Relays for Lesvedome

[pmlogg]

Hugh and Alan

Many thanks for the fault finding in the circuit so quickly.  Below I have redrawn, using my overly-complex layout as I have it on the computer as an image.  I've drawn in a connection from the battery 12V- to DC- on the Relay Board.  I've removed the DC- on the Relay Board to GND on the VM110 and also the 560 ohm resistor.  I've changed the wiring for the NO and NC on relays one and two, in compliance with Hugh's board, putting them on the opposite side of the diode from where I had them placed before.  I also reversed the battery symbol.

I drew in but then removed an extra connection from the 7ah battery 12V- to NO1 and NC2 as they seem already to be connected to 12V-.  Are those the GND connections that you think I need or are we talking about a separate ground line, e.g. to the metal plate that forms the housing of the Pulsar/Rigel system?  If so should the GND on the VM110 also be connected to that point?

For the moment I've left in the double pole switch to the 7Ah battery because otherwise there would seem not to be any 'off' position.  In my diagrams I've shown the 12V charging socket which needs to remain live for the solar panel to maintenance charge the 7Ah battery.  Would it not be better to isolate the two rotation circuits when not in use with a switch?

Concerning the Raspberry Pi to VM110 connection, that is via a USB to USB micro cable.  Not sure how I check grounding of that, that USB micro socket is the normal power in to the Raspberry Pi and I've not spotted any other GRD connection point on it.

Thanks, Peter

[pmlogg]

Alan

Re-reading your post, should I still have a link from DC- on the relay board to one of the GRD connections on the VM110?

Thanks

Peter

[hughgilhespie]

13 minutes ago, pmlogg said:

Alan

Re-reading your post, should I still have a link from DC- on the relay board to one of the GRD connections on the VM110?

Thanks

Peter

Yes!

Regards, Hugh

[pmlogg]

Hugh

Thanks, I'll put it back.

Peter

[pmlogg]

Alan

Just noticed that I've still shown 3 x 5K ohm resistors; I've corrected to show 1K ohm.

Thanks, Peter

[symmetal]

Here's the updated drawing with the amendments. Hugh's is much neater on showing the relay output modes. 

NO1 and NC2 are not connected to -12V, as they connect +12V to the motor when relay 4 is turned on.

It's best to have 1 common ground terminal close to the 7Ah battery and connect the high current Gnds to that point (7Ah -ve, NC1 and NO2, Rigel 0V, relay board DC-). Also connect one of the Gnd connections on the VM110 to this common ground too. The hall effect Gnd can go to one of the other Gnds on the VM110 board.

I'm sure the Pi USB 0V is connected to the VM110 Gnd on the board anyway but you can check that the azimuth sensor power negative terminal has continuity to VM110 Gnd when you've wired the Pi USB cable to it and the Pi is connected to the VM110 via the other USB cable.

Using the double pole switch on the battery doesn't give an off position on the battery, as it's connected to either your new system or the Rigel System. A separate on/off switch wired between the charger connection and the changeover switch would provide a battery off, or else use a 3 position on/off/on switch instead of the changeover switch.

Alan

Edited May 23, 2019 by symmetal

[pmlogg]

Alan

Thanks for that explanation and for doing more drawing for me.  I have clearly not drawn the DPDT switches correctly nor explained them.  The two I have are actually 3 position with the central one having no continuity.  That's what I meant by "off".  I suppose that the motor switch could have been a more basic two position switch as with the main switch at off there is no power flow anyway.

On the connections to battery's 12V- this comes back to my problems of interpreting some of the example diagrams accessed via Lesvedome e.g. https://github.com/tonygilkerson/astrobagel/wiki/Observatory-Automation-with-LesveDomeNet and http://eplumer.blogspot.com/2010/01/observatory-gets-automated-dome.html which show more direct contacts with 12V- than just ground. 

What I need to do I think is trace them all in my head to see that they are essentially all just ground connections.  Physically however what is that connection, if it's not to the battery's 12V-?  In buildings there are ground straps to earth but the rotating part of the dome is essentially insulated from earth by the plastic rollers on which it moves and the rubber tyres on the motor drive roller and guide roller which are in contact with the non-rotating walls.  So even the metal housing of the Pulsar/Rigel system is not a true ground. 

Thanks

Peter

[symmetal]

Treat the battery negative terminal and everything connected to it as Ground or 0 Volts. I assume all the equipment is fixed to the rotating roof so as you say there is no actual physical. connection to the earth for a true Ground connection. Ground in your instance is just a nominated point which you can call zero volts to which all the other voltages are referenced to. As your system has no physical ground connection you can't make any voltage measurements with your multimeter if you connect your black meter lead to a true ground like you will have in your home. So you have to nominate one voltage in your system to act as a ground for making measurements. The most logical is the battery negative terminal.

Talking in terms of 12V+ and 12V- could imply +12V and -12V which have 24 volts between then. Calling the 12V- terminal 0V or Ground for convenience, removes any ambiguity in labeling.

As your system is only 12V powered you don't need a physical ground or earth connection for safety reasons like you would with mains voltages. Ground and Earth are the same thing but it's more common to use Ground particularly with DC. 

From Wikipedia

Quote

In electrical engineering, ground or earth is the reference point in an electrical circuit from which voltages are measured, a common return path for electric current, or a direct physical connection to the earth.

The use of the term ground (or earth) is so common in electrical and electronics applications that circuits in portable electronic devices such as cell phones and media players as well as circuits in vehicles may be spoken of as having a "ground" connection without any actual connection to the Earth, despite "common" being a more appropriate term for such a connection. This is usually a large conductor attached to one side of the power supply (such as the "ground plane" on a printed circuit board) which serves as the common return path for current from many different components in the circuit.

Alan

Edited May 23, 2019 by symmetal

[pmlogg]

Alan

Thanks for the explanation.  I think it confirms that 0V is in effect the negative terminal of the battery.  So does that not mean that the connections from that point are not all, in effect, ground connections.  I've added numbers to my diagram for the purposes of going through this. 

Point 1 connects the negative terminal to ground on the VM110. You said that: "The three Gnd connections on the VM110 would be joined together on the board so there is no need for all three to be connected together to Gnd externally" so that connection would take all 3 of the VM110 GRDs back to the negative terminal?

Point 2 connects the negative terminal to DC- on the Relay Board so grounding it?

Point 3 connects the negative terminal to NO1 and NC2 so grounding them?

Point 4 connects DC- to GRD on the VM110.  Hugh says I need this but if the VM110 is already grounded from Point 1 why is that additional connection to the VM110 needed?

Point 5 connects the Hall Effect Switch to GRD on the VM110 so that by the internal connection on the VM110 takes that back to the negative terminal, grounding it?

From what you wrote about the EM14 would be grounded via the Raspberry Pi.  In turn, is the Raspberry Pi 5V grounded at the Power Bank?

On another point I'm confused over the connections to the output terminals of the Relay Board.  Ignoring its layout, is not my diagram similar to Hughes for the interconnections of NO4 and the protecting diode?  Is yours the same as Hugh's?

Thanks

Peter

[symmetal]

Peter,

Yes, the negative terminal of the battery can be called 0V or ground, whichever one you prefer, and anything that is connected to the battery negative terminal can be called 0V or ground as well.

Point 1: Correct

Point 2: Correct

Point 3: This grounds NC1 and NO2 which is what I think you meant to say and not NO1 and NC2

Point 4: Relay board DC-does need to be connected to a VM110 Ground,  but it already is by they connection in point 1 as you say. Hugh was just confirming that point and not saying that a second ground is needed to the VM110 though it could have been interpreted that way.

Point 5: Correct

The raspberry pi 0V power wire on the USB cable to the VM110 would be connected to the VM110 ground terminals on the board itself which is how the EM14 gets its ground. When you plug the powerbank into the raspberry  pi the powerbank will automatically connect its negative output terminal to ground too so no extra ground connections are needed there.

Your drawing of the relay board connections do look the same as mine and Hugh's drawings.

Alan

Edited May 24, 2019 by symmetal

[pmlogg]

Alan

That's great, I guess some of your and Hugh's knowledge must be rubbing off on me. I'm not as quick a learner as I once was I'm afraid.  Oops on Point 3, yes that was what I meant.

With all that confirmed I'll get on with the trial wiring.

Thanks, Peter

[symmetal]

On 23/05/2019 at 20:55, pmlogg said:

I have clearly not drawn the DPDT switches correctly nor explained them.  The two I have are actually 3 position with the central one having no continuity.  That's what I meant by "off".  I suppose that the motor switch could have been a more basic two position switch as with the main switch at off there is no power flow anyway.

Sorry, I didn't pick up on this point. You seem to have on/off/on 3 position switches anyway for the battery so that's fine. You still only need a single pole switch in the battery +ve connection as when that is in the centre 'off' position the battery is disconnected power wise. The battery -ve terminal can be permanently left connected to ground. You can if you wish leave it as you've drawn it and it will be fine. It's your choice. 

As you say the motor switch only needs to be a 2 pole changeover switch but if you're happy using the three position switch you have, there's no problem.

Alan

[hughgilhespie]

Hi Peter,

Another couple of comments about your circuit. First, you are showing a connection between DI3 and DO3. This is not needed for your application. Or - more accurately - it is only needed if you are going to use the Lesvedome ASCOM Switch driver. For the ordinary ASCOM rotation driver, it isn't needed. It won't do any harm and feel free to leave it in but I thought I would clarify a bit.

The second comment might be more important. You show a diode in the connection between DO2 (dome motor direction signal) and DI4 (tells the Lesvedome software what the dome motor direction actually is).  This connection is definitely needed for the circuit to work but the diode is not required. My setup works perfectly with just a piece of wire connection the two. If you look at the Charles Harrow schematic, which is considered the 'gold standard' for implementing the Lesvedome system, there is just such a direct connection. As we have discussed, the Velleman digital outputs are ACTIVE LOW, so the diode orientation is correct but it introduces a voltage drop of 0.5 - 1.0 volts and this MIGHT be enough to cause problems. Better to leave it out and just use a piece of wire. And think of all the money saved!!!

HTH.

Regards, Hugh

Harrow_Schematic.pdf

Edited May 25, 2019 by hughgilhespie
Added copy of Charles Harrow schematic

[symmetal]

The diode between O2 and I4 on the VM110 I believe is there to protect the input, if the output voltage (when the output is turned off) is pulled up above above the maximum voltage the input can handle (which is 30V). The VM110 can drive 48V relays (or a load up to 50V and 500mA) and in this case the input would likely break down without the diode.

As you are only using 12V relays this wouldn't happen so the diode can be safely left out and the O2 to I4 connection made with a piece of wire as Hugh suggests.

Alan

[pmlogg]

Hugh and Alan

Thanks for those two useful bits of simplification and great timing too.  I've been working on the wiring this afternoon and evening and have stopped for tonight just before getting to the connections requiring diodes and resistors.  So I will just take the wire I had put in between DI3 and DO3 and put it instead to link O2 and I4.

Thanks, Peter

 

[pmlogg]

Hugh and Alan

I didn't manage to finish last evening - soldering the few discrete components took longer than planned.  Also, I've been using cabling that I already had for my test wiring which probably a mistake as little of it is single core, and I don't have much of that, without stripping it from 2 and 3 wire cable at least partially to make connections - a slow process.  I have however been erring on the side of caution with cable diameter.

I have noted that the cable entry points of the VM110 and one side of the Relay board are relatively small, so not appropriate for thick wire.  What would be the minimum AWG rating/diameter you would recommend for the data and for the power connections - as I think I'd best buy some for a tidier final result.

Thanks, Peter

[hughgilhespie]

Hi Peter,

Have a look at this table. https://www.engineeringtoolbox.com/wire-gauges-d_419.html

For single core cable, 18 awg would be OK but 16 awg would be a bit safer. If the cost difference isn't too much I would recommend 16 awg wiring.

Regards, Hugh

[pmlogg]

Hugh

Thanks for that - I'll go for the 16 awg.

Peter

[symmetal]

For the data connections (all the wiring to the VM110 board) you can use thinner wire like 22 awg as none of it is high current. Only wiring going to the relay board COM, NO and NC contacts, motor, 12V battery, fuse, rigel system, and associated switches needs to be of 16 awg or similar (ie. any wire that can be carrying current going to the motor.)

Using multi-strand wire of a similar overall thickness to single core is fine. It's current rating is only slightly less that the single core and won't be a problem in your case.

Alan

Edited May 27, 2019 by symmetal

[pmlogg]

Alan

Thanks for that.  I've ordered some wire so when I go for the final wiring up that should make for a better and more tidy setup.

Peter

[pmlogg]

Alan and Hugh

That's the test wiring completed - but no power applied yet.  I'll not send an image as it's not a pretty sight at the moment.

I've not yet made the offset measurements for Poth, nor tried to start it in advance of inputting  them.

Thanks, Peter

[hughgilhespie]

Good Luck!!

The Lesvedome User Interface is a nice, simple way of testing if the hardware works - or not!

If you haven't already downloaded this it's automatically downloaded when you download the Lesvedome ASCOM driver. The User Interface is a stand alone program that is used to input the configuration parameters for the dome and azimuth sensor and to test the operation. After the original download you can find the UI program in C:\ProgramFiles (x86)\Common Files\ ASCOM\Dome\ASCOM.LesveDomeNet\ASCOM.LesveDomeNet.exe. I suggest you create a shortcut to the program as it is rather tucked away.

HTH

Regards, Hugh

[pmlogg]

Hugh

Thanks, I was wondering about testing. I'd noted reference to simuation ad buttons for it in Ascom Poth.

I have the Ascom Platform and Lesvedome installed, and I did find the setup buttons. for parameters  I've inputted the various measurements, offsets etc. after checking them in the dome.  One of the beds is currently covered with 'system' components including those I've wired up for testing.  I then had a pause when I was back to some hardware work. 

The Bourne EM14 encoder is housed in a Hammond 1590 aluminium enclosure, with the Hall Effect Switch on a bracket attached to its side.  In my first test of the box the encoder wheel was losing contact in places in just one quadrant of the dome.  I attached a spring to pull the wheel into contact.  That worked but I'd just used bits lying about.  I ordered a bit of stainless steel and a stainless spring to avoid rust, so that meant an evening of measuring, drilling etc.  An added complication was that one corner of the enclosure's back plate fractured.  Very impressed by the Canadian manufacturer who agreed that it shouldn't have happened and sent me a free replacement.  That also needed a couple of holes drilled and tapped. To simplify mounting of the Raspberry Pi, VM110, Relay Board and my little PCB with the 3 resistors and two diodes, I've cut a plastic cutting board to fit into the enclosure.  My thinking is to mount all the components to it and make all the connections before sliding them, as a unit, into the plastic enclosure box.

The next 4 days is taken up with a family visit so all is on hold until after that.

Thanks, Peter

[pmlogg]

Hugh and Alan

After farewells to family I went back and worked on the physical bits and pieces so only tonight did I power up for the first time (on the bench).  Power was getting through to the Pi, the VM110, the Relay Board and the Hall Switch - all had LEDs illuminated, and one relay clicked too.  Using Virtual Here the Pi was recognised as a hub, connected to the VM110.  Lesvedome and POTH both reported 'connected' to the Dome but when I gave it a position to slew to there was no relay clicks, so no motor activity and Lesvedom disconnected. It reported "Motor stalled".  Another issue was that the red LED on the Hall Effect Switch should only illuminate when passing a magnet. It stayed on with the magnet making no difference.  I've checked all the wiring as far as the socket on the enclosure for the Hall Switch and Encoder, all seemed OK, and as per the final diagram..  I'll have to leave opening it the enclosure up until Friday night - too tired now.

[hughgilhespie]

Hi Peter,

It sounds as if you are nearly there! The main thing I would recommend at this stage is don't try and connect everything at once. Just connect the Pi and use the Lesvedome User Interface to see if you can get the correct relays to operate. There was a recent change to Lesvedome - I quote

Hello Nicolàs

You probably use the beta version 6.0.1.20. With this version if the dome moves less than 3deg in 3 sec. The dome is declared as stalled and it disconnect.
If you don't want this feature use version 6.0.12

Clear skies,

Pierre

Pierre de Ponthiere (Belgium)
 

I think this explains why Lesvedome reported 'motor stalled' and disconnected itself.

 

I suggest you try disconnecting the Hall effect switch and seeing if that helps. Let us know how you get on this evening and I'm fairly sure it won't take much to get you going.

 

Regards, Hugh