Bench testing a diseqc switch?

Status
Please reply by conversation.

colbec

SatelliteGuys Pro
Original poster
Feb 5, 2007
354
0
Eastern Ontario, Canada
Sometimes when troubleshooting a setup the functioning of a switch comes into question. If you have a known working spare you just swap and test again. Now you have the doubtful one in hand and can test it thoroughly.

What I had in mind is a receiver set up with some satellites programmed in but not actually connected to any dishes/lnbs. A short coax from receiver to switch will be sending the required signals to the switch depending on which sat is selected. Can we now use a multimeter to reliably detect how well the switch is doing its job at each of the outputs?

Is the receiver necessary, is there a different way of doing this test more simply?
 
parts list: LED switch tester

Can we now use a multimeter to reliably detect how well the switch is doing its job at each of the outputs?
..., is there a different way of doing this test more simply?
You know, that is a really inspired idea.
Since the diseqc switches power to only one LNB at a time, you can see which port is getting the juice.
While not 100% guarantee that the switch will pass RF, at least it's better than nothing.

Collect the following components:
- 6 cheap coax connectors
- 1 length of coax, maybe 1 foot long
- 4 LEDs , regular for indoor, or high brightness LEDs for outdoor use
- 4 resistors, value = 4.7k ohm

- make up a test cable with a coax connector on each end.
. - alternately, use any old left-over TV cable with connectors - doesn't need to be satellite-rated.
- find the positive lead of the LED, attach it to a resistor, and to the center pin of a coax connector
. - connect the other lead of the LED to the coax connector shield
- plug one connector/LED/resistor unit into each diesqc input port
- attach switch output port to receiver with short cable
- program receiver to select diseqc 1, 2, 3, and 4
- for each input, select both verical and horizontal polarization.
- watch LEDs come on one by one, as you cycle through the simulated LNBs.

The same resistor/LED/connector can be used directly at the LNB connector of the receiver, to check for LNB voltage
Note that the brightness of the LNB changes from dim for Vert to bright for Horiz polarization

notes:
You can determine which lead of the LED needs positive by connecting it in series with the resistor to any 9 ... 20 volt power source.
Be sure the LED is connected in series with the resistor - attach the resistor the the coax center pin, and the spare lead of the LED to the coax shell.
Verify there are no shorts, and insulate the parts so none occur in use.

Voltage on the coax out to the LNB is always positive.
It's 12 volts for vertical polarization and 18 volts for horizontal.
That will cause the LED brightness to change.


edit: changed resistor value, and recommended LEDs
 
Last edited:
- 4 resistors, value = 1k ohm
...
If the LED is too bright, you might try 4.7k resistors.


If you start with 1k resistors, consider they should be at least 1/2 watt. At around 1.2k you can go down to 1/4 watt, but expect them to get a little warm.
 
Last edited:
oops

Yea, I'm sorry.
1k and 20 volts is around 400mw dissipation.

Probably best to start with 2.2k or 4.7k from the beginning.
If you can't see the LEDs (because you are outside) maybe the high brightness LEDs would be best.
On just 2 or 3 ma, they will be very bright.

To keep down to 5 ma max, at 20 volts, 4.7 k ohms would be 85mw dissipation. :cool:

Thanks for bring that up. - :up

I'll change the first post to recommend 4.7k :eureka
 
Great ideas, guys. Worth a try.

Since I am an economist rather than an electronics expert, I am struck by the fact that you can order a 5-10$ diseqc switch and find you have a dud, doa. So what do you do, send it back? You are into shipping or gas to return it, some poor tech guy gets to spend XXX$ to find out if it is dud or not and you get your replacement (or not). The dead switch has now doubled in price (or more). You might as well hold on to it particularly if only one lnb outlet dies.

So now put yourself in the place of the manufacturers. Knowing it is a cheap part your customer will likely just order another one. He may be unhappy, but he needs a switch. So what is your best strategy? Send the stuff out half tested with crossed fingers and hope they don't complain.

In addition, when things go awry with the dish farm it is helpful to sort out quickly what is working and what is not. The path from the signal to the TV has many stages, and the switch just happens to be one of the places things can go wrong and not be visible.
 
heh,, good points !

Considering they sell for $5 (or less), and considering the horror stories I've heard, I'd have to recommend buying two when you need one.
Keep the spare on the shelf for future debugging and replacement.

I was lucky when I got a Mercury II from eCanada earlier this summer.
Not only did they include one 4-port diseqc, but they also threw in two spares.
You think that's cheaper than dealing with customers who blow up their first switch? :eureka
Yea, it sound silly, but think about it from a dealer's point of view.

In a lengthy thread on the MyTheatre forum, where they discuss opening and modifying the switches to work better with Twinhan DVB cards, there is some discussion of the internal parts and what might fail.
I suppose if you had a big pile of switches, it might (barely) make sense to open them all up and fix 'em.
Of course for just one, it's too much work.

I'd put in a more robust transistor, or something to limit short-circuit current, so they'd be more tolerant to hot-swapping.
Dish switches take it day-in and day-out; they're just better made.
 
Use the receiver and check each port for voltage. Voltage should only be on the port you have selected.

my cheita switch's have power on all ports all the time. I havent checked my spaun 411 and 420, but the spaun 5x8's Ive got (5801) also have power on all sat ports all the time.
 
construction modification

my cheita switch's have power on all ports all the time.
... the spaun 5x8's Ive got (5801) also have power on all sat ports all the time.
Yea, but those are -good- quality switches which don't break, and are possibly powered by their own supply, too.
I've never had a bit of trouble with Dish DP-34's, either.
We are talking about the cheap-o diseqc switches everybody burns out at the drop of a hat. :rolleyes:

Remember too, that this test only shows that the switch supplies power to your LNB
Doesn't prove that the LNB signal is being fed to your receiver.
Makes me wish LNBs had LEDs on them . . . :eureka


Construction:
When I was on the way out to lunch earlier, I thought of one problem with assembly.
I was envisioning BNC connectors, not F connectors when I made the description in the earlier post.
BNCs have a little gold or silver pin you solder to the center conductor of your coax.
F connectors use the center conductor of the cable . . .
So, as a construction technique, you might want to get 1/2 watt resistors, and epoxy them in place in the connector body, making one of their leads the center conductor of your plug.
Solder the LED to the resistor lead protruding from the rear of the connector (+), and the last LED leg to the body of the connector.
Wrap the whole thing in electrical tape, shrink wrap, or maybe give it a clear epoxy overcoat.
Use your imagination. - :up


As for the MyTheatre forum link, they didn't come up when I was writing this, but I'm sure you have it right.
I know I posted a link a long time ago, and this SatGuy's page has it for sure.
 
my cheita switch's have power on all ports all the time. I havent checked my spaun 411 and 420, but the spaun 5x8's Ive got (5801) also have power on all sat ports all the time.

I am running a number of Chieta switches and this is good information to known. I don't like the idea of the power passing to each port all the time because it is too much load on PCI cards. Currently all my LNB's are powered by Multiswitches, so it should not be an issue for me.

Second idea on how to build a tester. Receiver - switch - Satellite meter with tone indicator (One of the satellite meters in the $99.00 range, that shows when a 22KHZ tone is applied).
 
let me send you my consulting rates . . . :)

I haven't looked inside any of my several $10 DBS sat meters to see how much spare room they have.
I'd think one might shoehorn in a little board that could either light a LED to indicate presence of 22khz, OR maybe put out the 22khz by itself.
But let's decide on which, and not do both.

I'm thinking the LED detector idea might be easier, thought certainly not a trivial circuit.
And by easier, I really mean that the first design would work without having to make a rev #2. :rolleyes:

A meter that would inject the tone would certainly be nice when dealing with those dreaded Unversal LNBs.
But last time I went looking for a bird on a universal, I just manually turned 22k on at my receiver.
Didn't think it was a really big deal.

And then, there are all those people here, who think the only way to find a bird with with your receiver and a display.
They probably couldn't care less about any tool that had a meter, no matter what else it did. :D


edit:
.....hrmmmm...... just thought up a clever solution.... well, best to save something for next time. :)
.
 
That reminds me...I have an old Sony 20" dish somewhere that has an LED for dish pointing on the LNB. Not sure what system it was used for. That seems like a good idea.
 
This thread has been very instructive. For the first six months of my fta addiction I had been using one 4x1 generic cheapo diseqc switch with zero problems. Not only was I tempting fate in having the switch between the motor and the receiver but I was hot swapping all the time with no ill effects at all. I felt immune from the danger. Fools step in...

Then the lightning strike zapped it a few weeks ago and I bought 3 new generic switches, all of which appeared to work fine out of the box. On testing with a multimeter according to suggestions I find that voltage is leaking out all over the place. A typical result will be that I set the receiver to deliver power on diseqc #4 and discover these voltages at the ports: #1 - 0.03, #2 - 0.06, #3 - 13.84 and #4 - 13.84, and all this on a switch that from other experiments appeared to be working ok!

Turns out that all of my new switches are now broken to one extent or another and each time I put it to use again another port could drop off the usable list. It makes tweaking and troubleshooting a nightmare.

The only switch I have right now that still works as expected is the one built into the BSC621-2. Is it ok if I hot swap this one?
 
Status
Please reply by conversation.

Weather?

Where I went on my summer vacation