Relays for Lesvedome

[pmlogg]

As part of my long-running project to automate my dome's rotation I think I may need to change plans again to bring the design more in line with what others have done, and in particular to use the relay arrangement set out in many of the circuit diagrams, a single relay board e.g. ELK 912 linked to a double relay board, e.g. ELK 924. However these two components only seem to be available from the U.S.  Could anyone suggest equivalents for each that might be stocked by UK suppliers?  Attached is a modified circuit diagram showing the relays, attached to a Velleman board, but with the manual rotation system left off.

[SlimPaling]

Farnell show that they seem to have the 912 and 924 in stock.

Their web site is : https://uk.farnell.com/

Cheers Mike

 

 

[pmlogg]

Mike

Thanks for that.  I see though that they would have to come from the US with a £15.95 shipping cost.  My most recent attempt to buy from the US had mixed results.  The parcel was rushed by FedEx Priority  from Texas through other states, Paris and finally to Scotland but I missed the delivery, having only been told arriving by 6pm when I was already at work.  It went back to the depot 40 miles away which then promptly shut not just for the 3-day bank holiday but added an extra day for good measure.  So, if I could get an equivalent in the UK I would prefer it.  I see relays listed on ebay that might do the job but my electronic knowledge is not good enough to differentiate which are good quality.

[pmlogg]

Mike

Just went through as far as the Checkout - final cost with the shipping would be £46.97.

[SlimPaling]

1 hour ago, pmlogg said:

Mike

Just went through as far as the Checkout - final cost with the shipping would be £46.97.

Why not give them a phone call and check the cost of shipping? .... these items are shown as being "in stock" ... but the shipping costs that they are asking is criminal !!! I hadn't noticed this part when I looked them up earlier.. These items don't weigh very much and should not need to cost that much to send them within the UK!

Alternatively phone up Radio Spares to see if they have an equivalent relay board? .... their shipping costs are included in the price of the item ... basically zero. I have ordered various bits & pieces from them recently and I have never paid any extra for shipping ... and things often arrive the next day.

Cheers Mike

[pmlogg]

Mike

I could call them but the stock is shown as being in the U.S. with the standard charge being applied.  I agree the cost, for such small components is ridiculous.  I have searched the RS site and can't see anything like the ESK components.  It does surprise me given that they come up straight away on Ebay, from China of course.  I'm still hoping someone may be know of a suitable equivalent to be had off the shelf in the UK.  We'll see.

Thanks, Peter

[Dr_Ju_ju]

From the published specs, almost any 12v SPDT\DPDT modular (pcb mounted) relays will do what you require. If you are handy with a soldering iron, you could easily knock-up your own...

[teoria_del_big_bang]

Would something like this do?

4 channel 12V relay board

 

Steve

[symmetal]

For wiring without needing soldering you can use these 12V 10A DPDT octal base relays with Relay socket. You can just use one set of contacts for the SPDT relay, so you only need one relay type. You only need to then get 1 spare relay which will fit either socket. 

The ELK-924 only needs 1mA to operate the relays as it has on board driver transistors while the RS relay needs 100mA. If the Velleman P8055 which provides a switched ground to operate the relays can't sink the relay coils 100mA, an interface driver transistor will need to be incorporated which makes the design slightly more involved, but can easily be built on a bit of veroboard or similar. The Dig Out1 can operate the SPDT relay directly so can sink enough current. If the Dig Out2 is similar you should be OK without a transistor interface.

Alan

Edited May 6, 2019 by symmetal
Typo

[pmlogg]

5 hours ago, teoria_del_big_bang said:

Would something like this do?

4 channel 12V relay board

 

Steve

Thanks for all that advice.  A board solution was what I had in mind to minimise extra circuitry but I will look at those options.

Peter

[hughgilhespie]

Hi pmlogg,

I am happy to help you if I can. However, there are a few questions that need answers so that you get the best help.

First off, are you actually going for a 'full' Lesvedome / ASCOM interface or just a manual solution for your dome rotation?

The second question is what sort of motor will you be using to rotate the dome. The choice of motor will determine what relays are suitable.

The good news is that the digital outputs from the Velleman K8055 board are buffered by a ULN2803 chip. This will handle relay drive currents up to 500 mA and voltages up to 50 volts which would be sufficient for driving almost any low voltage relay. The ULN2803 also incorporates protection diodes and the outputs can be connected directly to a relay coils. However, be aware that the power to drive the relay does not come from the Velleman board. You need to connect the Velleman 'Clamp' terminal to your own power supply.

There are literally hundreds of board mounted relay modules that come with screw terminals. The choice will depend on the current rating needed for the motor. To implement the circuit you have shown, a board with 4 spdt relays would do nicely. This https://www.ebay.co.uk/itm/Mini-4-Channel-DC-12V-Volt-30A-Relay-Module-Control-Board-High-Low-U0P3/113195531357?hash=item1a5afabc5d:g:OuoAAOSw441bCEZX is a 4-channel 30 Amp relay board that is the sort of thing that would  work for your application.

Let us know the answers and we can offer some more focused advice.

Regards, Hugh

 

 

 

[pmlogg]

Hugh

 My original model of Pulsar 2.7m dome is already motorised, but not automated, with an original Pulsar-source 'Rigel' system, battery powered and fitted to the rotating part of the dome.  It uses a 12V battery, which has a charging socket for solar powered charging.  I really only use that for maintenance/trickle charging and instead charge from the mains in the dome after each session.  The motor is a Parvalux PM4-LMS but I cannot at the moment find the specifics of its gearing and output speed.  I suspect that it will be one of the lower speed models (e.g. 1500 rpm motor speed) as that draws less amps (5.4 full load) so more suitable for battery operation than the higher motor speed versions (16.5 amps full load for the 4000rpm version).  The 4 relays on the Pulsar/Rigel Dome Controller board are Omron G5LE-1  1159W8.

My intention is to retain that system as the 'manual rotation' vs. computer controlled system but to add Lesvedome control for slaving to the mount.  My thinking is to use a pair of switches to isolate the Pulsar/Rigel controller board and switch the battery power and access to the motor to the Lesvedome system. 

One of the main issues was wireless communications to link the static computer with components on the dome.  The solution I'm attempting for that is to add a Raspberry Pi 3B, powered by a 50,000mAh power bank, running VirtualHereCloudHub.  That, with VirtualHere Client installed on the indoor computer provides wireless access to the usb port on the Raspberry Pi 3B and to the VM10 attached to it.  So Lesvedome runs on the indoor and sees the remote, dome-mounted VM110 (same as K8055) as if it was directly attached by a usb cable.  The Raspberry Pi needs no operating system nor ancillaries - it just runs the one program and performs a single function.  That link has been tested and works.

What that allows is a 'conventional' connection of the VM110 to the motor relays, optical encoder (Bourns EM14) and home position switch.  If I decide to move on to shutter motorising and automation the dome mounted components for that could also be direct connected to the VM110 for Lesvedome control. 

I'm suffering somewhat from my deviations from the Lesvedome standard set up.  I'd bought a cheap PCB mounted pair of relays that provide separate cable pair connections for CW and CCW commands - but the normal Lesvedome arrangement instead seems to use an SPDT relay (e.g. the ELK -912) and then a DPDT relay (e.g. the ELK-924) so it seems to me that I need to accept that I have the wrong relays and change either to the ELK ones, or similar, hence my question.  Thanks for confirming the same relay board that Mike did.  Could it replace both the dual relay ELK -924 (designate as a DPDT relay) and the ELK-912 SPDT relay?

I'm also planning to use a Hall Effect Switch rather than Reed Switch.  It's complication is that there are 3-wires rather than the two shown in most Lesvedome wiring diagrams.  My presumption is that the negative, non-signal line could be connected to ground.

Thanks

Peter

 

[symmetal]

The board that Hugh and Steve suggest has four independent SPST relays so one relay can be your SPST relay, and two relays with their trigger inputs connected together will become your DPDT relay.

A hall effect switch needs to be powered all the time to function, via the Vcc and Ground pins, while the Output pin is usually an open collector transistor (or open drain mosfet) like this

If the reed switch you want to replace with the hall effect switch is the Home Sensor switch then you just connect the output pin to Dig In2, Vcc pin to +12V and Ground pin to Gnd. Make sure you use a hall effect switch that can operate from 12V as some are 10V max working.

If the reed switch doesn't have one terminal connected to ground then it's not so simple to swap it with a hall effect switch.

Alan 

Edited May 8, 2019 by symmetal

[sloz1664]

Hi Peter,

I use the Levesdome system to run my Pulsar Dome. I used a std motor vehicle SPDT relay from a motor factors. I also use the hall effect switch for the home position. I tried magnetic reed switches but they were troublesome. You need to note that a 10 kΩ  pull up resistor is required between the 5v pin and the digital pin. I obtained my 5v supply from the K8055 board using the upper pins of either SK2 or SK3 5v outputs.

Steve

[hughgilhespie]

Hi Peter,

Many thanks for the detailed explanation of what you are trying to do.

I have no experience with Raspberry Pi's and their software so I can't comment on your proposed wifi link but as it is tested and working I am sure it will be fine.

So, if I am understanding correctly, you will install the VM110 board and the associated relays in the rotating part of the dome alongside the existing Rigel kit. Then you will have some sort of master switch that will allow you to use either the Rigel system manually or the Lesvedome software remotely to control the dome rotation - and hopefully - the shutter opening and closing at a future stage.

More questions I'm afraid. As far as I know, most people who implement the Lesvedome system use the Charles Harrow schematic with 6 dpdt relays. Is there a particular reason you are using an alternative circuit? The Harrow design would give you all the shutter control circuitry for the future. From my own (bitter) experience I can tell you that it is MUCH harder to try and modify things later than to try and include everything you might want from the outset.

Next the Bourns encoder. I am sure you know this but the VM110 board can only cope with a relatively low pulse rate from the encoder. The pulse rate of course depends on how many times the encoder turns per dome revolution and the dome rotation speed as well as the inherent ppr rating of the encoder. For my Pulsar set-up I needed an 8 ppr encoder. Anything above that was too fast for the VM110 board.

I would strongly suggest you do NOT use a Hall effect sensor for the home position sensor. These sensors have a high level of hysteresis and are not really suited to position measurements An optical switch would be much better. I used an Omron EE-SX3009-P1 switch. This came from RS I believe. This switch has a slot with an IR light emitting diode on one side and a phototransistor on the other. When anything interrupts the light path it triggers the switch. As the slot width is only 5 mm wide you are guaranteed a home position accuracy to better that 5 mm. As Alan explained, these switches, like the Hall effect switches, are active devices and need power connections as well as the switch output. The switch that I used operates at 5 volts as does the Bourns EM encoder. Bear in mind that you will need a suitable 12 volt to 5 volt converter to power these. Again, there are hundreds of suitable small buck-converters with screw terminals on fleabay.

How you actually make the stuff depends on your level of experience. My method (a very basic method indeed) is to first decide on what modules are needed and to get their dimensions. Then I do a drawing (I am very old so I use Microsoft Visio) that shows all the modules and where there connections are. Then move the modules around on the drawing until I have some sort of reasonable layout. Then I add the wire runs to the drawing. The idea is that you end up with a layout that can be made by screwing everything to a single sheet of Aluminium. Obviously you need to make sure that the aluminium sheet will fit somewhere in your dome. A quick aside  - I had a horror of drilling holes in my Pulsar, so I mounted everything on plywood panels that I glued to the inside of the dome. I use CT-1 construction adhesive, very good stuff!

So, I suggest that the next step for you is to finalise the design as best you can.

Step 1 : Which circuit will I be using?

Step 1 : What modules do I need? Generic at this stage, e.g. a relay module, a power supply module, VM110 board, master switch module.......

Step 2 :  Decide on the actual components - which relay module? which power supply module? ..................

Step 3 : Best thing is to actually buy them but if not get hold of the dimensions somehow. 

Step 4 : Work out how big your base plate can be and what size it will be then do a drawing of the shape.

Step 5 : Add simple drawings of all the modules and where there connections are

Step 6 : Fiddle about for hours moving things around until you're happy then add the wires / connections to the drawing.

Step 7 : Buy a bit of aluminium sheet - hopefully already cut to size - drill the holes to mount everything, and ....

Step 8 : Make it!!

I know this is all a horrible over-simplification but it really is doable. Let us know which circuit you want to use and I am sure we can come up with some more definite help.

Regards, Hugh

[pmlogg]

Hugh

I have not used the Charles Harrow diagram for a couple of reasons.  First it doesn't give me, as someone unfamiliar with the components, enough detail e.g. it shows the relays but not how to wire them up. Secondly it includes AC components which again confuses me as mine has to be all DC.  Finally, when at this stage I just want to wire up for the non-manual control circuit the inclusion of the shutter control, scope power and manual control again just serve to confuse me as I've not been able to figure out which components and connections I can leave out.  The diagram I posted is an edited form of one posted at this link:  https://github.com/tonygilkerson/astrobagel/wiki/Observatory-Automation-with-LesveDomeNet.  That showed me the actual connections and it was easier for me to edit out the manual control.  It does not show the azimuth sensor but others have posted or told me how they connect their EM14 encoders and it conform with the connections o the Harrow diagram - so I've not shown that on the diagram, trying to stick just to the parts where I was confused.  Likewise I left out all the connections to the Pulsar system.  Do you disagree with the connections shown in the diagram I posted i.e. would that circuit not work?

Views seem to very on the Hall Switches as you can see from the responses. I went for one for a couple of reasons.  The main one was that most Reed Switches seemed to be based on mains power rather 12V DC.  But I already had a working Hall Switch and magnets so I thought it worth trying.  I may need a buck converter for it as it's voltage range is 6-36V so too high to just take 5V from the VM110.  I do have the option to take 12V from the powerbank simultaneously with 5V.  If the sensor doesn't perform then I can try  what you suggest.

On the encoder I changed from the first one I thought to use as it's connections were non-standard.  I checked the Lesvedome spreadsheet and because my dome rotation is quite slow, and the EM14 version I've bought is 8ppr so I think it should be OK.  I do need to resolve at least one issue and that is keeping the encoder wheel in full contact with the dome wall - at the moment it comes off during part of one quadrant. But that's a mechanical issue I think I can deal with. 

I have an enclosure that I bought when Maplin was closing down.  It seems big enough for the components but as you say once I have them all I will tinker and if necessary get something else.  I too have been very wary of drilling the dome. Pulsar tells me just to use a good bit, suitable for steel.

Thanks for your advice to date.

Peter

 

[pmlogg]

Alan

Thanks for that advice on the relays and switch. I've modified my original posted image, replacing the ELK relays with the Ebay one (showing just the connections). Do you think I have got those connections to Relays 1, 3 and 4 on the board right?  I've not yet drawn in the Hall Switch.

Thanks

Peter

[pmlogg]

Steve

Thanks for that on the Hall Effect Switch.  Yes, I have added a resistor to the circuit. When I tried without it I got no signal, with the resistor the LED lights up and the full 12V passes through.  I do find it confusing that the circuit diagrams for the Reed switch don't show any power coming into the sensor, just showing connections to 12V negative and Digital Input 2.

Thanks

Peter

[sloz1664]

58 minutes ago, pmlogg said:

Steve

Thanks for that on the Hall Effect Switch.  Yes, I have added a resistor to the circuit. When I tried without it I got no signal, with the resistor the LED lights up and the full 12V passes through.  I do find it confusing that the circuit diagrams for the Reed switch don't show any power coming into the sensor, just showing connections to 12V negative and Digital Input 2.

Thanks

Peter

You don't need any power for a reed switch, it is purely a magnetic switch.

Steve

[pmlogg]

Steve

That makes sense, I was confused by references to 220V with some - but I guess that was just the line voltage for the alarm system to which it was fitted.  Silly me!

Peter

[symmetal]

6 hours ago, pmlogg said:

Alan

Thanks for that advice on the relays and switch. I've modified my original posted image, replacing the ELK relays with the Ebay one (showing just the connections). Do you think I have got those connections to Relays 1, 3 and 4 on the board right?  I've not yet drawn in the Hall Switch.

Thanks

Peter

Hi Peter,

There are a couple of things that need changing on your diagram.

The relay board needs to be permanently powered by the 12V. The DC- currently goes to Dig Out2 so the board is only powered when Dig Out2 goes low. If the board isn't powered then your SPST relay 4 can't be operated. This didn't matter on the ELK-924 board as it didn't affect the ELK-912 SPST relay power. So Dig Out2 should just go to the IN1- and IN2- connections and not DC-, and DC- should be connected to Gnd.

Dig Out1 switched the SPST relay on the ELK-912 board by grounding the coil -. On your new diagram you've shown it going to DC- and so connecting DigOut1 to DigOut2 which isn't what you want. Dig Out 1 needs to go to IN4- only, to enable the SPST relay 4 to be switched. Currently there is nothing switching relay 4. This is also why the relay board needs to be permanently powered, to allow relay 4 to be switched independently from relays 1 and 2.

I'm surprised that a pull up resistor is needed with the hall effect switch when connected as shown, as the Dig In2 input should have its own pull up resistor. The current reed switch just grounds Dig In2, and no pull up resistor is shown, which is all that happens when you connect the hall effect switch output to Dig In2 instead. There's no harm in adding another pull up resistor as long as it pulls up to the logic voltage the Velleman board uses.

The hall effect switch can be powered from 5V or 12V or any voltage within its specified ratings and won't have any effect on its operation. This is assuming you have one with an open collector output, which most are. This is a benefit as it isolates the output switching from the input power. Being open collector it would need a pull up resistor somewhere but this can be pulled up to any positive voltage which suits the circuit it's connected to and need not be the same as the hall effect switch power voltage.

Alan

Edited May 8, 2019 by symmetal

[pmlogg]

Alan

Many thanks for checking the circuit.  I've made those changes and also added in the optical encoder. Based on other users comments and diagrams I've put it's + terminal to clamp but if that is 12V, from what Hugh wrote I need to add a buck converter to drop that down to 5V.  I've not drawn that it.

I've still not drawn in the Hall switch, leaving for the moment the reed switch from the circuit I copied from.  The Hall Effect switch I have is 3 wire - with the positive (brown) and signal (black) showing potential across them when the magnet is sensed.  The resistor links black and brown, as I've not tried it attached to the Velleman board I don't know if the resistor is needed or not..  It needs power in 6-36V so in replacing the Reed Switch with it, should Black go to Digital Input 1, Blue to Ground and Brown to Clamp?

DomeWiringDiag9.bmp

[symmetal]

The Velleman Clamp input can be from 5V to 30V. I assume this feeds an on board regulator and that the board itself only uses 5V. (Not sure why it's called Clamp and not +V In). I assume also the Clamp power is needed if there is no USB connection otherwise the 70mA the board uses could come via the USB. Hugh could confirm this. The VM110N data sheet is not very clear on this. The digital inputs are all switches to Gnd and the digital outputs are all open collector so the board doesn't need to supply any real power to anything connected to it.

As you say you'll need 5V to power the rotary encoder so a buck converter or a 7805 regulator powered from the 12V would do. The encoder only needs a few mA. You could connect the clamp input to the 12V or the 5V as either would work for the Velleman board.

The Hall effect switch brown (+ve) just needs to go to any 12V connection available, so not necessarily the Clamp voltage. The Blue (-ve) to Gnd and the Black ( signal) to Dig In1 as you stated. When you put the magnet close to the sensor the open collector output transistor will switch on so pulling the output to Gnd. You would then measure about 12V across the pull up resistor you've connected, if you're powering the Hall effect from 12V. When you move the magnet away the output transistor turns off so no current flows through the pull up resistor so you measure no volts across it. As the Velleman digital inputs are just looking for a switch to ground I'm pretty sure they will have their own pull up resistor on the board inputs.

Alan

Edited May 9, 2019 by symmetal

[pmlogg]

Alan

Thanks again.  The 12V shown in my diagram is the 7Ah lead acid battery that has been powering the Pulsar/Rigel System.  To supply power to the VM110 my intention was for it to draw 5V via the usb cable connecting it to the Raspberry Pi.  The Pi has multiple usb so if necessary I could use two to increase the current available - but I had not expected that the VM110 and components connected to it would draw that much power.  The power supply to it is to be the large-capacity power bank.  It can simultaneously supply 5V and 12V. Although my original thinking had been that it would supply just the 5V it would not be hard to have two cables running from it rather than one. 

The Clamp certainly confuses me and I'm not certain what the voltage out from it would be. If it's 5V that would be OK for the EM14 but not the Hall Switch needing 6-36V, so an step-up e.g. Pololu U3V50F9? would be needed or a no power Reed Switch instead of the Hall Switch.  If it's supplying 12V then the Hall Switch would get what it needs but the  Buck Converter to the EM14 is needed.  I don't know which is the sounder solution.

[symmetal]

The Clamp connection is a voltage input (5-30V) to power the VM110 if you don't have a USB cable connected. At least that's how I see it. The VM110 data sheet shows it as a input so I doubt you could use it as an output to power something else. The VM110 draws 70mA which the USB connection should be able to supply without problems but you may as well connect the clamp input to your 12V battery supply as it's there.

My solution would be to connect your +12V battery supply to the clamp input and the hall effect switch +V terminal. Then use a 5V output buck converter also powered from the 12V battery and use the buck converter 5V output to power the rotary encoder. This way everything stays powered whether or not the Raspberry Pi USB is connected or not.

You haven't mentioned how the Raspberry Pi is powered. It could be powered from the buck converter 5V output to keep everything simple if you wish.

Reading your above post again will you be using the 7Ah lead acid as well as the power bank to power the project. If so, you could use the power bank 5V out to power the rotary encoder (and possibly the Pi) but a 5V buck converter off the 12V would be a neater solution I would think.

Alan