![[H-Bridge Diagram]](hb.gif)
It is likely that your declination drive motor runs off your 12VDC supply for your telescope. These motors generally require a very low operating current of a couple of hundred milliamps and this is ideal for a transistorized driver. You can check to see if your motor meets this low current requirement by placing a current meter in series with a 9V battery. Connect one lead of the meter to plus on the battery, the other lead of the meter to one of the DEC motor wires and the second meter lead to minus on the battery. It is best to use the 10A range on a digital voltmeter because the lower current ranges are fused to about 300mA and the fuse may blow. If your motor draws 200mA or less when running, the transistorized driver will work for your motor. The relay circuits shown on page 168 of the CCD Camera Cookbook are required for higher current motors.
To keep the parts count down, this circuit was designed for use with the 6402 CMOS UART chip. This chip has an output drive of 2.5mA. The AY-3-1015D chip can not source this much current and a transistor buffer must be added to each output pin. Our friend the DS0026 was selected as the H-bridge driver. An H-bridge driver is simply four switches that allow us to reverse the polarity to a motor. The motor is connected to across two outputs from the DS0026. If both outputs are low, or both outputs are high, the motor has no voltage across it and it will not turn. If one output is high and the other is low, the motor will have voltage across it and it will turn. If we switch the outputs to low and high, the voltage across the motor reverses and the motor turns the other way.
The DS0026 inputs can also be activated with a set of optional hand-control switches. These are connected to the DS0026 in a wired "OR" fashion. If the computer is controlling the motor, pressing the same switch on will have no effect, pressing the opposite switch will stop the motor until the computer releases its switch.
The DS0026 chip is protected from inductive transients by using a filter capacitor and two 10-ohm limit resistors. The 10-ohm resistors should be rated for 1/2 watts. If the outputs are shorted to ground or shorted together, these resistors will dissipate a lot of power and burn up. For protection, you may want to add a 1/4 amp slow-blow fuse in series with each output.
The outputs to the motor are differential and can not be connected to ground. If you use a 1/8" phone jack for connection to your declination motor, the outside of the jack must be insulated from ground. You can either house the serial controller in a plastic box or use fiber washers to isolate the jack.
Here is completed H-Bridge Circuit to drive DC declination slow-motion motors with the Cookbook camera. (Note: the circuit is a large GIF image. You may prefer to download rather than view this file.)
If you don't want to go with the relay control circuits given in the CCD Camera Cookbook on page 168, you can build a complete controller around the serial interface. (By the way, the 120VAC synchronous-motor-to-TTL converter on page 168 has the diode across the 2.2K ohm resistor to ground reversed in orientation. The anode should connect to ground as in the High Level RA Clock to TTL level Circuit). Our clock circuit will consist of a 555 timing chip that derives its stability from a resistor and a capacitor network. This RC network is temperature sensitive and it will drift. The reason we don't care about the drift is because the PEC function in the software can correct for an offset frequency. Over short periods of time, the clock does not drift appreciably.
The timing components were selected to give a free-running frequency of 480 cycles per second. This is eight times the normal 60 Hertz frequency we need. The 555 chip does not generate a square wave pulse and we need to use a counter to clean up the waveform. By using the higher frequency, smaller timing capacitors can also be used.
The 555 chip uses an internal voltage divider as a reference point for triggering the timing function. By changing this reference point, we can change the output frequency of the device. The UART chip, or a hand-control button connected to ground or +5V, is used to pull pin 5 of the timing chip higher or lower. This makes the timer run faster or slower depending on the value of the series resistor that is selected. You may want to select higher or lower values than the 10K and 22K ohm resistors in the design to give you more or less shift of the frequency. Also when using the AY-3-1015D UART, add a 3.3K pull-up resistor between pin 11 and +5V. This chip does not have enough high level drive.
Here is completed 60 Hertz Clock Circuit to drive AC synchronous motors. (Note: the circuit is a large GIF image. You may prefer to download rather than view this file.)
The output of the circuit is a TTL logic level signal which is compatible with the PEC divider circuit. You may need a different level to drive your 12V-to-120VAC converter circuit. This can be accomplished by using a level translator. An example of this is shown for an LX3 type drive hand-control jack.
Here is how to wire the circuit above to an LX3 hand-control jack. (Note: the circuit is a large GIF image. You may prefer to download rather than view this file.)
Making a battery or 12VDC power pack drive a 110VAC synchronous motor requires a circuit called an inverter. The motor requires an alternating current but the waveform does not have to be a sinewave like the one coming out of a wall plug. We already have an AC square pulse coming from the 60 Hertz TTL clock. If we use this to drive a transistor switch, we can chop 12VDC into 12V pulses. The pulses then can be stepped up to the correct voltage with a transformer.
Our circuit uses a 2N2222 transistor switch to invert the phase of the TTL clock by 180 degrees. Because the TTL clock is either at about 3.5V or 0V, it can only turn a chopper switch, the TIP120 transistor, on when it is at 3.5V. The inverted output allows a second chopper transistor to turn on when the other one is off. The chopper transistors connect to a center tapped transformer low voltage secondary winding. When one transistor turns on, it pulls current from the center tap through the winding. This induces a voltage on the high voltage primary side that swings in a positive direction. When this transistor turns off, the other transistor turns on and pulls current through the second half of the secondary winding. The current in the secondary winding is flowing opposite to the direction it was when the first switch turned on. This causes a negative voltage to appear on the high voltage primary side of the transformer. This circuit configuration is referred to as a push pull stage because one transistor pushes the output up and the other pulls it negative.
The voltage to the center tap on the secondary winding is provided from an LM317 regulator. This regulator is configured as a constant current source that sources 0.6 amps. This limits the current through the transformer windings and through the transistors so they don't burn up if the circuit is not clocked. The circuit also keeps the power requirements constant. If you would like to use less power, increase the values of the 1-ohm resistors. The current is set as 1.2V/R where R=2 ohms for our circuit. You could add a switch to decrease the running current by adding more resistance. Once the motor starts, it usually requires less current to keep running.
The resistor and capacitor on the 110VAC output of the inverter are used as an energy absorber to remove high voltage spikes and ringing that occur when the transistors switch.
The LM317 regulator should be mounted with a heat sink of about 2 square inches of area. The TIP120 transistors require small heatsinks of about 1 square inch of area. The tabs of these devices are electrically connected to the outputs of the parts and the metal heat sinks, tabs, and fasteners should not short to other electrical paths such as case ground (if you use a metal box). Also, the 110VAC output is just like the power coming from your wall socket and it is potentially lethal. Keep these wires insulated. You may want to connect the 110VAC output to a standard wall type receptacle if your telescope has a 110V plug connected to the synchronous motor.
Here is the circuit for 110VAC power supply. (Note: the circuit is a large GIF image. You may prefer to download rather than view this file.)
The PEC circuit divides the 60 Hertz synchronous motor clock rate by 144. The first stage divides by 16 and the second by nine (9 times 16 is 144). If you run 211/245Plus software, it is possible to select a count rate to match the number of teeth on your synchronous motor drive. Older telescopes sometimes used commercial worm gears with ratios not commonly found today. Also, you may have made your own worm gear with a custom number of teeth. The table below shows the PEC count that you can set the software to for different gear teeth.
Count Gear-Teeth RPM
-------------------------
100 359/360 1/4
125 288 1/5
150 240 1/6
200 180 1/8
225 160 1/9
250 144 1/10
300 120 1/12
375 96 1/15
400 90 1/16
-------------------------
The AP211/245 software will only toggle between 100 and 200 counts. This does not limit you to 359/360 or 180 tooth gears. The counters can be adjusted to divide by other integer ratios. This is accomplished by connecting the presets, pins 3, 4, 5, and 6, on the 74LS163 counter to ground or the pull-up resistor. The divide ratio for the 15 possible combinations are as follows:
Pin 6 5 4 3 Ratio
-------------------
G G G G 16
G G G P 15
G G P G 14
G G P P 13
G P G G 12
G P G P 11
G P P G 10
G P P P 9
P G G G 8
P G G P 7
P G P G 6
P G P P 5
P P G G 4
P P G P 3
P P P G 2
------------------
G = Pin Grounded
P = Pin Connected to Pull-Up Resistor
Select the divide ratio for your gear to yield a value of 100 or 200. Divide 51,840 or 25,920 by the number of teeth in your gear. This gives you the product of the two divide ratios needed for the 74LS163 chips. 25,920/144 teeth = 180. You need one divide-by-12 and one divide-by-15 counter to divide by 180. The 100 or 200 PEC count ratio of the AP211/245 software limits the possible teeth combinations to the integer ratios of the counters. The possible combinations of teeth are expanded by using the 211/245Plus software. Take the number of clock cycles in one day (5,184,000 for 60 Hertz) divide by the number of gear teeth. Then adjust the product of the PEC count and the two dividers to match.
In the rare instance that the 74LS04 inverters do not communicate with your computer, you may need to change to a true RS232 level shifter. Maxim makes a chip that converts 5V to plus and minus ten volts by using a charge pump converter. The chip also has two RS232-to-TTL receivers and two TTL -to-RS232 transmitters. Use the circuit given below in place of the 74LS04 connections to the UART and the serial cable.
Here is the circuit for RS 232 Level Shifter. (Note: the circuit is a large GIF image. You may prefer to download rather than view this file.)
Initial Setup
The first step to guiding is to get your serial interface correctly set up in the software. Go to the Options Menu and press R to bring up the Serial Port Menu. You will see a screen that looks like this:
C: COM1:300,N,8,1,BIN,RS,OP0,DS0 X: Reset Serial Interface raF: not bit0 raS: bit1 DecL: bit2 DecR: bit3
To select the serial port COM1 or COM2 that you are using, toggle the port by pressing C. Note that there is the word "not" in front of bit0. This indicates the right ascension command to bit zero (pin 12 on the UART) will activate the control when it is a logic zero. It will normally be a logic-1 when no change to the right ascension is needed. The remaining command bits are positive logic and they go from a logic-0 to a logic-1 state when the control becomes active. If you built the relay circuit in the book or the LX200 control circuit, press F to remove "not" from bit0. Exit the menu then press U to save the setup.
Next, your serial controller should be powered up and connected to your computer and telescope. The serial port will probably come up with one or more of the control bits set to the active state. This does not hurt your telescope but it will drift off of celestial objects. To reset the serial port, you have two choices. One is to get back into the Serial Port Menu and press X. The other is to enter into the Autoguide Menu (accessed from the main Menu) then exit. If you use the autoguide route, the serial port control will stay active and slow motion control is possible in the Find/Focus mode by pressing the arrow keys. In the AP211/245 programs, tap the arrow keys to send 0.1 second commands. In 211/245PLUS, you can hold the key down as needed. In either case, there is a delay in obtaining a new image after the key is released equal to the integration and acquisition time.
Selecting a guide star
Prior to selecting a guide star, rotate your camera until the slow motion controls cause the star images to move parallel to the edges of the display box. The guiding software can correct for any orientation of the camera, but aligning your camera will improve the guide. With the camera aligned, it is also easier to move stars in the image field with slow motion controls to a desired point in the field.
The Cookbook 211 and Cookbook 245 software operate the same with respect to guiding and the serial interface. The programs use the quarter-frame mode to obtain a fast readout of the center square (the 211 is in the center bottom). This quarter-frame is subsampled to generate an area for tracking. A 96 by 82 area is used for images acquired without dark frame subtraction (note: this area is the full size quarter-frame for the 211 but it is not for the 245). When dark frame subtraction is invoked, the frame size drops to 64 by 56 to reduce the computation time.
To track a star, it must have enough brightness to be above the noise of the CCD yet not so bright as to bloom or saturate. For most faint guide stars, the recommended integration time of one to two seconds is short enough to not require a dark frame subtraction. However when guiding on very faint stars, hot pixels will often be as large as the star's induced charge. Saving a dark frame of the same integration time then enabling the dark frame subtraction will allow guiding on very faint stars.
The long integration time of one to two seconds is recommended because of the slow response time of the system. Readout and calculations require about one second depending on your computer's speed. Using bright guide stars and short integration times will cause the system to respond to the short term atmospheric perturbations but with a long lag in response. This results in a worst guide than tracking the averaged position of the star.
When selecting a guide star, you should locate a star in the Find/Focus mode using a binned full frame. Move the star in the center of the quarter-frame area and change to the quarter-frame mode. Use the autoscaling feature to get an idea of the low and high range of the star. If you are using 211/245PLUS software, turn the autoscaling off after finding the range. Enter into the Select Options menu and set the low and high stretch to about 50 above the low value determined by the auto stretch. If the difference between the high and low value is less than 66, pick a value which is 2/3 the distance between the low and high value. Next, redraw the display. The star should appear as a spot on the display. If there are numerous speckles on the display, the hot pixels or dark noise is showing up. You may want to save a dark frame and set the dark subtract to on or you can increase the low (and high) level stretch until the noise is gone.
Once the guide star is in position and the low stretch is set, exit back to the Main menu and select the Autoguide menu. The following display should appear:
| T: 1.5
| R: 5.0
| W: 1.0
| Lo: 50
__________|___________ C: Cal
| I: No Lock
| + Z: Drift
| B: 0
| H: 9
| Angle: 0
raF: 0.000
+5 -12 23415 raS: 0.000
decX: 0.000
decY: 0.000
M: Switch DEC
D:
p: Learn PEC
P: stop PEC
@: PEC 200
E: exit
1 0
09:08:11
At first, the guider display looks overwhelming. Fortunately, most of the parameters are not needed for general purpose guiding. The integration time (T:) and low stretch (Lo:) clipping values should have already been set when you selected your guide star in the Find/Focus menu. The Lo: value may have changed if you had selected Autoscaling while in the Find/Focus menu and it must be returned to the previously selected clipping level.
The first two functions that you should enable are drift subtraction (Z: -Drift) and if you are using a dark frame (D: -Dark Ref). The drift subtraction keeps track of the electrical bias shift of the camera electronics and it corrects for the shift. This keeps the clipping point set above the noise which can cause errors when faint stars are used for tracking.
To keep the math calculations fast on the computer, integer math is used for determining the position of the star. The position information is given in tenths of pixels (pixel sizes are corrected from their rectangular format to a square equivalent pixel). The X-offset from the center and the Y-offset from the center are given in tenths of a pixel just below the cross hair display.
A pixel based format is independent of the magnification and to keep your guiding errors small, you must know how wide a pixel is in arc seconds. For the 211 camera a pixel is about 0.00028 times the focal length and for the 245 camera a pixel is about .00053 times the focal length. Using a 2500mm focal length scope, the X and Y errors of +5 and -12 tell us the star is 5/10 times 2500 times .00053 or 0.66 arc seconds to the right when using a 245 camera. The error is also 1.6 arc seconds down.
The star is allowed to drift from the center until it reaches a boundary of H: tenths-of-a-pixel from the center. A control command is then issued to drive the star back to the center. The dead band of H: in tenths of pixels keeps the controller from continuously dithering past the center. For the 245 camera, set H: to around the focal length divided by 275; and for the 211 camera, set H: to about the focal length divided by 525.
Before you can track on a star, the system must know how far a pulsed slow motion command (remember our baud rate time of 1/30 seconds per command pulse) will drive the scope with respect to the pixels on the CCD chip. The C: calibrate command tells the controller to drive the right ascension for R: seconds then back for R: seconds. Next the declination is driven for W: seconds then back for W: seconds. The preset values of 5 seconds and 1 second may not be the right values for your drive system. If needed, adjust the times so the star is not driven out of the quarter-frame view during calibration. Also, try to adjust them so at least ten pixels are traversed during calibration. If the star image does not move far enough during calibration, or the star goes off the field of view, the calibration will be in error.
After calibrating the system, press I: to start the tracking. The display will change from I: No Lock to I: Locked. The term "locked" does not mean the system is locked at the center but it is in the tracking mode and if you move the star with the slow motion controls, the system will try to bring it back to the center. Depending on your system, it may take a minute or two to initially drive the star to the center. Once the star is at the center, it is time to start your guided exposure.
Using Background PEC
Depending on the number of teeth on your gear drive, select either 100 or 200 with the @ key (or if you are using 211/245Plus software, enter the correct value). With your star centered in the cross hairs and the guider locked onto a star, press the small letter p to start the learning process. There are two numbers above the time display on the bottom of the screen. When the left most number is one, PEC is off. Pressing p will change this to a two and the computer will start learning the drive error. The right most number counts up to the @ value as the information is stored. When the worm period is completed, the left number will change to a three and the right counter will restart at zero and count up to the @ value as it replays the recorded error commands. The commands are replayed ahead of schedule by one counter time period to correct without an error due to delay.
The PEC function not only records the periodic error of the RA drive but it also records the drift of the RA due to alignment and frequency errors in the RA clock. The DEC commands are also replayed and the system corrects for drift in declination. My telescope is a Meade 10" LX3 and the worm drive has a periodic error of about 30 arc seconds. This error is typically reduced enough to get about two good unguided pictures integrated for two minutes per each eight minute period. I recommend retraining the PEC on a star near the object of interest at least once every hour and each time a new object is selected.
How good is the background PEC guiding? Here is a view of M51 which is two combined 2-minute exposures that are guided by background PEC only. Only dark frames have been subtracted and the image was stretched to show the detail of the galaxy. The stars near the edge of the field have comma due to the binocular RFT reducer I used to work at f/6.3.
Return to the Text DataBase Page
Return to the Cookbook Home Page
