Logo

Satellite Internet Forum.
Please note that this forum is now READ ONLY and closed for login and changes
The new forum is here Space and Satellite Communications forum

Welcome, Guest.
Welcome to this satellite broadband discussion forum.  Wherever you are and whatever your problem we are here to help each other. Connecting to the internet via satellite is not always easy but is critically important to those in remote places or with poor terrestrial infrastructure.  Both service providers and customers are encouraged to contribute.  Read the Forum rules.
      Satellite internet forum          

Offset Issue on Evolution Modems !

(Read 12905 times)
yazzdan
Ex Member


Dec 24th, 2009 at 7:07am  
There are so many Remotes which are doing Offset from 10000 - 18000 KHz. What will be the issue ! My LNB stability is between 0 - 10 kHz and modms are not login on Hub.
Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
***
Offline


Personal text from Profile,
Options, Top line

Posts: 2107
Reply #1 - Dec 24th, 2009 at 10:45am  
At the network hub the outlink transmit carrier frequency should be accurate. Make sure the master 10 MHz oscillator in the hub equipment is at a stable temperature. Many hubs use atomic or GPS frequency standards.

Once the hub outlink signal has passed via the satellite, the remote VSATs will see a frequency that is affected by drift in the satellite local oscillator and doppler shifts due to satellite movement.

Finally the VSAT modems will see the frequency affected by LNB local oscillator stability.  High stability PLL LNBs may have accuracies like 10 - 100kHz.  Cheap DRO LNBs may have errors up to to 500 kHz - 2 MHz.

If your VSAT modems are receiving a wide 27.5 Msps carrier then a 2 MHz tuning error (and modem receive acquisition range) is acceptable.  If you are trying to a receive a small 1 Msps carrier you need a frequency accuracy of 100 kHz or less.  

You need an LNB LO stability appropriate to the symbol rate of the outlink carrier. e.g. LNB stability spec = 10% of the symbol rate. Set the modem tuning acquisition range to match.

I hope this general advice helps.  If not, please explain your problem in more detail.  Modem rx acquisition range might typically be set to 1800000.  I believe this means 1.8 MHz.  I have never discovered if this is the total acquisition span, i.e. +/-900 kHz either side of nominal, or if the modem attempts to tune +/- 1.8 MHz either side of nominal.

All the above relates to the main outlink carrier from the hub to all the remote VSATs.

Your expression "so many Remotes which are doing Offset from 10000 - 18000 KHz" is strange.  Does 10000 - 18000 KHz really mean 10 to 18 MHz error ?  If you have a problem with your hub TDMA burst receiver please say.  Can you see the incoming bursts at the hub and are they at the correct frequency and level on an analyser ?  Put a test remote into CW transmit and check its frequency ? Try substituting another TDMA burst receiver ?

Best regards, Eric.
Back to top
« Last Edit: Dec 24th, 2009 at 12:20pm by Eric Johnston »  
 
IP Logged
 
TDMAMike
Ex Member


Reply #2 - Dec 24th, 2009 at 9:03pm  
I am unsure of your question. 

1. Stability of what?  Remote LNB or Hub LNB?
2.  Are the remotes reporting a Green receive light?  If so, are they reporting a Green Tx Light (LED)?
3. If they are reporting a Green TX, can you see them with a SPECAN at the hub?

Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #3 - Dec 30th, 2009 at 7:20am  
Dear M & Eric,

1.HUB LNB
2.  Both the Tx and Rx Lights Shows Green But Net Light is not going to Green after Adjusting from the Hub Down Converter from 10 Khz to 20 Khz than the Remotes became login again.
3. I am seeing it from iMonitor iDx UCP INFO it Shows the Frequency Offset from Above 10 KHz to 18 KHz.


Thanks For the Detail Knowledge . Sorry For the Expression Earlier mentioned is 10 to 18 KHz. Is there any specific Modem Series Issue ??
Back to top
 
 
IP Logged
 
TDMAMike
Ex Member


Reply #4 - Dec 30th, 2009 at 12:39pm  
Based on what you mentioned (that fixed it),  there was no specific modem issue, you just need to ensure that you have the downconverter stability set properly (in the Hub RFT).   The same applies for the  Remote's LNB (Remote RFT) ensure you have the stability set appropriately for the type of LNB you are using (DRO or PLL).


Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #5 - Dec 31st, 2009 at 6:01am  
Dear M,

we are using (PLL) LNB with the Stability of +/- 0 to 10 KHz.
1. Is there any effect on LNB if we change its Stability permanently From 10 KHz to 20 KHz ??

2. Out of 50 Terminals only 8 to 10 Terminals doing this offset problem but as i said earlier when i applied changes (From 10 KHz to 20 KHz ) to these specific 8 to 10 Modems , it logins in a min other than only Rx and Tx light status shows green and no response from NET LED if we didn't apply that Change.

3. Is there any issue related Frequency hopping or Frequency grooming ???

4. what will be the minimum value of Symbol Offset in this Case ?    

Back to top
 
 
IP Logged
 
Admin1
YaBB Administrator
*****
Offline


Personal text from: Profile,
Options, Top line

Posts: 844
Reply #6 - Dec 31st, 2009 at 1:31pm  
Quote:
we are using (PLL) LNB with the Stability of +/- 0 to 10 KHz.
1. Is there any effect on LNB if we change its Stability permanently From 10 KHz to 20 KHz ??


A PLL LNB with internal reference that stays within +/- 10 kHz is very good.  The frequency drift as seen by your receiver will exceed this due to frequency errors at the transmit end, satellite doppler and satellite conversion oscillator stability.  Also the claimed +/-10kHz for the LNB internal reference may be a little optimistic.   Note you cannot change the LNB stability from 10kHz to 20kHz.  The stability depends on the crystal temperature and drift characterics of the crystal and its surrounding components.

When you alter the LNB stability setting in the software you are altering the acquisition sweep range of the receiver.   The wider the range the better will the receiver be able to find and lock to the carrier.  In the case of the outlink - hub to remotes - a wide acquision range is normal e.g. 1800000 (1.8 MHz) to cope with cheap DRO LNBs, also since a remote only locks to the outlink once (when powered on) it does not matter if it takes a whole second or so to lock up.   In the case of the hub TDMA burst receiver the receiver must lock to each incoming burst extremely fast ( few microseconds ) so the initial frequency error must be low.  This is done by controlling all the remote site transmit frequencies, ideally so all bursts arrive at the same frequency (and level.) < some text deleted here > The hub software may tell you what frequency errors are being observed for incoming bursts.

Your hub will have some kind of 10 MHz frequency reference.  This may be a rack with ovenised crystal oscillators or atomic based from local rubidium standard, GPS or Radio.  This reference will feed a 10 MHz supply to all modems, BUCs and external reference PLL LNBs.

If you have ovenised crystal oscillators or are using the iDirect hub oscillator as the reference then it would be worth checking and adjusting this annually compared with an atomic derived standard. Adjusting an ovenised oscillator normally needs a trim tool or may be done in firmware by altering a bias voltage in an EPROM.

Regarding your other questions above, Mike is probably best placed to help.

Best regards, Eric.
Back to top
« Last Edit: Jan 5th, 2010 at 9:06am by Admin1 »  
WWW  
IP Logged
 
TDMAMike
Ex Member


Reply #7 - Jan 1st, 2010 at 7:04pm  
Quote:
Dear M,

we are using (PLL) LNB with the Stability of +/- 0 to 10 KHz.
1. Is there any effect on LNB if we change its Stability permanently From 10 KHz to 20 KHz ??

2. Out of 50 Terminals only 8 to 10 Terminals doing this offset problem but as i said earlier when i applied changes (From 10 KHz to 20 KHz ) to these specific 8 to 10 Modems , it logins in a min other than only Rx and Tx light status shows green and no response from NET LED if we didn't apply that Change.

3. Is there any issue related Frequency hopping or Frequency grooming ???

4. what will be the minimum value of Symbol Offset in this Case ?    


There are three parameters that play a role in frequency offset determination.
- Sweep Range:  Sweep range is determined based on the stability of the down converter. Default value is +/- 10000 Hz.
- Sweep Interval: Sweep Interval is provided by the NMS server and dependent on the upstream symbol rate.
- Sweep Method: there are two methods of sweeping, one is always enabled by default. The value of sweep method is set to 1 for fast acquisition.

Yes, there will be an effect when you modify your HUB downconverter stability setting from 10 to 20.  Are you certain that your Hub LNB stability (spec) is 10Khz?  Recommend you double check.  

When you modifed the downconverter stability from 10 to 20, did all the remotes enter the network?  If they did, and you were not observing any visible errors, it is likely a safe stability setting to use (although outside your stated specs), but I would recommend keeping things under observation over a 24-48hr period.  Pay close attention to when remotes leave the network, and ensure they come back into the network relatively quick.

It is doubtful your issues are related to Carrier Groom or Freq hop settings.  Are you using FH or Carrier groom?  How large are your upstream carriers?  

Minimum Symbol rate question:   If iMonitor is reporting symbol offset as high (>30000), that is clearly indicative of a timing issue.  Symbol offset is normally anywhere from single digits up to 300.  What are you presently seeing with regards to Symbol offset (remember each node will be different)?  

As you are probably aware, symbol offset is the value associated with the movement of the spacecraft. It is a timing correction value that is adjusted by the system as the spacecraft moves in its station box.  Again this is normally a low number.  

Back to top
« Last Edit: Jan 2nd, 2010 at 1:15pm by N/A »  
 
IP Logged
 
yazzdan
Ex Member


Reply #8 - Jan 4th, 2010 at 10:35am  
Dear M,

*Are you certain that your Hub LNB stability (spec) is 10kHz?  

Ans.Yes, i Have Checked it is 10 kHz.

-----------------------------------------------------------------------
*It is doubtful your issues are related to Carrier Groom or Freq hop settings.  Are you using FH or Carrier groom?  How large are your upstream carriers?    

Ans. Yes, We are on Frequency Hopping
------------------------------------------------------------------------
*What are you presently seeing with regards to Symbol offset (remember each node will be different)?  

Yes, it is Correct no issue in Symbol Offset
-------------------------------------------------------------------------

One thing I would like to be in your notice that in One RF Chain we are using iDirect iNFINITi as well as iDirect Evolution but issue is only on Evolution Network. ???

Back to top
« Last Edit: Jan 4th, 2010 at 11:36am by Admin1 »  
 
IP Logged
 
TDMAMike
Ex Member


Reply #9 - Jan 4th, 2010 at 11:46am  
Ref the downlink RF/IF chain (at the Hub): is that downnconverter shared?  ex: is that same 10kHz downconverter used by both networks?

I am not 100% certain, but I believe there were changes to the acquisition aperture for iDX.  Do you have iSupport?  If so, I recommend you call the TAC and ask them.  I say that because I am uncertain.

M
Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #10 - Jan 5th, 2010 at 7:07am  
*Ref the downlink RF/IF chain (at the Hub): is that downnconverter shared? 

Ans. yes the Down converter is Shared(10KHz)

Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #11 - Jan 5th, 2010 at 8:12am  
Dear M,

Do u think that the problem is due to the different acquisition aperture on both the networks.

I have checked and can verify that the acquisition aperture is slightly changed. I can get that changed to the same value for both the networks. Do u think that will help..???
Back to top
 
 
IP Logged
 
TDMAMike
Ex Member


Reply #12 - Jan 5th, 2010 at 11:42am  
It is possible.  How much do they differ?   If you have iSupport I suggest you call the TAC and inquire.
Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #13 - Jan 5th, 2010 at 12:23pm  
Dear M,

According to iDirect TAC, They said nothing to do with acquisition aperture !
Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
***
Offline


Personal text from Profile,
Options, Top line

Posts: 2107
Reply #14 - Jan 5th, 2010 at 12:39pm  
The original post said: "There are so many Remotes which are doing Offset from 10000 - 18000 KHz."

I presume you mean " 10000 - 18000 Hz" and that you are referring to the burst receive frequency offset at the hub TDMA receiver.

Can you make a list of all the remotes and the measured offset of each one ?

Does the collection of offset values make a pattern, e.g. normal distribution curve ?

Does the pattern drift with temperature or time of day ?

Best regards, Eric.
Back to top
« Last Edit: Jan 5th, 2010 at 1:41pm by Eric Johnston »  
 
IP Logged
 
yazzdan
Ex Member


Reply #15 - Jan 11th, 2010 at 5:04am  

Q.Can you make a list of all the remotes and the measured offset of each one ?

A.Yes , i have 10 Remotes that are doing problem

Q.Does the collection of offset values make a pattern, e.g. normal distribution curve ?

A.No, Not Exactly in a pattern, one of the remote have 18 KHz , other ones have 12 , 13 & 15 KHz..

Q.Does the pattern drift with temperature or time of day ?
A. In This Case, i want to check ..


Best regards, Eric.
Back to top
 
 
IP Logged
 
Adrian
Ex Member


Reply #16 - Jan 11th, 2010 at 9:24am  
The acquisition aperture can be changed from downconverter stability. You can modify the downconverter stability making the aquisition aperture wider (eg from 10 kHz to 20 kHz or 30 KHz).
Notice:
1.1. The configuration will be applied only to remote Hub side and the netmodems that are already in the network will remain in the network.
1.2 The acquisition time for the remote will be longer than usualy.
1.3 After acquisition (if it works)  you can change at the old value the downconverter stability.
1.4. For netmodem Infinnity, the acquisition in the network are made even if the aquisition range is narrow because after a period of time I think that change de mode aquisition from fast aquisition to wide acquisition.

I think this will not solve the problem but will give a clue about what happening.

Best regards,
Adrian
Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #17 - Jan 11th, 2010 at 9:49am  
Dear Adrian,

Thanks for the Reply

In this Case issue is that, i have already did this exercise on those remotes and the remotes became log in but if the modem is restarted, same exercise we have to do from hub side.

*** This Clue might help you to understand my problem.


Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
***
Offline


Personal text from Profile,
Options, Top line

Posts: 2107
Reply #18 - Jan 11th, 2010 at 10:32am  
Do all the remotes have the exactly the same outlink frequency, symbol rate and translation frequencies specified in their options files ?  Minor discrepances will not affect the outlink reception at the remotes but the remotes may use the actual outlink frequency and symbol rate and the programmed (text) frequency numbers to derive the exact remote transmit frequency.

I don't know how iDirect controls the remote return link frequencies.

In the hub software are all the remotes being told the same information regarding return link frequency into the TDMA line card ?

There was mention of two networks.  Are both networks in the same transponder ?  Satellites may use different conversion oscillators for transponders on opposite polarisations for example.

If you widen the acquisition window you will slow down the acquisition time.  The may decrease the traffic throughput significantly as fast acquisition is essential for short TDMA bursts.  The TDMA line card receiver must retune at the start of every burst.

Best regards, Eric.
Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #19 - Jan 12th, 2010 at 6:14am  
Q. In the hub software are all the remotes being told the same information regarding return link frequency into the TDMA line card ?

A. Yes.

Q.There was mention of two networks.  Are both networks in the same transponder ?  Satellites may use different conversion oscillators for transponders on opposite polarisations for example.

A. Yes, Both are on Same Transponder.

Dear Eric,
Let me tell you the Senerio , We have two Live Hubs (iDirect infinity & Evolution) on same RF Chain. iDirect Infinity network is working smooth and no problem of frequency offset but on the other end there is evolution network with same RF Chainn In that system some of the remotes have offset issue.
P.S. Both networks have same parameters but issue come up with Evolution. like Adrain Said ...
-----------------------------------------------------------------------
The acquisition aperture can be changed from downconverter stability. You can modify the downconverter stability making the aquisition aperture wider (eg from 10 kHz to 20 kHz or 30 KHz).
Notice: 
1.1. The configuration will be applied only to remote Hub side and the netmodems that are already in the network will remain in the network.
1.2 The acquisition time for the remote will be longer than usualy.
1.3 After acquisition (if it works)  you can change at the old value the downconverter stability.
1.4. For netmodem Infinnity, the acquisition in the network are made even if the aquisition range is narrow because after a period of time I think that change de mode aquisition from fast aquisition to wide acquisition.
------------------------------------------------------------------------
we always do this exercise and make forcely to login the modems.

i though Adrain will come up with some conclusion waiting for his responce...
Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
***
Offline


Personal text from Profile,
Options, Top line

Posts: 2107
Reply #20 - Jan 12th, 2010 at 9:51am  
Do both networks use the same outlink carrier ?

I don't have any experience with iDirect so please if anyone knows how iDirect controls the remote transmit frequency and how the TDMA receiver line card deals with off-frequency remotes, please help.

Best regards, Eric.
Back to top
 
 
IP Logged
 
yazzdan
Ex Member


Reply #21 - Jan 12th, 2010 at 10:03am  
Q. Do both networks use the same outlink carrier ?
A. Yes, both the outbound carriers are diiferent.

Back to top
 
 
IP Logged
 
Adrian
Ex Member


Reply #22 - Jan 12th, 2010 at 3:28pm  
I had a similar problem with the terminals in my network but sincerely I think you have a different situation.
Anyway I solve the problem in this way (after many attempts like swap downconverter with backup downconverter and the entire receiver chain, including the Rx HLC with standby HLC).
From Hub Chassis, I swap the Reference Clock Module (Master) with the reference clock Module (Slave), and  the frequency offset from the netmodems was more stable than in the first configuration, but not enough. So I conect an 10 MHz external reference to RCM and the network work fine.
Notice
The external reference clock must be very stable (Not working with 10 MHz from netmodem iDirect).
The only explanation what I have is the out of calibration RCM.

Other suggestion:
1. Check if the BUC from the remotes are similar for the sites what working and different for the site with problem.
(maybe are differencies between radio equipment configuration for infinnity and evolution remotes) and the upstream Tx frequency offset are to big.In this case you can put the stability for hub downconverter at a convenient value (40 kHz) and mentain this value to see what is hapend.
2. Try to swap Rx HLC with a spare one (or Tx card) because HLC have a FLL loop. Maybe you have a problem with demodulator.


Best regards,
Adrian




Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
***
Offline


Personal text from Profile,
Options, Top line

Posts: 2107
Reply #23 - Jan 13th, 2010 at 9:27am  
I don't know the iDirect hardware but the Reference Clock Module (Master) and reference clock Module (Slave) probably have an adjustment possibility to set the frequency.  Note the adjustment position now and then try adjusting it.

Typically you need a trim tool to adjust some capacitor or resistor inside the crystal oven.  Don't adjust unless the unit has been powered on for many hours in its normal operating position. Ideally adjust the frequency by comparison with an atomic based reference, e.g. GPS or radio standard. Aim for 0.01 Hz accuracy at 10 MHz. i.e 1 slip every 100 seconds on an XY scope.

Best regards, Eric.
Back to top
 
 
IP Logged
 
Adrian
Ex Member


Reply #24 - Jan 13th, 2010 at 12:24pm  
Dear Mr. Eric,

In theory you have perfectly right.
But practical it's very difficulty to adjust RCM frequency.
That  because I don't have any  circuit diagram for RCM and it have a lot of component on the printed circuit that don't have a clue what it do. I see where is the crystal oscillator but it isn't enough.
It is neccesary to know at least a pinout for the  RCM board equipment and it very hard to adjust the frequency when you don't know where is the pinout for voltage supply or output reference signal.
It isn't possible to make this operation with the RCM in use because you don't have any additional space to adjust any component. I can't even to measure the reference signal because the risk to damage the board is very high. I think it is necesarry an aditional printed circuit that allow you to do that outside the RCM. And the things begin to be more complicated because you need the same type connector  like that are used by hub chassis.

Otherwise, you need a very good instrument for measure frequency and at this precision it is hard to get one because is very expensive.

First I chose to check if this is the real problem and after that I will  decide what are the best way to solve the problem.

Best regards,
Adrian


Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
***
Offline


Personal text from Profile,
Options, Top line

Posts: 2107
Reply #25 - Jan 13th, 2010 at 3:16pm  
I'm not sure quite what Master and Slave means but if in fact, there is a main and standby master oscillator, then does switching to the standby oscillator make any difference to the offset values ?

Best regards, Eric.
Back to top
 
 
IP Logged
 
Adrian
Ex Member


Reply #26 - Jan 14th, 2010 at 9:25am  
According to iDirect if power is applied to a chassis that has both RCM installed, either RCM can become the chassis RCM master. The first RCM to become operational will determine if another RCM is installed in the chassis and will arbitrate to determine wich one will become the master RCM.

I notice the RCM slave track to 10 MHz reference signal from RCM master. This is accomplished by Frequency-Locked-Loop circuit on the RCM.
If the 10 MHz reference signal from master RCM isn't calibrated and have an offset frequency the master RCM don't consider that failure and remain master.

Best regards,
Adrian
Back to top
 
 
IP Logged
 
jobione
Ex Member


Reply #27 - Jun 7th, 2012 at 2:42pm  
Does anyone have any additional info on this issue?

It seems we are having the exact same problem with the evolutions and not with our infinitys.
Back to top
 
 
IP Logged