Author Topic: Battery performance in the cold - D5300/D5500, D7500, D500 - test challenge  (Read 2587 times)

Řivind Třien

  • NG Member
  • *
  • Posts: 1026
  • Fairbanks, Alaska
Another update: During the above test and tracking attempt a few nights later where temperature dropped to -30°C, while the battery bank and camera worked without a hitch, my SkyTracker stopped working abruptly. First night this happened about half way, and the remote release cable appeared to be snagged between the rotating platform and the tracker, but the next colder night it only lasted for about 40 minutes before it stopped. I recalled some previous problems when it got that cold.

I then got the idea to also power the tracker from the same 9V outlet via the super-capacitor, and a quick test revealed that power consumption of the tracker is very low, only 70mA drawn from the battery bank. Last night the cold weather at -34°C appeared to open up enough with stars visible for a tracking test test closer to midnight. Of course right after everything was set up and aligned clouds started drifting in and with that it warmed up towards -30°C during the test. Clouds were still transparent enough to see some stars to check if tracking worked. Except that the camera hit the polar scope at some point (ouch, luckily I caught it after only two non-tracked frames so no damage done), the SkyTracker and the D7100 worked OK though the 4hour test. So as long as the tracker has power it might still work at these low temperatures. I will now incorporate the extra power outlet for the tracker in the final design. It is anyway very awkward to change the AA batteries on the tracker (needs pliers/leatherman tool to get the battery holder out), so using the 9-12V power inlet might be a good solution in spite of the extra cable clutter.
Řivind Třien

Seapy

  • NG Member
  • *
  • Posts: 729
Very interesting Řivind, I am following this with great interest.
Robert C. P.
South Cumbria, UK

Řivind Třien

  • NG Member
  • *
  • Posts: 1026
  • Fairbanks, Alaska

In a further modification to the super-capacitor circuit I included an extra schottky diode at the input to prevent charge in the super capacitor to go back to the source. This also brings the voltage very slightly down, which is safer with the 9V rating of the super capacitor:




To avoid extra cable clutter I got the idea to build the super capacitor circuit into the SkyTracker. At first sight of the internals, the space looks tight, showing the large worm drive to the right, the DC motor with its clock drive and an optical encoder that provides feedback on speed to the upper left, and the control board to the lower left.



But underneath the control board I found some unused space (image rotated 90° clockwise compared to last one).




My hard-wired circuit board looks a little odd, as the components had to clear the transistors etc. underneath the control board.




It was finally insulated on the bottom and simply hot-glued glued in place, so it looks a little ugly. I can still have batteries installed in the tracker as a backup solution for the tracker.




A difficult part of the modification was drilling the hole for the power outlet to the camera in the tracker at the right location. Generous amount of plastic bags and masking tape was used to prevent metal filings getting to the sensitive clockwork mechanism.




I also changed the cable to the DC converter with a flexible silicone insulated one I had in stock, as the original one cracked after only a couple of uses. A similar cable runs to the connector for the fake EN-EL15 battery (hotglue is your friend). The DC converter fits on the side of the Tomo battery bank contained in a modified ThinTank pouch that I got as a freebie at some point.




The complete setup (except intervalometer not connected to the camera yet). The ThinkTank pouch with the battery bank is conveniently attached with its Velcro attachment to the tripod leg.




Řivind Třien

Erik Lund

  • Global Moderator
  • **
  • Posts: 4793
  • Copenhagen
    • ErikLund.com
Very nice DIY job!
Erik Lund

Akira

  • Homo jezoensis
  • NG Supporter
  • **
  • Posts: 8119
  • Tokyo, Japan
    • Akira on flickr
This is neat, Řivind!
"The eye is blind if the mind is absent." - Confucius

"Limitation is inspiration." - Akira

CS

  • NG Supporter
  • **
  • Posts: 832
  • You ARE NikonGear
Oivind, I showed your handiwork to my friend that does astrophotography, because I knew that it would interest him. His comment below:

"I’d like to see some results…
My experience is that unguided trackers have way too much wobble, unless you spend the $$$ to get precision ground worm and ring gears."
 
Examples of images using the modified rig, along with your comments, would be very nice.  :)

Carl

Řivind Třien

  • NG Member
  • *
  • Posts: 1026
  • Fairbanks, Alaska
Thaks for the enthusiastic comments, Erik and Akira.

Carl, If fixed the wobble in the Skytracker already at the start by removing a washer in the rotating base. Both Armando and me have shown plenty of results with 300mm focal length on the SkyTracker Classic (ver. 2) in the Night Sky shots thread, most of mine based on stacking 60 second exposures, for instance this page, :
http://nikongear.net/revival/index.php/topic,1992.msg130037.html#msg130037
The last image of the North America Nebula is at 2500 pixels horizontal resolution if you open it in a new tab.

Here is an image recorded using the battery bank to power the D5100 with 300mm and the tracker with super-capacitor at -40°C, also at 2500 pixels horizontal resolution if you open it in a new tab. The super capacitor circuit was still on the prototype stage, so not built into the tracker yet, just contained in a Ziploc bag. The battery bank lasted for 2.5 to 3 hours before the Li-Ion batteries got too cold and the battery bank shut down. It was taken inside to warm up for 15-30 minutes and then lasted until the end for another 2.5 hours, including recording about 90 darks at the end when the target got out of range. Light pollution was a little too strong to get a good view of the Jellyfish nebula to the center left, but I like how the Shoe Buckle Cluster at the right edge came out.


Nikon D5100 with 300mm f/4 PF @ f/4.5 front aperture, ca. 66 minute total integration time at ISO 400. Jellyfish Nebula (IC 443) left, and Shoe Buckle Cluster (M35) right edge.

A lot of advanced trackers that allow guided exposures stop working at much higher temperature. The mounts owned by our university's Physics department  did not even work at -10°C during a recent star party.

I was tracking last night with the D500, but have not gotten to process them yet. Lately I only need to throw away about 10% of the frames due to mis-tracking with careful polar alignment, depending a little how far away from the pole I image. The periodic error curve has a flat portion and then a bump when it reaches the end of the worm drive at 9-10 min intervals where a frame or two sometimes need to be discarded. Lots of those using guided tracking track for much longer times, sometimes 5-10 min with similar magnification or even more, but due to light pollution I would blow out the stars with that long exposure.
Řivind Třien

CS

  • NG Supporter
  • **
  • Posts: 832
  • You ARE NikonGear
Thaks for the enthusiastic comments, Erik and Akira.

Carl, If fixed the wobble in the Skytracker already at the start by removing a washer in the rotating base. Both Armando and me have shown plenty of results with 300mm focal length on the SkyTracker Classic (ver. 2) in the Night Sky shots thread, most of mine based on stacking 60 second exposures, for instance this page, :
http://nikongear.net/revival/index.php/topic,1992.msg130037.html#msg130037
I was tracking last night with the D500, but have not gotten to process them yet. Lately I only need to throw away about 10% of the frames due to mis-tracking with careful polar alignment, depending a little how far away from the pole I image. The periodic error curve has a flat portion and then a bump when it reaches the end of the worm drive at 9-10 min intervals where a frame or two sometimes need to be discarded. Lots of those using guided tracking track for much longer times, sometimes 5-10 min with similar magnification or even more, but due to light pollution I would blow out the stars with that long exposure.

Thanks, Oivind, I will pass that on to my friend.  :)
Carl

Řivind Třien

  • NG Member
  • *
  • Posts: 1026
  • Fairbanks, Alaska
Carl, I edited in parallel, please see the last part of my post.   :)
Řivind Třien

CS

  • NG Supporter
  • **
  • Posts: 832
  • You ARE NikonGear
Carl, I edited in parallel, please see the last part of my post.   :)

No problem, Oivind. More to come....... :)
Carl

Řivind Třien

  • NG Member
  • *
  • Posts: 1026
  • Fairbanks, Alaska

An analysis of the D500 power consumption:
The peak current draw of the D500 is even higher than the D7100, exceeding 3A. However the final version of the super capacitor circuit nicely caps the current draw from the DC-DC converter at ca. 1A. This curve is with the Nikon 12-24mm mounted, so a worst case scenario (see below) rather than minimum current draw without lens:


Blue line current draw by D500, red line current draw from DC-DC converter, teal line voltage supply to D500.

The 8.5 V supply in idle mode of the D500 registers at 98-100% battery power. The voltage dip is partly due to that I am passing the supply though the 0.1 Ohm resistor I use to measure the current draw, and does not affect the battery state display. Another reason is voltage drop over the diodes, about 0.25V per diode and increasing somewhat with the current draw. This voltage drop is partly by design, as it will help draw current from the super capacitor rather than the power bank.

Here is a chart comparing static types of current draw by D5100, D7100 and D500. Each state was measured over several seconds with a highly accurate voltage meter.



While the D5100 generally use draws little more than 2/3 of that of D7100, what is striking is the minimal % increase in current draw of D500 over D7100, generally in the 2-8% range. The exceptions are Live view  exposure where the screen back light stays on with D500 opposed to D7100. The latter draws about the same current during live view exposure as one in viewfinder mode. Also video recording use a little more on D500, however I had digital image stabilization turned on in D500. Image review with paging though frames as fast as possible also is higher, which could be due to D500 being able to page though images faster, perhaps in combination with higher display resolution.

Using high ISO has no effect on current draw during a 30 second exposure. I also tried longer exposures and turned High ISO noise reduction on and off, but there was no difference. so Peter, your shorter battery life at long exposures and high ISO mentioned on page 1 seems to have other causes.

If AFS lenses are attached, current draw even when idle might increase 30-40%, however active VR is only a 10% increase in the 300PF, about the same as continuous AF operation. The 12-24mm has higher idle current draw but uses little extra during continuous AF operation. My custom chipped 55mm f/3.5 micro did not draw any measurable power compared to my second non-chipped copy. (It likely just powers up to communicate when mounted).

There is a tiny bit of identical current draw in standby mode and with the power switch off. It is a respectable 167 uA (mikro Ampere) in D500, considering that both the top display and the viewfinder mask both receive power. In the D5100 the standby/off state draws almost 1mA, which could drain the battery in 47 days.

If we integrate the curve of the shutter actuation (one with the 55 micro used as base) it is possible to calculate how much extra power a short exposure need in Ampere-seconds, and set up a budget of current drawn based on the above table.



In this particular calculation I have assumed that the camera stays on for 15 sec at each use. D500 cuts power when battery reports 25% in D7100 (see below), so this must be taken into account with respect to which capacity is available.  It is clear that what costs power is not the shutter actuations, which is only a minor fraction, but how long the camera stays on on each use. Also when the D500 is new, and one sit down learning and setting up the menus, one could run the battery down in about 5 hours (How time flies...). So what has been said about mirror-less cameras is also true for DSLRs. (Increase on-time per use to 60 seconds, and the frames per charge will go down dramatically.) Nikon must be taking some liberties when calculating CIPA standard exposures (which is a fairly loose standard). The camera is supposed to stay on for 30 sec for each exposure with screen on, while my calculations will result in the specified 1240 exposures if I set on time per exposure to 21.5 sec and no live view.

A comparison of the battery reporting in D7100 and D500 (inserting the same battery at different charge levels in both) reveals an almost linear relationship between the two, and independent of battery temperature. D500 seems to be more picky on cutoff voltage and will report empty when D7100 reports around 25% charge left. Note that this graph is just a battery meter test and does not imply that D500 use more average power. But the chosen voltage cutoff level might be a consequence of the need to sustain peak current draw. Keep in mind that the single digit series with the same frame rate use a battery with higher voltage. Previous bodies needed a battery grip to achieve this high frame rate while D500 can do without it.




I also let batteries at two different charge levels cool down for several hours outside before quickly inserting them into D500 and D7100 to check reported battery level:



The reported charge levels by the battery meters show a nice slow decline down to about -20°C, but then the capacity start to drop off more abruptly. The conclusion is that one should go out with battery fully charged at these temperatures, and not expect to have more than 50% capacity available. It remains to be seen how the D500 will perform below -30°C. The situation might be better than this chart indicates, as the camera usually stays in an insulated bag and takes time to cool down. And in very active use the camera will heat itself too, and might not reach ambient temperature.
 

Řivind Třien

CS

  • NG Supporter
  • **
  • Posts: 832
  • You ARE NikonGear
I see that as a beyond worse case scenario for folks like me that live in milder temperature zones. So, those in this area should expect useable battery capacities somewhere north of 50% in cold weather.
Carl

Peter Connan

  • NG Member
  • *
  • Posts: 886
  • You ARE NikonGear
Thank you very much for this very interesting post Řivind.

I would love to understand what is causing my low battery life, as I am not changing any other settings except the ISO?

It could be different batteries, but they are both equally new, and since I always cycle them, they should have done nearly the same number of cycles.
But at some point I will re-run my much less scientific tests using both batteries for both high and low ISO tests.

Řivind Třien

  • NG Member
  • *
  • Posts: 1026
  • Fairbanks, Alaska
Thank you very much for this very interesting post Řivind.

I would love to understand what is causing my low battery life, as I am not changing any other settings except the ISO?

It could be different batteries, but they are both equally new, and since I always cycle them, they should have done nearly the same number of cycles.
But at some point I will re-run my much less scientific tests using both batteries for both high and low ISO tests.

Glad if it can be useful, Peter. I must admit I am a little at loss with respect to your battery life problems at high ISO only. One could think of power consumed by the attached lens, however then the problems would not be fixed by imaging at ISO 400, if the same lens was still in use. An approach would be to do some measurements of the current draw similar to what I did above. You only would need to do static measurements, as the mirror movements is a negligible portion of the energy used during long exposures. A USB power meter could be used for the current draw measurements, but would require the DC-DC converter to get from 5V to 9V, and then there is the problem with buffering the power peaks.

The freezer test would reflect both the current consumption and the battery thresholds. You could possibly check if the threshold is affected by the ISO setting, but it does not seem very likely, and I think I checked that.

I must say that the D500 is a lot more pleasant to work with for astro photography beyond the flip screen. The live view appears more sensitive so that I can use live view directly when focusing with the Bahtinov mask on prominent stars - with D7100 I rarely could see the spikes and had to do test exposures. This also makes it easier to navigate to the target, as more stars to navigate by can be seen. I can in addition see more stars in the viewfinder if I choose to use that.
Řivind Třien

Peter Connan

  • NG Member
  • *
  • Posts: 886
  • You ARE NikonGear
I am not an electronics expert at all, but since the lens and camera are set manual focus and VR is switched off the lens would consume very little or no power?

Although to be honest I don't even know if this lens has an electronic aperture (it is the Tamron 15-30mm f2.8 ).

Unfortunately I do not have the tools to measure current draw to these levels, nor do I have the power adapter to provide external power, and since I very much doubt I would use it ever again (almost all my shooting happens away from external power sources), I hesitate to buy one just for a test.

I will test again using both batteries, but to be honest it's not that much of a deal since I have found a method that works sufficiently for my needs.
Thanks again for your assistance!