FocusStar Controller

FocusStar Controller

Introduction

I am a big believer in using the computer to achieve focus for astrophotography. This is likely a result of my complete inability to manually find focus – I always think there may be a better position if I try some more, and am nagged by the thought I may have missed the optimum spot. I love the FocusMax approach of reaching focus, although it took quite a bit of practice to get consistent results.

I have used an Optec TCF-S focuser for several years with both a Celestron Nexstar 11” and a Meade 12” Classic; I highly recommend the Optec for durability, convenience, and accurate results. I have also used Meade’s microfocuser on the 12” Classic with reasonable results, although it does flaky things occasionally. I also routinely use a Robofocus on the Meade for rough focus; I have been somewhat disappointed by it due to its slow speed and large backlash.

Unfortunately, my beautiful Optec does not work on my Tak Sky90; when it is in the optics train I cannot get my camera (SBig 2000) to come to focus. So, I have been looking for a reasonably cheap alternative focuser for awhile. I thought I had a great plan with the Shoestring Astronomy FCUSB product paired with the Orion focuser motor. Unfortunately, while the FCUSB operated great, the motor did not have the power to lift the camera. It would probably work well with DSI/LPI cameras, but I might expect it to have trouble even with DSLRs.

Then, I ran across the FocusStar product and decided to give it a try. This review describes my results. I should point out that my overall goal is to perform astrophotography using CCDAutopilot (which I have just started to use). Thus, I am looking for the high reliability needed to allow automated operation, rather than accepting having to futz with things all the time.

Product Features

The FocusStar comes in a solid control unit, and includes a hefty stepper motor (not the small Orion type of DC motor). The motor is significantly larger than the Robofocus unit. Some features that appealed to me:

The motor is an absolute position stepper motor, not a DC pulse motor.
The software supports ASCOM, so I can run it under FocusMax.
The unit supports temperature compensation, although I did not purchase the temperature probe (and did not test this feature).
The unit supports two motors, a main and auxiliary. Again, I did not test this capability although I could see it being very useful down the road.
The unit communicates with the computer via USB, which is simpler to deal with (compared to RS232 devices) on current computers.

Mounting the Motor

FocusStar supplies brackets ($100) for a couple of common telescopes, but not for the Tak90. I am not particularly good at building hardware, I eventually managed to cobble together a setup that takes advantage of the tapped holes in the Losmandy rings I use to mount the Tak (see image).



The motor is mounted to a thin aluminum plate, which attaches to the smaller wood block on the left. This block rotates around a pivot bolt into the main wood block, allowing the motor position be adjusted to mate up with the focuser knob. This also allows the motor to be disengaged from the coupling.



One useful feature I managed to implement was a neat coupling between the motor and the focuser shaft. An annoying problem is that once the motor is attached to the focuser shaft you cannot manually turn the focuser; the motor is locked in place. The only way to move the motor is through the controller, either using the buttons on the box or using software. This can be a pain, since sometimes you want to do a quick adjustment to get in the approximate range desired. It can take a long time for the motor to turn through the entire focuser range! This coupling allows me to disengage the motor by pivoting it away from the scope, move the focuser by hand, then reengage the motor into the coupling. The small images show the inexpensive coupling I used (obtained from McMaster Carr). In the upper shot the coupling is partially assembled, showing how the three pieces slide together. The bottom view shows them all together. You can order different hole sizes in the end pieces to match the motor and focuser shafts. It is not as easy as I had hoped to re-engage the motor, but it can be done. The coupling seems to have a surprisingly small amount of backlash.

I originally attached the motor to the primary focuser shaft. This resulted in a focusing “sweet spot” of approximately 20-30 steps. This is probably usable, but I was concerned that precision might be a problem. Instead, the image shows the motor attached to the fine adjustment (10:1) knob of a FeatherTouch focuser. Now the sweet spot is 200-300 pulses, so I am comfortable that I can be fairly precise. Of course, using the fine adjustment knob has the drawback of taking roughly forever to move the focuser a large distance, another reason for wanting the decoupling feature on the motor.

General Operation

I first installed the software and tested the motor at home on my laptop running Vista. The initial installation gave no errors, but did not work correctly. The control program could not connect to the USB controller. I found the vendor supplying the USB drivers, downloaded their drivers, and re-installed them. This seemed to fix the problem. It may be that re-installing the FocusStar supplied drivers would have worked; the version numbers on the drivers I downloaded seemed to be the same.

The control unit has a small display that can be used to set some menu options; I did not use this, preferring to set everything from the software. The display does not show the current motor position as I would expect. When the motor is being pulsed from the computer there is a flashing display indicating “pulses are happening”.

The controller allows manual positioning of the motor via two buttons. As shipped, one of the buttons does not work – it causes the motor to stutter in place rather than turn. This is a pain, but since I generally do everything through software I can operate OK for now. I will need to send the unit back for repair.

At one point in the evening, I was running CCDAutopilot to generate a number of focusing runs using FocusMax. The system ran along merrily until FocusStar ran into a USB communication error. The error was never communicated to FocusMax or CCDAutopilot; Autopilot ran along for another couple of hours trying to reach focus and creating trashed images. Each night I received at least one crash, sometimes 3 or 4.

Several people indicated a problem with the unit running in cold weather. Sure enough, as soon as I got outside the controller could not communicate with the software (the evening started at minus 10-deg). I wrapped a heating tape around the controller, which allowed it to work. I found this strange. I have lots of other electronic gear which has no problem with the cold (well, except for the LCD displays on some of the SCT hand controllers –they still operate, you just can’t see the display), even though they were not necessarily intended for that environment. Surely it cannot be a surprise that this product is used outsideJ!

FocusStar Software Program

The controller program also had a few problems (OK, I am a bit biased here, since I have had a career writing user software).

The main control buttons are unclear as to how they move the motor- are they supplying a set number of pulses (perhaps from the Small/ Med/ Large selections?), or causing the motor to run continuously? The larger buttons seem to do one function, while the smaller button (between the “In” button and the position label) may do the other. I can never remember which does which; they need a Tool tip or better labeling to make it clear.

Apparently the Step Size slider changes the number of pulses assigned to the small, medium, or large movement. Their values are also assignable on the configuration screen. The two sets of values were not always consistent, so I was never comfortable how many pulses were being sent to the motor.

There is also little indication as to when the motor has completed its move. If you cannot see the motor it is hard to tell when it is done. For example, FocusMax has the nice convention of coloring the button while the motor is moving, indicating the action is active. The screen needs some visual feedback indicating the move is in progress.

Connecting to the motor is accomplished by selecting the menu choice Focuser/Connected, which is strange wording (Connected sounds like a status, not an action). Once connected, the menu label does not change (to Disconnect, for example)- it still says Connected. Selecting “Connected” a second time disconnects the motor. While a logical arrangement, the labeling is confusing.

A configuration screen allows one to set up various options, including the usual definition of In and Out directions, backlash, etc. When I tried to change the In/Out directions, it did not work correctly the first few times I tried. It seemed to need some combination of changing the option, exiting the program, and restarting the program.

The configuration screen allows the definition of Preset positions, as in Robofocus. Unfortunately, this did not work; I kept getting the useful error message “Invalid Row Index”.

As seen in the screen shot above, the program starts out with garbage data in the position field. Only when the controller connects is the correct position reflected in the field. This is not a problem, but another sign of sloppy coding.

ASCOM/FocusMax Operation

The ASCOM driver is not a separate piece of software; the same FocusStar program serves as the ASCOM driver. The ASCOM Chooser selects either the Main or Auxiliary FocusStar Driver to run one of the two motors supported by the controller. The settings screen brings up the same FocusStar screen shown above.

Using the motor from FocusMax worked accurately. Focus seemed to be repeatable and accurate. The VCurves produced were very nice and repeatable, indicating no slippage of the motor or other issues. The motor had no problem lifting the SBig camera. In general, I found the ASCOM operation to be more consistent than using the FocusStar program directly.

However, a couple of oddities did rear their ugly head.

When using the Jog screen to reposition the focuser, the move does not actually occur. For example, if I tell FocusMax to jog to absolute position 32000, the “Move To” button flashes briefly and immediately returns. The Focus screen briefly flashes a "Moving to 32000" message, and the reported position on the Focus screen does not change. If I tell it again to move to 32000, now the button stays red, the "Moving to 32000" message stays there, and the Focus position gradually updates until it reaches 32000.

Sometimes I jogged the position and the motor moved (at least somewhat, it is unclear if it completed the move) and the Focus screen in FocusMax displayed an incorrect position. Repeating the move resulted in a correct display.

FocusStar and FocusMax are not communicating cleanly. I got in the habit of doing all moves twice, using only absolute positioning. I could never tell if a relative move (move out 1000 pulses) completed successfully, so I avoided using them.

On the other hand, moves initiated by FocusMax during VCurve and focusing runs seemed to complete correctly. Only manual moves through the Jog screen had problems.

As mentioned above, FocusStar somehow messed up its USB communication with the controller. This status was never communicated to FocusMax or CCDAutopilot, which continued merrily trying to focus for about an hour. FocusStar displayed a message box "USB communication Error", but since FocusStar was minimized the message box never appeared. Worse, FocusStar kept repeating the error- every time you clicked OK, it popped up the same error box. It never recovered, and I could not exit the program. Eventually I had to 1) kill FocusStar from Windows Task Manager, and 2) Reset the controller by unplugging its power.

Summary

Overall, I was disappointed with the overall quality of the product. It still has a lot of little flaws. Quality issues keep the product from really shining. It should work fine in an interactive, hands-on environment where the user can regularly work around the various problems that pop up. However, it is not robust enough for running automated sessions throughout the night.

Similar stepper motor based products are the LazyFocus (www.lazyfocus.com), and Robofocus (www.homedome.com/robofoc1.htm). Both of these systems cost a bit more and are more mature products. I have found the Robofocus to be more reliable, although slower, subject to more backlash due to its use of a belt drive, and unable to disengage the motor as designed. I have no experience with the LazyFocus product.