Daves info is spot on but I'll add some. You can check out ccdcalc.exe to find how various scopes and CCD combo's go together and part of the info given is Critical Focus Zone (CFZ) per combo.
As Dave said shorter FL scope have smaller CFZ...so you want lots of steps from your focuser in this zone...not too many as it will be verrrrry slow! Longer FL scopes...not so critical.
You can calculate various stepper combos on your focuser by Circumference C=πD (lets use D=5mm, the pinion for the drawtube diameter, for arbitrary sake, then each revolution of the pinion = 15.7mm movement). If I choose a stepper that has 2040 steps per revolution (one I used, has internal gearing) then I get each 2040C=π5 ~ .007699mm (~7.7 micron per step, 15.7/2040).
In a scope that needs CFZ = 20 microns then this is not good I'll need a stepper with more steps per revolution either thru smaller step angle or bigger gearing...or even a smaller DT pinion.
If you connect via a 10:1 reduction pinion drive then obviously you will get 10 times more step accuracy per step "point 7 microns" per step (depends on the pinion). This is probably just acceptable for this small CFZ. Given lets talk an F7.5 scope as an average scope then it's CFZ is ~123 microns so 7 microns per step is fine.
Anyway hope this helps, try not to choose a stepper that has too many steps per revolution as movement is slow...but more accurate.... Have a look around this may help you decide on choosing. No one can tell you how many steps you need in the CFZ...the more the better but too many and the focus movement rate is very very slow!
hope this helps.
Hi,
I didn't really get the math, anyone able to do a quick calc for an F4?
As per Dave's data, the DoF is 48 microns. what resolution would I need with a geared stepper if i'm coupling to the main focuser shaft, not the 10:1 reduction shaft?
I'm using a moonlite and the reduction knob can't lift the load.
It has to be a unipolar motor 12V or for the old system 5V.
I am using the cheap stepper from Fleabay and it drives my GSO RC8 10:1 focusser with ease and i have it set very tight to hold a QHY9M with filter wheel and OAG.
I think the link on Dave's website for the sharpsky kit is to an old driver version (3.14), and for some reason that driver isn't working for me on my new astro laptop.
I think the newest version is 3.5 (as this version was on my old astro laptop) and it works fine. I am trying to get the 3.5 version to see if that gets my focuser working properly.
I thought my focuser may be broken but I hooked it up to my old astro lappy ( driver version 3.5) and it works fine.
I have no idea why the 3.14 version doesn't work on my new lappy so I am just trying to narrow the problem down.
Cheers
Jeremy
I think the link on Dave's website for the sharpsky kit is to an old driver version (3.14), and for some reason that driver isn't working for me on my new astro laptop.
I think the newest version is 3.5 (as this version was on my old astro laptop) and it works fine. I am trying to get the 3.5 version to see if that gets my focuser working properly.
I thought my focuser may be broken but I hooked it up to my old astro lappy ( driver version 3.5) and it works fine.
I have no idea why the 3.14 version doesn't work on my new lappy so I am just trying to narrow the problem down.
Cheers
Jeremy
I had trouble too when I tried to upgrade - went back to the old version and everything working again. Also thought it might have broken - auugh - sure do rely on this thing .
I didn't really get the math, anyone able to do a quick calc for an F4?
As per Dave's data, the DoF is 48 microns. what resolution would I need with a geared stepper if i'm coupling to the main focuser shaft, not the 10:1 reduction shaft?
I'm using a moonlite and the reduction knob can't lift the load.
can a bipolar stepper be wired to work with this?
Cheers
Alistair
The load "not lifting" (slipping) sounds like the crayford pinion is slipping on the DT, needs tension/friction adjustment to fix this.
A ML pinion is ~ 4.75mm dia (well mine is) thus C=πD (D=4.75mm) thus each time you turn you pinion, 1 revolution, the DT will move the circumference of the pinion = 14.922 mm.
So you want to have a motor that moves this in small steps…a lot of small steps (aka a stepper motor) in a fashion that is finer than your finger can do and automatically!
So pick a stepper that has a lot of steps per revolution…the original Dave used is 2040 steps per revolution. Couple this to the pinion to move it. We know 1 full revolution = 14.922 mm now divide this by the number of steps = 14.922/2040 = .00731mm per step movement of the DT or 7.3 microns per step.
Your given CFZ for an F4 ~ 48 microns (ccdcalc states this as ~35 microns, so they are reasonably in agreement) 48/7.3 = 6.57 steps in the CFZ.
This is not enough, you need a stepper motor with more steps per revolution especially with filters etc.
Some people use reduction gearboxes like the McLennan gearbox (have I got that name right?) to achieve this, or use the 10:1 pinion or maybe even purchase a Hurst LSG35 stepper motor with a very high number steps per revolution.
If I was you I’d get the best stepper I could as you need high step resolution (or use a gearbox), maybe the HurstLSG35012E98P, 14400 steps per revolution, 12V. This would give you 14.922/14400 = .00103 mm per step = 1 microns per step! Now we are talking! Thus CFZ 48/1 = 48 steps in CFZ for your system.
I cannot tell you how many steps in the CFZ you need but more is better and I suggest at least ~50 or more maybe 100 to give filter change step accuracy.
Lastly I think you’ll need a unipolar stepper…the Hurst and fleaby are mostly these types.
its clear now.
the reduction knob slipping is slippage at the reduction knob and not the main shaft.
it doesn't slip when I turn the main shaft. I've already increased the pressure and Ron also confirmed that that's why they couple the moonlite motors directly to the main shaft and not the reduction one.
I had a look at ebay and most geared steppers were 4096 steps/rev.
The moonlite steppers have a resolution of 2 micron.
with the awful seeing that I have, I'm not going to need that fine a resolution, but I'll see what I can fit.
Cheers
Alistair
Edit: just realized I also have this stepper and gearbox and whilst this is 125:1, it'll give me a resolution of 2 microns as the output shaft dia is 4mm, step angle is 7.5deg which should be enough and this is rated for 5V and a current draw of 550mA. sweet. https://australia.rs-online.com/web/...otors/3514647/
is it possible to run a 12v stepper with this circuit? if so, I also have a 250:1 gearbox but the matching motors are all 12v unipolar. http://australia.rs-online.com/web/p...otors/3514574/ is a matching one for the gearbox.
sorry about the multiple edits
Last edited by alistairsam; 22-01-2014 at 02:13 PM.
Alistair those motors look ok (except read below) and do have some nice gearing ratios...this is exactly what you want...a method to change gearing as you change scopes!
ML steppers are capable of "micro stepping", thus they can get down to 2 microns per step, this system cannot do micro stepping but that's not an issue with the right motor or gearing.
Just look at the specs for the motors as the stepper driver is capable of driving 500mA per coil but...this is de-rated for duty of cycle, thus probably at best 150mA~ 200mA per coil continuous operation. Ok the stepper isn't operating all the time but be aware of trying to drive coils that are too low in ohms, try to limit ~ 15~200mA per coil from motor selection.
I use Hurst which all appear to be ~ 65ohm per coil = 12/65 = 184mA per coil...this is sweet, no heating on package but I wouldn't go too much under this is coil resistance.
The 1st Mclennen motor is almost exact coil Rx as mine, the P542-M481U is a 5V system, sorry cannot use, but their pdf shows the P542-M482U, a 12V system...this is the one to get. Sadly doesn't show up on RS site!
These McLennen motors are listed (yes I know they are 5V just going into some more details on coil rx for the motors) is 52.2 ohms per coil and this is ~ 230mA...starting to get up a bit in current draw..."but"...since the system has a Schottky diode protection on the PCB, then available voltage is ~ 11.6V thus in reality is ~ 220mA per coil, and should ...mmm just make it ok as it's not being used all the time. Maybe a small heat-sink glued to the driver IC will help alleviate any concerns using these motors. you can also consider that a coil when being driven with current, heats the coils and their actual resistance rises ...meaning you should be ok at this I draw.
IMHO...go for the 12V McLennen system with different gear ratios...looks real nice indeed!
Had a quick look over, very impressed with the soldering. Nice job Brendan!
Haven't manage to get a motor yet, see if I can find one somewhere just to test it out
cheers,
Bram
Thanks Bram, just be careful inserting the PIC.
Soldering really is an art and unfortunately I'm at the low end of the art gene pool but have done more than enough to still do passable joints.
Most soldering these days from mass production is the non lead solder...which is really metal glue, not solder! Lead/tin solder 60/40 is still the best by far.
All seems to work nicely - Thank you Brendan and Dave!!
Drilled the holes in the box and mounted it. Then I realised I forgot the holes for the leds .... so for now I don't have any leds (yet).
I found a stepper motor (unipolar, 12V, 48 steps), and manage to insert the leads into the 9-pin connector. I had to view the video-manual to workout the manual control button, which works like a treat by the way.
I finally finished assembling my controller, thanks to Dave and Brendan for their help.
I wanted to reduce the cables on my OTA so I did away with the 3pin DIN and mounted the temp sensor in the controller box and soldered it to the board directly. works fine for my setup.
I also found a 2.5mm male to 2.1mm female DC power socket adaptor. this allows you to use the standard 2.1mm power adaptors if the socket is 2.5mm.
I still need to tweak the step period setting for my moonlite stepper, but I did have a few questions,
1. if you set the step period via the software, does the pic store the value and use it with the manual control? mine has a bit of a lag with the manual control and it doesnt recognize the pulses if rotate the knob too fast.
I will add two spst micro switches next to the encoder so I can use them instead of turning the dial. just a personal preference.
2. how is the temp compensation intended to work between subs?
I use maxim and ccd autopilot so will be refocusing between filters, but if I'm taking say 10x15 min subs for the same filter and the temp drops, will it adjust the motor during the sub exposure? is it independent of maxim or focusmax's control?
anyone using the moonlite 2.5inch newt focuser with the stepper? if so, could you post what step period works best with it?
I've tried between 4 and 35. 4 is the fastest without jitter, 10 was suggested by Dave and is smooth. gets progressively slower and at 35, there is no movement at all.
mine is an F4, so I will need the smallest mechanical step possible with the moonlite stepper.
Brendan, your PCB layout is very good, thanks for letting me use it and thanks again to Dave for the PIC.
Good job, looks great - I can answer your questions :
If you set the step period in software it is stored in the controller (EEPROM). This value is applied to both manual and PC based control inputs. The manual control was intended for fine tuning the focus rather than large movements. Therefore, if the manual control is spun faster than the software debounce period input will be lost.
When temperature compensation occurs is generally a function of the client software rather than the focuser driver. The focuser driver has no knowledge of what the camera is doing and so the client enables/disables temperature compensation (or should) when subs are being taken. So when I sub is underway temperature compensation should be disabled.
Just to clarify, the step period does not change the physical step size delivered by the motor but instead the length of time energy is delivered to the motor. The physical step size of the motor is a function of the motor design and the attached gearbox. The 35BYJ motor delivers 2048 steps per revolution and this is fixed with the current firmware design. As you change the step period the motor will speed up or slow down but the number of steps per revolution stays the same. Each motor will tend to have a sweet spot where speed and torque are optimum.
Ok, I have made a few modifications to the SharpSky firmware. To be honest one of these changes should already have been in but somehow slipped through the revision net. I have enhanced the way the temperature probe is being read by the PIC micro to make the read less invasive. The current V4.0 has an issue whereby reading of the temperature sensor interrupts the motor move (this should have been fixed). Also the sensor read was disturbing the operation of the manual control.
The new V4.1 firmware load fixes the motor interruption issue and makes the manual control much more responsive.
Also for those that have built a focuser using a development card and a PIC18F4550 device a new firmware image is available on the original website : www.dt-space.co.uk/SharpSky
Quick reflash instructions :
- Download the PC bootloader from the website
- Download the new firmware load (right click and save as)
- Run the bootloader PC application
- Push the manual control button & while pushed power up SharpSky. The bootloader will see the PIC and respond
- From the bootloader Pick the new firmware HEX file you have saved in the second step.
- Program & verify
- Reset device
Good to go you are now running V4.1 and this will be displayed in the ASCOM setup page for the driver.
Any questions or problems with the reflash please fire at will
Hi Dave & other Sharpsky users...just finished a kit for netwolf (Fahim), Alistair made his own. Alistair supplied the parts I just the goon who slap it together.
One thing ...I noticed Alistair sent me a 3 pin pre-moulded mini DIN lead. This a real good idea in that the 3 PIN Mini DIN plugs I had been using were pretty poor quality and didn't fit well due to the backshell case diameter. Also had a few suffer from "pushed pins" and had to toss them. This pre-moulded lead is much neater/smaller.
But, you need to splice the pre-moulded lead to the sensor. Well I did that buy slitting the lead open and stagger splicing and heatshrink, right up near the head of the sensor and covered it back up with the original outer layer and heatshrink up to the probe. The result was ok, better than before.
However, I did see guys over at Homebrewtalk forum using DS18B20's and stainless steel caps and making their own sealed sensors. Which I though would suit these pre-moulded leads real nice. Obviously they use a plain non-sealed DS18B20 and heatshrink over the legs and pot the device into the cap. I was thinking JB weld...it's non conductive electrically and should be thermally conductive.
Anyway just regurgitating this idea. Other may learn from my trials.
.
