Telescope Reviews: Re: Some rambling, feedback solicited

Groz,

I had the SSII before I got the QSI520. I had 2 SSIIs because I had readout problems with the first one. Those problems did not go away with the replacement so I sent both back. I think the readout problems may be specific to my situation, but another problem I had with the SSII was that it left a streak/shadow behind bright stars. I don't think the linearity was very good either but I confess I didn't measure it. The streak I saw from stars was clearly enough to ruin photometry at the mmag level.

The shortest transiter I know takes ~75 minutes to transit. Tack on an hour before and after for a baseline and you're looking at 3+ hours of data. Now, the question about background subtraction...If I'm awake, I'll run LC without background subtraction to see how things are going. I can see transits in them. The thermal noise in my QSI is uniform except near the edges. Comparative photometry will handle this case very well. The noise is ~equal under all the stars so mostly what is hurt is my S/N. I do have a few light curves where my background subtraction did not remove a few hot pixels properly. As stars drifted over these regions, I saw changes in the light curve that looked like transits but weren't real. There are some details here that need careful consideration but might allow you to cut some corners. For me, flat-fielding is a bigger concern because my stars do drift across the chip a good distance in 3 hours and dust motes are very problematic.

Filter wheels: Sigh...I's like to solve this one too. If all you're interested in is seeing a transit, the sampling rate can be quite low. Take TrES-3 with an ~75 minute transit. a 10 minute sample rate should give ~7 data points during transit--which should be acceptable to identify a transit. From a science standpoint, your measured mid-transit time and duration may not be accurate enough to see the sort of timing wobble that may indicate another planet in the system. For reference, I've measured TReS-3 with a 120s sampling rate and the model fit to my data gave +-6.5min on the duration and +-2.5min on the mid transit time.

Another aspect of filters: 3 hours means the star transits 45 degrees of sky! I usually use an R as extinction is not as great in R but my count rate can still change >50% over the course of the evening! A number of transiters are 10+ magnitude. TrES-3 is 12.4. With my 11 inch scope and an R band filter, I still need 120 seconds to get to the end of the linear region of my chip. That's at F/4 and binning my chip 3x3! My sampling rate is rarely faster than 2 minutes even with a single filter!

I think I'm going to do multiband measurements by putting a piggy back scope on the CPC1100 to measure the second band. My mount tracks well enough--even in alt/az--to keep the target on the chip over the course of a transit. Tight stars are not desirable for photometry--a lot of people defocus to increase S/N--so not using the piggy scope as a guide scope should be OK. Instead, it will acquire a second band--and will be using my DSI ProI to do it so we'll see how well that works!

I have a star analyzer. I haven't tried spectroscopy on a transiter yet, but it will take a long exposure. You're spreading the light over the length of the chip to get maximum resolution. Also, on my system I need the star analyzer about 100mm from the chip to get full resolution. If I put my filters there, I'll vignette something fierce! I have to put them about 25mm from the chip now and that's far enough to vignette at f/4. So there's one of my trade offs. If I image at slow f ratios, the vignetting from my filter position is less, but I don't have as wide a field of view (harder to find comparison stars) exposure times are longer and I risk star trailing and losing the field due to tracking errors. I'm hoping the wedge will alleviate these problems a bit! Since you're EQ already, this might not be as bad for you.

You can answer a lot of questions by setting up one night and imaging a star for a couple of ours. Doesn't matter if there's a transit or not--just see how stable your equipment is over the course of a few hours. You're looking for a flat "curve" with as little scatter in the data as possible. I spent many evenings doing this before I even attempted to measure an exoplanet transit. I've found that you can recognize a transit even if the depth of the transit is less than half the scatter you're seeing. I've measured an 11mmag dip with a scatter of ~24mmag. The curve isn't pretty and wouldn't look good in a journal--but I can clearly see the transit!

I put enough words in there that hopefully something is useful to ya

Brian

--------------------
CPC 1100
Orion 120st
Milburn Wedge
QSI 520i
Meade DSI Pro

Post Extras