Quantar Simulcast Issue

[WmHPowell]

I'm looking for some Quantar "engineering level" help re: an interesting simulcast issue.
I live in an area where I can hear several of our high band simulcast Quantars. The whole thing was installed and set up by Motorola including GPS stabilized time bases.
I'm monitoring with a true "monitor": wide band IF and little limiting.
When the dispatcher drops a dead carrier I hear little in the way of hetrodyne or "grunge" as it should be.
However, when the dispatcher drops alert tones I hear a hetrodyne that decreases in frequency over the duration of the tone.
My guess is that the tone is somehow pulling one of the VCOs in a Quantar exciter because of a lack of DC restoration in the modulator: a capacitor is charging and slightly shifting frequency.
I consider this to be abnormal and undesirable behavior - especially in a $y$tem of thi$ caliber.
I haven't done any field tests yet. I suppose I can set up 2 service monitors: one to receive in the AM mode and the other to provide a reference carrier and then send tone to each transmitter, in turn. That, at least would let me isolate the problem to one, two, ?? radios.
Has anyone else experienced a problem like this?
Any Motorola engineers out there? Our local tech is also baffled so
I'm reaching out for ideas.
Thanks,

[Bill_G]

Have you brought this up on MTUG?

http://www.motorola.com/business/v/inde ... trefresh=1

[Bill_G]

I missed the part about VHF. Forget MTUG though they would be helpful. Tell us more about the system. How old is it? How many sites? How does audio ship out: telco, microwave, ip? How does voice sound? Are you using tx pl or csq?

[WmHPowell]

Thanks for the link to MTUG!

> Tell us more about the system. How old is it?
One channel about 4 years, the other 2 less than a year

> How many sites?
4 total

> How does audio ship out: telco, microwave, ip?
Private microwave. Uses DSM-II cards in IMACS channel bands. Everything synched via GPS locked Rubidium clocks.

> How does voice sound?
Fine except for right in the center of the overlap zones.

> Are you using tx pl or csq?
Right now PL supplied via USCI at central site. Will be switching to DCS when we narrowband, also supplied at the central site.

This is a strictly by the Motorola book DSM-II simulcast system!

NOTE: What I am concerned about is that the hetrodyne pitch (tone) CHANGES with the length of the transmission - during the first 3 to 4 seconds. It moves OFF frequency when the tone is first applied and GRADUALLY moves back.
It is repeatable and the tone frequency does not matter. I suppose a drawn out "HELLOOOOOO" would do about the same thing.
I'd like to get a firm handle on this before we start building out our 10 channel / 10 site T band system.

Thanks
Bill Powell

[Bill_G]

Okay. You said IMACS. Are the channel banks IMACs upgraded with DSM cards, or were they sold as Tensr channel banks with DSM cards already installed? Is the microwave and Tensr new along with the Quantars? Do all four sites overlap? Can this problem be heard everywhere? Have you tried disabling one site at a time to isolate the problem?

[WmHPowell]

> You said IMACS. Are the channel banks IMACs upgraded with DSM cards, or were they sold as Tensr channel banks with DSM cards already installed?
Yes and no. Original alert channel was Moto installed. DSM for 2 additional channels installed later.

> Is the microwave and Tensr new along with the Quantars?
No - Microwave, TenSr, DSM cards and Quantars for one channel all installed at same time - same purchase from Moto.
We later purchased additional DSM-II cards and Quantars to convert 2 additional channels to simulcast.

> Do all four sites overlap?
More or less: terrain is far from flat. POverlapping coverage is the major reason we are going simulcast; fill in one-site shadows.

> Can this problem be heard everywhere?
Everywhere I've listened.

> Have you tried disabling one site at a time to isolate the problem?
Next on the list to do but requires taking down a very busy and vital channel: tone dispatch for all fire and squad pagers!

Thanks,
Bill Powell

[ard099]

What are the paging code frequencies?

[WmHPowell]

Why?
It doesn't matter in the issue.

[xmo]

Does the system use Trak 9100 clocks?

If so, are the Quantars fed from 9106 Frequency Distribution Modules or from 9107 or 9109 modules?

Have you verified that each station is programmed correctly for external reference and locked to it? You can disconnect the reference coax and the station should go into alarm [synth out of lock].

Have you performed the simulcast netting at each station?

Do you have access to a 35670A or 35665A?

[fineshot1]

Possible FSB10432 problem? - Note: This is not the entire FSB - please see the actual Motorola FSB for
appendix A info.

FSB NUMBER: FSB10432
Government and Public Safety APC: 509
1301 E. Algonquin Road DATE: Jun-10
Schaumburg, IL. 60196 EXPIRES: 30-Jun-11

SUBJECT: ASTRO Simulcast timing difference between Quantar stations equipped with EPIC V and EPIC II control boards.

MODEL / SYSTEM AFFECTED:
All ASTRO Simulcast sites using Quantar EPIC V stations mixed with Quantar EPIC II stations.

SYMPTOM:
At ASTRO Simulcast sites; EPIC V signal launch is 8.2μs late comparing to EPIC II.

CAUSE:
The EPIC V control board introduces 8.2μs delay in the ASTRO Simulcast path compared to the EPIC II control board, which may lead to sub-optimal audio quality in the overlap areas. To ensure maximum audio quality, transmission from each site must be synchronized and delay differences should be avoided.

RESOLUTION:
If the ASTRO Simulcast Test Pattern alignment was correctly performed before and 8.2μs delay was already added to the EPIC II boards, no action needs to be taken. If this alignment has not yet been completed or a EPIC II/EPIC V board has been added as a replacement then the 8.2μs delay must be added to the Simulcast Launch Time Offset of any station that has an EPIC II board in it.
Note: See Appendix A for EPIC Identification chart
SEVERITY RECOMMENDATION:
Medium - Perform at next scheduled maintenance
PARTS REQUIRED (HARDWARE/SOFTWARE):
NA
LABOR ALLOWANCE
This is an informational bulletin. No labor warranty is implied, intended or authorized.

[ard099]

Its a good possibility that you have a problem with a VCO, and what XMO suggested is a good idea too. The reason I asked for the paging frequencies is because they do matter in a simulcast system. The higher in frequency the tone is, the more cricital the timing becomes in the system. I was just curious if you were running plectron tones or not. Here's a pretty cool phase angle calculator I found if you want take a look.

http://www.sengpielaudio.com/calculator ... yphase.htm

[WmHPowell]

Guess it's a good time to refocus on the issue I'm concerned about.
When monitoring multiple sites,
when a tone - any tone is applied, the carrier frequency of one (or more) station(s) "jumps" (pulls, shifts . . .) OFF of the unmodulated carrier frequency and then S L O W L Y (~2 to 3 seconds) "drifts" back to center.
Other than this one issue, all 3 systems are working per specs and sound great!

As I previously mentioned, it sounds like a cap is being charged / discharged (pick one) by the tone, which is pulling the carrier as it slowly charges / discharges. Those of you familiar with the term "DC restoration" will understand what I'm pointing at.

So - I'm not hearing where this is a known problem then. My next step will have to be to isolate the problem to 1 or more sites and then play musical exciter boards to narrow it down.

Sure wish I had a lead to someone higher than the local dealer, who is a great guy but not technical at the level I need.
Thanks, Bill Powell

[xmo]

"...when a tone - any tone is applied, the carrier frequency of one (or more) station(s) "jumps" (pulls, shifts . . .) ..."
___________________________________________________________________________________________________

What are you using to measure the carrier frequency shift and just how far does it shift?

[WmHPowell]

A calibrated ear, of course.
Would be near impossible to measure what with PL and the page tone too.
Suppose I could break out the HP audio spectrum analyzer we were forced to buy with the system and see if I can catch it.
It sounds like as much as 20 HZ but, with all the tones beating against each other, who knows.
Bill Powell

[Bill_G]

Ah - so you're hearing a low dub dub dub dub beat note? Sounds like PL deviation is off. Someone should go back through each station and perform the compensation alignment, then check levels. In simulcast deviation is more important than freq. 5hz difference can add a growl.

[xmo]

Bill is correct, an error in setting mod-comp by as little as a single 'click' can result in degraded simulcast performance.

You should perform the mod-comp & fine level optimization process as outlined in the GPS Simulcast manual using the Agilent DSA. Be sure you have the newest manual version containing the June 2004 SMR-6724 pages [or newer].

Other things you could check:

If you really think that one of the carriers is drifting you could try leaving the transmitters keyed prior to the dispatch transmission. You could do that with the E&M setting of the DSM cards through the channel bank interface - or you could just activate the local PTT on the Quantars from their front panel. That might show if there is any association of the problem with drift at key-up.

A faulty DSM-II card is another possibility.

T-1 timing settings in the channel banks should also be checked. One channel bank at the prime site should be master and source time to the rest. An error here can cause T-1 slips with undesireable results.

[WmHPowell]

Lets try this one more time.

I hear a mass of tones on a wideband monitor (little limiting - little or no FM capture), the product of the paging tone, PL and small frequency offsets between transmitters (in the LOW Hz range).
Since the PL and page tones come from the SAME source, there can't (and isn't) any possibility of them changing from site to site.
This leaves only the heterodyned caused by carrier frequency differences and THAT "should" be stable (and they are - without page tones) since they are all synthesized from GPS stabilized rubidium standards.
However, when a page tone is first applied, one or more carrier frequency is changing ~20 to 50 Hz (by guess, remember the colophony of tones and beets), slowly decreasing back to where it was with no tone applied.
Therefore I surmise that one or more exciter is being pulled by modulation and, with a tone (steady-state modulation) the carrier shift is audibly apparent.
Having received no information that would make me believe that this is a trait of a Quantar transmitter and, failing to reproduce the effect on either the local or spare transmitters, it is not, apparently endemic to the Quantar and, in all likelihood, either a configuration issue or a defective exciter.
Mow, to find it (them?) and how to describe the problem to Motorola.
Thanks all for your help. I’ll try to remember to post an update when I manager to unearth the critter and squoosh it.

[Bill_G]

You're right - it is not a common and well documented problem in the model. I'd isolate the station(s), go through the basic alignment procedures again, and troubleshoot this by the numbers. Definitely keep us posted.