Strange Mercury behavior

[wejones]

At least I think this is Mercury behavior, as it's possible that the Motor and/or DiseqC switch may also be involved.

Anyway, I have a 90CM/Sg2100/QPH031 and a fixed Primestar dish/Kul1, side by side, both feeding a 4x1 DiseqC switch going to my Mercury.
Both dishes were aimed at AMC3 a week ago, and were giving me good signals on the PBS channels. Then, we got about a foot of snow, which shouldn't have hurt anything, as nothing stuck to the dishes, but over a few days, I noticed that I had lost signal on both dishes. I assumed that my fixed dish had shifted, since it is mounted on a portable mount on a log that I can move around to get a view of different sats, so it made sense that the snow may have moved the log. However it didn't make sense that the 90CM dish would have moved, since it's on a pole sunk in concrete.
So I decided to see if I could tweak the reception a bit. I've been using DiseqC-1.2 to move the motor, so I started bumping the motor east and west, hoping to get reception. I was having problems finding the signal, so I moved the dish to a couple sats west of AMC3, and they came in fine, but back to AMC3, and nothing.
THEN, I remembered. At some point in time, I had inserted a satellite east of AMC3, and I remembered the bug in the Mercury firmware, that was first discovered by Bill190, ie see:
http://www.sadoun.net/forums/fortec-...ury+diseqc-1.2

ie the Mercury seems to have a problem whereby the DiseqC1.2 position numbers get scrambled, and it will send the dish to the wrong satellite.

So I decided that I must have scrambled the AMC3 position somehow. So I decided to switch back to USALS, which I had not been using since my motor's zero is slightly off. But USALS usually gets me close enough to see the signal. So I switched to USALS. Got signal back on AMC3, but it was poor, since the zero is off. So I decided to alter my longitude to get it synced right. My actual longitude is about 70.78, so I decided to experiment by increasing/decreasing this value by small amounts. However THEN, I ran into yet ANOTHER bug of the Mercury, which again was first noticed by Bill190, ie that in some menus, the signal/quality meters don't work. Ie if I'm in that USALS menu, and have a signal/quality reading, and make ANY change to ANYTHING, but in this case, make a small change to the longitude, then I completely lose the signal/quality meter readings. This makes it very hard to make fine adjustments. In the past, I've had to make an adjustment, and instead of being able to use the meter in the menu I've been in, I've had to back completely out of the menus, and see how the change has altered my reception, then go back into the menus, etc, etc. THis is very time consuming, so I decided to speed things up by using another receiver. I fed the Mercury's passthru to my Twinhan, and used the Thmover.exe program to monitor the signal/quality, figuring that even if the meters on the Mercury weren't working, that I could use the Twinhan to measure the signal. WRONG! It turns out that it isn't that the METERS don't work on the Mercury, it's that the Mercury completely kills the signal somehow, when in those menus. HOWEVER, I DID find out, that I could usually (not always) get the signal to come back, by switching transponders, {usually it had to be a transponder of the opposite polarity} for example, if I changed my longitude to 71.0, I'd lose all signal/quality when I left the longitude selection (which is when the Mercury would say 'moving to sat position'). However if I'd go up to the transponder, change to another transponder, then back to the transponder I was previously on, I'd usually get the meters functioning again, and the meters on my Twinhan via Thmover also went to zero then came back, showing that it wasn't just that the meters weren't working, but that I wasn't getting any signal.
Anyway, I worked on the darn thing for nearly a half hour, and ended up getting a strong signal when I'd set my longitude at 71.7. I turned off Thmover, and started up TSREADER on my computer, and was getting good reception on the PBS signal I was tuning on. So I thought GOOD, this seems to work. So I left that 71.7 longitude in there, and backed out of the USALS menu, and instantly, the signal/quality went away on the Mercury, and I lost reception on the Twinhan!?!?!?!? I went back into the USALS menu, and it was still set on 71.7, so I started over again, adjusting longitude. I couldn't find signal anywhere. Put the longitude back at my actual 70.8. Sent the motor to zero, and back to AMC3.... finally got back to the poor reception I started with an hour ago. Started to go through the same process of adjusting the longitude, but this time it ended up showing the best reception when I had the longitude set at about 70.1, ie in the other direction?!?!? Since my original intent in all of this, was simply to get some reception on AMC3 so I could watch a program on PBS, I finally just quit... this time when I exited the menus, I WAS able to get back and was able to view programming out of the menus. So I never did figure out what was going on.
Another weird thing that was going on during all of this, was that initially, when looking for signal, I was using the 12180V/30000 transponder, but wasn't finding anything, so I switched to the 12163H/4444 transponder, and I finally found the signal using that transponder. However after peaking on that transponder, I switched back to 12180V, and it wasn't there, however after peaking on that for a while, I got a good signal on that, and tried switching back to 12163H, and IT wasn't there?!?!? I eventually got the thing working, and am able to switch back and forth between the various transponders, but this darn thing was doing WEIRD THINGS.

Looking back at what I observed, I'm almost convinced that the problem may not be completely with the Mercury, but may be partially related to the SG2100 motor, and/or the DiseqC switch. Ie I'm thinking that somehow I may be losing lnbf voltage to the LNBF whenever the motor starts to move, and/or whenever I switch polarity. Perhaps my long (~250') coax run is making the lnbf voltage marginal whenever the motor is trying to move. I've previously noticed that some DiseqC switches also seem to decrease lnbf voltage even with short (10') coax runs, so it's possible that the DiseqC switch might also be involved. I've also noticed, that at least with my old Ultra and Lifetime, that whenever the receiver loses lock, that it will try to send out DiseqC signals, apparently trying to make sure that the switch is in the proper position. I'm half wondering if the receiver also sends out DiseqC-1.2 commands to the motor when it loses lock, but I doubt that it does this.
But bottom line is that something is seriously wrong with the Mercury firmware, and it just got worse for me over the past week, for no apparent reason. I am really confused at what is going on with the dumb receiver. I'm tempted to switch back to my old Lifetime receiver, which has a functional DiseqC-1.2 capability, reprogram the motor, then use a binary editor to re-program the Mercury's memory to match those settings. Seems like a lot of work, but anything is easier than fighting with this dumb Mercury losing signal every time I make a change.
Bad thing is that I may have to wait till spring to play with the thing, since I'm up to my lnbs in snow and ice here. I may just re-aim my fixed Primestar dish, and forget about the 90CM until spring.

[Larry Acklin]

I come from the IT side of the computer industry. Ever since flash memory (firmware that can be re-programmed) not eprom or eeprom came to be used, we have always had a low level of this type of wierdness going on. Sometimes a router or firewall will lock up or do wierd things, and a re-flash fixes it. Sometimes a server exibits bizzare behavior system wide and a bios flash/update fixes it also. I have not re-programmed my Mercury II with the latest Fortec flash due to warranty issues, but I bet a re-flash could help.

I bought a couple of e-bay specials (Fortec lifetime ultra and Pansat 2500) with "other persons's code" and had all sorts of goofy behavior, till I Jtagged one to factory code and the other with serial loading. Both are functioning well now on factory code.


Larry

[wejones]

Originally Posted by Larry Acklin

...... I have not re-programmed my Mercury II with the latest Fortec flash due to warranty issues, but I bet a re-flash could help.

....

Larry

I re-flashed mine with what was the latest versions right after I bought it, but I haven't upgraded since, because people have been saying that the newer firmware is worse than the original. I was kind of waiting to see if Fortec ever announced that they had corrected the bugs that they've been told about before I got a new version. But if this continues, I may do what you say, and re-flash it with the same version I put in before.
Interesting.... I have 3 Fortec receivers, a Lifetime, a Lifetime Ultra, and a Mercury II, and that seems to be the order of reliability. Ie I've had no problems at all with the Lifetime, I've had a bunch of problems with the Ultra, and even more problems with the Mercury. Unfortunately, the Mercury has the best picture quality of the three, or I wouldn't bother even using it. But it does seem like Fortec is spending less and less time debugging their firmware, and are more and more interested in just jumping into a new receiver........ which brings up the question..... what ever happened to the North American version of their new HD receiver. That was supposed to become available in September, but it never did. They must be having problems with it, and have devoted all their resources to figuring that out.

[Larry Acklin]

Evidence by the fact that the Pansat 2500A and Fortec Lifetime Ultra are the exact same receiver (board) made by (I think) Hundai, it follows that the code is developed overseas by the same batch of coders.

If you look at the $130 retail price point, there can not be too much profit for Fortec to pay much for code, much less do it in house.

Even if they had $30 profit after expenses (no way) it would take a bunch of sales to pay the programmer.

Pure speculation, but I bet they buy the code customised to their brand, and have very little input as to bug fixes or special features.

Even Microsoft is held hostage to some extent by marketing modules (e.g. Hyperterm and others) rather than writing.

Larry

[wejones]

Just as an update on my strange symptoms... I'm starting to think that my diseqC switch may be a large part of my problem. I went out to my fixed Primestar dish, and even though I had it selected at the Mercury, on a channel I was receiving strong last week, I was getting no LNBF voltage out at the dish. Confused, I came back inside to make sure I had the parameters set correctly. While standing there looking at my TV.... POP the channel came on..... strong signal ?!?!? I connected the receiver output to a TV that was out in my garage that I can hear from the dish, and turned the volume up loud. Went back outside, and started moving the dish around... no change. I completely blocked the LNBF.... it was still playing the PBS channel (Montanna). I even disconnected the COAX.... still receiving the channel.
At this point, it became obvious that even though the Mercury was selecting the Primestar dish, that I was actually receiving signal from my 90CM dish, that I also had aimed at AMC3, and the obvious possibilities are that either the DiseqC switch is bad, or that the Mercury is not sending the proper DiseqC signals to the switch. I've going to have to first bypass the switch (in order to re-aim the Primestar dish), then I think I'm going to switch to my old Lifetime receiver, to see if IT is able to control the DiseqC switch, and if that doesn't work, I'll replace the switch (I may actually try that first).

One weird aspect of my problem, is that even when I've switched the input over to my 90CM dish, I am initially not able to get reception on the 12180V/30000 transponder, but if I switch to one of the horizontal channels, like Montanna, it pops in, and I can then switch back to the vertical transponder, and watch that for a few minutes, but it will then pop out (go from a 70% signal quality, down to zero), and the only way to get the signal back is to switch to horizontal for a while then back to the vertical channel. Really weird, and suggests that this is a voltage thing, ie the lnbf isn't getting enough voltage, and is switching polarities on it's own somehow. That's another thing I need to do, ie monitor the lnbf voltage out at the dish, while this is happening. With a foot of snow out at the dishes though, it isn't a particularly good time to be doing a lot of experimentation.

But bottom line is that I'm not sure whether the problem is with my Mercury, the motor, coax, or the switch, but the switch is certainly part of the problem.

[Larry Acklin]

I'm starting to hate Diseqc switches. PIN diodes do not take any overvoltages well, and static even less. I've popped one port on 2 seperate switches just by forgetting to remove power, and unhooking coax. My impression is there is also lots of leakage. I had the same situation- and could get usable quality numbers even with no coax on the port I selected. The stupid switch leaked so much signal that I was very confused.

I have another that only cleanly switches between ports 2 and 4

I'd love to find a brand or model that isn't so touchy. (without having to spend $100 in Canada)

Even reed relays are better than this technology.

Larry

[wejones]

Originally Posted by Larry Acklin

I'm starting to hate Diseqc switches. PIN diodes do not take any overvoltages well, and static even less. I've popped one port on 2 seperate switches just by forgetting to remove power, and unhooking coax. My impression is there is also lots of leakage. I had the same situation- and could get usable quality numbers even with no coax on the port I selected. The stupid switch leaked so much signal that I was very confused.

I have another that only cleanly switches between ports 2 and 4

I'd love to find a brand or model that isn't so touchy. (without having to spend $100 in Canada)

Even reed relays are better than this technology.

Larry

Yeah, I agree. A couple years ago, I bought 4 different diseqC switches, and none of them worked, right out of the box. The particular switch I'm using is one of those 4, however for some strange reason, about a year ago I decided to try it again, and for some reason, it had come back to life, and was working, so I've had it out on my 90CM for the past year, and it seemed to be working fine, but I guess it reverted to it's original condition.

I can't remember what the original symptoms of this particular switch originally were. Sone of the switches wouldn't switch to some of the ports, but one of the switches didn't pass DC properly, which is similar to my current problems, so I"m guessing that maybe that's the one I'm using now.
But yes, I wish there was a source of quality DiseqC switches. My first DiseqC switch cost me close to $30 I think, and it is still working fine on another receiver, but these 4 more recent switches all cost me less than $7 each, and I guess you get what you pay for, however there doesn't seem to be any place you can get the quality ones anymore. I wouldn't mind going back to $30 switches if we could rely on the quality.

[wejones]

Another update of my experiments. To make a long story short, my Mercury is pretty much working again.... I think....

I was preparing to switch over to try another receiver. I had been connecting the Mercury's passthru to my Broadlogic PCI receiver. So I disconnected that, and connected it to my Twinhan, which I was going to then bypass the Mercury via connecting a barrel connector between input and output of the Mercury. WA-LA.... Got reception from 90CM again, on both H and V. I had about as strong a signal as I ever get on AMC3. I thought that perhaps the Broadlogic was sending out diseqC signals, but it wasn't, and I'm using a DC-block, so it shouldn't be interacting re voltage either. I then connected this passthru to my Twinhan.

I thought that perhaps the switch just decided to come back to life again, so I went into the Antenna setup of the Mercury, which was set on the same transponder that I was currently watching, and I had the positioner on "OFF", so it wouldn't try to move the dish. SNAP... signal went away, both on the Mercury AND on the Twinhan, which I was also monitoring via the passthru. Switched back to the horizontal transponder... signal back. switched back to the Vertical transponder. Signal back. So things were still strange.

I also found that even though I got signal back on vertical polarity, that if I watched it for a while, it would eventually go snap, and I'd lose signal, but I could always get it back by switching to horizontal then back to vertical. Strange.

At this point, I disconnected the Mercury, and connected up only to my Twinhan. Reception was solid, both H and V. And I could move to sats using DiseqC-1.2, which was the original problem I had with the Mercury, ie I couldn't get to AMC3 via DiseqC-1.2. With the Twinhan, however, using Tsreader, I couldn't find any satellite via USALS, although I've never used USALS with Tsreader, so I'm not sure how well it works. But anyway, the 90CM seemed to work OK with the Twinhan, so I tried and tried to get USALS to work via altering my station longitude, but I just couldn't find anything. Finally, I decided to see where USALS is actually sending the dish.
First, I sent the dish to AMC3 via diseqC-1.2, then went outside and checked the rotation angle on the motor. Since my longitude is 70.8, the USALS angle for AMC3 should be about 18.1 for me. I checked the motor, and when aimed at AMC3 USALS@=18 , which was just about perfect.
Then I used USALS to send the dish to AMC3, and the USALS@=25 !! Ie WAY OFF. I know that my ZERO is off a couple degrees, but this is 7 deg off.
Then, I used USALS to send the dish to 70.8, which should be my zero, and it went to USALS@=8.5 ?!?! Again, way off. I thought that perhaps Tsreader wasn't calculating USALS offsets correctly, so I re-connected the Mercury. Went into DiseqC-1.2 mode. The Mercury was currently on AMC3, so it said that it was moving to AMC3. I told the Mercury to goto 0 , and the USALS@=17 !!!???!!! {I think that it actually went to AMC3, and then never moved. Ie later I tried several times to get the Mercury to go to zero in DiseqC-1.2 mode, and it never once moved at all.}

I then switched to USALS mode, and told it to goto 0 , and the USALS@= -2 . Ie this was the original error in my zero that I had noticed when I first got my Mercury. This makes it pretty clear that TSREADER doesn't work in USALS mode.

Then I told the Mercury to goto AMC3 in USALS mode, and the USALS@= 16.5 {Ie this is wrong, but correct when considering the 2 degree error in my zero}

The Mercury was still giving me problems, in that it often wouldn't come up on vertical transponders, and the signal would kick out, etc.

I then switched to my Fortec Lifetime receiver.

In DiseqC-1.2 mode, I told it to goto ref, and the result was USALS@= -2 , which is my original zero error. Ie DiseqC-1.2 goto zero works with the Lifetime, but not with the Mercury.

Switched to DiseqC-1.2 mode, and told it to goto 29 (which was my OLD DiseqC-1.2 number for AMC3, and it took me to AMC3 just like it should.)

Switched to USALS mode, and changed my longitude from 70.8 to 68.8 to compensate for the 2 degree zero error, and told it to goto AMC3, and it took me to AMC3 just like it should.

I then went back to the Mercury, changed the USALS longitude to 68.8, and now USALS works fine. DiseqC-1.2 mode seems to be working now too, which it wasn't before. So now, I'm not sure if the Mercury has just fixed itself due to use, or if the warm dry day, has dried out my connections out at the dish, and the Mercury is now able to handle everything. But now everything seems to be working again. I even re-aimed my Primestar, and the once broken DiseqC switch is once again switching to the Primestar when requested.

So I'm confused. There were definately times when things were working with other receivers (Twinhan and Lifetime) and not working with the Mercury, however since it's working now, I suspect that the Mercury's problems were at least partially caused by moisture or cold out at the dish.

Anyway, I guess I'll have to wait for this thing to conk out again, before I can troubleshoot more. I'm tempted to try the new firmware, but I think I'll wait on that too, since other people had problems with it.

[wejones]

Well, I've been semi-sucessfully using the Mercury for the last few days. It's been sort of working on both my Primesar fixed dish and the 90CM/SG2100/QPH-031 , and switching through a 4x1 DiseqC switch has been working.
HOWEVER, from time to time, I'll lose reception. Ie I've been keeping it on the AMC3 12180V/30000 transponder. It will work fine with a strong signal for hours, but usually, after several hours it will suddenly lose signal, and it will NOT come back on it's own. To get signal back, all I had to do was switch to a horizontal polarity channel, then immediately back to the vertical transponder, and reception would be back, and would last again for several hours. But this made it difficult to schedule recordings much into the future obviously.

ANYWAY, due to people in the other thread saying that the version 1.53 firmware correcting other problems, I decided to try that version, and see if it would help MY problem.

So I backed up the channel info, and loaded in ver 1.53, then reset to factory conditions, and recovered my channels.

NOW, while I still receive the horizontal channels fine, I cannot get the vertical channels at all. Ie, switch to a horizontal channel, and back to vertical, and now the vertical transponders no longer come in. I've gone back and forth many times, but I can't get the vertical transponders back.

I hooked up my old Fortec Lifetime, and both polarities work fine with it.

I haven't done much trouble-shooting yet, but I'm starting to suspect that the Mercury might not be putting out enough lnbf voltage.
I checked the lnbf voltage at the receiver using my Channel Master signal meter, which has an LNBF voltage meter, and the Lifetime is putting out 20V on horizontal, and 14V on vertical.

I then re-connected my Mercury, again through my Channel Master meter, and NOW I'm receiving the vertical channels again, and the Mercury is putting out 14.5V on vertical, and 19.5V on horizontal.

What I started to suspect was that perhaps the 14.5V might actually be too much voltage, and my lnbf was switching over to horizontal. So to check this out, I connected my Lifetime to the pass-thru of the Mercury, took the Channel Master out of the circuit and connected the Mercury directly to the dish coax again, put the lifetime on a horizontal channel, and tuned the Mercury to a vertical transponder........ NOTHING, on either receiver. Ie it is NOT that the LNBF is in the wrong polarity, because the Lifetime now doesn't see anything on either polarity coming from the Mercury.

So WHY does the Mercury now work when going through the signal meter, but does NOT work when connected directly to the dish coax???
The only things that seem possible, are that :
(1) somehow something is causing the Mercury to completely shut down when connected directly to the dish coax, or:
(2) for some reason when connected directly, it is sending out distorted DiseqC commands to the switch, which is causing the switch to connect one of my other 3 lnbs (ie the circular port on 1, an un-used C-band lnb on 3, or my Primestar lnbf on 4, instead of my linear QPH031 port on 2).

Re (1), I can't think of ANY reason why the Mercury would shut down when on the low voltage polarity, so (1) doesn't really sound likely, unless there is something really weird about the firmware... but the firmware is definately involved here, because when I changed from 1.33 to 1.53, it stopped working.

Re (2), this sounds possible, but the only way I can prove or disprove this is to figure out which port switch I'm actually on. Perhaps I'll go out to the dish, and connect my meter to port 1 on the DiseqC switch, to see if it is getting power.


Anyway, this is all very strange.


ADDITIONAL INFO:
I was just about to go out and check voltages, and put my TV back onto the Mercury source, and NOW the Mercury is working again on vertical, even though I had been trying for the past 20 minutes to get it to work. The only thing I had done since the last time I checked, was to switch my Lifetime from horizontal to vertical, and back to horizontal. Just to check, I just switched the Mercury to horizontal (works) and back to vertical, and now it doesn't work again. This is all VERY WEIRD. It's possible that it is related to my DiseqC switch, but why is the Lifetime able to control it, and the Mercury with 1.33 firmware able to usually control it, but the Mercury with 1.53 can't??? WEIRD.

EDIT: More info. Just took my meter out to the dish. And sure enough, the DiseqC switch is actually on the #1 port, which is where I have the circular port attached. Not sure what's causing all of this, but I"m tempted to revert back to the 1.33 firmware, and switch the #1 and #2 ports on my switch, because 99% of the time I'm using the linear port anyway. Although I guess this means I'll have to change all the settings in 3 different receivers. But I think I may just slave the Mercury off the Lifetime for a while, in hopes that Fortec might come out with a new firmware version that fixes all this.