Choosing a Fast NAND gate
Looking through the data sheets for NAND gates found the following timings:-
| Part No | Chip | Manufacture | Speed | Package | Price | ||||
|---|---|---|---|---|---|---|---|---|---|
| Min | Typ | Max | |||||||
| 1085297 | 74HCT00D | NXP | 10nS | SOIC | 0.221 | ||||
| 381755 | 74HCT00N | NXP | 10nS | DIP | 0.667 | ||||
| 9590889 | 74HC00 | Ti | 8nS | DIP | 0.31 | ||||
| 1105915 | SN74AHCIG00 | Ti | 5,2 | 8 mA | SOT-23-5 | 1.78 | |||
| 1287697 | SN74LVCIG007 SN74LVC1G00DBVR | Ti | 4.0 | 32mA | SOT-23-5 | 1.98 | |||
| SN74LVC1G00-EP | Ti | 1.0 | 4.0 | ||||||
| SN74LVC1G00DBVR | Ti | 1.0 | 4.0 | SOT-23-5 | |||||
The Single 2-input Positive NAND gates display the fastest times.
Here is the chip that has been chosen:-
Full Data Sheet for SN74LVC1G00 Single 2-Input Positive-NAND Gate & Timing
|
Circuit Description
There are 2 inputs,
- A19 - Selects either the top half of the 1M memory space for the flash chip or the bottom half for the RAM.
- /CE - Selects the 1M card.
Truth Table of half a 74139 using two outputs
| Inputs Enable Select | Outputs | |||
|---|---|---|---|---|
| /E /CE | A1 | A0 A19 | /Y1 /CE1 FLASH | /Y0 /CE0 RAM |
| 0 | 0 | 0 | 1 | 0 |
| 0 | 0 | 1 | 0 | 1 |
| 1 | X | X | 1 | 1 |
Logic Diagram
Original
Cut Down Version using two inputs and two outputs
This ignores A1 GND line.
The Z88 only uses one half of this dual decode chip. This circuit may be made with
- 3 inverters
- 2 NAND gates
The inverters are slower than the NAND gates, so it was decided to use 5 NAND gates using 3 of the gates as inverters.
Logic Lab Test v Truth Table
The logic was checked with thanks to http://www.neuroproductions.be/logic-lab/ for the simulator.
| Inputs Enable Select | Outputs | |||
|---|---|---|---|---|
| /E /CE | A0 A19 | /Y1 /CE1 FLASH | /Y0 /CE0 RAM | |
| 00 | 0 | 0 | 1 | 0 |
| 01 | 0 | 1 | 0 | 1 |
| 02 | 1 | 0 | 1 | 1 |
| 03 | 1 | 1 | 1 | 1 |
00
| 01
|
02
| 03
|
They both agree.
Circuit Diagram
Circuit updated 15:04 6/9/16
Building a prototype board internally
A prototype board was built inside the Flash card..
07/09/16
It was very difficult to solder onto the legs of the inverted small chips. As a result, this prototype failed to work.
Building the circuit on a breadboard externally
Components required
|
|
10 sot to IC adaptors were obtained in addition to the 5 NAND gates.
The circuit could now be bread-boarded and tested outside the Z88 card case.
The four signal and power lines can be seen connecting the card to the breadboard.
Plan view showing all the connections.
Oscilloscope Readings
This produced the signals required.
This trace shows the Flash chip being selected (/CE1) in 3nS.
The RAM (/CE) would be the same timings.
Signals not the same abbreviations as the circuit
The names of the signals on the scope are not all the same as shown in the circuit.
| Signal | Scope | Circuit |
|---|---|---|
| A19 | A19 | A19 |
| /CE1 | _CE1 | /CE1 |
| /CE0 | _CE | /CE0 |
Printed Circuit Board
A PCB has been laid out (just in case Tony's prototype works), but this was never produced due to ongoing development.
Main points
- A letterbox slot has been made in the centre of the HD1, HD2, (which are connected both sides,) so that either direct soldering or wire links may be used to connect the signals from the 512K/512K card to the PCB.
- The pads of the footprint of the SOT-23-5 have been made longer, to enable easier soldering of the small parts.
- The bottom Left Hand corner of the PCB matches the shape of the 512K/512K Card for easy alignment.
- C2 on the 512K/512K, needs to be removed and put onto the PCB, in the same space. This is to make the PCB a bit larger for stability.
The 7 layers
Here are the layers printed on A4 sheets. The board is small, zooming in is generally helpful.
(updated 19:16 07/09/2016)
Results
The latest software was tried again in a standard 512K/512K Flash/RAM Card and Vic could not make the games software go wrong. The prototype board was packed and sent to Mr T who had a failing Z88 and 512K/512K Flash/RAM Card. He confirmed that it was the decoder chip that was causing the fault and suggested that another decoder could be found.