Don't throw away dead switches

[B.J.]

I know that I have said this before, but I had yet another example of this, so I figured that it was worth posting again.
I used a 2 port diseqC A/B tone burst switch for many years with no problems, but I needed more ports so I bought a 4 port diseqC switch. I used this for a short while before it died, so I bought 3 more 4 port switches, all of different brands. None of the 3 new switches worked completely, although one worked on 2 ports, but two of them were basically completely dead right out of the box. I later got another switch with a receiver. It worked for a while, but then died, so I decided to try once again with the other ones that had previously died, but which I hadn't thrown out yet. Sure enough, the first one I picked up worked fine. Used it for a few months, then IT died, so I grabbed another of the dead switches, sure enough it worked. Used it for several months, until it died, grabbed another dead switch, it worked fine.
About a week ago, I went to watch sat, and NOTHING. Went out and exchanged for another dead switch. Everything worked fine. Today, went to watch sat, only one port out of 4 was working. Plugged in another dead switch.... working fine. So by now, I have reincarnated every one of my dead switches at least twice, except for one that I took apart to experiment with

Now, the disclaimer.... I'm pretty sure that my recent failures were due to moisture, since I don't hermetically seal any of my outside connections, and the last 2 failures were during or after rainstorms, BUT several of the failures, including the 3 that were dead right out of the box obviously weren't weather related. {Instead of sealing my switch, and other connections, I keep them in a box that's waterproof on the top, but open to the air on the bottom, so the switches are still exposed to humidity.}

But the bottom line is .......
IF your diseqC switch "dies" DON'T THROW IT AWAY, it may come back to life later.

 

[guapoharry]

Some of my "dead switches" were due to "barely loose" cable connections.

 

[Stargazer]

It may not be that. It could be sometihng else going on there. I seen something like this happen with the older DP-34 switches. You replace the switch (or switched ports on the switch), all worked well for a while then it would do it again. What I have found is that if you disconnect it then reconnect it then it would work fine again. This is what I would try in your case. Disconnect the switch completely and instead of replacing it with a different one entirely, put the same one back on it. I bet it will work again. Could be poor ground?

 

[Mr Tony]

I agree BJ

I only throw away switches if they are destroyed

I have a couple 4x1 diseqc's that port 4 is shot...but 1-3 work fine

 

[Ironsides]

I guess

I guess I am very lucky as I have never lost a diseQc switch due to weather. My PrimeStars at located just outside the window. Less than 20 feet from my receivers and my switches are located inside about 6 feet from my receivers. I have never had a bad one out of the box so again I am lucky. I have used the same switches about 5 years. I just recently added a 3rd for my Sami because I replaced the LNB with the DMX741 and instead of using the 22 hz switching I used the 2nd coax and a DiseQc Switch. I use Chieta Or Spaun. I did have a Pansat switch located outside but it never failed to work.

Personally, I think the length of coax involved will effect a switch. If it's a long run, 50 feet or more then some times the switches will not work on all the ports. The Pansat switch had troubles with a 75 feet run and would only work on ports 2&4 but when I moved it up closer and it was less than 20 feet from the receiver it worked fine on all ports.

I wouldn't throw away a switch either unless it was destroyed. If it doesn't work it could be because it's to far away from the receiver simple as that. I have found they work flawlessly with a 6 or 10 foot jumper. Even at 25 feet I have never had a problem...

So I guess that makes me lucky...

 

[tvropro]

My switches are inside so weather don't affect them. I'd rather run the coax and keep the electronics inside if I can.

 

[eurosport]

I have a couple 4x1 diseqc's that port 4 is shot...but 1-3 work fine

I still have my first diseqc switch around here somewhere. Similar to yours, but 1, 2 and 4 work, and port 3 took a crap!

 

[AcWxRadar]

I agree BJ

I only throw away switches if they are destroyed

I have a couple 4x1 diseqc's that port 4 is shot...but 1-3 work fine

I still have my first diseqc switch around here somewhere. Similar to yours, but 1, 2 and 4 work, and port 3 took a crap!

This is a very interesting discussion. I, like B.J., do not toss my suspected bad switches. Why I keep them is kind of an unknown to even myself. But, allow me to explain something that I have personally observed and see if it rings any bells with all of you.

First of all, I haven't been utilizing the four port DiSEqC switches much in the past few years as I have set up my motorized dish and the application of a switch was no longer a necessity for me. But, going back in time a little bit, I had a number of fixed point dishes which were combined using an array of SW21 and 4X1 DiSEqC switches.

During that time, I found several of the 4X1 DiSEqCs to fail on me. Sometimes, as B.J. mentioned, they were defective right out of the box! Other times, they seemed to work decently initially, but after a few weeks or a couple of months, problems started to appear.

Port #4 seemed to always be the first to fail. Sometimes it would just seem to die a quick and obvious death and at other times, it would simply slowly degrade. The signal strength and quality level would diminish gradually over time until it would finally become unusable. If I continued to use the switch "as is" then I would notice that port #3 would follow the same path of demise as port #4 did.

I never continued to use the switch after these two ports failed, so I don't know if port #2 would follow suit in numerical order.

I may just be trying to look for an electronic conspiracy theory here, but maybe there is something else to it. I just cannot explain why so many different 4X1 DiSEqC switches failed in the same pattern for me. Perhaps you guys have witnessed this, too?

I need to mention that all my switches were protected from direct exposure to the elements, but were still outside (in a Hoffman box with a gasket seal). Therefore, no direct rain or sunlight or intense heat was at hand, but still, it was not like being inside in a controlled environment.

As I mentioned, maybe I am looking for some "conspiracy" theory here and I am using that term as a tongue-in-cheek comment, but from what I have read of your comments compared to my own experiences, there appears to be a pattern here.

From B.J.'s comments, I get an impression that the switches which are failing are exhibiting an almost uncanny resemblance to the operation of a FET. I mean in the sense that, over time, the gate control voltage seems to become residually controlling and diminishing the current capacity of the junction until it eventually no longer can maintain the connectivity between the source and the drain.

I am certain that this is not a valid explanation of the problem, but the symptoms seem to want to act as if it were so. Hence, the ability for the switches to be resurected after time passed. It sounds like a capacitive charge is building up on the gate that is preventing the channel from opening fully and grows increasingly worse. Then, after taking the switch out of service for a length of time, that charge bleeds away and the switch acts like it was new once again.

As far as I have been informed, most of these switches are constructed with PIN diode junctions and therefore my theory doesn't really follow the guidelines of the operation of these junctions exactly. I cannot explain why these switches are reacting this way. However, I think that you can identify with how I perceive the symptoms.

Again, I don't intend to say that I am able to explain the reasons why, I just know that it is occuring and what it seems to emulate on the surface or what it appears to be at first glance seems to have a certain pattern and therefore it should be explainable.

I say that this is a very interesting subject as you all seem to have noticed many of the same things that I have also observed.

RADAR

 

[sidha]

A HEXFET handles the DC switching [ultra low resistance ON state]. The RF signal passes through forward biased PIN diodes. I swap the bad FETs out when I have similar dud ported switches accumulated. The little FETs are llimited for current and can run hot to the touch. Will even burn your finger when there is a 22k switch Y-ing off one of the exit ports. If it's a garden variety dual port 22k switch that doesn't switch DC, but feeds two ports continually. This doubles the load on the FET for that port of the DiSEqC switch. Not cool. [no pun]
Good idea in that situation to insert the 22k units first. But, then the DiSEqC switch probably won't work. Due to capacitors in the 22k that filter the DC and degrade the digital command burst. Thus, select heavy duty 22k switches that are known to work in that situation. Or, eliminate the electrolytic capacitor filter situated on the DC line inside the switch. Which I have done, to allow the data to pass, unaffected. Haven't upgraded the HEXFETS to larger devices able handle higher current and heat more efficiently, but I will. If an FET runs hot, due to higher current demand; most likely to give trouble in hot weather, as they cook, enclosed like in an oven. Of course, a shorted load on an exit port will blow an FET, instantly. But cause no damage to a 22k switch of simpler design.

 

[pendragon]

I have a design policy of never using receiver voltage to power LNBs or motors in my FTA system. While this means the receivers must still supply power to DiSEqC switches and in some cases unpowered multiswitches, it does reduce the load on the receivers and the switches themselves. Power to the LNBs is done by external or internal power supplies that feed the 'inputs' of the multiswitches, and motors are wired to isolate the DiSEqC commands from the actual motor power.

So far I am up to 16 DiSEqC 4x1 and 8 DiSEqC 8x1 switches in my matrix and have not experienced a failure. This scheme also means I can be a bit more cavalier about hot-plugging LNBs, which was another reason for doing this.