Cometary photometry is not simple and some rules must are important as it has characteristics quite different than classical star photometry.
Comets move in the sky and finding each night proper reference stars for photometry is not obvious. As routine we ask to check in advance the field to where the comet will be that night to look for reference stars in the field of view. Sometimes you need extra images of nearby fields to get reference stars.
The use of filters is a basic requirement and also it is absolutely necessary to pre-process (dark fame subtraction flat field correction) in proper way.
A sequence of images (not single short shots) is then always recommended to increase the S/N.

Of course the ideal solution would be a relatively large telescope at an high mountain site but we experienced that even with small (not too sophisticated) equipments and from suburban sites you can get useful data on bright comets. Of course this can be done in stable clear nights only and relatively long exposures.

Basic information

Here you find some more detailed information about what is needed:

1. Telescope : any telescope is potentially useful for photometry. In the first year of trials the GOC observers contributed with telescopes ranging from 7.5 cm to 40 cm of aperture. Keep in mind that refracting telescopes can have problems of chromatism when coupled with CCDs due to the wide spectral sensitivity. Filters can reduce this defect, but a check about quality must be done.
2. CCD : the main characteristic of a CCD is to have a linear response. Usually CCDs without anti-blooming are linear within 1-3%. CCDs with antiblooming need a test to check how the response is and in what range it is linear. A calibration of the response of the sensor on the other hand is always recommended.
3. Resolution : CCDs with larger pixels are usually better for photometry as they allow a larger full well capacity and a wider dynamic range. The recommended scale on an average is between 1 and 3 arcsec/pixel. An upper useful limit is estimated around 4-5 arcsec/pixel, while the useful lower one depends from seeing.
4. Exposures time has to be set in order to never reach the CCD camera saturation level for both the comet and the reference stars. As a rule is suggested to average many images, aligned on the nucleus, in order to improve the signal to noise ratio. Faint comets need to stack a larger number of images and a longer total integration time.
5. Bright stars superimposed to the coma will lead to unreliable Af results. In this case, if the star is in the outer part of the coma, you can limit the measuring window to the inner unaffected region. Usually, thanks to the high proper motion of comets, one can wait the comet moves far enough from the unwanted star to get the coma free from contamination. For this reason too is always a good practice to check the star-field and comet position in advance.
6. Observing site : ideally we should observe from an high mountain site but on the other hand most amateur astronomers live in suburban sites. We found that experienced observers, using better nights, and with the proper technique, can get good photometry also from suburban sites.
7. Filters : For a general monitoring of comets an R(Cosuins or Bessel) photometric filter is recommended. An I (Cousins or Bessel) filter is also good, specially from suburban sites where it helps to reduce the noise from light pollution and haze. In specific cases and for very bright comets, more selective narrowband filters (647 or 650 nm,10 nm FWHM) provide usually a better result.
8. The observer must have a software allowing to perform simple basic pre-processing procedures (dark frame subtraction, flat field correction, alignment, average and sum of images). The photometric measurements and Af[rho] quantity determination are made by means of a dedicated software (Winafrho) realized for the CARA project. Winafrho is provided for free to all active observers.
9. One crucial step is to obtain the best possible flat field correction. The flat field must be obtained in a well standardized manner in order to grant the highest grade of reproducibility. Sky flat can be done in twilight using a proper procedure [ see for instance http://astronomiadigitale.blogspot.com/ by Martino Nicolini – in italian language]. You can also built a “flat-box” that allow you to get a flat field at any time.
10. Flat field time exposures don’t have to be too short as CCDs with mechanical shutter can introduce systematic effects, also CCDs with the so called “electronic shutter” can introduce noise if a too strong light fall on the sensor during the image download. A time exposure of at least 5 seconds is recommended
11. Flat field exposure should be enough to cover about 60-70% of the dynamic range (CCD useful linear response range), and each filter needs his own flat field.
12. A flat-box can be easily built. A simple cheap solution is for instance a wood box (or a polystirol box if you need a lightweight one), with some leds within and one or more opaline plexiglass screen in front. Other solutions are possible providing they grant an uniform illumination and standard reproducible results. The box size must be larger than the tube of the telescope. Avoid to use uncontrolled variable methods that introduces random noise in the images. The flat correction still add noise to the image, but it is important that the benefit is much greater than the damage. If well done it gives great improvements. The flat field correction is absolutely necessary for photometry.
13. Images must always be pre-processed, i.e. dark frame and flat field corrected.
14. Dark frame must be obtained at the same temperature and exposure time than comet images, and it is recommended to average several dark frames (ideally at least 10 or more) as this reduces noise and improve accuracy. Flat field must be dark frame subtracted and also in this case as above it is recommended to average several flat field images and dark-flat fields (10 or more each).
15. Exposure time for the comparison stars should be possibly greater that 5-10 seconds. In case of very short time exposures (bright reference stars) the atmospheric scintillation becomes an important factor, so it is better to average at many images. For very short exposures verify also the CCD shutter timing on both operations, specially if it is a mechanical one that could introduce systematic errors.
16. Relatively bright reference stars usually should be preferred when possible as they are better monitored for variability, provide an higher signal to noise ratio (higher accuracy on measurements). Furthermore the magnitude and color indexes from the catalogues are usually more accurate for bright stars than for fainter ones. Usually stars between 8-13 magnitude are fine for small telescopes. If needed use different time exposures fot the comet and the stars to avoid saturation.

Usefull links:

http://www.cfa.harvard.edu/icq/icq.html

http://www.observatorij.org/