Addresses of grating companies:

China Daheng Corp.
Edmund Scientific
Richardson Grating Labs (now called ThermoRGL)
Diffracton Products, Inc; Dr. Edward Leibhardt, President, 9416 West Bull Vally Rd.,P.O. Box 1030
Woodstock, IL 60098, USA; Phone: (815)338-6768; Fax: (815)338-7167

Comments by Fred Veio on buying diffraction gratings

Amateurs who look at this web site are likely to eventually need to buy a grating for a SHS. But they must buy a 90% theoretical resolution grating in order for the SHS to work well. They must be warned that a good grating will cost about $350 or more for size 32x30mm ruled area with 1200 gr-mm, 5000A (500nm, green) blazed is adequate, even 6000A (600nm, orange red) or 7500A(750am, red) are adequate. You should give nanometers, nm, and angstroms.  Catalogs about 20 years ago used angstroms, but recently they use nanometers, which can be confusing to a beginner.

Selection of a plane grating for a spectrohelioscope: There are a few ways to judge the quality of a replica grating for a spectrohelioscope. You need a 90% theoretical resolution for the grating so that the SHS performs properly. A 45% to 50% resolution grating is never recommended, the latter replicas are about half the price of a 90% grating .For example, a 90% resolution grating of 32x30mm ruled area with 1200 gr/mm will cost about $350 or more. A 50% resolution grating  for the same area and grooves/mm will cost about $170 or less.

Thickness of the grating: A 50% resolution grating will be 5mm to 10mm thick for 25x25mm to 50x50mm ruled area. A 90% resolution grating will be 10mm thick for 25x25mm to50x50mm ruled area. A larger grating will be 16mm thick.

What glass is used: Float glass will be for 50% resolution gratings. BK7 and pyrex will be for 90% resolution gratings.

Price of the grating compared to a known high quality 90% resolution grating: A 50% resolution grating will be about half the price of a 90% grating. The price itself will be a big important factor in evaluating the grating.

How the grating was made: A 50% resolution grating can be part of a larger replica.The latter is cut sawed into smaller pieces. A 90% grating is made individually from a master grating, never cut sawed.

How is the replica made?:  First, the master grating is put in a vacuum chamber. A very thin layer of oil is sprayed over the surface. This is the parting agent. Then the master grating is coated with an aluminum coating. The master grating with Al coating is taken out of the vacuum chamber. Now a layer of epoxy is put over the Al coating, a glass substrate is put on top of the epoxy,  all is allowed to cure for about 20 hours.Finally the master grating and replica are separated. The replica will be exactly equal to the master grating if proper methods are employed.


The passband for the  H alpha line should be about 0.6A wide for best contrast to the solar disk detail. This means that the grating must resolve about two to three times better than 0.6A, namely about 0.2 to 0.3A. The resolution of a grating is calculated by adding up all the lines and dividing the latter into the wavelength used. So a 32x30mm ruled area with 1200 gr/mm is about 38,000 total lines, and that figure divided into 0.6A will yield about 0.2A in the first order. It is that simple. 

For a SHS and a powerful solar spectroscope, use a 90% theoretical resolution grating. For an average solar spectroscope, one can compromise with a 50% resolution grating, save much money, and still have a good view of the solar spectrum.

Master gratings used by a company are 90% theoretical resolution or better. If the grating catalog of the company states that the replica is made from a master, that does not mean that the replica is equal to the master in quality. In other words, if the master grating is 90% resolution, a high quality replica will be equal to the master, namely 90% or better. And an average quality  replica will be about 45% resolution due to the method used.  The wording is the catalog can be vague.

Reflection Grating Efficiency Curves;    Fredrick N. Veio, 2002
 The following values were taken from reflection grating charts on the web site of the Richardson Grating Laboratory. It is now called ThermoRGL. Littrow test condition. H alpha line is 6563A wavelength. Solar spectrum has visible colors of violet at about 4000A wavelength, blue at 4500A, green at 5200A, yellow at 5900A, orange-red at 6500A, and red beyond 7500A wavelength.
  The violet CaII lines are H at 3934A and K  at 3968A wavelength, blue H beta at 4861A, green Mg of b1, b2, b4 at 5184A, 5173A, 5167A respectively, green Mg b3 is two close Fe lines at 5169A, two D 2 and D 1 yellow Na at 5190A and 5196A, D 3 yellow He at 5176A, orange-red (red for short) H alpha at 6563A.
 A 600 gr/mm grating has orders of 1, 2, 3, 4, 5, 6, 7 and violet of the 8th (80 deg angle).
A 1200 gr/mm grating has orders of 1, 2, 3 and violet of the 4th (80 deg angle).
An 1800 gr/mm grating has a full visible first order and violet-blue-green of the second.
A 2100 gr/mm grating has a full visible first order of violet-blue-green-yellow-orange red. 
The following are examples for comparison in regards to blazed wavelength. It is not too critical in most cases what blazed wavelength to employ. A grating blazed at 4000A wavelength is bright for violet, a bit less for the rest of the spectrum. A grating blazed at 5000A is bright for green but reasonably good for the rest of the spectrum. A grating blazed at 6000A is bright for the orange-red. A grating blazed at one micron is bright for the  red but still good from the  orange-red to the deep red, not so good for the violet-blue. Grating blazed for 4000A, 5000A and 6000A are the usual choice, but the one micron blazed grating has special characteristics of usage. See below. 
First order will be used almost all the time for the spectrohelioscope mode. But in the spectroscope mode, you can use the first up to the fifth order of a 600 gr/mm grating as most practical; use first and second orders for a 1200 gr/mm grating; use first and part of the second order for an 1800 gr/mm grating. Higher orders give more resolving power by the grating and produce a much longer solar spectrum, thereby spreading out lengthwise the spectrum in order to see finer spectral details.
 Here are several gratings for comparison.
 1200 gr/mm           4000A   78%            average in first order about 60%  
4000A blazed        5000A    59                            
first order               6000A   38                                
                              7000A     47            
1200 gr/mm            4000A   70%           average in first order about 70%
5000A blazed         5000A   82            
first order                6000A   71              average in second order about 10%
                               7000A     63            
1200 gr/mm            4000A   40%           average in first order about 70%
6000A blazed   5000A   72            
first order                6000A   77              average in second order about 20%
                               7000A   68            
1200 gr/mm             4000A   10%          average first order about 60%  
one micron blaz.    5000A    32            
first order                6000A    60            
                               7000A    58             average second order  from 4000A
                               8000A    60             7000A is about 50%    
                               9000A    67            
                             10000A    68            
1800 gr/mm             4000A   60%            average first order 65%    
5000A blazed          5000A   75            
first order                 6000A 68            
                                7000A   64            
2160 gr/mm             4000A   57%            average first order about 56%
5000A blazed          5000A   58            
first order                 6000A   59            
                                7000A   56            
 The point to all the above presentation is that all is not critical about a certain grating just for the H alpha line. One does not have to have exactly this or that blazed wavelength. There is some leeway here and there.
 There is NOT a lot of sun light that enters the human eye in the spectrohelioscope mode. The sun image is focused by the telescope on the entrance slit, but the slit is very narrow.About 99% of the sun light does not enter the entrance slit. The remaining one per cent goes to the spectroscope lens, bounces off the grating as a spectrum, back through the spec lens and to the exit slit. The spectrum is greatly spread lengthwise. Only a narrow section of the solar spectrum passes through the exit slit at a greatly reduced level, which is very safe to the human eye. The spectroscope mode is just as safe as the spectrohelioscope mode.
March 15, 2002
Edmund Scientific sells holographic transmission grating in the form of a thick film. Sheets are about 150x 300mm, if I remember correctly. Each sheet costs about $10 USA. They have two types, namely about 25,000 lines/inch and about 14,000 lines/inch. You can cut a small section about 30x30mm and mount in a cardboard slide of about 50x50mm. Hold up to a light source and see a rainbow of light. Do not hold up to the sun in the sky. You can have a white card,or newspaper, on the ground. Let sun light reflect off the paper and into the grating.Or use a cloud in the sky.
For a simple spectroscope, use a cardboard box about 150x150mm or larger. Or a piece of cardboard tubing about 150 to 200mm length. For best visual results, you must have a clean sharp slit made out of two razor blades, or blades from a pencil sharpener.The clean slit must be about 10 to 25microns wide in order to have some easily visible lines in the solar spectrum, or other light source in the home or out in the street that you look at. Never make a slit just out of paper. It is impossible to get a decent slit that way. Maybe you get a few crude, visible lines seen, but never much beyong that. Try to make a clean slit, does not have to be extreemely sharp, can be dull sharp, not that critical at all.
 You mount two blades on a piece of metal, a piece of wood, thick cardboard, whatever.Have one blade fixed with tape. The other blade is held in place with a piece of tape. Put the mounted blades up to a light bulb. Look through the slit towards the light.. Adjust the width of the slit by finger pressure. When the slit is barely visible, this is about 10 microns wide. When the slit is easily visible, this is about 25 microns wide. Do not make the slit wider than 25 microns.You do not need exactly this or that, not that critical, just get it close is ok.
 Remember that the strong lines in the solar spectrum are thick and wide, and they will be easy to see. So the 25 microns slilt width will be ok.  The fainter lines are narrower and will be more difficult to see because the slit width of about 25 microns is a bit too wide. So close the slit to about ten microns and the fainter solar spectrum lines will sharpen up somewhat.
All the above is for a transmisson spectroscope of basic needs. You can buy a small reflecton grating about 12x25mm from Edmund. Get  1200 gr/mm at least, preferably 1800 gr/mm. Have a small box about 200x200x200mm placed on the table. Cut two separate holes on tope of the box. Put the grating inside on the bottom of the box. One hole  ff for the mounted slits as discussed above. The other hole off to the side is for the eye. The sun light via a white card, or home light source, will pass through the slit, down to the grating, reflect off the grating and out to the second hole into the eye.You will see spectral lines. With a home florescent light, you will see emisson lines.
The two above basic spectroscopes are easy to build. For much better spectral detail, buy some short f.l. achromats, the longer the better, because the spectrum is stretched out longer, spreading out the solar detail. So about 200mm f.l. achromats is ok for a start. But about 500mm f.l. achromats will be better.The two separate achromats mounted in separate tubing eliminates a problem of stray light. The whole spectroscope does not have to be 100% light tight, about 99% light tight is ok.The f.l. of the eyepiece should be about 30mm to 40mm. A high powered eyepiece, say about 6mm f.l., does not work to best advantage. Always remember that the spectral lines are a reflection of the slit itself. My 119 page book on spectrohelioscopes had discussion in the back pages.
Cheers, Fred Veio