How to Hack a SuperstarTM
by
Cliff Kelley
Step 1. Buy the parts on the parts list.
Parts list: Approximately $200
BAE SuperstarTM receiver (not the least expensive one anymore) PN220-0610011
I/O Board and parts from JDR microdevices JDR-PR10 and JDR-PR10-PK
7400 or 7404 ICs for inversion
14 pin socket
wire-wrap wire AWG 30
Step 2. Following the instructions provided assemble the JDR board. Recall we are using the I/O part so install the headers for I/O D0-D15 and for all of the sel signals A-O. Also, be sure to connect SEL K to ground. This indicates that the I/O transfers are all 16 bit. A circuit diagram drawn by Jenna from the University of California Riverside is shown below. I've tried to make it readable but have had no luck, if you would like the full sized version please email me at cwkelley@earthlink.net. Thanks Jenna!
Step 3. Remove chips from SuperstarTM. The photos below will help you identify the parts and orient you with the receiver layout. I used an x-acto knife and "sawed" through the leads on one side of the RAM & PROM chips and 3 sides of the ARM chip. I sawed from the top being careful to avoid damaging the circuit board. With one set of leads left on one side of each chip I bent the chip and remaining leads back and forth until they broke. Then I used a soldering iron to "clean" off the parts of the leads still attached to the boards. Check to be sure that there are no solder bridges between the connectors. The ARM and PROM pads are not critical .
Step 4. Remove the SuperCap from the board. I did this to gain access to some traces under the capacitor. I used a pliers and rotated/wiggled it back and forth until both leads broke. Then I cleaned up the connection holes with a soldering iron. The photos below show how the board should now look.
Step 5. Add a 7400 or 7404 to the board along with preparing the mounting holes, the rf bulkhead connector and cable. I drilled 4 holes in the JDR board to match the receiver board mounting holes in each corner. To avoid shorts I used nylon screws, nuts and spacers to mount the receiver to the interface board. The board rf connector is an MCX type. I like to work with SMA connectors so I built a short rf cable to go from the bulkhead connector to the receiver board with the connector in the back. The general arrangement is shown below. Connect power and ground to the 7400 or 7404 chip.
Step 6. Using the diagram below cut the traces and add the ALE and NCS jumpers. I found a large unused pad near the ALE line which I used so that soldering to it would be easier. I used AWG 30 wire wrap wire for the jumpers on the board. This wire is small enough to fit through the tiny through-hole pads. Since the pads for the RAM chips are the largest on the board and they are already connected to the data and address lines of the GP2021 they are very convenient.
Step 7. Cross connect the lowest order data bits to the address lines on the receiver board. I used an adhesive paper to cover the traces under the RAM chips and soldered bare wire for the leads from A8 through A5 over to D6 through D3. I used wire wrap with insulation on for A4-D2 and A3-D1. Since A2 and D0 are next to each other I bridged them with a piece of bare wire.
Step 8. Using wire-wrap or other small wire connect D0-D15 from the JDR I/O board to the receiver RAM pads.
Step 9. Using wire-wrap or other small wire connect sel H to the READ pad, connect sel M to the WREN pad, and sel D to the NCS pad. Connect sel E to the inverter chip and connect the inverted signal to the ALE pad.
Step 10. Using the diagram below make up a 2x10 2mm pitch header. Then connect it to the receiver board and solder the power supply and ground lines.
Step 11. Connect the rf cable to the board. These final photographs show the end result.
Step 12. Install into computer and run the gpstest software. In both of my attempts to build the interface the only problems have been cold solder joints. The test program first tests the data lines. If a problem occurs it displays what was sent to the board and what was returned. By looking at the bit patterns it should be obvious which bit is either stuck high, low, or cross connected. In addition one should be aware that there is a potential for IO port conflicts. Thus far only ethernet network cards have been found to have this problem.
Updated February 5, 2003