IS there an Option to retune the tuned Channel on Errors? (1 Viewer)

[mm1352000]

Hi again MaxT

This problem is strange and frustrating! I'm sorry the last build didn't fix it.

The test build that you just tried should have been continually checking for signal lock while the motor was moving. Since the tuner didn't lock, we have proved that the solution is not simply a matter of waiting for the motor to stop moving. It is something else.

Let us be very clear: my understanding is that the problem only occurs on change of satellite (when the motor has to move). You can successfully tune to any channel after TV Server starts. You can also successfully tune to other channels on the same satellite.

Perhaps the switch command needs to be resent after the motor has stopped? If this is the case then we have a problem because we can't tell when the motor has stopped moving!

Seems that something blocks the signal lock while dvb: LockedInOnSignal waiting ..... Is it possible to run this after motor timeout?

So you want to do nothing until the motor timeout has expired? Okay then. I'll post two more test builds:
1. First, a version that does what you want - it will do nothing until the motor timeout has expired. At that point I will resend the DiSEqC command and move on to check for lock.
2. A slight modification to the previous version to record the tuner statistics (locked, present, quality, strength) while TV Server is doing "LockedInOnSignal waiting...".

Two questions for you:
1. Are you able to check whether the motor is actually moving to the correct position when the tuner fails to lock?
2. When you perform the attempt to tune after lock failure, does the motor move at all?

 

[mm1352000]

Okay, here you go!

[Edit: attachments removed to save space on my account]

 

[MaxT]

Hi again mm1352000!

Actually I cannot ask or demand anything from you and I am really grateful that you are helping!

As to the situation:

Indeed in majority of cases the problem occurs on sat change and it rarely occurd with the very first tune (to the same sat) after TV Server start, however with your builds with diseqc command resending I've never experienced it, and of cource I can tune to other channels on the same sat without an issue.

As to the switch command and the motor move, I think that ideally this could look like:

1 command to move the motor to the position with a fake parameters for the LNB so that high voltage (18V) is provided to the motor (to speed up the move)
2 angular delta in sats positions multiplied by the MotorTimeout. This timeout should be user controlled since it depends not only on particular motor but on user's longtitude and for different users the actual turn angle between thу same satellites can differ.
3 real LNB switch command sent twice (in any case, irrespectively whether motor moves or not) just as a control shoot with another user controlled Timeout and the TuneTimeout starts running.

Since 2 represents quite an issue - to read strings, etc. this can be replaced by a fixed MotorTimeout.

I think that for example ProgDVB usues the same logic - by default it uses 1000 ms timeout per turn degree (MotorTimeout), for TwinHan cards resends commands 2 times (adjustable) with 200 ms timeout which is also adjustable.

Do you think this if this is logical?

Refards,

MaxT

ups, sorry I've seen your post only after posted mine - was speaking over the phone. Will test are write back

 

[MaxT]

Ver. 1 is WORKING!!! Several tunes back and forth so far!!!

Ver. 2 is not

However seems that V1 wait for the motor timeout to pass irrespectively whether there is a need to move the dish or not (last 2 tunes were on the same sat). Did not check this with V2 since it failed to tune at all :sorry:

Also you've asked:
- whether I am able to check whether the motor is actually moving to the correct position when the tuner fails to lock. I think I can since I see the disc which is on my balcony.

- whether the motor move at all when I perform the attempt to tune after lock failure. Not sure I understand the question, but if you ask whether the motor moves in those cases when the tune fails, the answer is yes, the motor moves - I hear and or see it.

Probably I will continue testing V1!

Regards,

MaxT

 

[mm1352000]

Ver. 1 is WORKING!!! Several tunes back and forth so far!!!

GREAT

Ver. 2 is not

This is not surprising. It is the same as the first version except with some logging added.

However seems that V1 wait for the motor timeout to pass irrespectively whether there is a need to move the dish or not (last 2 tunes were on the same sat). Did not check this with V2 since it failed to tune at all :sorry:

I'll fix that. Give me 5 minutes and I'll post a new version that fixes the unnecessary delay. As I said above, I didn't expect V2 to work so no worries about that.

- whether the motor move at all when I perform the attempt to tune after lock failure. Not sure I understand the question, but if you ask whether the motor moves in those cases when the tune fails, the answer is yes, the motor moves - I hear and or see it.:

What I mean is, when you change satellite, the motor moves but the tuner fails to lock. Does the motor move again (maybe even just a little adjustment) when you attempt to tune again immediately after the tuner fails to lock.

Check back soon

 

[MaxT]

What I mean is, when you change satellite, the motor moves but the tuner fails to lock. Does the motor move again (maybe even just a little adjustment) when you attempt to tune again immediately after the tuner fails to lock.

I never heard/seen that move - I often open balcony so I perfectly hear the motor and know that when I give the comand to go to 0 it moves a little bit around 0, but this does not happen with sat positions.

Regards,

MaxT

 

[mm1352000]

Sorry for the delay. Here is the new version...

I'm off to bed now. It is 1:00am - too late to do anything more on this tonight...

Some "food for thought": right now the patch is a hack. I will have to think very hard about what we can do to create a patch that would be accepted by the team. I think the only option may be something like what you suggested earlier:

1. Find the distance that the motor has to move.
2. Send the switch command (if any).
3. Send a fake 18V (horizontal polarity) tune request so that the motor moves fast.
4. Wait for distance * motor movement multiplier.
5. Resend the switch command (if any).
6. Check for lock.

If the motor doesn't have to move, then all you have to do is send the normal switch command (if there is one).
The tricky parts are step 1 and step 3. We've already discussed the difficulty of step 1. It can be done, but will take careful thought. Step 3 is trickier to do because of the way the code is. To be completely honest, I think it is unlikely that the patch would be done in time to get added to 1.2.0 so I may have to keep building custom versions for you.

Alright, let me know if there is a problem with this release and I'll try and take a look again tomorrow. It would be good if you could test that scrambled channels work too.

[Edit: attachment removed to save space on my account]

 

[MaxT]

Hi mm1352000!

Was a bit overloaded today, so tested V3 just recently. Unfortunately it does not work as expected- have to press tune for the second time . Tested twice: with motor\tune timeouts 33\28 (or smth like this) and 30\5. It seems that no retune command is sent and have not seen the delay for the motor movement.

Also just realised - it seems that initially MP sends the tune command to LNB and then to the motor and then tries to lock the signal. But does not the command to the motor overrides the command to LNB and therefore signal just physically cannot be locked or these commands do not overlap?

If so, the way you suggest in general is the only way to make thing working rather than a hack, hack is only item 3 to speed up/unify motor speed and it adds convenience rather than fuctionallity. So it is possible to give it up although it would be nice to have. What do you think?

BTW after this you will be extremely close to USALS implementation which so many dream about

Regards,

MaxT

 

[mm1352000]

Sorry, in a rush right now. Try enabling the pause graph option...

 

[MaxT]

Hi!

As you suggested, I tried pause graph and managed to tune which was quite strange - only PC reboot, following TV server separate restart and without motor move. So not sure whether pause helped actually.

Further noticed that when later tuning to another channel on the same sat logs show MotorTimout delay and when tuning from one sat to another log shows that no need to move motor and therefore no MotorTimeout while motor actually moves!

After further TV Server restarts (tried stop/pause graph back and forth) could not tune at all.
Will now reboot, try and edit this post.

So it seems that second and further tunes on the same sat are working. Move to another sat does not untill trying another channel which forces MotorTimeout. Seems that reverce logic is needed.

Regards,

MaxT