marknorton Posted March 11, 2009 Share #1 Posted March 11, 2009 Advertisement (gone after registration) There's been some interest shown in how the discreet shutter wind mode affects power consumption and with it battery life. I thought therefore I'd put together another "Anatomy" thread to show you all you need to know about the M8's power consumption. This time, we need some help from my assistants in the form of a Tektronix Digital Phosphor Oscilloscope and a laboratory power supply to replace the battery. The 'scope allows us to monitor the current drain into the M8 while it's doing various things and get a better idea on what costs and what doesn't when it comes to battery life. The M8 uses a Lithium-Ion battery which has a nominal voltage of 3.7v and around 4.2v when freshly charged. These batteries can supply large amounts of current which is why every battery pack has a protection circuit inside it to protect it in the event of over-temperature, short circuit on discharge and over-voltage and over-current on charge. The battery is provided with a safe envelope of operation and the protection circuit prevents operation outside that envelope. How do we measure current? Ohm's law, one of the fundamentals, tells us that when we pass a current through a resistor, a voltage is developed across it. So, to measure the M8 current, I have simply added a 0.1 ohm 1% precision resistor in series with the camera and the oscilloscope monitors the voltage across the resistor. With that 0.1 ohm resistor, a voltage of 50mV across it means the camera is drawing 0.5A from the battery. For this test, I replaced the connections from the Li-Ion cell itself with connections to a regulated power supply with its own over-current protection. The lab power supply uses remote sensing which means the voltage drop across the resistor is outside the regulation loop and does not affect the voltage seen by the camera which, for these tests, was a constant 4.2v. So, to get started, here's a trace of the camera current consumption, starting with the camera idling: 1/30 second, normal wind, DNG only, no review. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! You can see from the labels at the bottom that the horizontal scale is 1 second per division, so the whole scan lasted 10 seconds. I started the scan, half pressed the shutter after 2 seconds, fully after 4. The initial peak is the shutter opening, the next peak is the shutter closing, the data being read out of the sensor and the motor starting; the current remains high while winding against the resistance of the shutter and than falls away as the motor returns to its home position; following that, there's a characteristic pattern of the camera writing the DNG to the SD card and we'll see it's different from when a JPEG is written. With the shutter re-cocked and the data written to the card, the camera returns to idling state in two stages. Again, we'll see the small pulse at the end is characteristic. You can see that the current consumption is about 200mA while the camera is idling, about 600mA when the shutter release is half pressed (presumably, sensor and DSP are now running) with a sharp rise to kick the shutter off and a peak current of 1.8A to start the motor wind, decreasing to about 1.35A during the wind, and back down to around 0.75A during the image processing. The total power consumed by the camera when powered at a constant voltage is V * the integral of current over time, effectively, the area under the line; the higher it goes and the longer it does it, the greater the power consumption and the shorter the battery life. You can see immediately that keeping your finger on the shutter release half-depressed increases the power consumption significantly. And the blue line? That's a second channel on the oscilloscope connected to the flash contact with an external pull-up to pick up the pulse which would fire the flash gun. The camera was set to 1st Curtain, so if we expand the horizontal scale, you can see the pulse is about 10mS wide... If we back up in time to the release point, we can see about a 70mS delay between the shutter release magnet being fired and the camera recognising the shutter is open and firing the flash. If we now set the camera to expose the flash on the 2nd curtain, you can see the flash pulse is shifted to the right compared to the 1st Curtain - it fires later in the exposure, just before the shutter closes. Measure the time between the start of the first pulse and the end of the second and you get about 33mS, or 1/30 second, the exposure time set. If we now go back to the original and, this time, delay the wind using the discreet function until after the write to the SD card is done, this is what we get: The extra current to wind the shutter is moved to the end, and, depending on how long you delay the wind, the current stays higher for longer. It's clear then there is some modest price to pay for using discreet mode, the electronics is kept alive for longer and uses more power. Setting the camera to continuous, here's what happens when you take 8 shots in a row; the time scale has been changed, so the total sequence takes about 15 seconds. You can see how the "domed" pattern we see when writing DNGs to the card is superimposed on the exposure/motor wind pattern. Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! You can see from the labels at the bottom that the horizontal scale is 1 second per division, so the whole scan lasted 10 seconds. I started the scan, half pressed the shutter after 2 seconds, fully after 4. The initial peak is the shutter opening, the next peak is the shutter closing, the data being read out of the sensor and the motor starting; the current remains high while winding against the resistance of the shutter and than falls away as the motor returns to its home position; following that, there's a characteristic pattern of the camera writing the DNG to the SD card and we'll see it's different from when a JPEG is written. With the shutter re-cocked and the data written to the card, the camera returns to idling state in two stages. Again, we'll see the small pulse at the end is characteristic. You can see that the current consumption is about 200mA while the camera is idling, about 600mA when the shutter release is half pressed (presumably, sensor and DSP are now running) with a sharp rise to kick the shutter off and a peak current of 1.8A to start the motor wind, decreasing to about 1.35A during the wind, and back down to around 0.75A during the image processing. The total power consumed by the camera when powered at a constant voltage is V * the integral of current over time, effectively, the area under the line; the higher it goes and the longer it does it, the greater the power consumption and the shorter the battery life. You can see immediately that keeping your finger on the shutter release half-depressed increases the power consumption significantly. And the blue line? That's a second channel on the oscilloscope connected to the flash contact with an external pull-up to pick up the pulse which would fire the flash gun. The camera was set to 1st Curtain, so if we expand the horizontal scale, you can see the pulse is about 10mS wide... If we back up in time to the release point, we can see about a 70mS delay between the shutter release magnet being fired and the camera recognising the shutter is open and firing the flash. If we now set the camera to expose the flash on the 2nd curtain, you can see the flash pulse is shifted to the right compared to the 1st Curtain - it fires later in the exposure, just before the shutter closes. Measure the time between the start of the first pulse and the end of the second and you get about 33mS, or 1/30 second, the exposure time set. If we now go back to the original and, this time, delay the wind using the discreet function until after the write to the SD card is done, this is what we get: The extra current to wind the shutter is moved to the end, and, depending on how long you delay the wind, the current stays higher for longer. It's clear then there is some modest price to pay for using discreet mode, the electronics is kept alive for longer and uses more power. Setting the camera to continuous, here's what happens when you take 8 shots in a row; the time scale has been changed, so the total sequence takes about 15 seconds. You can see how the "domed" pattern we see when writing DNGs to the card is superimposed on the exposure/motor wind pattern. ' data-webShareUrl='https://www.l-camera-forum.com/topic/79157-anatomy-of-the-leica-m8s-power-consumption/?do=findComment&comment=837191'>More sharing options...
Advertisement Posted March 11, 2009 Posted March 11, 2009 Hi marknorton, Take a look here Anatomy of the Leica M8's Power Consumption. I'm sure you'll find what you were looking for!
marknorton Posted March 11, 2009 Author Share #2 Posted March 11, 2009 We know that the camera takes longer to deal with JPEGs and here's the evidence. First, a basic JPEG of 1Mb takes about 5.25 seconds: Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Changing to JPEG Fine and 10Mb takes about 6.2 seconds: Changing to DNG + JPEG Fine takes the longest, around 8 seconds. Notice how the characteristic pattern of the JPEG processing precedes the pattern for the DNG. So, that's what happens when you take images. What about menus and image review? Here's an example of pressing the SET key twice, change the ISO 3 clicks up, press SET and dismiss the menu by pressing the shutter release. You can see how each button press, each click of the thumb wheel causes a blip in the current drawn, probably due to the refresh of the display; when the menu is dismissed, there's additional work, probably writing the menu contents away to flash memory. The LCD backlight seems to take around 150mA, not so much. It's similar when displaying images. Here's image review with the rapidly repeating blips when an arrow key is held down to scroll through the image. Notice how even brief pauses in user activity cause something to be shutdown to save energy... Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! Changing to JPEG Fine and 10Mb takes about 6.2 seconds: Changing to DNG + JPEG Fine takes the longest, around 8 seconds. Notice how the characteristic pattern of the JPEG processing precedes the pattern for the DNG. So, that's what happens when you take images. What about menus and image review? Here's an example of pressing the SET key twice, change the ISO 3 clicks up, press SET and dismiss the menu by pressing the shutter release. You can see how each button press, each click of the thumb wheel causes a blip in the current drawn, probably due to the refresh of the display; when the menu is dismissed, there's additional work, probably writing the menu contents away to flash memory. The LCD backlight seems to take around 150mA, not so much. It's similar when displaying images. Here's image review with the rapidly repeating blips when an arrow key is held down to scroll through the image. Notice how even brief pauses in user activity cause something to be shutdown to save energy... ' data-webShareUrl='https://www.l-camera-forum.com/topic/79157-anatomy-of-the-leica-m8s-power-consumption/?do=findComment&comment=837195'>More sharing options...
marknorton Posted March 11, 2009 Author Share #3 Posted March 11, 2009 Finally, something I noticed which, if I was responsible for the design, I'd want to look at. On this early camera, there's a huge current spike when you first switch on, up to 3.5A: Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Zoom in and it looks like some sort of damped oscillation; quite a narrow pulse, 1-2mS but the size of it surprised me and [speculation] I'm wondering whether this was responsible for the sudden death syndrome some cameras had. Maybe the infamous T2 is damaged by this transient?[/speculation]. That's it. The next "Anatomy" (when I find the time) will take a closer look at how the camera control processor handles the shutter. Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! Zoom in and it looks like some sort of damped oscillation; quite a narrow pulse, 1-2mS but the size of it surprised me and [speculation] I'm wondering whether this was responsible for the sudden death syndrome some cameras had. Maybe the infamous T2 is damaged by this transient?[/speculation]. That's it. The next "Anatomy" (when I find the time) will take a closer look at how the camera control processor handles the shutter. ' data-webShareUrl='https://www.l-camera-forum.com/topic/79157-anatomy-of-the-leica-m8s-power-consumption/?do=findComment&comment=837199'>More sharing options...
george + Posted March 11, 2009 Share #4 Posted March 11, 2009 Mark, you are incredible. Many thanks for the valuable information. George Link to post Share on other sites More sharing options...
jtorrents Posted March 11, 2009 Share #5 Posted March 11, 2009 astonishing! thanks for the explanation. Link to post Share on other sites More sharing options...
miklosphoto Posted March 11, 2009 Share #6 Posted March 11, 2009 Can some one translate this into plain English in one paragraph? Link to post Share on other sites More sharing options...
Guest Posted March 11, 2009 Share #7 Posted March 11, 2009 Advertisement (gone after registration) Thank you very much Mark. Brian Link to post Share on other sites More sharing options...
sfokevin Posted March 11, 2009 Share #8 Posted March 11, 2009 Can some one translate this into plain English in one paragraph? Yes - An Executive Summary would be much appreciated... Link to post Share on other sites More sharing options...
fotografr Posted March 11, 2009 Share #9 Posted March 11, 2009 Can some one translate this into plain English in one paragraph? --Discreet mode uses more battery power, increasing with the amount of time you delay the shutter cocking. Link to post Share on other sites More sharing options...
marknorton Posted March 11, 2009 Author Share #10 Posted March 11, 2009 --Discreet mode uses more battery power, increasing with the amount of time you delay the shutter cocking. Excellent, A++ !! I think though that choosing to use JPEG will have a bigger impact. Link to post Share on other sites More sharing options...
fotografr Posted March 11, 2009 Share #11 Posted March 11, 2009 Mark--It makes perfect sense to me that the spike when turning the camera on could have caused the sudden death problems. I'd call that a very well-educated guess on your part. Link to post Share on other sites More sharing options...
firststream Posted March 11, 2009 Share #12 Posted March 11, 2009 Mark, we certainly appreciate the effort you went through for this! I'm very interested in using the M8 for astro-photography. My attempts at 2 minute exposures resulted in a complete battery drain after just a handful of images. It may be caused by the built-in noise reduction. Would it be possible for you to test for battery drain for exposures requiring built-in noise reduction processes, so as to figure out why the excessive drain occurs? Also, given the battery dies within 20-30 minutes of long exposures, I would be very interested in your AC power conversion set-up, and whether we laymen could imitate it at home. Thanks, Jeff Link to post Share on other sites More sharing options...
marknorton Posted March 11, 2009 Author Share #13 Posted March 11, 2009 Jeff, the current drain during long exposure is about the same as the normal image processing, so what is about 3 seconds using DNG at 1/30 becomes, about 3 + 2 * exposure time. So, if you're running 30 second exposures, the processing bit is something like 50 - 60 times a normal exposure. You can visualise the total drain simply by estimating the area under the line. Doesn't surprise me then you are seeing the battery life you are. As regards the external power, I cut open a standard Leica battery, disconnected the Li-Ion cell and connected the wires to my external supply. You then have to route the wires out through the bottom of the camera which I did by drilling a hole in one of my unused base plates. You cannot leave the base plate off because of light leakage and the switch which detects it. You need to use a much better grade power unit than a typical "wall-wart". Ideally, am industrial supply of 30VA which can be adjusted to 4.1v. Extreme care is called for. I do this stuff all the time, but you would not want to cook the camera's electronics. Link to post Share on other sites More sharing options...
firststream Posted March 11, 2009 Share #14 Posted March 11, 2009 Yes - that explains the drain. Thanks. As to the juryrigging, I'll wait until one of my batteries finally dies. You make it sound so simple - and I hope it is - but I often find what is easy for people like you is a real puzzle to people like me! Regardless, I shall make the attempt in the near future.Cheers. Link to post Share on other sites More sharing options...
carstenw Posted March 11, 2009 Share #15 Posted March 11, 2009 Another great analysis! Thanks! Link to post Share on other sites More sharing options...
Ecar Posted March 11, 2009 Share #16 Posted March 11, 2009 Very informative! Thanks for shedding some light on the mysteries of our favourite toy:p Link to post Share on other sites More sharing options...
cbretteville Posted March 11, 2009 Share #17 Posted March 11, 2009 Excellent! Thanks Mark. Cheers, - Carl Link to post Share on other sites More sharing options...
turner Posted March 12, 2009 Share #18 Posted March 12, 2009 Mark, Thank you for this information! Link to post Share on other sites More sharing options...
xrogers Posted March 12, 2009 Share #19 Posted March 12, 2009 What fun! Looking forward to future installments. Thanks, Clyde Link to post Share on other sites More sharing options...
adan Posted March 12, 2009 Share #20 Posted March 12, 2009 How about comparing the power use for the loud, fast, heavily sprung M8 shutter vs. the quieter, slower M8.2 (or upgrade) shutter. Can you get hold of 2 "lab animals"? It would be interesting to see if less power is needed in the M8.2 to power the shutter springs into submission. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.