Introduction
This alignment procedure is an alternate from the factory manual ver 1.1, page 13.
It uses levels that are more in line with the system's specifications and definitions.
This might work for you as well. Otherwise, the factory manual and procedure is downloadable from their web site.
The voter takes in eight separate signals from remote receivers. It "listens" to a audio portion of those eight inputs and
converts them to DC levels to drive a microprocessor. Quieting for a channel is indirectly proportional to this DC level.
Therefore a voltage around 0 is either a full quite signal or no signal (standby condition).
A voltage around 3-5 is a noisy or full noise (open squelch) signal/channel.
(Obviously there's no signal with full noise, but we are calling it that for sake of explanation, here).
When converted inside the microprocessor and viewed on the front panel screen these figures are reversed and shown in three digits.
For example an inactive channel will show a "000". A rather noisy signal will be around a "100", while a full quite signal will be
around a "255" on the panel. A graph for one of the channels was ploted, which is on the first page of this article.
Keep in mind the voter is looking at flat audio from the receivers.
Issues and Notes
All SRG's equipment is based on active low PTTs and CORs (cos).
That means they are normally relaxed (Hi-Z) in standby mode and go resistive low during activity.
This is done with collector outputs of NPN transistors.
In standby mode (no activity) the RVS-8 COR inputs have to be a logic high.
Apparently, their engineering overlooked this condition for this mode, because there are no pull-up resistors in the unit; only pull-down.
Since there are no (internal) resistor pull-ups on the RVS-8, an interface strip was built to solve this problem.
A common 12v, fused, buss feeds 10K resistors, one for each channel of the RVS-8.
There were some errors, then later, changes, to some of the component circuit designators.
For example most of the adjustment pot's label were changed.
Original designators were R17 through R24 for the "AUD" adjustments and R41 through R48 for the "Noise" adjustments for schematic Revision 2.0.
These were change to VR1 through VR8 and VR9 through VR16 respectively for schematic Revision 3.2.
The latter makes better sense to the Author as well. Therefore, the following alignment will reflect the last change for these adjustments.
This is an 8-channel voter. There are 8 adjustment pots for the 8 audio inputs called "AUD-x".
There are also 8 adjustment pots for the 8 AF-DC conversion levels called "Noise-x".
You will be adjusting the "AUD"s first, then the "Noise"s. Do this for each of the 8 channels.
Refer to the factory manual for the locations these test points on the PCB.
For this alignment it's assumed you have the RVS-8 on the bench with a test receiver. A single test receiver can be used to set up each channel.
Otherwise, the same procedure can be performed at the site, using the existing equipment (receivers).
Presently, SRGs system has the voter at the MCP, therefore, it's one in the same, for location and test equipment access.
Alignment
Connect up the unit with a link (or local) receiver; audio and COR, and powered up in standby mode.
Input an on frequency, RF signal around a -60 dbm, with a 1 KHz tone, deviated at 3.0 KHz. Maintain this tone level for all the channels.
This will produce a good, quiet "test" signal to align the unit.
Connect an AC voltmeter, calibrated in dbm, in "bridge" mode (high impedance) to the first AUD test point.
Adjust the first channel; AF pot VR1 for a -5 dbm level (bridged) at "AUD 1", (TP11).
Drop the signal generator's RF input to the receiver and open the receiver's (carrier) squelch to produce full noise into this channel.
Adjust this channel; Noise pot VR9 for + 3.5 volts, DC at test point "A/D0", (TP3). This is your noise reference.
It's a DC reference level, indicating full receiver noise.
Repeat these steps for the rest of the ports (2 ~ 8). This will get the unit aligned close enough for the remote downlinks.
This is especially true for SRG's equipment because of the tight tolerance for frequency response and level of the equipment.
In other words, the remote "circuits" will be good-to-go.
If you have installed the unit at the operational site/rack you could re-align all eight channels for each individual receiver
connected to each port. (the tone level and DC reference). This did not seem to improve, what already had, a good operating voting setup.
However, for conventional (non SRG) systems this may optimize the system.
The manual's procedure has you set the pots too close of a "drop-out" of the tone on each test point as observed with a scope, therefore,
the tone level standard was changed to promote stability of the voter. Loosely converting the P-Peak figures they talk about to
something more useful, it appears this practice lowers the levels by about 10 db.
This also provides an easy reference of -5 dbm for everything.
The second (noise) adjustment works fine with this change as well. (still the +3.5 VDC figure)
Tips: The second to the last step is assuming you wish the voter's channels weighted evenly. In the case you have a problem-input receiver
being interfered with power line arcing (which the voter does not seem to respond to), you could "desensitize" a channel into not "wanting" to
vote as much. By turning up the DC bias (Noise pot) the voter will skip over that channel unless it's real quite. In that case is good,
since a real quite channel means the user is close to that remote receiver, therefore, should override that arcing, thus, the problem
becoming transparent. Later, repairs, such as locating the arcing source and fixing it can be done in the summer, then you
can restore that channel's bias to normal.
Another tip, (pun intended) is to modify the test prob tip so it will "sit" in the PCBs eyelite. One of the issues not covered above is
each "AUD" test point (TP) on the board is too close to the related adjustment pot, plus the fact these TP holes are not large enough
for a conventional pin to be insterted, to free up a hand. The Author ground down this tip for this purpose.
If you wish a much closer view of this one image click on it. It's not a perfect grind but gets the job done.
In the case of the "Noise-x" TP and pots, they are properly spaced apart for easy adjustment. Front and back views, here.
As mentioned the "AUD-x" TPs are not so friendly. Set the level for a -5 dbm; the Author used a "flat meter" or otherwise known as the
H/P 3555B meter.
For the rest of operational features, the manual is adequate and correct. Happy voting!
Back to: http://www.ak2o.org/srg/tech.html
![[SRG home Direction]](images/srghome.gif)
