Building a GOTO Mount

[jason udallParticipant @jasonudall57142]

Having championed the ramps previously. ..
The ramps isn’t ” arduino” compatible any more than than any circuit anything compatible.

The pin out fits a particular arduino model
The ramps basically acts to breakout these pins and add some hardware interface. ( sockets for stepper drivers…three 5 amp fets…
Five rc servo ports…and more.

And other bits/bobs..
If you are expecting a shield with handy-dandy library ..sorry

[Neil WyattModerator @neilwyatt]

What I mean is RAMPS a shield for the Arduino mega (which sounds like compatible to me) and all the libraries for use with the Mega will be available as well as the Arduino IDE.

I'll be programming mine in AVR assembler from scratch using studio 6.2. I wouldn't recommend that to others – its the path I choose because it is one of my interests.

If the end product is of use to others I'll be willing to share it, but I'm not going to build in extra complexity because someone else might want to use it for something else in the future.

Neil

P.S. if you think hanging a FET of a port pin is dodgy, what about this from AVRfreaks:

[jason udallParticipant @jasonudall57142]

Neil..and what I mean is that the ramps isn’t exclusively arduino…conveniently yes but not to be overlooked if using pi. Pic.or otherwise..

[Ian PParticipant @ianp]

Posted by Neil Wyatt on 25/11/2015 11:50:40:

Posted by Ian Phillips on 25/11/2015 09:54:11:

Posted by Ian Phillips on 23/11/2015 21:35:25:

Neil

Curiosity has got the better of me, is the purpose of JP4 or the switch so that you can test the PSU?

IanP

Bumping my own post, but what does the switch do?

Er.. turn it on and off?

N.

So does the PSU sense that its output is being shorted so it shuts down?

Ian P

[AjohnwParticipant @ajohnw51620]

Freaks is the right thing to have on the end of the name. It's interesting to put a current sense on the supply on a circuit like that but few people have anything suitable – high frequency hall effect oscilloscope probe as both mosfets will be on for short periods as they are switched. They have sort of done something about the problems that is likely to cause on the supply almost certainly adding more parts than if the drive had been done correctly.

90nSec rise times are good at generating RFI especially with long lengths of wire on them. No need so why switch that quickly unless it's needed. The actual rise times might turn out faster that that. I'd hope that the data sheets have pulse thermal resistance info otherwise guess.

The absolute maximum rating of the i/o pins on the processor is 40ma however the max for the chip is 200ma. They point out that these aren't what might be called real absolute ratings as long term use might damage the device or cause premature failure and later reduce them to 50% as far as the pins are concerned.

Often on electronic items spec's are given in regions that they are expected to work in. As far as output is concerned that's 3ma but on this sort of part that can be exceeded providing the totals are looked at. The i/o conditions that the device can stand are given in condition 3&4 on page 317 of the full data. Essentially 100ma per port by the look of it and 20ma per pin – they don't mention the chip limit when driving which is likely to be 100ma max in practical terms with near 10 running the processor.

The reason for putting something in the way when driving mosfets is the capacitance in them. It's possible for the processor pin to get volts forced onto it back through the mosfet. The ports do have clamp diodes but it looks like they are for ESD protection only as they don't specify a max current through them. Processors turn into thyristors if the port input voltage range is exceeded. A puff of smoke, a nasty smell comes out and a small hole appears in the package. Not terribly likely in this case but it can happen. Pull down resistor can help a bit even the usual 20k because it may happen when they are inputs. Looks like high side drivers ideally need pull downs as well. I'd have thought that they would build them into the chip as no one wants to add more bits. It's annoying. Lots of things need them to define what happens during reset and before the post is set to the correct state.

John

–

[Neil WyattModerator @neilwyatt]

Posted by Ian Phillips on 25/11/2015 15:22:48:

Posted by Neil Wyatt on 25/11/2015 11:50:40:

Posted by Ian Phillips on 25/11/2015 09:54:11:

Posted by Ian Phillips on 23/11/2015 21:35:25:

Neil

Curiosity has got the better of me, is the purpose of JP4 or the switch so that you can test the PSU?

IanP

Bumping my own post, but what does the switch do?

Er.. turn it on and off?

N.

So does the PSU sense that its output is being shorted so it shuts down?

Ian P

You are making this complicated.

Two wires from the pads go to a switch, and two more wires go from the switch to a battery.

Neil

[jason udallParticipant @jasonudall57142]

Looking at that from avr freaks..

One…its in effect a dimmer.

It even sniffs the ac…so the sw could switch (both fets on) at zero crossing…
Bit confused over the IR sensors with switch to short out but ho hum

[Neil WyattModerator @neilwyatt]

Whatever the datasheets say, in practice AVRs are pretty bombproof. I think I've blown one up so far. They can do quite surprising things…

This:

Plus this:

Makes this (actually this is an earlier version, that ran out of flash for the interpreter and more esoteric display driver functions):

Which does this (amongst other things):

Later version isn't very pretty… but it is a work in progress.

[Ian PParticipant @ianp]

Neil

Not trying to make it complicated, just that it looked to me as if the two wires went to the poles of a switch. If it had said 'From Switch' or 'from PSU' I probably would have twigged.

As the power supply is not shown and you (sensibly) use the same symbol for all the points connected to 12v I thought 'JP4' was used for either a switch or a jumper. Presumably the normal on/off switch would be on the input side of the PSU but putting the connector on the board gives you the choice later.

Now, studying your PDF in greater detail I realise that all your connectors have the JP prefix.

Ian P

[AjohnwParticipant @ajohnw51620]

Posted by jason udall on 25/11/2015 16:29:29:
Looking at that from avr freaks..

One…its in effect a dimmer.

It even sniffs the ac…so the sw could switch (both fets on) at zero crossing…
Bit confused over the IR sensors with switch to short out but ho hum

LOL I missed that aspect and something else about it too. How the load is connected.

As to the rest automotive stuff is produced in rather large numbers and some one some where eventually finds the flaws. Not a problem if just a few but frowned on anyway as it could cost a fortune to put right and has at times. The other aspect is that if some one spends 24 years on them most things crop up one way or another. Personally if something means adding a couple of resistors or using a better drive arrangement that might cost a couple of quid more so what especially in the hobby area.

Unused pins on them have the usual suggestions. Activate the internal pull up resistors or better still in some circumstances use external pull ups and pull down but direct connect to a rail is discouraged.

If a board was ever available I would be happy providing that any spare pins could,be used if needed.

Atmel do have a zero crossing application note that shows that the input pin diodes can be used as clamps via high value resistors. A rather low current as I suspected. We have had mains spikes round here that have blown dimmers and bulbs. Not sure how it would cope with that. They can be in the Kv range, more than one of them at times. MOV's have killed that problem.

John

–

[Neil WyattModerator @neilwyatt]

Posted by Ian Phillips on 25/11/2015 16:39:28:

Now, studying your PDF in greater detail I realise that all your connectors have the JP prefix.

To be fair on me, its really just a working diagram to generate the PCB, so I'm happy to flaunt conventions or use the wrong part with the right footprint, for example. If I use a 'pin header' to get a series of pads at 0.1" spacing, it calls it a 'jumper'. Often I wire direct into them, especially for power leads.

It took me a while to understand 'nets', if you use these instead of a wire to join up power pins etc. then you don't need to physically join them up on the circuit diagram. This is what I have done with +5, +12 and GND.

The board is a right dog's dinner at the moment.

[Neil WyattModerator @neilwyatt]

First stab at a PCB.

I will need to increase the stand off a bit as 100% accuracy in lining up top and bottom hole isn't guaranteed.

[Neil WyattModerator @neilwyatt]

Well, I went to make the PCB today – and sods law says I only have singel-sided board!

Now cramming it onto a single layer.

[Neil WyattModerator @neilwyatt]

Single layer PCB:

[Neil WyattModerator @neilwyatt]

Quite a bit of progress, but I haven't been posting it.

I have redesigned the PCB as instead if a 16×2 alphanumeric display, I will use a 64×128 green graphic display from an old project instead which can display a lot more information, including proper lists of targets.

I have a nice project box made up, with five colour-coded XLR socket on the back: RA, DEC and focus steppers, guide port and dew heaters. On the front the display, chunky brightness control (for ease of use in a hurry), power and an XLR socket for the control handset. The handset is a nice small project box with five coloured push bottoms – I found out why they were cheap on eBay – the bodies have a tendency to melt during soldering!

Found a lovely little unipolar stepper for focus and it had six solder pads, two joined together. I split them and it's magically become a bipolar stepper so I can run it from a polulu, but that's right down the road.

Not much more I can do with hardware until get some molex connectors – I don't want a board tied to the box with dozens of wires. The steppers are fitted to their 'gearboxes' and just await wiring. So for the time being I've started on the software side.

Here's a pic of my 'splash screen' for now.

Neil

[Neil WyattModerator @neilwyatt]

Basic box is working now, solved a problem with excess current consumption – one of the Pololu stepper boards was swallowing 65mA extra following every reset. Swapped it out and all three boards only take about 7mA each when doing nothing. Mind you the display is greedy – but the LEDs will be 'turned down' for night use!

Some screen grabs of work so far. Coding for calculating sidereal time to an accuracy of about 3-4 seconds over the whole century and the position of Polaris to half a degree over the same period were challenging in 8-bit assembler.

I'll share some pics of the case inside when I've done a bit more wiring.

Neil

[Michael GilliganParticipant @michaelgilligan61133]

Impressive work, Neil

Just one little negative point, if I may … It is generally accepted that Red is a better colour for use in a dark environment.

MichaelG.

[Bodger BrianParticipant @bodgerbrian]

Posted by Neil Wyatt on 25/11/2015 19:13:11

It took me a while to understand 'nets', if you use these instead of a wire to join up power pins etc. then you don't need to physically join them up on the circuit diagram. This is what I have done with +5, +12 and GND.

I hesitate to use the word 'always', so I'll settle for 'generally' or 'usually' when I say that any connection between two or more points is called a net (some software even assigns a net name to a pin that has no connection but that's another story). The type of connection that you're referring to is often called a 'global net' or 'global' for short.

Apologies for digging up an older post…..

Brian

[Neil WyattModerator @neilwyatt]

Posted by Michael Gilligan on 02/05/2016 21:29:15:

Impressive work, Neil

Just one little negative point, if I may … It is generally accepted that Red is a better colour for use in a dark environment.

MichaelG.

Try finding a RED 64 x128 LCD display! Swapping out the LEDs on these waveguide -type displays isn't a trivial exercise.

That said, there is a counter-argument that as the eye is more sensitive to green than any other colour, you can set a green display much dimmer than a red one and a display doesn't have to be very brightly illuminated at all to be readable at night – much dimmer than my red torch. My existing unit uses a 2×16 green LCD and it doesn't cause me any problems.

And having said all that… I have a QVGA full colour display that I might drop in instead, which would do a very good job of red text/lines on a black background. But then I will probably want to add an SD card to display colour pictures of the objects…

Neil

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 02/05/2016 22:03:40:

That said, there is a counter-argument that as the eye is more sensitive to green than any other colour, you can set a green display much dimmer than a red one and a display doesn't have to be very brightly illuminated at all to be readable at night – much dimmer than my red torch. My existing unit uses a 2×16 green LCD and it doesn't cause me any problems.

.

The problem is more to do with Loss [vs. retention] of Dark Adaptation.

But; if dim Green works for you, who am I to argue.

[You may not be popular at a typical 'Star Party' though.]

MichaelG.

[AjohnwParticipant @ajohnw51620]

I don't know if you are going to get into actually making a mount as well Neil but another thread reminded me of something. Many people make their own worm wheels but Celestron used another technique on my first slightly larger scope, 8" sct. They used a spur gear and drove it with a worm. It worked rather well. They also used the same arrangement on larger ones.

There are several gear pitches that are very close to threads that can be easily cut. Even if there is a few thou difference it's probably going to require less lapping than a typical home made worm wheel. It would also be possible to lap the gear against others based on different primes. That's how pro telescope drives used to be made.

I never had to take it apart but what I would expect to see is a worm of larger diameter than the thickness of the gear angled to account for it's helix angle.

All but the most expensive mounts have lots of spur gears in them anyway so why not use one directly. There is also a fad for toothed belts at the moment but as anything will stretch to some extent they don't make much sense really. The cogs will have periodic errors anyway. Chain drive might make more sense.

John

–

[Neil WyattModerator @neilwyatt]

Mount is next year's project

I suspect periodic error is mainly an issue for the worm – a theoretical 10-minute sub on a 144 tooth gear would only use on tooth & space on the wheel, but a whole turn of the worm.

Neil

[AjohnwParticipant @ajohnw51620]

I like fork mounts Neil. No counterweight so far more rigid and lighter. The 8" SCT I mentioned was sold due to aperture fever. I went for 10" which is still around but have regretted selling the 8" many times so recently bought another. I'm encouraged by astro shots I have seen when used with a focal reducer.

I've also added a good quality 6" newtonian as a coma corrector will sort that out. It's fairly happy on an EQ3 with an EQ5 tripod but as I mentioned before if I added weight to the end of the newtonian I'm pretty sure it could be fork mounted and be even more stable – on day maybe and then look at directly driven spur gears.

I've recently sold a couple of other scopes, both apo's so the money went into more workshop gear and another lens for my Nikon V2, the long zoom for it. I'm aiming to try M 4/3 on astro but the local council recently messed things up for me. LED street lighting. It was lp sodium and a simple filter killed that completely even just being a couple of miles south of the centre of B'ham. Not much good looking north but south was fine.

John

–

[Neil WyattModerator @neilwyatt]

Despite lack of posting, I have done a fair bit recently.

The mount now appears to be performing it's GOTO moves (just a minor issue of clockwise and anticlockwise getting mixed up, but that's a minor tweak.

More detail to follow soon…

Neil

[john carruthersParticipant @johncarruthers46255]

I've fettled a few mounts that use a friction drive rather than the traditional wormwheels. No backlash, but a small chance of slipping under acceleration, which can be ramped up and down of course.
One chap in Ireland has used a large diameter small pitched toothed belt. Then there are the 'caterpillar' drives, sectors, drum/wire etc…
So many solutions to the same problem

I find the worm bearings on mass produced mounts a major source of PE, better quality bearings or angular contact bearings can make a measurable improvement. Also worm end float needs better control.
A lot of worms have their cetre holes so a conical set screw on one or both ends can help.

The belt conversion kits are a good idea too, doing away with the spur backlash.

Edited By john carruthers on 24/06/2016 08:12:48

[Author]

Posts