Antenna Switch...

[acyddrop]

I'm think of doing a new project.. An antenna switch. I know this *should* be an easy project, but like any good engineer who ignores K.I.S.S. I'm going to over engineer it. I'm think I'll use an Adruino Uno R3 from Adafruit as the controller. I'll probably have to design a shield of my own, since the Adruino control lines put out +5V, and the relays I'm looking at (AC 10amp 240v) the coils use 24v so I'll need some small relays that run on 5v to trip the 24v to the bigger relays. I will keep looking because this solution even for an over engineered project is sloppy. Ultimately I'm looking at a 16x2 LCD display with back light that also has a few buttons that can be programmed. My idea is, instead of using LEDs to indicate your selected antenna (which would be too easy, and far too easy compared to a rotary switch) I'll display your selected Antenna on the LCD panel. With some additional programming I bet I could display forward and reflected power and possibly even SWR. Also I believe I'll build in a dry 300 watt dummy load into the whole thing.

Another advantage with the Adruino design, over a purely mechanical one is the easy inclusion of a two rig setup as opposed to just one (In fact you could have more than two). I don't think you could (at least not easily) allow two transceivers to be connected and assign them to two separate antennas and use them both at once. You probably could actually but maybe not with the Uno R3. I'll have to give that some thought.

The beauty of this is simple, even though at first this design will be a local switch (ie: up by the transceivers) it will be built as two separate units from the get go. By doing this you can EASILY adapt the entire design to be a remote antenna switch. I'm not sure of final cost, but I can't imagine it'll be too expensive. The relay collection will be the single most expensive portion. I'm planning to start with just a 3 position switch + dummy load.

Right now, I've found SPDT 10A 240vac relays which should be plenty for 1Kw or more, the only problem I have right now is all the relays I'm seeing so far use 12V or 24V (or both) coils, which means the Adruinos 5+ vDC lines can't energize the coils directly (and by directly I don't mean directly, I mean +5vDC isn't enough to move a 12v or 24v coil, I don't mean without the diodes). If anyone knows of any relays that can do 240vAC @ 10A with a +5vDC coil let me know. Otherwise you're looking at a mess of tiny relays to step up to work your big boys, it'd certainly sound fantastic (I like relay chatter). Personally I think this project has potential, in conception it's better than any other switch solution I can think of, it might even have commercial viability.

 

[ab3a]

I'm a bit confused as to what you're doing. The way you write this, it seems as if you're starting with the solution and you're looking for the problem.

If your goal is to consolidate antenna transmission lines to one quality piece that goes up a tower, then I understand what you need.

First, let me point out that instead of using relays you might want to investigate using PIN diodes. Solid state is always better.

Second, allow me to point out that you can always use a 2N7000-ish part to key a relay.

Third, if you need to communicate with the Arduino over any distance, be sure to use a properly shielded cable with a reasonable driver. It is commonplace to use RS-485 drivers with optical and galvanic isolation.

That's just a start. Engineering something like this is often made to sound a lot simpler than it really is.

 

[digitalanalog]

You could use this (or another mfg)to get that voltage you want to drive the coil(s)
DC DC Inverter Module Artesyn Emerson BXA40 24S05 18 36 V in 5V 8A Out | eBay

 

[acyddrop]

I'm a bit confused as to what you're doing. The way you write this, it seems as if you're starting with the solution and you're looking for the problem.

I'm not exactly sure what you mean? I did this to get the kind of feedback you've given, it's sort of my version of an RFC to get the input of people smarter than I am.

First, let me point out that instead of using relays you might want to investigate using PIN diodes. Solid state is always better.

Solid state is NOT always better. SSRs need to be attached to heatsinks which just added to the size of this thing. While an SSR might be more easily driven driven from an Arduino, it still requires some sort of external power source to transistor to supply enough current to strip the SSR. I mostly understand the point I think you're trying to make, but saying "solid state" is always better is flat out wrong. I happen to be attached to my 4CX1000's (as an example of the wrongness [or at least my extreme disagreement] of your statement).

Second, allow me to point out that you can always use a 2N7000-ish part to key a relay.

You're right I should of though of that before even writing. Good catch!

Third, if you need to communicate with the Arduino over any distance, be sure to use a properly shielded cable with a reasonable driver. It is commonplace to use RS-485 drivers with optical and galvanic isolation.

Now that's something to think about, when I was originally conceiving this.. I thought of putting the Arduino in the house as the control unit and sending the correct relay it's coil current/etc. But if I put the Arduino in the remote unite directly I might be able to move I2C over a pair of cables. That's something to think about...

That's just a start. Engineering something like this is often made to sound a lot simpler than it really is.

Believe me, I know. Even simple projects (which this really isn't) turn into more trouble than you realized they would be. I'm hoping consolidating a lot of the control (brains?) into a microcontroller will simplify some of the design work.

 

[acyddrop]

I could use that for a different project (unrelated to ham radio), thanks for the link!

You could use this (or another mfg)to get that voltage you want to drive the coil(s)
DC DC Inverter Module Artesyn Emerson BXA40 24S05 18 36 V in 5V 8A Out | eBay

 

[ab3a]

Ok, you didn't state that you wanted to put 1500 watts on this. You are quite right, PIN diodes at that power level are expensive.

Second, single-ended communications protocols such as I2C is NOT a good idea for any distance. Use RS-485 drivers, perhaps something like an LTC1484. See LTC1484 - Low Power RS485 Transceiver with Receiver Fail-Safe - Linear Technology

Third, remember that if you run lower voltages such as 12 volts across any distance, there will be voltage drop. You would probably do better to run line voltage to a power supply at the top of the tower to power your arduino up there. You could set up a second Arduino at the station to communicate with it. That way, you're not spending ridiculous money on cables.

Another thought is that instead of using 120 volts, use a 48 volt positive ground telecom power source. A set of four batteries in your station could power your rig and also could be used to power the Arduino without too much voltage drop, and the batteries could be floated with a charger you buy off of E-Bay. There is a reason the telephone companies still use power systems like like to this day. At the Arduino, use an inverter to step the voltage down.

I have seen a lot of blown up gear in my career. There is no shortage of idiots who design around unprotected communications systems and then they act all surprised when the first crash of lightning within 20 miles brings their system down. Do yourself a favor and use the protection.

 

[acyddrop]

Ok, you didn't state that you wanted to put 1500 watts on this. You are quite right, PIN diodes at that power level are expensive.

I suggested that a 10A at 240V relay would do quite nicely at 1kw or better, on this subject I found a suitable relay board (shield) already built. It uses 10A 240V relays and costs $20, I'm not sure if these relays are truly up to the task but for $20 I'm willing to find out. It's complete with opto-isolators and all the diodes etc. Saves me some time.

Second, single-ended communications protocols such as I2C is NOT a good idea for any distance. Use RS-485 drivers, perhaps something like an LTC1484. See LTC1484 - Low Power RS485 Transceiver with Receiver Fail-Safe - Linear Technology

Yeah I2C is only good for about 1 meter before it conks out. I also love your suggestion of using RS-485 for distance it's extremely rugged. I'm going to use An I2C to UART to RS-485, since the front panel unit I'm using (keyboard + 16x2 LCD) is I2C. That should work quite nicely I think.

Third, remember that if you run lower voltages such as 12 volts across any distance, there will be voltage drop. You would probably do better to run line voltage to a power supply at the top of the tower to power your arduino up there. You could set up a second Arduino at the station to communicate with it. That way, you're not spending ridiculous money on cables.

Another thought is that instead of using 120 volts, use a 48 volt positive ground telecom power source. A set of four batteries in your station could power your rig and also could be used to power the Arduino without too much voltage drop, and the batteries could be floated with a charger you buy off of E-Bay. There is a reason the telephone companies still use power systems like like to this day. At the Arduino, use an inverter to step the voltage down.

Yeah it's why Edison lost the battle to power our homes with "his" DC. I'm thinking that powering it the way you suggested is actually a mighty fine suggestion. One could run 48VAC from a local transformer that could provide power to the local electronics (few) and the remote unit. Maybe supply 2-3A out the line the relays don't need very much juice to operate (nor does anything else out at that end) so could probably power the whole thing with 1 - 1.2amps at 5V.

I have seen a lot of blown up gear in my career. There is no shortage of idiots who design around unprotected communications systems and then they act all surprised when the first crash of lightning within 20 miles brings their system down. Do yourself a favor and use the protection.

Believe me I know all about over design and making things more well built than they probably need to be. But that is good advice for everyone/anyone reading this thread.

 

[zz0468]

Pay attention to the relay contacts, and the unnecessary extra lead-length from the relay armature to the external contacts that many relays have. That makes them ill-suited for RF switching, but depending on the construction of the relay, some of that can be eliminated. You'll see what I mean when you look closely at the typical relay, and you'll see a wire running from the base of the relay, to the top, where it connects to the armature.

What will you do with the unterminated port on the relay? Leaving it hanging will leave it susceptible to coupling to the other contacts, and reduce the amount of isolation it provides. Do a bit of research on other remote antenna relays and see what other people are doing... and why.

Consider some type of fail safe operation so that if all relays fail to energize, that some sort of load (antenna or otherwise) is presented to the radio by default.

Lightning and RFI protection! Having an Arduino controller in the same box with a bunch or relays switching a kilowatt of RF will require considerable care. You want to keep RF out of the Arduino, and Arduino out of the RF. You won't be happy if, when you turn on power to your new switch for the first time, the receiver S meter pins at 60db over S9 of noise. Or if your relays all get sent to random positions when the Arduino goes bananas with all that RF floating around in the box.

 

[acyddrop]

Pay attention to the relay contacts, and the unnecessary extra lead-length from the relay armature to the external contacts that many relays have. That makes them ill-suited for RF switching, but depending on the construction of the relay, some of that can be eliminated. You'll see what I mean when you look closely at the typical relay, and you'll see a wire running from the base of the relay, to the top, where it connects to the armature.

Yeah I'm going to take them all apart and see if that is something fixable or not, or if it even needs to be. These are the smaller relay types with much shorter leads (or none at all) so it might be OK. Good advice though, I knew this one.

What will you do with the unterminated port on the relay? Leaving it hanging will leave it susceptible to coupling to the other contacts, and reduce the amount of isolation it provides. Do a bit of research on other remote antenna relays and see what other people are doing... and why.

I haven't entirely decided, I'm think a diode or some such to ground, or similar way to get that under control

Consider some type of fail safe operation so that if all relays fail to energize, that some sort of load (antenna or otherwise) is presented to the radio by default.

I hadn't thought of this one, and that is an excellent suggestion I think the power off position of the relay will default to maybe some sort of dummy load. Which also begs the question how do you know if the relays aren't functional. I think including some kind of power indicator back at the head end makes a lot of sense. Excellent point, thanks!

Lightning and RFI protection! Having an Arduino controller in the same box with a bunch or relays switching a kilowatt of RF will require considerable care. You want to keep RF out of the Arduino, and Arduino out of the RF. You won't be happy if, when you turn on power to your new switch for the first time, the receiver S meter pins at 60db over S9 of noise. Or if your relays all get sent to random positions when the Arduino goes bananas with all that RF floating around in the box.

Lightening protection is in line on all of my coax, but a grounding wing nut for this thing makes a whole helluva lot of sense too! As far as the RFI from the Arduino I'd thought of that one and was going to use some sort of Faraday cage arrangement, something like the old classic radios did. I could build some sort of gas discharge tube arrangement into this rig, not sure how effective that would be though. Thanks for some great suggestions I'd not thought of!

 

[ab3a]

At the end of the day, the stuff that bites you with endeavors like this are

1. Grounding
2. Power supplies
3. Lack of shielding
4. Poor decoupling and poor isolation of critical parts

It's not hard to design a relay controller, but it is hard to design one that works consistently in an application like this.

I also suggest using a real coaxial switch instead of a plain relay. See
RF Coaxial Relays "N" Connectors
for examples of what I'm talking about.

 

[zz0468]

I haven't entirely decided, I'm think a diode or some such to ground, or similar way to get that under control

No! You don'e want any diodes in the RF path, unless you try for the previously mentioned pin diode switch.

Terminate them to a 50 ohm load. so, what you might end up with is a SPDT relay for each antenna port, and one for each radio port.

Since you mentioned being able to have two radios, with each having the ability to switch to any antenna, you're looking at some sort of matrix switching arrangement. It's not overly complicated, but it would take some thought. The Arduino logic will keep the relays from getting switched into some illegal configuration, like the two radios tied together.

I'm pretty certain that this sort of thing has been done. Look through back issues of QST or QEX and see how it's been done in the past.

Sounds like a fun project.

 

[acyddrop]

A pair of these at 100ohms at 100watts will be my dummy load http://www.vishay.com/docs/50051/lto100.pdf just need some heatsink material for them. In fact I have a few of these hanging around right now, I was going to make a dry 100w dummy load project. That's super easy, I think maybe I'll post that up here when I get a chance. I think these resistors are $9/ea at 5% tolerance and around $15 or so at 1%. You could of course play with the values and make a dry dummy load up to 300-500 watts I think. I wonder how they'd work in mineral oil...