Best op-amp and component choices for simple DIY solar tracker circuit design

There are many cheap eBay solar trackers online.  But, I've decided to attempt to diy one myself with components I have laying around.  However I'll need to purchase an opamp to act as my comparator and I'm not sure what to buy or what else I'll need.http://imgur.com/UFiFyDS
Here's a link to the circuit design.  Any suggestions would be very appreciated.  Especially when it comes to which specific components would work best and or changes to the design.  I built it intentionally as simple as I could with the goal in mind to keep it as cheap and easy to build as possible.The  solar panels are represented by the ac and DC power sources on the left of the image.  
Jon

Hi Jon-I take it that the solar panels connected to the comparator are not the main solar panels, but are maybe just one or two individual cells.  Is that correct?  What is the maximum voltage they could produce in direct sunlight?At the end of the day, how do you plan to get the tracker back to the starting position? How big is the motor?If you can provide a little more info I'll try to offer some suggestions.-Rick

Rick you are correct.  The solar panels will be two small individual panels in a slightly angled V or I'll place them flat with a metal separator in the middle.
The voltage in the small solar panels is set for 5v maximum in this example but I suspect I could just as easily use 1v solar panels.
As for getting the system to reset.  The assembly will be mounted on the East side of the larger panels.  I live in between mountains so the initial sun is fairly high in the horizon.  If I use a reflective separator it should reset at or shortly after sunrise,
I'll fiddle with the stop switches to ensure a good early morning reset.
As for the motors.  I salvaged some cheap rs445pd motors.  They're 12-24vdc. 3800-7600rpm.  I'll have to do quite the reduction to make them have enough torque but this setup is a prototype for a larger system I hope to build later.

OK- I'll try to do a quick sketch along with part numbers when I get home tonight.  Any preference between surface mount or through-hole for the comparator chip?
-Rick

Rick, surface mount would be easiest.  Please and thank you.  

I meant to say through hole.  Hah

Here's your circuit- cheap, simple, old school.
Solar tracker schematic
This is how it works: Starting from the left, I used low-pass filters to get rid of any noise (or spikes caused by lightning flashes) on the signals from the two solar cells.  I fed those two signals into an LM211 or LM311 comparator.  I chose that comparator mainly because it's cheap, readily available, and will work at up to 36 volts.  In this case we really don't care that much about speed, offset, or bias current. The 1 meg resistor adds a bit of positive feedback so the comparator will have some hysteresis.  This makes sure that it will "snap" on or off and not chatter around in between.
The motor you mentioned doesn't draw much current, and Darlington power transistors are cheap, so I did away with the relay. That transistor will pass up to 8 amps.  If you plan on drawing much more than half an amp I'd recommend adding a small heat sink to the transistor.  You'll notice that just to the right of the DC input there are two capacitors in parallel. This is mainly to deal with the noise that all brushed motors generate. Make sure to locate the .1uF capacitor as close to the comparator chip as you can.
One last thing- If you find that the motor turns on while the solar cells are dark, you may have to add a resistor between pin 5 (balance) and either supply or ground.  You'll have to experiment with this- keep the resistor value as high as possible while still assuring that the motor does not run unintentionally.
-Rick 

Rick,  thanks for the help.  I'll order everything and build a mock up.   I'll make a video and submit the link in a few weeks.  
Jon

I look forward to seeing the video.  
I think this will be a quick and easy way to get a "test bed" going, but I'm thinking the best long-term solution will be to use an 8-bit  microcontroller like a PIC18F1220 (less than a buck), and add bidirectional motor control (easy with a single motor driver chip, under a buck), and limit switches. In my opinion, that's what it will take to make sure the motor moves the panels back to the starting position every night.  After you do some initial testing using the sketch I provided, if you want to try the microcontroller approach, just let me know- I'll be happy to help.-Rick

Rick, will do.  It sounds like an interesting idea.