Modifying RGB Filters for Improved Autoguiding

by Allan Guthmiller

Have you ever had difficulty locating a guide star that is bright enough to penetrate through the RGB filters, allowing your guiding system to operate properly? If so, this tutorial is for you. I own a 20” f/5 Newtonian and a SBIG ST-2000XM with a standard CFW-8. Locating a suitable guide star is not much of a problem though the 20”, particularly with the clear filter. However, this radically changes when the Red, Green, or Blue filter is positioned over the guide chip. Instantly this wonderful looking guide star virtually disappears, causing the ADU readout to drop well below minimum operating conditions. In many cases, the guide integration time has to be substantially increase, sometimes as much as 20 seconds, in order to bring the ADU back to a workable level. After struggling with this on several occasions, I came up with a “gutsy” solution:

Why not cut a hole through the filter directly over the guide chip to allow unfiltered light to fall upon it?

After exploring various options to accomplish this, including chemically erasing or buffing off the coating with polishing compounds, the best approach was to simply cut a notch into the filter, extending over the guide chip. This was a bit tricky because I didn't want to cut too deeply as to interfere with the filter light over the main imaging chip. There was only about a 1.5 mm gap between the two to work with. The best way to define theses dimensions was to obtain a diagram of the camera from the manufacturer before cutting the filters. In this tutorial the ST-2000XM used as the example.

The first step is to remove the lid over the filter wheel assembly and lift out the filter carousel. Be sure to place all the tiny screws and other parts into a small container. I recommend that you wear photographic cotton gloves while handling the filters. These can be purchased at your local photo supply store, and will help prevent finger marks. Unscrew the ringlet and remove the filter.

Be extremely careful not to allow the screwdriver to slip. It could cost you $169.00 for a new filter!

Sandwich the filter between a layer of lens paper and blocks of balsawood. The balsa blocks will act as a cushion to help prevent the filter from cracking. Secure in a vise or holding device. Do not apply too much pressure, only enough to keep the filter from moving.

Get Ready To Grind!

I use a handheld, variable speed, multipurpose drill tool with a special diamond coated grinding bit. This can be purchased at your local hardware or hobby store. Before proceeding, I recommend that you wear protective gear while cutting glass. Creating a slot can be somewhat nerve wracking, so take your time and go slow. Set the grinding tool speed to medium-low for best performance. Do not operate at high RPM. This will cause large fragments of glass to chip off and may damage the filter. Take your time making small cuts by applying light pressure onto the filter for about 2-3 seconds. Lift the bit away from the filter and dip into a cup of water. This will cool the diamonds and remove excess glass debris. It is very important to never allow the diamond bit or the glass to get too hot. It could ruin the diamonds and damage the filter. Periodically measure your progress by checking the diagram as a guide. It takes about 30 minutes to complete each filter.

Go for it!

Remove the filter from the vise and clean with water. Pat the filter dry (never rub) with generic toilet tissue. Place the filter back into the holder and secure with the ringlet. Do not tighten at this time. Wait until all the filters are completed, align in the carousel, then tighten. Place the carousel back into the camera to check if enough glass was removed over the guide chip. Center your eye over the guide chip and check to see if the entire guide chip is visible. If it is not, remove the filter and grind a bit more. Do not get too aggressive. You don't want to cut too deep as to interfere with the main imaging chip. If indeed the filter was cut too deeply, there is some saving grace. I noticed there is about 1mm of slop within the filter holder that may make the different between forking out $169 for a new filter or not.

The notched filters work great for the guide chip, however, there is considerable amount of diffraction onto the imaging chip, even though the notch does not overextend. This is partly due to the convergence of the focus. This defect could not be completely removed by flat fields. The image to the left is a 15 minute exposure of NGC 4725 using the green filter. Notice the halo at the upper center.

To correct this problem I developed a collimating baffle made of black paper, which blocks the diffraction onto the main chip. The wings of the baffle are glued onto the non-coated surface of the filter using cyanoacrylate glue (super glue), which can be purchased at your local hobby store. I recommend medium viscosity.

The adjacent photo shows the filter wheel upside down demonstrating the collimating baffles. It should be noted that the baffles do not extend beyond the filter holder. If so, they could brush on the back plate of the CFW assembly and inhibit the carousel wheel from rotating. Different cameras may vary.

The image at the left is a 15 minute exposure of M74 taken with the green filter using the collimating baffle. Notice the baffle removed the halo at the upper center. In its place is a slight dark gradient. This was easily removed by applying a flat field image.

This is the combined result. Now the image appears more homogenous across the field and any hidden data surfaces.

This is a LRGB image of M74 resulting from the described filter modification technique. The exposure times are: L=60, R=30, G=15, and B=15 minutes. RGB's were binned at 2x2.

Another finding involved the guide star itself. With the glass removed, the focus had slightly shifted. This did not seem to effect the overall guiding performance, but did reduce the ADU output by about 50%. This is okay, but might become an issue if the guide star is faint to begin with. To correct this I glued a 1mm thick piece of clear Lexan window over the notch. This redirected the focus back to normal and eliminated the drop in ADU.

The results can be seen on the adjacent photo. These are images taken from the guide chip. Frame # 1 demonstrates a typical star through the clear filter. Frame #2 is the same star through the “notched” green filter without the Lexan window. Frame #3 shows the same star through the green filter with the Lexan window installed. Note the focus is back to normal.

I have photographed several objects utilizing this filter mod. Included on this page are a few samples.

I cannot express enough to the reader how pleasing it is to choose a faint guide star and never have to increase the integration time to compensate for the RGB filters. If I can see it, I can guide on it. My exposures are kept short-n-sweet, and never will I have to be concerned about potential guide errors due to increased times. I hope that this tutorial will be extremely beneficial to all, especially those with smaller instruments, where locating a suitable guide might be difficult. I also believe that this filter mod will be beneficial for AO7 users as well. Have fun and enjoy!