Electronics Engineering

If you want to know a printed circuit board is designed here is the explanation in simple terms.

1. First you decide on the components which are in the design.

2. Need a software in which you can get symbols (both schematic symbols and physical symbols) of standard components and which also allows you to create your own symbols for new components which are not in the standard tool library.

3. In schematic capture tool, you will use symbols of components and also connect the pins of components with wire lines to complete the schematic drawing. Based on complexity and size of the design this can be few pages or hundreds of pages.

4. In the layout tool you will have to make a board outline (based on the shape/size of PCB you want to design) and define number of layers etc.

5. Schematics need to be exported to layout using the board outline you designed. Result will be that all the physical symbols will appear in the layout tool along with un-routed wire connections.

6. Now you can place components on the PCB based on floor plan (where you want to place what components) and start routing the connections. Decide thickness of routes or traces based on impedance and current these nets need to carry. Based on how many layers have been defined and needed as per design traces can run through multiple layers

7. When routing is complete, with help of tool you need to create PCB fabrication data for PCB fabricator.

These are top level steps involved in a PCB design. There are finer steps like back annotation (process to sync up layout and schematics) which also important. Many tools give options for auto placement, auto routing which can aid in the design. These used with proper constraints and care can help speed up the design process. Check for PCB tool vendor websites.

- Neeraj

Go to www.circuitboards.com scroll down to the links area and click free cad software by cadsoft. My friend told me the other PCB designing software mentioned makes you pay for there software if you want it manufactured outside of there facility. Basically locking you into there price or no price. Cadsoft is free up to a certain layer count I believe.

To build a 64 to 1 multiplexer using cascaded 8 to 1 multiplexer, use nine 8 to 1's. Connect the first 8 to each of the 64 inputs, then connect the ninth to the outputs of the first eight. Connect the three address lines of the eight together to form 3 of the address lines. Connect the three address lines of the ninth to form the other three, for a total of 6 address lines selecting 1 of 64 inputs.

This is a lot of logic. Fan-in and fan-out may be considerations. If you are trying to scan 64 switches, there may be a better way using an 8-bit output connected to a switch matrix (with diodes if you need more than one at a time close-able) and then connected to an 8-bit input. Even better, consider the 8279 keyboard/display controller.

Absolutely not, if you are a residential consumer. Capacitors, when used for power-factor correction, reduce the supply current but NOT the amount of energy consumed and you pay for energy, NOT current! Energy meters measure the supply voltage and the in-phase (resistive) component of the supply current, so power-factor correction is irrelevant to residential metering. The video referred to in the original answer shows current being reduced but, as explained, this has absolutely no relevance to your electricity bill. In fact, the video is complete and utter nonsense! Capacitors used in this way are a complete scam.

Power-factor correct ONLY applies to large commercial and industrial consumers because they are penalised for unnecessarily-low power factors because the resulting unnecessarily-high currents require the utility companies to install larger transformers, switchgear, and cables.

Another Answer

Yes, if you are a big customer and you are paying a premium on your power charges through having a poor power factor on the site. Otherwise a capacitor does not reduce the kWh consumption of your appliances, but it might reduce the current you draw, but your meter measures energy used and not the current. For a domestic customer power factor correction will almost certainly notreduce your charges.

The data rate (C) is equal to the bandwidth (B) times the logarithm base 2 of 1 plus the signal-to-noise ratio (S/N) (how much interference is introduced in the transmission of data)

C = B x log2(1 + S/N)

So your data rate is directly proportional to your bandwidth. If you increase your bandwidth, your data rate will also increase provided the signal-to-noise ratio isn't affected.

how to draw timing diagram?discuss the various steps

You first need to understand the machine cycles of 8085

The status signals are as follows

IO/M(bar) :--- 1 IO 0 Memory

S1 | S0 | Process

-----------------------------------------------------------

0 | 0 | Halt

0 | 1 | Write

1 | 0 | Read

1 | 1 | Opcode fetch

1)Opcode fetch ( Compulsory Machine cycle)

This cycle requires 4 T-states.

1st T state ALE is high and lower byte of address from PC(Program Counter) is placed on the multiplexed data/address bus.

In the second T-state, after checking the status of READY pin, RD(bar) goes low the opcode is placed on the data bus, This state continues in the 3rd T-State.

The fourth T-state is used by the uP to decode the instruction and to generate the relevant control signals. The state of the address bus is unspecified( This T-state is used by some DMA controllers to transfer data in hidden/transperant mode)

IO/M_ = 0 S1=1 S0=1

2)Memory read(for 1 byte)

Three T states, similar to the first 3 T states of opcode fetch( as first 3 states of opcode fetch is effectively memory read)

IO/M_ 0 S1 = 1 S0 = 0

3) Memory Write(for 1 byte)

Similar to Write but instead of RD bar WR bar is used. Also the data stays on the bus a little longer than READ*.

IO/M_ 0 S1 = 0 S0 = 1

4) & 5) IO write and read

Simlar to the above two, only IO/M_ = 1

These are the basic machine cycles you will require to draw timing diagrams for most instructions. There are additional cycles such as INTA bar and Bus idle. If anyone requires diagrams for these cycles, message me and i will explain them later.

Also some instructions like CALL require 6 T-state Opcode fetch. For this you can draw the 4 T state Opcode fetch but 4th T state extended to the fifth and sixth T state.

------------------------------------------------------------------------------------------

Now, to draw the timing diagram for any instruction you need to understand what exactly the instruction does. I will explain a few. If you need a specific instruction, msg me.

A) MOV A,B

Draw only opcode fetch as no further memory acces is required as operands specified in registers only

B) MVI A,32H

Draw opcode fetch and memory read as operand(1 byte) has to be fetched from memory

C) LXI H, 2000H

Draw Opcode Fetch and two memory Reads as two bytes, 00H and 20H, (lower byte fetched first) have to be read from memory.

D) STA 2000H

This instruction stores the value of accumulator(8 bit) at the location specified.

Opcode fetch + Memory read * 2 (byte address) + Memory write * 1(1 byte)

i.e 13 T-states 4+3+3+3

During the memory write the address bus contains the address fetched by the memory read cycle earlier

E) CALL addresss(can be specifed in terms of a label)

During a call instruction the uP pushes the current value of program counter(16 bit ie 2 byte) to the stack and then copies the new value from the memory(specified in the instruction)

6 T state Opcode fetch

+ Memory write * 2 (PC pushed to stack)

+ Memory read * 2 (New value of PC fetched from memory)

ie 6 + 3 + 3 + 3 + 3 = 18 T-states

Note that during the memory write cycle the address bus contains the address of the top of the stack(Stack Pointer)

F)JMP 16-bit address

3 Cycles as Follows

4 T-State Opcode Fetch

+ 2 * Memory Read ( 16 bit = 2 bytes)

ie 4 + 3 + 3 = 10 T-states.

Note that separate cycle is not required for loading the address into the PC as PC is a register.