DISEqC MultiSwitch failover switch

[Archive4]

Thankfully, I have just received conformation from Star Fleet Command that I have full interrupted access to the Enterprise " E " universal translator for at least three hours. Now I have have a slight hope of what this thread is all about!

Jokes aside, I have watched this from post one, but it was above my knowledge, but I am learning from it

 

[David Jashi]

Diseqc 1.1 switches are normally used OUTPUTside of a diseqc 1.0 switch, and would therefore also be applicable in your case (but probably not for the TBS6909? Not sure how they manage diseqc control), assuming the receivers (TBS6910) can give diseqc 1.1 commands (that could even be easier than building some kind of controller?)

Now, that's exactly the problem and reason why I had to deal with this kludge - The multiswitch I am using (or should I call it matrix switch) uses DISEqC 2.0 for selection of one of 4 input dishes and 0/22kHz+14/18V for band/polarization. Therefore, whenever I send command to first DISEqC, the second one reacts too (they are both committed).

 

[Vipersan]

Thankfully, I have just received conformation from Star Fleet Command that I have full interrupted access to the Enterprise " E " universal translator for at least three hours. Now I have have a slight hope of what this thread is all about!

Jokes aside, I have watched this from post one, but it was above my knowledge, but I am learning from it

You're not alone mate ..
I use one of the most complex receiver/lnb/positioner diseqc setups on our group ..but the thread has me confused also.

 

[a33]

Therefore, whenever I send command to first DISEqC, the second one reacts too (they are both committed).

I don't understand your problem here. If you use a diseqc 1.1 switch between a receiver and the two multiswitches, then only one multiswitch gets power of the 1.1-switch; the other one doesn't get power so can't react to the diseqc1.0 command.

So, if by any chance multiswitch #1 (at diseqc 1.1, #1) is out of order, you could choose diseqc 1.1 #2 in your receiver, and a connection to multiswitch #2 would be made, with again all the possibilities for the diseqc 1.0-switching.
Isn't that what you were after?
I guess 0/12volt switches would be cheaper, and could be switched collectively (one collective command-line for all switches). But for individual control, a diseqc 1.1 switch works wonderfully.

Only for the TBS6909's you would have the diseqc 1.1-problem I guess; and as you have cascaded the four input-cables (if I understand correctly) I would guess you only need four 0/12volt switches (or homemade frankenswitches) at the beginning of that input-line?
Edit: Well, thinking back, I'm puzzeled here: If a TBS6909 has 4 quattro inputs, how does it choose one of the four satellites? Or are 16 cables leading to one TBS6909? Not sure if I understand correctly how you cabled your TBS6909's....

Greetz,
A33

 

[David Jashi]

I don't understand your problem here. If you use a diseqc 1.1 switch between a receiver and the two multiswitches, then only one multiswitch gets power of the 1.1-switch; the other one doesn't get power so can't react to the diseqc1.0 command.

Here are those switches:

They always have power - either from 18V PSU, or at least one of receivers.

So, if by any chance multiswitch #1 (at diseqc 1.1, #1) is out of order, you could choose diseqc 1.1 #2 in your receiver, and a connection to multiswitch #2 would be made, with again all the possibilities for the diseqc 1.0-switching.

Multiswitch will also react to this command and switch this very input to position 2.

Isn't that what you were after?
I guess 0/12volt switches would be cheaper, and could be switched collectively (one collective command-line for all switches). But for individual control, a diseqc 1.1 switch works wonderfully.

Only for the TBS6909's you would have the diseqc 1.1-problem I guess; and as you have cascaded the four input-cables (if I understand correctly) I would guess you only need four 0/12volt switches (or homemade frankenswitches) at the beginning of that input-line?

If for those I will definitely need external control, as they don't pass DISEqC commands out.

 

[a33]

About uncommitted ones - all multiswitches I ever saw in action or heard of are committed ones, so that's not quite an option.

I hope I have made clear by now, that you would need extra 2/1 diseqc 1.1 switches?

As for 0/12V switch - thought of that, but how will I stop it from passing further? Remember, the switch in question is a first one in chain.

I don't understand this question? Or have I answered it by writing that the 0/12 volts need a separate cable?

Edit: Ah, you have posted almost simultaniously. Maybe we crossed....

 

[David Jashi]

I hope I have made clear by now, that you would need extra 2/1 diseqc 1.1 switches?

As I mentioned above, commands for 1/2 switches will cascade to multiswitch, and it will react too, won't it?

I don't understand this question? Or have I answered it by writing that the 0/12 volts need a separate cable?

It depends on what do you mean by 0/12 switch. If something like Frankenswitch - then my comment was invalid.

 

[a33]

Multiswitch will also react to this command and switch this very input to position 2.

Why in heaven's name would a diseqc 1.0 multiswitch react to a diseqc 1.1 command?
It wouldn't and it couldn't!

As you wrote: the multiswitches are on COMMITTED diseqc, so they can't react to UNCOMMITTED diseqc-commands!

Or where lies the misunderstanding, here?

Greetz,
A33

 

[David Jashi]

Why in heaven's name would a diseqc 1.0 multiswitch react to a diseqc 1.1 command?

Now it's a bit clearer. The only way to check if this works is an experiment.

So, as I understand, I should pass only 1.1 commands to my multiswitch (I will have to check, whether I can do basic switching without using anything from 1.0 subset, which I doubt - they are backward compatible) and when it comes to switching to another multiswitch - use 1.0 commands, and then return multiswitch port to valid position by sending 1.1 command again?

This will work only if I find unintersecting subsets of commands for them.

UPDATE: 1.1, of course, 2.0 won't do definitely.

 

[Archive4]

Sorry to intercede again, but if I follow this so far, a 2.0 command could and should affect a 1.0 switch, but a 1.1 switch would ( and should ) ignore all commands by a x.0? ( sounds a bit like SWMBO )
Either way, reminds me of my early electronics training with and gates nand gates etc.
I think the best solution would be for you two to get round the dining room table and bolt the bits together!
May I be so bold as to suggest that A33 goes to David´s place, as I suspect the outcome......

Have released the Enterprise back to service, translator didn´t help

 

[David Jashi]

Sorry to intercede again, but if I follow this so far, a 2.0 command could and should affect a 1.0 switch, but a 1.1 switch would ( and should ) ignore all commands by a x.0? ( sounds a bit like SWMBO )

You are right:

Either way, reminds me of my early electronics training with and gates nand gates etc.
I think the best solution would be for you two to get round the dining room table and bolt the bits together!
May I be so bold as to suggest that A33 goes to Davis´s place, as I suspect the outcome......
Have released the Enterprise back to service, translator didn´t help

Well, it's much easier to get visa to my country, so I guess it's the better way out. ( by the way, it's not THAT Georgia, it's smaller one next to Turkey)

 

[David Jashi]

...but a 1.1 switch would ( and should ) ignore all commands by a x.0?

Will it be "all", that's the question I must check. If not, its a dead end.

 

[Archive4]

( by the way, it's not THAT Georgia, it's smaller one next to Turkey)

That´s the one thing the whole forum understood, without it being explained

 

[Vipersan]

In the past ..most of my more exotic disqec switch config lashups have been a bit hit n miss ...
Not so tricky any longer thanks to this rather excellent project I found a while back..
Might I suggest building one yourself ?
It will prove what signals are generated and where they reach ..
I find mine very useful indeed..
Diseqc Decoder project ..
Of course ..you then need an accurate diseqc hex table in order to interpret the results ..
..and sadly not all receiver manufacturers bother to follow Eutelsats protocols exactly ..
Many take shortcuts ...which is where a lot of the problems and confusion can arise.
(If anyone finds a concise dieqc byte analysis lookup chart - please post it here ...it would save keep digging into Eutels lengthy pdf)
rgds
VS

 

[a33]

Firstly:
a 0/12volt switch looks like this: 0 /12v switch for sale
or this: http://rik.dn.ua/st/foto_st01/b21_switch_0_12_v_visa.jpg
or this: 0/12V Switch SW-02/E
or this: 150709. SW-02/E 0/12V Switch, 13-18& 12V Switch & 22KHz Gen_Sevis Technology Ltd.
and so on.....
Some even mention 5 volts instead of 12 volt.


Secondly: there is no real need to experiment with the diseqc protocol I think. Though many satellite watchers don't fully understand the diseqc protocol, the systematics behind it are not that difficult.
The matrix you copied out of wikipedia (or somewhere else) is somewhat misleading. Don't rely on wikipedia concerning diseqc; go to the eutelsat website (the developers! e.g. http://www.eutelsat.com/files/contributed/satellites/pdf/Diseqc/Reference docs/bus_spec.pdf from the website Technical support - DiSEqC specifications - Eutelsat ), or the spaun website ("diseqc for technicians": http://www.spaun.de/files/8df79_en_DiSEqC_for_Technicians.pdf ).
I was misled back when I started with satellite reception, believing what I read about diseqc switches at some websites; but they were written by people who didn't really understand diseqc themselves... It took me quite some time to discover how wrong the information was, as 'the primary effect of first knowledge' is quite persistent!

Diseqc can be used parallel, or serial (cascading).
There are UNDEPENDANT sets of commands: committed (diseqc 1.0), uncommitted (diseqc 1.1), motorcommands (diseqc 1.2) and bidirectional commands (diseqc 2.x).
I don't think I have ever seen diseqc 2.x operational though: it is a beautiful possibility, but hardly implemented I think. (If a switch mentions diseqc 2.0, it is by no means a guarantee that it operates at 2.0-level. Many times it understands the command allright, but cannot give the wanted reply, so then it is a 1.0-switch really!)
Diseqc has much more possibilities than normally used.

A diseqc device reacts only to commands that are meant for that specific device, and ignores other commands. So a committed (multi)switch can NEVER react to an uncommitted command.
The commands in diseqc are in fact like computer commands: they are telling specific diseqc parts to react in a specific way and maybe (in diseqc 2.x) are asked to give an answer.
The commands and replies are transported by means of a 22kHz-signal.

At a satellite change, a receiver gives off a diseqc command for the diseqc part that must change its position. Non-changeing diseqc parts usually are not adressed with a (repeated) diseqc command.
In cascaded switching, a diseqc command is only executed if the command is really received. Most receivers give diseqc commands in the order (sometimes adjustable!): diseqc x.1 commands --- diseqc x.0 commands --- and thirdly (I believe) diseqc1.2 commands. That is why the most commonly used order of diseqc devices is also in that way, but by no means obligatory! You can use any order you like, if your receiver is able to modify order of commands or repeating of commands.

For your situation: No problem whatsoever to put diseqc 1.1 switch (or 0/12v switch) between receivers and duplicate multiswitches, if your receivers support diseqc 1.1.

Greetz,
A33

So, as I understand, I should pass only 1.1 commands to my multiswitch (I will have to check, whether I can do basic switching without using anything from 1.0 subset, which I doubt - they are backward compatible) and when it comes to switching to another multiswitch - use 1.0 commands, and then return multiswitch port to valid position by sending 1.1 command again?

This will work only if I find unintersecting subsets of commands for them.

UPDATE: 1.1, of course, 2.0 won't do definitely.

This text of you did not really make sense to me, alas....
I think you have some reading to do!

 

[David Jashi]

Firstly:
a 0/12volt switch looks like this: 0 /12v switch for sale
or this: http://rik.dn.ua/st/foto_st01/b21_switch_0_12_v_visa.jpg
or this: 0/12V Switch SW-02/E
or this: 150709. SW-02/E 0/12V Switch, 13-18& 12V Switch & 22KHz Gen_Sevis Technology Ltd.
and so on.....
Some even mention 5 volts instead of 12 volt.

You could save us both hell of a time by posting those pictures earlier.

This is factory-made thing that does EXACTLY what I need without any extra fiddling and risk. It switches 2 RF inputs with an EXTERNAL (not embedded, like DISEqC) signal. Marvelous. I'll simply buy 32 if those and connect them in parallel to whatever relay I will find easy to connect to one of PCs: USB one or chinese Arduino compatible board with relay. And it's much cheaper, than modifying 32 DISEqC switches manually.

Secondly: there is no real need to experiment with the diseqc protocol I think. Though many satellite watchers don't fully understand the diseqc protocol, the systematics behind it are not that difficult.

That's why I'm here to ask, ain't I?

The matrix you copied out of wikipedia (or somewhere else) is somewhat misleading. Don't rely on wikipedia concerning diseqc; go to the eutelsat website (the developers! e.g. http://www.eutelsat.com/files/contributed/satellites/pdf/Diseqc/Reference docs/bus_spec.pdf from the website Technical support - DiSEqC specifications - Eutelsat ),

Thanks, I was looking for that one for last two days.

This text did not make sense to me, alas....
I think you have some reading to do!

Yes, and that's why I came here. On the other hand, you saved me from reading by posting those pictures above. No need to modify receiver's software to send out extra commands, if I can simply handle it with one relay in a separate process.

 

[a33]

You could save us both hell of a time by posting those pictures earlier.

Well, to be honest:
You could have saved me a lot of time if you would have told earlier that you didn't know what a 0/12v-switch is exactly, or if you would have googled on it yourself!

Furthermore: it's nice to be of help! That's what this forum is for!


When you like my posts, I like to be liked ;););););)

Btw,
for testing reasons you could also first order just ONE 0/12V switch; to test if it really does the job. I just read during my googling that someone had problems with a certain type of 0/12V switch; it would be bad to have bought 32 bad switches....

Greetz,
A33

 

[David Jashi]

for testing reasons you could also first order just ONE 0/12V switch; to test if it really does the job.

Of course, and different models from different suppliers too.

 

[Archive4]

The reason the like button was not there was because the forum software does not allow self gratification, also, even though you may change the post, the original is still in the e-mail alert , although no longer visible to others. ( Which is why the rest of you think I´ve gone mad ),
But you know what I mean!

 

[a33]

May I be so bold as to suggest that A33 goes to David´s place, as I suspect the outcome......

Have released the Enterprise back to service, translator didn´t help

Rainy night in Georgia?
One day I'll fly away!!!!

(does this link work? Never tried it before....)