Modifications to Parallel Port Interface
The Discovery Series K-2805 Parallel Port Interface as distributed by Dick Smith electronics has proved to be a reliable and versatile introduction into PC interface and analogue conversion. As suggested in the assembly instructions, I packaged mine in a "Zippy" box and provided two forms of connection between the unit and my experiments. One was a row of bannana plug sockets for each of the analogue inputs and outputs and one for ground. The other was two 25pin female "D" connector sockets which provided connections to all of the analogue and digital outputs and inputs. It was through this connector that I used this interface to control the PC Controlled Voltage or Current Source for battery charging.
The following diagram shows the pin connections for the "D" type connector I used:
The program (charger) and circuits for PC Controlled Voltage or Current Source will (should) work with the Interface unit as it was sold. In the case of the program, the default options are for the standard Interface unit, but, as I had modified my Interface unit before finishing the program, I have never tried the program on a standard Inferface unit.
The following are modifications designed to make the Inferface unit more versatile. I have not shown the circuit of the Inferface unit as I do no wish to break any copywrite laws and the quality and value of the kit is such that making your own interface unit from scratch would seem a little foolish anyway.
Firstly, I wanted to run the unit off either the plug pack that charges my laptop PC and, for more reliability in power failure, the laptop PC's own battery (which in my case is nominally 5V). The diagram below shows this and other changes:
As can be seen, the Vb which supplies power to the Analogue output opamps is taken now off the 5V rail instead of from the 7.5 to 15 volt supply (labelled here as 9V). This limits the analogue output levels marginally but also allows the unit to operate from the laptops 4.8V battery (make sure use fuse this inside the laptop itself).
Since the supply to analogue inputs and outputs is only nominally 5V, some means of calibration is necessary. I chose to use the A10 analogue input to IC4 to measure a 2.5V reference voltage. If this input registers 128 out of the 0 to 255 input range, then the supply voltage is 5.0V. This input was formerly used to measure the level of the 9V internal battery. The reference diode LM336 has an a trim connection (10K trim pot between Vout and ground) which if left unconnected, as I have, can (in theory) result in a reference voltage between 2.44V and 2.54V. My reference diode has given 4.9V unconnected so for the 0.02V resolution of analogue inputs, this has caused no problems.
The 1K resistor connected +5V to pin 1 on SK5 is to connect to pin 19 on SK7 and SK8 described above.
The remaining circuit is largely external to the Inferface unit but is housed in the same box so that it can provide a plug pack supply voltage. Leaving the laptop charging for long periods wasn't going to cycle the batteries in the laptop (they loose capacity if they're not worked regularly), so I have included the facility to periodically and automatically switch the laptop (but not necessarily the Interface unit) to internal power. By temporarily disconnecting the plug pack power to the Interface unit (using the Power On function intended to connect or disconnect the 9V battery), the laptop battery voltage can be measured by referencing it to the 2.5V reference voltage. This can be used to determine when the laptop should be returned to mains power; this depends on how confident you are of the mains supply and the importance on the function the laptop is monitoring.
I have used the <init> pin of the printer port to control the relay switching the laptop plug pack power to the laptop. This proves to be a somewhat contrary pin who's default level seems depend on the particular BIOS found in a given PC. Hence the provision of SW1 to select a normally open, normally closed, or the desperation "lets leave it" on position.
Speaking of connections to the printer port, I noted there were three digital inputs unused on the printer port. I've used these as shown below:
The final modification was used to speed up conversions where only a particular output IC was being used. This enables data to be loaded into all three output ICs at the same time rather than in serial. An option in my program allows the system to operate without this modification if necessary, but some speed is lost.
This modification assumes that J1, J2 & J3 are linked as shown on the circuit provided with the kit for individual loading of the ICs.
I hope the above modifications will increase your enjoyment of this great kit.