Moded / Unmoded Nikon D3000 CCD temperature comparison: The red scale shows the CCD sensor heating of an unmoded Nikon D3000. To made the measurements, a NTC temperature probe has been added to the CCD original chassis, near to the CCD but not in direct contact with it, so the real temperature should be a bit more. In two hours the CCD heats up to 25ºC over ambient temperature, if we could made a measure directly to the CCD rear surface probably we could get 30ºC over ambient temperature. This heating should be much similar for all DSLR including Canon models, we can get an idea of the camera power eficiency taking a look to the manufacturer data about battery duration that, in the case of a Nikon D3000 can make 550 continous shots, half of them with flash fired, this is the CIPA standard for DSLR power efficiency. In my test, with an used 7,2V 1000maH battery i made 1h45m of continous exposures of 15m with a rest interval of 5s on a Nikon D40X, that is pretty good. Blue scale shows the temperature drop of the same Nikon D3000 moded and cooled, in both cases, moded and unmoded versions i made continous exposures of 15 minutes with a 5s rest interval between shots at the same time the camera powered and cooled. In two hours the CCD temperature of the moded version is 45ºC lower than the unmoded. The efective cooling when the camera is taking long exposures continously is at least 20ºC below the ambient for indefinitelly time. At less ambient temperature and higher the ambient humidity more temperature drop.
CCD Temperature on camera power off This other graph shows CCD temperature drop with the camera power off and cooling module working 100% of time, in 5 minutes CCD temperature drops 19ºC below ambient temperature, 25ºC below in 15 minutes and 30ºC below  in two hours.
Dark test D3000 With the moded D3000 i did 4 astrophotography typical exposures at ISO 400-800 of 5, 10 and 20 minutes with the CCD temp at -10ºC and other 4 with the CCD temp at 35ºC. In each RAW file has been cropped the same area of 1926 x 1296 px 1:1 and then resized to 600 x 404 px 1:1 To discern how much te noise increase with the CCD temperature i streched the imput levels from 255 to 30 in photoshop, this is that astrophotography software makes during the picture process, the file size atached on each frame refer to the unstreched original RAW file. Seems the noise begings to be much persistent with the CCD at 35ºC on ISO 800 frames and very low on all frames with the CCD at -10ºC. This results are similar on Nikon D40X, D60 and D3000 cause they have the same CCD Sony SuperHad ICX493AQA with 6,0 um pixel size but  are even better on Nikon D40 with higher pixel size Sony SuperHad ICX453AQ (7,8um)  
D3000 Dark test - Imput levels 255 (no change)                            (Click to enlarge)
D3000 Dark test - Imput levels from 255 to 30                           (Click to enlarge)
Amp Glow D3000 The feared “Amp Glow” effect that no modder or manufacturer like to tell about is caused, as Space Telescope Science Institute says, by the IR light emited by sensor read amplifiers. This effect, as far as i know, is present on all imaging devices, on the high priced cameras and in the Hubble Space Telescope:  
Hubble Space Telescope Amplifier Glow (
Nikon D3000 Amp Glow - Original RAW file resized - No changes (Click to enlarge)
Seems the Amp Glow effect has low dependence of ISO and temperature and high dependence of the exposure time, is very similar on Nikon D40X, D60 and D3000 but notoriusly smaller on Nikon D40. I suspect that astronomy dedicated cameras fix this effect adding calibration frames memorized on their hardware automatically but we can too fix it on a moded DSLR easily if we have a precise measurement of the CCD sensor temperature as we will see in the next step, adding software calibration frames give us more freedom to fix the effect.  
Test caption
Amp Glow fix - Light test D3000 This is a single RAW frame no cropped - no resized from a moded Nikon D3000 of 40 minutes exposure at ISO100 taken with a very low ambient light with CCD temp at 0ºC. Once registered and dark substracted on Deep Space Stacker the effect is properly fixed.  
1x 40min ISO100 0ºC Dark RAW Original
Amp Glow Fixed
1x 40min ISO100 0ºC Light test Nikon 18-55 Lens f8
1x 40min ISO100 0ºC Light registered on DSS
Nikon CCD - Canon CMOS unmoded sensor specs comparison (data from
6Mpx CCD ICX453AQ 23,7x15,6mm 7,8um
     Full SNR    (at min ISO)
Color sensivity    (at min ISO)     
Tonal range (at min ISO)   
Dynamic range    (at min ISO)     
Unmoded color response (Before IR filter removing)
     Sensor data
10Mpx CCD ICX493AQA 23,6x15,8mm 6,0um
10Mpx CCD ICX493AQA 23,6x15,8mm 6,0um
10Mpx CCD ICX493AQA 23,6x15,8mm 6,0um
0,47 1 0,66
 Minimum ISO*
ISO 200*
* Real measured ISO vary from manufacturer ISO - Data from * Nikon D40 ISO 100 option unavailable on this model. * Unable to find Canon CMOS sensor model.
ISO 100
ISO 100
ISO 100
0,56 1 0,72
0,56 1 0,72
0,45 1 0,80
This data has been taken for the minimum ISO each camera can shot, on this way we can get a near to true comparison because the market trend is to program a lower ISO than manually selected ISO on the camera, this may be the reason because sometimes we heard someone to say that new DSLRs has less noise than olds at higher ISO. Taking a look to measurements, we can see that an EOS 1000D with ISO1600 selected has a really measured ISO of 1380, the 1D Mark III at ISO 6400 selected has a measured ISO of 3645 and EOS 40D with ISO 3200 selected has a measured ISO of 2160. At minimum ISO the difference between selected ISO and measured ISO is the minimum, as we can see on As we can see in the comparison below, Nikon CCD specs are better than their Canon equivalent models and has better SNR than professional and semi-professional DSLRs like 1D Mark III and EOS 40D
-CCD sensor temperature drop 45ºC or more below unmoded version -Up to 30ºC below ambient temperature when camera power off, -20ºC in 5 minutes,  -25ºC in 15 minutes, -30ºC in two hours. -Up to 20ºC below ambient temperature taking continuous long exposures indefinitely,  -18ºC in 5 minutes, -20ºC in two hours. -Minimum safe selectable temperature -10ºC (adjustable) -CCD temp regulation precision +/- 1ºC of selected -External shutter plug type Canon C1 compatible (audio jack 2,5mm) -Autonomous camera, dont need a PC to work, optionally can be automated via USB  bus with picture autodischarging, DSUSB bulb mode compatible. -Effective anti-dew pasive system:  Camera power off: 12ºC below dew point or better  Camera StandBy:  No dew in any condition -Back focus = 46mm -Power: DC Jack 5,5x2,1mm 12 - 14 VDC, one power plug for cooling module and  camera funtions with battery automatic recharging, no discharging, no overcharging -Consumption at 12 VDC: Cooling module at 100% 2,8A. Camera long exposure 0,2A -Weight: Cooling module: 235g. Camera body: 485g. Battery 50g. TOTAL 770g*
6Mpx CMOS ? 22,7x15,1mm 7,2um
10Mpx CMOS ? 22,2x14,8mm 5,6um
10Mpx CMOS ? 22,0x15mm 5,6um
10Mpx CMOS ? 22,2x14,8mm 5,7um
0,52 1 0,81
ISO 100
ISO 100
ISO 100
ISO 100
0,50 1 0,63
0,48 1 0,67
0,40 1 0,73
10Mpx CMOS ? 28,7x18,7mm 7,2um
ISO 100
0,43 1 0,73
Night Sky Test - M33 Triangulum Galaxy  
Optics: Newton 10” f4 + GPU Aplanatic coma corrector Camera: Nikon D3000 Peltier cooled at -10ºC Integration: Only 3 x 300s Lights ISO400, 1Dark, 1Bias, 1Flat Mount: NEQ6 Pro Syntrek - Unguided - Control with BTDir adapter. Bustillo de Chaves (Valladolid)  19-01-2017 Some village lights.  
Only 3x300s ISO400 (Click to enlarge)
This is not exactly a great picture by process and final result but yes it is attending to de low exposure and integration times (only 3 x 300s lights), the low noise, unguided frames and the street lights of a 50 population village. Where we can get the better results on a cooled DSLR is on long exposure times of 10 minutes and higher. For the next challenge i hope to have the guiding system ready on this Newton 10” f4 to get the full perfomance of this great camera mod of Nikon. A little mistake caused the wrong framed view that it had to be rotated 60º to right instead of 60º to left.