Use a Sony alpha 900 DSLR for astrophotography on a cold winter's night and you will soon discover the infoLITHIUM NP-FM500H battery packs, even when fully charged, last only a couple of hours, if you're lucky. It doesn't matter where you buy them, or how much you pay for them. It doesn't matter if they are Sony oem's or cheap Hong Kong knock-offs, they discharge in the same short interval, especially when its freezing.


The battery grip VG-C90AM is a possible solution.

It holds two battery packs, so it will last about 4 hours. But you'll need a spare pair of packs, and a means of recharging both in less than 4 hours. The vertical grip costs about 365, and a pair of Sony oem NP-FM500H infoLITHIUM packs about 160.


Browse thru' Sony's website looking for an external power supply unit for a Sony alpha 900 and you will be led to this accessory, the AC-VQ900AM AC adapter and charger. It is a smart charger, able to recharge a pair of infoLITHIUM NP-FM500H battery packs, sequentially, within 4 hours. The charger is bulky, 5" x 5".3 x 1".9 and weighs 18ozs, too clumsy for an external PSU dangling from your telescope mounted DSLR. It also costs about £160, although Calumet sell them for £130. Just a tad OTT for powering your alpha 900 on BULB throughout a freezing night. According to Sony the unit has to kept in a warm, dry place. Definitely not the answer. But I'll give Sony their due, they certainly know how to make their money back on the peripherals. Lets do the "arithmetic" (ITS NOT MATH!): Grip £365 + 4 packs @ £80 + Smart Charger @ £160 = £845. Hardly a sensible way to address the problem.


If you've already gone down this route, but sensibly been deterred by the expense, then follow me as I dig deeper. Buried in the Pre-Cambrian layers of their website is the AC-PW10AM AC Adapter. No mention made on the Sony alpha 900 brochure's accessory page, this unit looks more promising.

Its clearly designed to be used with the alpha 900 & it looks like you can keep it dry at night without too much difficulty. It's also compact, 5" x 2".5 x 1".4 & weighs only 11oz. Warehouse Express had it in stock @ £80 + p&p. It is available in the USA for about $100, but by the time you've paid Shipping, Duty & VAT & Clearance and Handling fees, that'll be closer to £100, and it won't come with a UK mains plug, so you'll need a US adapter. That's the bad news. The good news is a compatible UK mains PSU is available from Hong Kong for only £20 & a replacement from digi-Quick @ £14 + p&p.

If you can live with the mains input this is probably the simplest solution. You can dispense with the infoLITHIUM NP-FM500H battery pack and power the camera via a 240VAC supply. OK in an observatory or back garden, not ideal for a remote site, unless you can interface it via a 12VDC cigar lighter mains inverter.


Ideally what is required is an external PSU and instead of an infoLITHIUM battery pack in the battery chamber, just a dummy battery pack. The camera will operate with the battery door open, but there is a difficulty in making such a PSU function with the camera. When the a900 is switched on the BIONZ processors boot up and look for a handshake from the infoLITHIUM battery cct board built into the pack. When it doesn't receive the handshake the processors shut down and the camera will not accept inputs.

My initial approach was made in ignorance of this firmware requirement. With the assistance of Pete Warrington we dismantled a duff battery pack. The cells were live, but their charge retention was poor so it made a handy sacrifice. Pete split open the thin plastic casing using a Stanley knife blade, held flat against the case weld seem, and gently driven in with a hammer. Once the induction weld had been partially split, the blade was put in the knife handle, and the tip run along the seem, carefully cutting through without penetrating the thin case wall. Don't be cowed by all the H&S crap posted on the web and printed on the packs &c - there is no possibility of shorting the cells. The only thing that can go wrong is you scoring too deep and damaging the cct bd. You're going to discard the cct bd, so that won't matter. Try to keep the score line within the induction seem weld line, and cut through precisely and neatly. The case may need dressing with a Swiss file before you Araldite it back together after the mod.

Inside the casing are a pair of Li-ion batteries, taped together with a plastic separator holding their casings apart, and a sticky membrane bonding them to the curved part of the casing. The pack has three terminals. The outer pair are marked + & -. Tag welded to the terminals is a small integrated cct. board. This takes a 3.7VDC supply from one of the batteries, and is connected across the power output terminals and also to the centre terminal. It is the centre terminal that transmits and receives the charge state & infoLITHIUM handshake. The BIONZ processors keep a record of how long the camera is powered up, and the charge state of the pack. (The camera also stores data on how many times the battery is replaced and how many hours the camera is on - talk about Big Brother!)

The Li-ion cells are Sanyo UR18500FK nominal voltage 3.7VDC with a rated charging time 2.5hrs. Data sheet Safety sheet If a NP-FM500H fails it can be rebuilt using these cells Li-ion 18500 Cylindrical Rechargeable Cell: 3.7V 1400mAh (5.18Wh) --- UL listed (0.42)Li-Ion 18500 7.4 V 1400mah (10.36Wh) battery module with Protection IC (LC18500S2R1WR) (0.84)

UR18500FK Li-ion cells

This photograph shows why it is impossible to short the cells in a NP-FM500H infoLITHIUM battery pack. You are looking at the lower end of the cells, the end opposite the info. cct. bd. Note the cells are tag welded together! The opposite ends of each cell are in turn tag welded to the info. cct. bd. Note also the score line across the left cell which has had its pink plastic sheath removed. This was where Pete's Stanley knife blade dug into the end of the cell as he cut through the plastic casing. Remember this when you follow our example. Everything you have ever read about not dismantling these types of battery pack is a lie, except for one thing, if somehow you do manage to short the pair of cells, (& it won't be by drawing a knife blade across them) they will ignite, and the fumes from a Lithium fire are toxic.

The cct bd itself comprises IC002 a M166Q0394CKX smd microcontroller, one of the C16x Family of Siemens 16-Bit CMOS Single-Chip Microcontrollers. The smd's IC0001 WB8E3 & IC0003 WB8E4 are ram chips used by the microprocessor's bootstrap loader. It reads the BIONZ processor boot sector when the camera is turned on. This is what constitutes the info handshake. NPT-GMK-T is a class of Panasonic spec. built-up multi-layered cct bd. IB-051-970-21 is the Sony cct. bd. assy. #. A & B are the corresponding halves of the cct. bd. + (on output left) & - (on output right) solder terminals . EP-GW is the centre terminal id.

Sony NP-FM500H info cct bd output side

Sony NP-FM500H info cct bd input side

What we did was disconnect the batteries from the cct board. I tried making a power connection across the cct. board, but without the centre 3.7VDC pickup there was no output voltage across the terminals. I then disconnected the cct board altogether, and fitted a Maplin tip bipolar plug from their Digital Camera AC Adapter L69BQ with leads soldered directly to the power terminal tags. The void was filled with potting compound, Maplin PX700K and the case Araldited back together, ensuring the two halves lined up precisely.

The battery pack cells supply 8.4VDC under zero load. This Maplin PSU can supply 8.4VDC, which drops to 7.4VDC under load from the a900. When I inserted the dummy battery into the camera chamber, and turned it on, the BIONZ processors booted and the LCD info screen came up, but ≈12 seconds later flagged an error message <for infoLITHIUM battery only> and shut down.


After discussing the problem with Pete Warrington, I decided upon a more devious approach, a proxy dummy battery. I bought a very cheap NP-FM500H compatible pack WIN-SYFM500HB for 11 via Amazon from importer PowerPlanet in Jersey, & Gaffa-taped it to the Maplin PSU & wired the output terminals in parallel from the PSU via the pack to the dummy battery. I then ran a separate third lead from the centre terminal of the live pack to the centre terminal of the dummy battery.

(Don't try soldering the wires directly to the live battery pack terminals. The terminals will drain heat from the soldering iron, melt the casing and may cause a lithium fire. Fit 2.0mm pins into the terminals with the wires already soldered to them. DC power plug Maplin L49AY with fitted internal brass pin is a transition fit. Solder each connection across both poles of the plug.)

This time when the camera was switched on and the BIONZ processors booted, they got the handshake from the infoLITHIUM battery they were waiting for, albeit it took a little longer. But because the live battery pack was not in the chamber, the heat from the pack could not raise the temperature of the sensor. Strapping the live battery pack to the PSU also keeps it warm and extends its operating time. The PSU supplies the power, and also charges the pack (although not efficiently but that doesn't matter). What the live battery pack does is send the info signal to the camera so the BIONZ processors can monitor the charge state, and battery charge life.

The leads are about 4ft. long, and there is no significant voltage drop or info signal dropout. The PSU is plugged into a Powertank with a mains inverter. It powers the camera all night without any problems.

This dummy battery solution will work with Sony alpha 450; 500; 550; 700; 850; 900 DSLR's. You need two battery packs at £11 each and the Maplin L69BQ AC PSU @ £20.

Having arrived at a working mains input dummy battery solution, I then tackled the problem of a 12VDV input. Maplin do a cigar lighter mains inverter, 75W peak, 60W continuous, A85HZ for £18.


The consensus following discussion about DSLR dark frame noise was that a dummy battery psu could not cause the sensor to warm up during a dark frame imaging session, whereas a battery pack resident in the chamber would warm the sensor. To put this hypothesis to the test I conducted the following dark frame analysis.



The following sequence of dark frames were taken using an Intervalometer:

infoLITHIUM BATTERY NP-FM500H 7 x 300s consecutive NR off & 7 x 300s @ 300s interval NR off

SONY AC-PW10AM - no infoLITHIUM BATTERY 7 x 300s consecutive NR off & 7 x 300s @ 300s interval NR off

DUMMY BATTERY EXT PSU 7 X 300s consecutive NR off & 7 x 300s @ 300s intervals NR off

The .arw RAW image files were decompressed in Graphic Converter v6.5 running on a G4MDD PPC MAC OS X 10.3.9. The .arw files were opened in as 6048x4032 unclipped. There was no editing of colour space, resolution or interpolation, or rescaling of EXIF data. The decompressed 32 bit file sizes were all 98,512,128 bytes. The pixel map and histogram of each image was read, the noise count and RGB %'s recorded and transferred to a spreadsheet. The RGB% values were expressed as a fraction of the 24bit raw file allocation & normalized against the sensor pixel count 24,611,840 & plotted on a bar chart.

The red channel is more noisy than the G & B channels.

Using infoLITHIUM NP-FM500H battery during each 7 consecutive exposure batch, noise increased, probably because the sensor was getting warmer. R channel noise increased steadily from 1.10% to 1.60% in the 35 minute 300s sequence, & from 1.60% to 1.90% in the next 70 minutes.

R channel noise increased slightly thereafter until it plateaued at 2.10% during the AC-PW10AM powered sequence. The alternative power modes seem to make little difference. After the noise plateaued the noise in each dark frame was between 1.95% & 2.10%, implying there was little point in averaging multiple dark frames an hour into an imaging session.


Whether the camera was powered by an internal battery pack, a DC IN external psu or a dummy battery psu, noise increased during the dark frame run for 140 minutes, after which it remained constant. The battery may be a source of heat, but it is more likely that it is the Exmor cmos sensor and the Bionz processors that are the heat source.

If you have access to a mains supply the most convenient power source is the AC-PW10AM AC Adapter or a compatible/replacement. If you are reliant on a 12VDC supply from an automobile or power tank, then the proxy dummy battery option is the best option.

Chris Lord

This page was created by SimpleText2Html 1.0.2 on 10-JUN-2010

Top of Page
Return to Home Page