Using the Trimble SV-6 GPS Receiver

These GPS units were available from Tanner Electronics in Dallas at the end of 2002. They were used in an OEM application and came mounted in a metal trough. They were also supplied with a (non waterproof) antenna and a photocopy of how to get them going. Experience has shown that the supplied documentation sv6cm3.pdf and the following notes together may be enough to get one going.

Firmware and Protocol ID

The document shows how the firmware revision and protocol version of your GPS can be identified. Figure 3 is a view of the underside of the board; to see it the unit has to be taken out of its trough. Yes they are not all the same. My initial purchase of four had only two with identical firmware and protocols. . This information will be vital later

Interface

The first item of business is to build a converter to change the TTL level signals supplied by the unit to the RS232 levels the computer expects. I decided to use the MAXIM MAX232 rather than the MAX233 suggested in the document as they were cheaper and more readily available. The only advantage of the MAX233 is that it needs no external capacitors; the disadvantage is that it is in a bigger package.

Power

The unit is powered by 5V so a 7805 regulator was used. Due to the units power needs the regulator needs mounting in the box to dissipate heat.

Housing

Originally the unit was housed in a lidless metal tray with signals being fed in and out via feedthroughs. This would not be good enough for general use, so a sealed metal box was sought. This resulted in using a 6x3x2" diecast box which is just big enough. These are available in a number of varieties with very similar outside dimensions; make sure the one chosen does not have internal PCB mounting ribs or mounting the antenna jack through the box sidewall will be a problem

Testing

Initially the unit was connected via the regulator to a 12V  supply and just the MAX232 IC interface circuit.  It was then connected to the serial port on the laptop. A suitable AMPLIFIED (not passive) GPS antenna was also connected. Upon initial switch on nothing will appear from the rs232 port of any significance as the unit is Motorola Binary mode. From the protocol identifier recorded earlier choose the appropriate file to upload to the unit to turn it to NMEA operation.

The files for protocol versions  30323 31116, 34055, 34057 37424 37425 are all in file sv6_nmea.zip

These files cover all the protocol versions I have seen.

The .txt files supplied with the files help decide which one to use. As an example 2 of my units said 30323-61 on the protocol label. To turn that into NMEA output I executed the file 30323-62.exe

As an initial check of the success of the operation if a terminal program is now run on the com port with 4800-8-N-1 settings, NMEA strings should be seen as follows. Due to the OEM nature of the unit the GPVTG and GPGGA SENTENCES ARE THE ONLY ONES THAT CAN BE OUTPUT.

$GPGGA,,,,,,0,0,,,,,,,*56
$GPGGA,,,,,,0,0,,,,,,,*56
$GPGGA,,,,,,0,0,,,,,,,*56
$GPGGA,,,,,,0,1,,,,,,,*57
$GPGGA,,,,,,0,1,,,,,,,*57
$GPGGA,,,,,,0,1,,,,,,,*57
$GPGGA,,,,,,0,1,,,,,,,*57
$GPGGA,,,,,,0,2,,,,,,,*54
$GPGGA,,,,,,0,2,,,,,,,*54
$GPGGA,,,,,,0,2,,,,,,,*54
$GPGGA,020240.5,3309.944,N,09639.154,W,1,3,005.1,,M,-024,M,,*56
$GPVTG,000,T,354,M,00.0,N,00.0,K*4C
$GPGGA,020242.0,3309.949,N,09639.138,W,1,3,005.1,,M,-024,M,,*56
$GPVTG,039,T,033,M,00.1,N,00.2,K*47
$GPGGA,020243.0,3309.952,N,09639.134,W,1,3,005.1,,M,-024,M,,*51
$GPVTG,002,T,356,M,00.0,N,00.0,K*4C
$GPGGA,020244.0,3309.954,N,09639.131,W,1,3,005.1,,M,-024,M,,*55
$GPVTG,000,T,354,M,00.0,N,00.0,K*4C
 

The times are UTC, not GPS
 

1 PPS output

Next attention were turned to the 1PPS outputs. According to the spec sheet the 1PPS output is open collector, so pin 6 of the connector on the board was connected to +5V via a 10k resistor. Initially no pulse could be seen on an oscilloscope. However after resorting to the old "trace intensity to maximum and turn all the shack lights out" trick, a very narrow pulse 5uS wide was detected. Not surprisingly even after feeding the pulse through the MAX232 chip the pulse was completely missed by the serial port of my computer and my NMEATIME program. Obviously the pulse needed stretching. Remembering that the TAPR TAC-2 succesfully used a CMOS 555 to stretch the pulse to 100mS the same circuit was employed. The circuit was breadboarded and tested and solved the problem.

Final design

As the unit had now been succesfully interfaced  the final complete circuit was drawn up as shown in Fig 1.

LED1 is also driven by the 100mS wide 1PPS signal to indicate activity  R6 is mounted off the PCB. V1 is a 3.5V NiCd pack which is used to battery backup the receiver, which retains the Satellite Ephimerisis data making the unit find the satellites quicker on power up.

A small PCB was then developed to contain most of the components, the voltage regulator excluded. The track layout is shown in Fig 2, the component overlay in Fig 3.

Conclusions

The SV6 GPS unit has adequate performance for basic location determination and for time locking computers with programs such as NMEATIME. Due to the inability to put the unit into fixed position mode it is not very useful for generating the 1PPS signal required to drive GPS locked frequency Standards. It can also be made into a digital clock, see sv6_clock.htm