Spectroscopy & Spectrography

Comparison of spectra of 12V Compact Fluorescent Light Bulbs

The 12V Compact Fluorescent Light Bulbs I have purchased come in two flavours – 2700K and 6400K (later is daylight). Tonight, I took spectra of both types using my CCDSPEC spectrometer to determine whether the peaks shown on the spectra were the same. It turns out that they are and hence can be readily used for calibration purposes on my spectrometers. The main difference between the two is the intensity of the peaks with the 2700K light bulb having much lower intensities than the 6400K light bulbs.

Information on different colour temperatures of compact fluorescent bulbs can be found at http://www.lamptech.co.uk/Documents/FL%20Colours.htm


2700K spectrum:

6400K spectrum:

Re-analysis using RSPEC Software of Vega spectrum from 4/8/2018

This is a re-analysis using RSPEC software of the Vega spectrum I took 4/8/2018 using my Sky Watcher Equinox 80mm Pro telescope & CCDSPEC spectroscope, hand guided on Manfrotto mount.


CCDSPEC spectrum of Vega 4/8/2018

I have now taken out a trial subscription to RSPEC amateur astronomy spectroscopy software https://www.rspec-astro.com/

The following is the original spectrum taken using Nebulosity 6 software:

This is an x-projection of the above from PCSpectra software supplied with the CCDSPEC spectrometer:

The following screenshot shows the above spectrum loaded and displayed in RSPEC software (below):

In the above diagram, I calibrated the x-axis using a Vega spectrum from the internet, using two points. I have created the following Vega spectrum for calibration purposes based on this internet data. Note that I have annotated the lines using Angstroms as RSPEC uses Angstroms on its x-axis and it all gets rather confusing if one graph is in Angstroms and the other in nanometres as it is not obvious that the difference in scale exists on the x-axis and the graphs do not line up.

RSPEC allows reference spectra to be compared to user spectra. Here, I have loaded up a reference spectrum in blue at the same time as my own Vega spectrum in red (below):

Having calibrated my spectrum against two lines on the reference spectrum I created above from other internet resources, there is excellent correlation between the lines on my spectrum and the RSPEC reference spectrum (no – I did not use the RSPEC spectrum to create my calibration Vega spectrum!)

Another interesting feature in RSPEC is to use it to identify elemental lines – using this feature on my spectrum:

I think that the closest elemental match to the lines seen in my graph is found when both Star Type A is ticked (Vega is A0V) and also Hydrogen Balmer series:

The Compact Fluorescent Bulb Spectrum as a standard calibration spectrum for low resolution spectrometers in astronomy, and comparisons with alternative calibration light standards

I have annotated the following compact fluorescent light spectrum with the wavelengths of the main peaks, in order to make the diagram most useful for spectrometer calibration purposes. I am indebted to Wikipedia for the source information on which this graph is based.


The above is useful for calibrating my CCDSPEC and Science Surplus DIY Spectrometers and other similar spectrometers.

The process involves me taking a spectrum and then identifying the lines on it and calibrating the spectrometer using the process in the relevant software package. This is a spectrum from the Science Surplus DIY Spectrometer of a Compact Fluorescent Light:

The compact fluorescent spectrum can be used to help identify lines on other calibration lights, such as one made from a RELCO neon fluorescent bulb starter:

Further enhancement of calibration can be achieved using alternative calibration standards such as the solar spectrum – here I have annotated that spectrum with the Fraunhofer elemental lines:

Another alternative to calibration standard lights is to use an LED light – this one provides lines in red, green and blue, and is from on a variable colour strip light I purchased from ebay:

Compare the above to a commercially sold white LED calibration light:


Three LED lamp calibration light

I purchased the following multi-coloured LED strip light from ebay – Ed at the last RAG meeting asked whether coloured LEDs could be used as a calibration light for spectrometry – it seemed like a good idea! This one comes with a neat little remote control to allow it to be placed some distance away (ideal for focusing a telescope on it and then turning on and off and changing colours).


I made a calibration light out of the strip light above and a piece of Perspex made to diffuse light (off ebay designed for light boxes). I could then take spectra using my CCDSPEC spectrometer:

The first three spectra show that red, green, and blue are all very pure colours from the relevant LEDs in the strip:

Spectrum of LED strip light – green LED (below):

Spectrum of LED strip light – red LED (below):

Spectrum of LED strip light – blue LED (below):

Producing multiple lines on a single spectrum:

The above graphs demonstrate that the LEDs individually produce pure colours. However, for calibration purposes, it is useful to be able to have multiple lines on a single spectrum. This is where this particular colour-changing LED strip is useful – the strip uses three different coloured LEDs as above but allows you to select various combinations of LEDs to produce variety of “colours” – each of this is a mixture of LEDs switched on together in varying amounts.

The following are three examples of combinations – there are a total of 16 colours available (16 combinations):

Spectra from the three examples of colour combinations:

X-projections from spectra from the three examples of colour combinations:

White colour on the LED strip:

Probably the most useful combination is when all three LEDs are turned on together “white” – donated by a convenient “W” button. The spectrum from this produces three convenient peaks for calibration (below):


Comparing spectra from compact fluorescent lights in lounge, portable 230V desk lamp and portable 12V calibration light

In previous posts, I have discussed the various calibration lights I have tried for my spectrometers. Each time I come back ultimately to fluorescent lights – these have typical spectra and easy to identify peaks and are ideal for relatively low resolution spectrometers such as my CCDSPEC and Science Surplus DIY Spectrometers.

They are available as long bulbs or more compact portable bulbs called Compact Fluorescent Lights (CFL). The spectra are similar.

Initially I used a compact fluorescent light in my study/lounge but then realised I could do with a portable arrangement to take outside into the field. I obtained a simple desk lamp and used 230V CFL bulbs powered using an inverter on my leisure battery. This is not the safest method – 230V can cause harm – but then I found a caravan online shop selling off its last 12V CFL bulbs. I did not realise these were sold in 12V varieties so I have obtained a number of these and built myself a lamp holder than plugs into a 12V cigar-type lighter plug socket so it can be powered from any standard 12V telescope power supply.

In today’s post, I am comparing the spectra on the three sources of CFL bulbs – the ceiling lights in my lounge, 240V desk lamp and homemade 12V mobile calibration system. The question I need to answer is whether all three produce similar spectra – in which case the 12V homemade setup will be my preferred calibration light as it is mobile and safe.

I used my CCDSPEC spectrometer for today’s tests.


Taking spectrum from 240V desk lamp CFL arrangement (below):


Taking spectrum from 12V desk lamp CFL arrangement (below):


Comparing the three spectra:

Spectrum from 240V CFL ceiling light (below):

Spectrum from 240V CFL desk lamp (below):

Spectrum from 12V CFL homemade lamp (below – here I have altered the exposure to give longer and shorter exposures to show that this made minimal difference to the spectrum as long as it did not reach maximum (16000 on intensity at which point peaks broadened and became less useful for calibration as exact wavelengths difficult to read from graph)):


To assist in comparison between the three spectra, I have repeated them below but this time one after the other:

It is amazing how similar these bulbs are – from different manufacturers but still providing same peaks with slight differences only in intensity. It shows that the similar chemical makeup of their constituent gases and that the emission spectra of those gases do not vary.

My conclusion – any of these bulbs can be used for calibration of my spectroscopes and hence I can happily use the 12V outside in the field.

Spectra taken with CCDSPEC Spectrometer @ Rosliston Forestry Centre RAG meeting 14/9/2018

I demonstrated my CCDSPEC spectrometer at the mid-monthly RAG meeting at the Forestry Centre tonight.


CCDSPEC Spectrometer:

The following spectra were all taken using the CCDSPEC spectrometer during the RAG meeting. The spectra are direct from the CCDSPEC spectrometer without telescope or other optical aid.

I started by taking a spectrum of one of the long fluorescent ceiling bulbs in the building (below):

For comparison the following is a spectrum I took from a compact fluorescent bulb in our study on 3/8/2018 (below). The same lines are clearly evident:

The following is another comparison spectrum from Wikipedia – I have labelled it with the wavelengths of the main lines:

I then took a spectrum of my calibration light made from the neon bulb in a RELCO starter bulb:

We also compared the above to a spectrum from the LED projector at the forestry centre (below):

Ed Mann came up with the idea that lasers might work well as calibration lights.

He had two laser pens with him – blue and green – so we tried taking a spectrum of those laser pens reflected off a white piece of paper (below):

First attempt at calibrating RELCO starter spectrum against solar spectrum

This post is a follow up to this previous one:

Calibrating spectrometers – using solar spectrum as gold standard

Following is solar spectrum recorded today (cloudy sky, CCDSPEC):

I went through process of calibrating it against solar Fraunhofer lines – so following has wavelength on x-axis post-calibration:

I then took a spectrum of the RELCO starter lamp using the CCDSPEC (same set up as above):

And used the solar calibration file as a calibration file for the RELCO starter spectrum to give the following calibrated image – of course any errors in the solar calibration will now have also affected the calibration of this image…..

My next step is to compare the wavelengths of the lines on the spectrum above to those listed at http://www.ursusmajor.ch/downloads/sques-relco-sc480-calibration-lines-5.0.pdf – hopefully, even if my spectrum is out by a couple of namometers in places, I will be able to identify the lines in my spectrum and hence their precise frequencies – which would then allow me to recalibrate the spectrum but this time against the correct frequencies for the RELCO starter bulb.


Calibrating spectrometers – using solar spectrum as gold standard

I have described previously how I made or obtained calibration light sources using compact fluorescent bulbs, LEDs and RELCO starter bulbs – however in each case I need to be able to identify accurately which lines are present on the spectra from those light sources before I can effectively use them for calibrating my spectrometers.

For some – such as compact fluorescent bulbs – information on the main lines is accessible online.

However, the RELCO starter bulbs has turned out to be more difficult. I have managed to identify a few lines only so far:

First attempt at determing frequencies of lines on my spectrum of RELCO Glow starter lamp taken with Science Surplus DIY Spectrometer

In this post, I will describe my attempt to use to the solar spectrum to identify further lines in the RELCO starter spectrum.

The solar spectrum has well defined lines (Fraunhofer lines) and detailed information on these lines and their frequencies is readily accessible online. It is very easy to obtain a solar spectrum during the day by pointing a spectrometer at blue or even cloudy sky (do not point directly at the sun).

I have therefore had a go at preparing a typical solar spectrum in graphical format similar to that generated by my spectrometer software and labelling it with main Fraunhofer lines, in order to use this to help me determine what the wavelengths are of lines on my light sources.

The idea is that I can then use blue sky (I can do this easily during daytime) to obtain a solar spectrum with my spectrometers and then use same set up for obtaining a spectrum of the RELCO and LED light sources so that I can compare the two and use the graph below to determine the wavelengths.

Hopefully this will work!


The following will take you to what followed on from the above – my first attempt to calibrate my homemade REKCO starter based calibration light using the solar spectrum:

First attempt at calibrating RELCO starter spectrum against solar spectrum

First attempt at determing frequencies of lines on my spectrum of RELCO Glow starter lamp taken with Science Surplus DIY Spectrometer

This post follows from this post in which I took the spectrum used in the analysis below:

First calibration spectra from RELCO neon starter bulb (RECLO SC 480 S53993) taken with Science Surplus DIY Spectrometer 9/9/2018

I compared the spectrum I took using my Science Surplus DIY Spectrometer of my modified RELCO SC 480 starter bulb calibration light with the spectrum found in http://www.ursusmajor.ch/downloads/sques-relco-sc480-calibration-lines-5.0.pdf

I think I have managed to identify a few of the lines.

The spectra in the image below from top to bottom are:

I have attempted to link matching lines together across the spectra.


Although I have had some success identifying a few lines above, I decided to see if I could use the Fraunhofer lines in the solar spectrum to help me identify others.

In the following post, I have tried to do this:

Calibrating spectrometers – using solar spectrum as gold standard