There is a problem with locked channels in pip, they can be seen freely without the pin request. The same is in the channel list which uses pip functionality for channel preview.
dm900 + experimental image 2018/11/27

I found out that PIN mechanism doesn't work at all with my FBC twin tuner (it's labeled as 8S2I FBC Tuner v03). I have assigned one (doesn't matter the same or different) user band per each of my 4 unicable II LNBs. The same setup works on my PC dvb card with PIN active and PC card is always able tune a channel. But receiver still ends up with tuning failed message. When I set exactly the same user band for the receiver and for PC card and then tune a channel on PC card, the receiver can tune channels with the same frequency. Immediately after I switch off PC software the signal on receiver drops to zero and I get tuning failed message again.

For testing purpose I have replaced FBC tuner in my dm900 by old one from dm800se (single tuner) and everything immediately works rock stable. I can select single band from each LNB and together with PIN it works correctly. If I disable PIN protection mechanism, as expected, tuner is not able tune anything (it works exactly the same as my PC card). Now I am sure that PIN mechanism is crucial for correct function of my LNB equipment.

Back to my original problem - something seems to be wrong with either my FBC tuner (but it works with any legacy LNB and even single unicable LNB) or (most likely) with its driver, especially with the processing of PIN protection of user bands.
Is it possible to shift this information to the Ghost in order to look at this particular question? Thanks.

Edit:
The single tuner has the same issue, it only took him longer to lock the LNBs. So also here PIN mechanism doesn't work properly.

It's not classic timing problem. How can you tweak exact values while default time varies by almost 300%? I don't need to tweak the time after diseqc change or time after voltage drop (only these 2 can be set) for some specific switch to be able complete switching process. Time for diseqc command (18V-diseqc-14V) is the issue, especially inconsistency of its duration. I can insert the pulse pictures from the scope if you want. This is separate issue from the next and maybe it's not the root cause of my issue (but may relate to it, it's at least strange behavior).

The plugin can set only one unicable LNB at a time so it is not much useful for me. In enhanced mode it also crashes the enigma during start with the message:

Traceback (most recent call last):
enigma2[236]:   File "/usr/lib/enigma2/python/mytest.py", line 82, in <module>
enigma2[236]:     from Components.DreamInfoHandler import ImageDefaultInstaller
enigma2[236]:   File "/usr/lib/enigma2/python/Components/DreamInfoHandler.py", line 4, in <module>
enigma2[236]:     from Components.NimManager import nimmanager
enigma2[236]:   File "/usr/lib/enigma2/python/Components/NimManager.py", line 2067, in <module>
enigma2[236]:     InitNimManager(nimmanager)
enigma2[236]:   File "/usr/lib/enigma2/python/Components/NimManager.py", line 1719, in InitNimManager
enigma2[236]:     doc = xml.etree.cElementTree.parse(eEnv.resolve("${sysconfdir}/enigma2/unicable.xml"))
enigma2[236]:   File "<string>", line 62, in parse
enigma2[236]:   File "<string>", line 38, in parse
enigma2[236]: cElementTree.ParseError: mismatched tag: line 17, column 3

where on the line 17 there is only </lnb> which definitely matches previous <lnb>. Anyway, as I wrote, it's not the correct tool for me, I guess.

I also checked NimManager.py file which parse system config (and also unicable.xml). It simply checks pin usability and once used then assign the pin for lnb (not for the user band as it should). I wonder how it could pass more pins for a single lnb once the parser doesn't support it. As you wrote earlier, it would be nice to support it. I would say that once it claims that it supports unicable II/jess, it should support also complete pin logic. There are cases (like mine) where it's used and necessary. I've asked the manufacturer for advice and I get the message that system must support unicable II in all aspects otherwise they cannot guarantee correct functionality. It is quite expected reaction from their side, and obviously they're right, specification is clear.

There is something wrong with the latest version of dm900 drivers regarding unicable II (issue starts with October's update, but I am not quite sure about exact time). I found out that after specific time my 4 Inverto LNBs lock to single frequency (not to user band, what's strange that all 4 LNBs lock at the same frequency on which active tuner has been tuned) and I cannot tune any channel with another frequency since then. I always get "tuning failed" message. In rare cases even power off of the LNBs doesn't help (usually does). And now a little explanation why I wrote this.

A few days ago I was able to tweak the settings of PC tuning application for my PC dvb card and since then this PC card always works. It is always able to tune any frequency from any LNB even when my "LNB 4-block" is in locked state. Moreover, it unlocks it and since then it's possible to tune channels on dm900 again. It means that once the LNBs stuck on single frequency then I am able to unlock them by starting of PC tuning application. I'm almost sure that there is something wrong with the tuning in the enigma2/driver. I can see difference in PIN possibilities (PC card has specific PIN for specific user band, once I deactivate PIN then I am not able to tune it too) and yesterday I also tested the outgoing signal on my dm900 on digital scope. There is quite wide tolerance for power pulse for diseqc command. I saw that standard timing of sequence 18V-diseqc-14V should be approx. 100ms but from time to time there is almost 3-times longer time period (280ms) and regularly varies between these two values. It can be simply replicated and measured this way. I guess that there must be some side branch (or waiting loop for a driver?) which causes such difference in timing. I don't see any difference in PC tuning application, there is still the same switching interval 150ms, neither shorter nor longer. I guess that 120ms should be enough but in my case the PIN functionality is crucial otherwise I am not able to switch the channel once the LNBs lock. I use the latest experimental version from 2018/11/27, FBC S2 tuner (not S2X version) and 4x Inverto IDLU-32UL40-UNBOO-OPP LNBs. If you want, I can test it.

I tested also Unicable Editor plugin but it didn't find unicable.xml file anywhere in my /etc/ folder and it's subfolders. Maybe it works for matrix setup or switch only but my configuration is 4x LNB with combiners, splitters and power inserters.

I've got 4 lnbs with programmable pins with the similar configuration as you have mentioned with pin active. But every time I want to add a new device I have to re-program all 4 lnbs to allow new device to use it's own pin. I tried to set the pin all the same but it works the same like no pin at all. For me there is a strange situation with no pin active. Even when I have specific user band assigned to each from 24 tuners, from time to time I got message "Tuning failed" and I have to turn off the power inserters (which are necessary for 4 lnbs). Then I can tune the same frequency again. It happens at least once a day. With pin active I have no issue.
I will try your plugin, thanks for suggestion.

Zitat von gutemine

even if you give a user multiple scrs you would not confuse him with multiple pins.

But then there is no reason to show 1 PIN for let say 4 user bands. Why? It can be valid only for 1 of them, am I wrong? That's what's confusing me more. I guess PIN is not intended for all users and these who need it won't be confused.

Zitat von h2402

Is possible to drive an USALS motor with DCSS lnb for FBC tuner on dm920 ?

I guess yes. I use different lnb with usals motor and all is working fine here.

Zitat von gutemine

DreamOS has a superior fully dynamic FBC configuration and Tunermanagement, so yes, you can define all 4 Tuner entries of 2 FBC Tuners for a different position and if you have 8 Unicable Frequencies on each LNB you will be able to receive 8 Tranponders at the same time on all 4 postions.

If not, then I'm pretty sure that Ghost will be pleased to fix it, because this was one of the strategic investments that DP has done in DreamOS to have a fully dynamic Tunermanagement and Unicable frequency allocation.

One thing which is not 100% correct here is a PIN logic. PIN is a security mechanism which prevents the usage of already used/assigned user band. Every user band has its own PIN (fixed or specified by user). Once I can select/use more user bands for a single unicable LNB then I should be able to specify more PINs (for each selected user band). Now I can specify only one common PIN in the interface. So the only one working option atm is to set the PIN option inactive (or active without the real impact, wrong PIN = not working lock).

Zitat von gutemine

DreamOS has a superior fully dynamic FBC configuration and Tunermanagement, so yes, you can define all 4 Tuner entries of 2 FBC Tuners for a different position and if you have 8 Unicable Frequencies on each LNB you will be able to receive 8 Tranponders at the same time on all 4 postions.

If not, then I'm pretty sure that Ghost will be pleased to fix it, because this was one of the strategic investments that DP has done in DreamOS to have a fully dynamic Tunermanagement and Unicable frequency allocation.

I can confirm that it works like this. Good job, guys.

OK, it seems that storing of unicable slot has been solved in 4.3.1r8.

Now I am on 4.3.1r9 and I cannot select the same unicable slot number for two satellites even when lnbs are proprietary programmed to specific chained lnb position (like A for 19.2E, B for 13E, see the attachment) and even with different pin assigned to each tuner. It means that I have to choose unique slot number for each tuner and lnb combination. The result is a half of the capacity of slots (16 instead of 32) for each lnb. If I use 4 lnbs (as a maximum for chained lnbs) then only 8 slots per each lnb are available. What somehow doesn't work is the assignment of sat position to specific unicable chained lnb A, B, C or D. I can compare it with my G60 Panasonic TV (unicable I). There I can select only one slot and assign one sat to A and another one to B chained lnb (without use of any pin). Then in a status window of the programmer I can see that this simple TV tuner uses for both lnbs the same slot number without any issue and properly tune both sat positions. If I try this with dm900 it always ends up with Tuning failed message. It looks like the tuning doesn't distinguish the correct sat position/lnb and then it tries to tune both lnbs at the same time. I hope you got my point.

4.3.1r7-2017-05-23

Edit:
Now I can see that similar problem has been fixed also in 4.3.1r8.
I will try to update the system and then I'll give you a shout.

I found a bug in unicable II/JESS setup. In case that I want to use 2 unicable II lnbs (like IDLU-32UL40-UNBOO-OPP) for different satellites at the same time and I want to set them up it is not possible to save settings for both. Once I set one lnb for one satellite and try to set the second one for another satellite I cannot save unicable slots for both. Once I press the OK button to save settings (green button does not work for storing in this part of menu) and open them again then the values for one of them are back to original. Only possible way is to directly edit settings file. I use oe2.5 experimental image for DM900uhd.