMOS sensor. The advantage of the Xenon flash is its ability to capture fast action, however, if using a CMOS sensor a shutter is also required. The Xenon is well suited for the high end digital still camera and larger cameras where space and pricing permits. On the other hand, the high current LED flash solution is better suited for the camera phone and small pocket camera where space is at a premium. Technical information comparing the high current supercapacitor LED flash vs Xenon flash may be found on the Cap-XX website under News, white paper "Comparison of xenon flash and high current LEDs for photo flash in camera phones" Dave McIntosh Applications Engineer, Cap-XX.

at 9/4/2008 3:57:01 AM, AT said:
Hi Dave, It looks like you have a great understanding of cameras. I went to your web site and read a few articles. It looks like supercapacitors have an exceptional future in mobile phone cameras. i am sure many leading mobile phone companies are trying hard to include them in their upcoming phones. Do you think they can be used improving battery run time in low voltage high power devices?

at 9/4/2008 6:03:40 PM, MrSkruggs said:
A nice article. Are there any safety issues with prolonged (100ms), high intensity light as described? Mom always said, don't look at the sun.

at 9/5/2008 12:40:57 AM, Dave said:
AT, please contact us directly and we will look at your design to see if we can help out.

at 9/5/2008 12:45:16 AM, Dave said:
MrSkruggs, I don't know of any safety issues, the light energy 100mS high current LED solution vs 200uS Xenon is virtually the same and Xenon flashes have been around for a long time.

at 9/8/2008 6:09:08 AM, atom said:
It takes a time to charge the Xenon flash cap after discharging it when the picture is done. I wonder what causes that. Is it SuperCap''s series resistance? As far as I remember according to the datasheet it is equal to 50ish ohms roughly at around 10kHz.

at 11/20/2008 3:54:16 PM, Tim said:
Mito, Your point is well taken, but if these articles are written to "...keep us informed of technology trends...", then the author is at least 2 decades behind. I used a 3.3F 5.5v cap to retain memory on a project I did in college in 1993. How are we supposed to be informed of 'new technology trends' if the author doesn't know what is new technology? For someone at this suppsoed level to maken such a silly statement about supercaps is ridiculous. Half the idiot teenagers out there with there annoying car stereos are using giant supercaps to stiffen their 12v amplifier power. Curious that those teenagers have seen them but a seasoned BSEE hasn't (?).

at 11/22/2008 3:41:24 AM, Dody said:
There was no sup

PowerSource

Profile

RSS Feed

Recent Posts

Recent Comments

Most Commented On

Archives