Spectra TAC Voter Notes

Spectra TAC Voter Notes
By Kevin Custer W3KKC

Information provided by several folks, assembled here as an overview / guide on using the Motorola Spectra TAC Voter in a non-Motorola environment and possibly running it without the need of the Status Tone / 24 hour link.

Quick Jump Links:
Kevin Custer's initial inquiry.
Skipps Reply
Johnny Sichert's Reply
Glenn Hochwalt's Reply

The Spectra TAC Total Area Coverage Comparator Reference Manual, 6881039E50 is still available from Motorola, brand new, for $12.25. The manual I ordered and received in February 2007 has revisions through 12/31/03 and is revision A.

Kevin Custer's initial inquiry on the repeater-builder mailing list: I'm looking for someone that has modified the Spectra TAC comparator for use outside the Motorola environment that has done away with the need of continuous links and a status tone, i.e.: modified it for COS input. I need some opinions from folks that may be using the (newer) Motorola Spectra TAC Comparator as a stand-alone voter, using RF links that don't transmit all the time. I have some theories on what to do, but I thought I would see what others have done.

Some dialog followed and this is the best of the gleaned parts...

Q: What didn't you like about running the Spactra-TAC voter without the tone generators on the originating end?

Skipp wrote: I've done it... you really need to make and install the status tone generators. It was just a lot easier to use the tone generator adapters. Plus we were using a combination of original method operation and local-distant located receivers with the adapters. (It was a mixed system). The adapters were pretty easy to build once you had the information.

Someone wrote: Kevin, Now that I understand what you are trying to do, be advised that just because you have the link receivers co-located with the comparator your remote site link transmitters do not have to transmit all the time.

Kevin wrote: What keeps the Comparator from failing, then?

Skipp wrote: What I call the local adapter method is just a local COS/COR operated status tone generator. The adapters were made to use non-Motorola receivers into the voting system, but you can easily locate the adapter at the voter end and just trigger it with the link receiver COS/COR.
Call the status tone a non signal present idle tone that is removed with an active signal. It's used to detect the link status/quality and set the SQM AGC action for best signal-to-noise voting.

Here is a PDF of an aftermarket (CSC) Spectra TAC Status Tone Encoder.

You can use one tone generator for all the comparator cards. Picture a SPDT relay with the armature to the comparator module, the normally open to the link receiver audio, and the normally closed to the tone generator. Or use a solid state equivalent to the relay.

Skipp wrote: Also know that Motorola came out with a modification to the Spectra TAC Current Generator Module, which improves the vote performance quite a bit. I also have that writeup if you burn to have it.

Kevin wrote: Yes, please.. It may be the same mod that Dan talked about (SRN-1068), but it would be interesting to see if they are the same mod.

Skipp wrote: The update was related to running lots of SQM (Signal Quality Module) units and the current generator becoming unstable. The fix was the addition of a small daughter board, with an updated current generator. If you're only running a few modules in one shelf, you probably won't notice the problem. If you're running more than one voter shelf, you should think about the mod if the programmed type of operation has hunting problems.

Here is a PDF of the Motorola Voter Update document talked about.

Then comes John Sichert's reply:

Kevin Custer, at 08:54 AM 1/20/06, wrote:

John,
Jeff DePolo said you might have some information I am looking for...

I'm looking for someone that has modified the Spectra TAC comparator for use outside the Motorola environment that has done away with the need of continuous links and a status tone, i.e.: modified it for COS input. I need some opinions from folks that may be using the (newer) Motorola Spectra TAC Comparator as a stand-alone voter, using RF links that don't transmit all the time. I have some theories on what to do, but I thought I would see what others have done.

Thanks for any assistance you can provide...

----------------------------------------------

John Sichert KA3LAO wrote:

I was going to post this sometime back, but someone made me mad when I tried to help him.
He basically wanted the same information. When I told him what to do, he said, I don't want to modify it, it was too much trouble.
At least, that was the drift I got.
So, I crawled back in my shell.

Remove the AGC shunt diodes.
Where one of the diodes were, add a 330 ohm resistor.
Remove the tone detector toroid. From the center terminal of the coil, add a 10k resistor to the spare pin on the edge connector.
I am doing this from memory, I believe it is pin 9. Whatever, there is only one pin that is unused.
That is the COR input, active high. Ground it, the module goes idle; open the connection, the module shows an unsquelched indication.
To set levels, put 1000 Hz at 5 kHz in the remote receiver. The level across the 600 ohm terminals should be -14 dBm.
The comparator is very sensitive to distortion and audio response. I kept the response between links, +/- 2 dB from 50-4,000 Hz.

This is what I found worked best:
No hang time at the remote site.
The remote site TX must be true FM, that means mod the TA-451 with a MV-1604 diode and associated resistors & caps.
That thing will be flat as a pancake. I like to decode PL at the link RX. Even though I have done it since 1984, I have been told, it cannot be done.
Add the cap across the short tail cap in the remote MICOR RX, to prevent unwanted squelch closing. It is not usually needed on the UHF link RX.
The voter is designed for de-emphasized audio. I have never tried one flat.
Just in case you are using Spectra TAC RXs, replace the 25k pots on the Audio Control Module if they are on a White substrate, they drift in value.

To check the amplifier string in the SQM.
Put a 3000 Hz tone into all the SQMs, (parallel the 600 ohm inputs). Use a low level to start with, monitor the Signal Quality line of each module.
Bring up the level, verify the modules all track with one another to about 11 volts. If a module does not make it to 11 volts, it is defective.
Good luck finding the trouble, lots of "thick film hybrids".

Now the good news,
If you have it set up correctly, the users should never know it voted, or worse, what RX is voted.
The system using this setup has been operational since 1984.

I took me about 5 years of grief to find all this out.
It only took me 15 minutes to tell you.

You have my permission to straighten this out and reprint it. (If you think it was worth anything) Just give me credit.
I don't have time to tell you how poor it will work if you don't follow these guidelines. But I can tell you a few that have not listened.

John

SpectraTac Signal Quality Module Modification For COR Operation

This modification will permit Spectra TAC comparators to be used in the conventional "status tone" type of operation but will also permit remote receivers to be linked in over RF paths using carrier operated relay detection instead of status tone. This mod will eliminate the need for link transmitters to be constantly keyed. Every effort was made to keep this modification simple and easy to reverse if so desired.

Steps one, two and three eliminate the status tone level controlled agc and replaces it with a fixed value. In normal operation using phone line linking the agc is required to keep the levels constant under varying phone line conditions. Using RF links are this is not normally necessary as the losses are constant.

Steps four and five isolate the status tone detector and allow it to become a simple dc switch controlled by the cor input.

Steps six and seven provide diode isolation of the detector and activity checker and route them out to an unused pin (21) on the edge connector to be used as the cor input (active low).

Below are the seven steps required for this modification. Please refer to both the schematic and board layouts to view the modified areas. Any input as to help make this a better modification is always welcome.

Steps:

Remove C9 and replace with a 22K ohm resistor.
Jumper U4a pins 1 and 2.
Remove or cut Q2 collector lead.
Remove CR6.
Remove R66.
Add a 1N914 or similar diode from the base of Q6 to edge connector pin 21.
Add a 1N914 or similar diode from U3 pin 8 to edge connector pin 21.

*Note---Check your board for a female connector on pin 21. This will be the new COR input. Some boards do not have the female pin on unused terminals, if this is the case remove a pin from an old module and install at position 21.

Hi Kevin:

You may have gotten other answers and ideas long before this. I was following some links today and came across your page regarding RF linking receivers back to a Spectra TAC comparator where the links do not transmit 24/7.

Here is how I set up some sites for my fire department:

1. The remote receiver/link combinations are GM300 or M10 radios (VHF system/UHF link) connected with RICKs. I like using the RICKs rather than just tying the radios together for the flexibility and level control ability. They operate as simple unidirectional cross-band repeaters. (When we narrowband, we'll be switching to M1225s or something similar.)

2. The comparator end is Spectra TAC receivers which generate status tone for the comparator. They are normal in all respects. (When we narrowband, we'll replace the micor decks with M1225s; I've done this with several Spectra TAC receivers already in another system.)

3. The links do not transmit unless the remote receiver unsquelches, and the links do not transmit any status tone.

4. We use an ICR squelch gate in the MICOR Repeater so that it will default to In Cabinet Repeat if the comparator should fail. It never has. (When we narrowband, we'll replace the MICOR with an MTR2000. It will be a drop-in replacement. There is a little trick to getting pseudo-ICR with an MTR2000.)

Admittedly, it takes about 300-400ms for one of these links to come up and get through multiple PL encodes/decodes, first at the remote receiver, and second at the link receiver, then key the repeater with TRC. It is a matter of discipline for system users to not get in a hurry and let the system work. If the user presses PTT and then takes a breath, that is enough time for everything to come up.

It works very well, with minimal false votes, although if two receivers are full quieting it can hunt a little. During a 2-tone page-out sequence all receivers are fully quieted, and you can hear a false vote ocassionally on the first tone. In our case, the first tone is pretty low frequency, and voters don't like low frequency tones anyway.

I have some friends that have a multi-site RF linked system for a Sheriffs department that is running all narrowband, links and all. It is pretty critical to set up in narrowband, but it works well for them.

We're doing it this way because we want off-the-shelf stuff with minimal special modifications.

The exception is the Spectra TAC receiver deck replacement mod which is actually VERY straight forward, requiring nothing more than a buffer transistor on the COR output of the new receiver. (matter of fact, I think it's your little buffer circuit) It plugs directly onto the backplane of the Spectra TAC receiver with no mods to to the receiver chassis, using the old connector and wiring from the micor deck plug-in strip.

The good part about doing it this way is that a remote receiver failure will not lock up the system until a comparator channel times out, but the bad part is that you do not know that a receiver has failed. Since our comparator is co-located with the repeater, we have to go out there and check it every once in a while anyway, and in some areas of our district, we will know RIGHT AWAY that a receiver is not working.

We have not had any significant failures in 6 years with this setup.

That's how we do it.

Best Regards,

Greg Carttar, CHS-V
3rd St. R & D Production Services
Special Event and Disaster Communications

Safety Officer 307, Communications Officer,
Fire Service Instructor I, HazMat Technician
Central Taney County Fire Protection District

Another reader offered this information:

A common situation is when a receiver is co-located with the comparator that the local receiver is almost never voted. This is because the local pair of wires passes high frequency noise content that leased phone lines do not. The Signal Quality Module (SQM) "sees" that extra noise and decides that other receivers are better when they really aren't.

This problem can affect any Spectra TAC system where different types of connectivity are utilized for the various receivers.

Motorola provided a solution to the problem in the form of a voice frequency filter module, sometimes called a roofing filter. These could be factory ordered as option C366AD-SP or the modules could be ordered as QRN8498A. Each module takes up a single slot in a comparator chassis. The incoming wireline is connected to the filter module and the output of the module then connects to the SQM input. Thanks to Greg Carttar, we now have documentation and a photo of the board.

Text, concept, images, etc., Copyright John Sichert, Jeff DePolo, Skipp May, Dan Babilla, Glenn Hochwalt, Greg Carttar, and Kevin Custer, January 2007 and date of last update.
All Rights Reserved, including that of paper and web publication elsewhere.