Widefield Triple Imaging Rig

[Gina]

Done some testing of the Arduino sketch and found a couple of errors which I've now put right.  Next stage will be to wire up the stepper motors and check that they work.

// Gina's triple remote focussing system using PC control - 2014-12-24 1630// File name - astro_focuser_half_step_triple_WF_02//// Stepper motors used are 28BYJ-48 with ULN2003A drivers// 4 phase, 8-beat motor, geared down by a factor of 64. // Step angle is 5.625/64 degrees.//////////////////////////////////////////////////declare variables for the motor pinsint motorPin1a = 2;	// Blue   - 28BYJ48 pin 1int motorPin2a = 3;	// Pink   - 28BYJ48 pin 2int motorPin3a = 4;	// Yellow - 28BYJ48 pin 3int motorPin4a = 5;	// Orange - 28BYJ48 pin 4                        // Red    - 28BYJ48 pin 5 (VCC)int motorPin1b = 6;	// Blue   - 28BYJ48 pin 1int motorPin2b = 7;	// Pink   - 28BYJ48 pin 2int motorPin3b = 8;	// Yellow - 28BYJ48 pin 3int motorPin4b = 10;	// Orange - 28BYJ48 pin 4//int motorPin1c = 11;	// Blue   - 28BYJ48 pin 1int motorPin2c = 12;	// Pink   - 28BYJ48 pin 2int motorPin3c = 13;	// Yellow - 28BYJ48 pin 3int motorPin4c = 9;	// Orange - 28BYJ48 pin 4////////////////////////////int incomingByte = 0;   // for incoming serial dataint Focuser = 0;int FocusSpeed = 1;int motorSpeed = 1200;  //variable to set stepper speedint lookup[8] = {B01000, B01100, B00100, B00110, B00010, B00011, B00001, B01001};//////////////////////////////////int focusCount[3] = {0, 0, 0};  // int stepSize = 5;  // how many motor steps correspond to one focussing step - this may need changing after testing//////////////////////////////////////////////////////////////////////////////void setup() {  //declare the motor pins as outputs  pinMode(motorPin1a, OUTPUT);  pinMode(motorPin2a, OUTPUT);  pinMode(motorPin3a, OUTPUT);  pinMode(motorPin4a, OUTPUT);  pinMode(motorPin1b, OUTPUT);  pinMode(motorPin2b, OUTPUT);  pinMode(motorPin3b, OUTPUT);  pinMode(motorPin4b, OUTPUT);  pinMode(motorPin1c, OUTPUT);  pinMode(motorPin2c, OUTPUT);  pinMode(motorPin3c, OUTPUT);  pinMode(motorPin4c, OUTPUT);  Serial.begin(9600);}////////////////////////////////////////////////////////////////////////////////set pins to ULN2003 high in sequence from 1 to 4//delay "motorSpeed" between each pin setting (to determine speed)void moveBackward(){  for (int s = 0; s < stepSize; s++) {    for(int i = 0; i < 8; i++)    {      setOutput(i);      delayMicroseconds(motorSpeed);    }  }  focusCount[Focuser]--;//  sendFocusCount ();}///////////////////////////void moveForward(){  for (int s = 0; s < stepSize; s++) {    for(int i = 7; i >= 0; i--)    {      setOutput(i);      delayMicroseconds(motorSpeed);    }  }  focusCount[Focuser]++;//  sendFocusCount ();}////////////////////////void setOutput(int out){  switch(Focuser) {    case 0 :  {      digitalWrite(motorPin1a, bitRead(lookup[out], 0));      digitalWrite(motorPin2a, bitRead(lookup[out], 1));      digitalWrite(motorPin3a, bitRead(lookup[out], 2));      digitalWrite(motorPin4a, bitRead(lookup[out], 3));      break;  }    case 1 :  {      digitalWrite(motorPin1b, bitRead(lookup[out], 0));      digitalWrite(motorPin2b, bitRead(lookup[out], 1));      digitalWrite(motorPin3b, bitRead(lookup[out], 2));      digitalWrite(motorPin4b, bitRead(lookup[out], 3));      break; }    case 2 :  {      digitalWrite(motorPin1c, bitRead(lookup[out], 0));      digitalWrite(motorPin2c, bitRead(lookup[out], 1));      digitalWrite(motorPin3c, bitRead(lookup[out], 2));      digitalWrite(motorPin4c, bitRead(lookup[out], 3)); }  }}///////////////////////void moveManyF(int count) {    for (int i = (count -1); i >= 0; i--) { moveForward(); }  sendFocusCount ();}   //void moveManyB(int count) {    for (int i = (count -1); i >= 0; i--) { moveBackward(); }  sendFocusCount ();}   ///////////////////////void sendFocusCount (){  Serial.println(focusCount[Focuser]);}void loop() {  if (Serial.available())  {  incomingByte = Serial.read();  switch (incomingByte) {    case 49://      Serial.println("Ha Focuser Selected");      Focuser = 0;      sendFocusCount ();      break;    case 50://      Serial.println("OIII Focuser Selected ");      Focuser = 1;      sendFocusCount ();      break;    case 51://      Serial.println("SII Focuser Selected");      Focuser = 2;      sendFocusCount ();      break;    case 52: //     Serial.println("Unused Focuser Selected");  // Not used in this version (only in 4 focuser maodel)      break;    case 53: //     Serial.println("Single Step Focussing");      FocusSpeed = 1;      break;    case 54: //     Serial.println("10 Step Focussing");      FocusSpeed = 10;            break;    case 55://      Serial.println("100 Step Focussing");      FocusSpeed = 100;      break;    case 56://      Serial.print("< ");      moveManyB(FocusSpeed);      break;    case 57://      Serial.print("> ");      moveManyF(FocusSpeed);      break;    default:       // if nothing else matches, do the default      // default is optional      Serial.print("Unrecognised Command: ");      Serial.println(incomingByte, DEC); }  }}

[Gina]

Have now finished the circuit board for the triple imaging rig remote focussing system with Arduino Nano plus ULN2003As and have connected up the three stepper motors and wires to my bench PSU.  The sketch and VB app work fine   I need to add the turn off code that I have in the four focuser version as the 12v steppers get warm if left turned on with the13.8v supply.  (Or drop the voltage a bit.) 

With the focus step to stepper step ratio set to 5, the 100 focus counts turn the stepper motor shaft one revolution.  This will certainly be suffient as a "large" focus change   I may want to reduce the size of the focus step but I think 100 steps will still provide a good very coarse focussing mode.  In fact I think the x10 mode may well be coarse enough but we'll see

[Gina]

Been looking at the focus drive again with a view to arranging an alternative to timing belts.  The shorter focal length lenses could easily take a 3D printed quadrant gear but the 200mm ones have a gap of only 3mm.  Fortunately the angle of focussing action required is quite small so I guess a thickness of 2mm in a wide part round the focus sleeve would be alright.  I think spur gears would be better than timing belts since no tension or pressure is required.  The downside is possible backlash but by approaching focus from the same side each time gets round that.  The stepper motors have a gearbox which has a small amount of backlash anyway.  This amounts to around 3-4 steps of the motor itself.

Here's a rough sketch of the arrangement.

[Gina]

Focus ring quadrant gear and stepper motor pinion.  These give a reduction ratio of 100:8  ie. 12.5x.  Much better than I could achieve with timing belt and pulley.

[Gina]

I have at last managed to buy an Asahi Takumar 105mm f2.8 lens that I've been looking for for ages   That completes the set   I now have 3 200mm f4 lenses, 3 105mm f2.8 lenses and 3 55mm f1.8m lenses.  Though one of the 200mm is a Vivitar rather than an Asahi Takumar, Super Takumar or Super Multicoated Takumar (SMC).  Best lenses will be used for Ha and OIII and poorer ones for SII though in practice I doubt there's much difference.  With the 3 460EX cameras, I think the 105mm lenses will be very useful.

[Gina]

Fitting everything together, I am finding the current gap between the focus sleeves on the 200mm lenses is very small for the focus gears and I think a redesign may be required with more spacing.  I think I shall leave this until I get a larger 3D printer working than my little UP! and concentrate on the 105mm lenses which are a lot smaller.  I think this will enable me to get a triple widefield imaging rig working quicker.

[Gina]

Designed and printed the focus quadrant gears for the 105mm lenses this afternoon.  Plent of clearance with these lenses

[Gina]

Getting back to this project and trying to sort out where I'd go to   I have a 3D printer working well enough to print parts including the turret parse that I had to print in sections before when I had just the little UP! Plus 2 printer withe a build volume of around 130mm cube.  I can now cope with up to around 300mm in each direction.

[Gina]

This photo shows one of the turret sections on the bed of my large 3D printer and the ability to print a larger turret if required (in one piece).

[Horwig]

Wondered what had become of this one, welcome back.

H

[Gina]

Thank you   The lack of 3D printing size has been the problem as I said.  I have been thinking about this project and I'm now looking at a larger turret with three lenses and cameras in a circle as before but with the guide camera and lens in the middle rather than using OAG.  I have an 85mm lens that I think should be fine for the LodeStar guide camera.  With this arrangement I'm thinking of a central aluminium tube that the triple imager can rotate on and possibly doing away with the external framework.

[Gina]

I've looked out the 85mm FL lens - it's a Jupiter-9 85mm f2 made in USSR.  It's neither small nor lightweight though with the focus ring measuring around 65mm.  This is rather bigger than I would like to put in the central tube.  Of course, I could have a quad turret with three imaging systems and one guiding.  OTOH guiding doesn't need the high quality lens that's wanted for imaging and a lower quality, maybe smaller aperture lens should be adequate.

[Gina]

I've been looking for CS mount lenses that will fit directly onto the LodeStar camera but for the focal length I'm thinking of these seem to be quite expensive.   Mind you, compared with the cost of the cameras this is really insignificant I guess   OTOH I bought the cameras when I had more funds available.  Pretty much any lens with sufficient back focus and a coverage of 1/3 inch (8mm) should do for guiding and it goes against the grain to use anything costing more than necessary.

[ChrisLX200]

Well if you got a CS>Canon (or whatever) adapter you can fit DSLR camera lenses. A 50mm prime lens would give you an apparent much longer focal length than 50mm fitted to the Lodestar (maybe 200mm? - wild guess!). The image will also be much better than a cheap CS/C lens, and a 50mm prime wouldn't be much bigger than the 65mm you mentioned. I've played with camera lenses on my QHY5L-II and Altair GPCAM but I was looking for a wider field not a smaller one :-)  Just a thought.

ChrisH

[Gina]

Thanks Chris   Yes, 50mm FL on LodeStar would equate to something like 100mm on the Atik 460EX.  I have several standard Canon lenses that I got when I was collecting EOS DSLRs for sensor destruction - err I mean debayering   One of them is stuck at 55mm FL otherwise OK.  I could use that in a quad turret.  I guess a Nifty Fifty would be smaller but would need buying

[ChrisLX200]

I bought my adapter off Ebay straight from HK, only cost about £8 IIRC (costs a lot more from UK suppliers). It didn't actually take long to arrive, less than 2 weeks.

ChrisH

[Gina]

Thanks Chris

A quad turret would be possible but the disadvantage is that the balance would change when the turret angle is changed.  The main reason for using the turret design was that with all the same lenses and cameras the balance would not be affected by turning the turret to compose the image within the frame.

This screenshot from SketchUp shows a turret plate with 70mm lens holes and central 30mm.  The overall diameter is 220mm - well within my 290mm limit.

[Gina]

With a 65mm lens in the centre and 70mm in a circle the plate would need to be 250mm diameter (that's about 10") and with the 105mm lenses which are 55mm diameter the OD would be 190mm.

[Horwig]

Gina, do you really want to rotate the guider on the turret? would phd not then need re-calibrating between runs?

BTW, I got a genuine cs to nikon adaptor from the states on ebay for £6.75, postage cost me more

Huw

[Horwig]

Oh no, I think I've had a particularly dense moment, the turret does not rotate does it, ignore my garble

H

[Gina]

It does rotate but not during an imaging session.  The idea is to be able the orientate the cameras to make best use of the frame shape but rather than rotate each camera individually and use a mechanism to ensure that all images were similarly framed I decided to align the cameras once and for all and then turning the whole turret will turn the cameras whilst maintaining precise alignment between them.  But I agree that guiding is best done with the guide camera axis aligned with either the RA or Dec axes of the mount.  In fact I think I'll keep the guide camera rotation fixed - I could arrange the turret to rotate on the tube and fix the guide camera and lens to the inside of the tube.  The turret could be clamped to the tube one it had been rotated and before the imaging session began.

I wanted the guide system to be as close to and attached to the imaging system as possible to avoid relative movement during imaging.  Otherwise I could have the guide system separate.  It might not be so critical with widefield though.

[Gina]

I might try the finder scope off an ST80 as the guider lens - it's about the same size as the LodeStar.  Something like 100mm FL and f4.  PHD prefers a diffuse star to a pin-point sharp one and high quality lenses/scopes are often defocussed slightly to improve guiding.  I see no point in using a high quality lens and then deliberately reducing it.

[Horwig]

PHD prefers a diffuse star to a pin-point sharp one and high quality lenses/scopes are often defocussed slightly to improve guiding.

Too true, this has been my argument against Maxim's guiding algorithm, and for PHDs.

PHD seems to be able to find the centroid of any smudge of pixels, maxim is dead picky.

I use an oag, Maxim could not cope at all with it, PHD just got on with guiding, never missing a beat.

I've used a 105mm lens to guide a 400 imaging scope no probs.

Huw

[Gina]

Using the lens and tube from the ST80 finder will need just a sleeve adapter to connect the tube to the LodeStar camera with focussing achieved by sliding the LodeStar in or out as I do ATM on the OAG for scope use.  The whole assembly would be placed inside an aluminium tube with alignment screws to enable the guide scope to be pointed at the best guide star available.  The ali tube would be attached solidly to the mount and the imaging turret would rotate on it.

Here is a SketchUp model of the guide scope.

[Gina]

Here is part of the ali tube shown by the adapter - it will be longer in practice.  I might go for thicker tube - this is 16SWG.  Link here.