tracing through digital circuits IC chips

Thread Starter

relicmarks

Joined Oct 13, 2006
355
When tracing through these components:
Encoders
Decoders
Flip Flops
Counters
Shift Registers
Multiplexer
Demultiplexer

How do u know if they are functioning properly? or how can i trace through each one of these components ? how would u guys do this to verify them?

Its hard to use a DVM meter or oscilloscope sinse they have multiple inputs and outputs , plus i would need alot of signal generators for the logic inputs , resets, enable inputs

All the new PCB's are surface mount IC chips so you can't use Logic clips or logic monitor because they are only made for dual in-line package IC chips not for surface mount or through hole IC chips

How can i "STEP" a counter and a shift register through a complete cycle and verify inputs, outputs, resets, clears, and other signals?

How can i "STEP" through a Encoder and decoder inputs and outputs?

The Encoders outputs are not binary but numerical or address lines? or LCD displays

So you can't use a DVM meter on the outputs of a Encoder or inputs of a decoder because its not logical binary states but numerical address codes? instead of using a DVM meter can i use a LCD display and probe the outputs of a encoder?
 

beenthere

Joined Apr 20, 2004
15,819
Beside the logic analyzer, you need the data sheet for each IC so you will be able to tell how to tell if each one is functioning properly.

You should be able to find clips that will let you attach your o'scope probes to the devices. I find that a dual trace scope with an external trigger will let you visualize most timing relationships. Most logic circuits are synchronous, so things happen at definite intervals.
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
I find that a dual trace scope with an external trigger will let you visualize most timing relationships.

1.) What do i hook up to the external trigger on the o-scope?

2.) How do u find timing relationship with the o-scope?

Most logic circuits are synchronous, so things happen at definite intervals.

1.) What do u mean by this? how would u know the intervals?

2.) How would u know if its "triggering" on the leading edge or falling edge?

You need a logic analyzer of course.

1.) Ya , most companys i work for don't have logic analyers or timing charts so i have to figure it out old school manually, my managers mostly just tell me this is basic digital electronics, but i just tell them back that i'm not the designer of the circuit so how do i know what the inputs,outputs etc are suppost to look like and what the function and results are suppost to be
and how is this circuit going to work or what its suppost to do.


Timing problem List/rules::
1.) propagation delay can be as high as 65ns or higher
2.) Missed counts
3.) out of sequence counts
4.) clock pulse frequency increases the timing propagation delay gets worse
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
data sheet for each IC so you will be able to tell how to tell if each one is functioning properly.


1.) I tried that too , but there is so many different ways a flip flop can be wired or a counter

SR Flip-Flop
Switch Debouncing
Clocked RS Flip-Flop
Delay Flip-Flop
JK Flip-Flop
Master-Slave Flip-Flop
Ripple Counters
The Decade Counter
The S-R latch
The gated S-R latch
The D latch
Edge-triggered latches: Flip-Flops
The J-K flip-flop
Asynchronous flip-flop inputs
R-S flip-flops, D flip-flops, and J-K flip-flops
synchronous and asynchronous counters.
ripple counters, up/down counters, self-stopping counters and frequency dividers.
serial load shift registers and parallel load shift registers
universal shift registers.

The data sheets just tell if the IC chip is working but not if the network or circuit is working if they are wired in these ways to perform a certain function
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
Encoder and Multiplexer IC chip testing with a O-scope::

I used channel#1 on the input of a encoder IC chip , i see low or high level states

I used channel#2 on the outputs of a encoder IC chip , the signals are encoded its hard to understand the displayed signal on the O-scope

Would using a Seven Segment LCD display be but to use on the outputs of a encoder IC chip be better than a O-scope?

The same with Multiplexer IC chips , that output displayed on a O-scope looks weird and encoded, what would be better to use than a O-scope or logic analyzer?

The Inputs of a Multiplxer IC chips on a O-scope look fine because they are 1 or 0 zeros Highs and Lows but the output is encoded
 

Papabravo

Joined Feb 24, 2006
21,225
Truly, your situation is very nearly hopeless. If your company won't spend money on tools then they deserve the results they get. Ask more specific questions and we'll do what we can to help.
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
using a logic analzer, which "mode" do i use to monitor the outputs of encoders and multiplexers?

Since the outputs of encoders and multiplexers are address code or encoded code you can't use a DVM meter or oscilloscope right ? how would u monitor the signals?

I printed out the data sheets of each IC chips , than i test each input and output using a DVM meter, logic probe, oscilloscope but how can i "CYCLE" the inputs,resets,enables,outputs all at once?

A logic analyers doesn't have multiple signal generators its just for monitoring , so what can i use to provide multiple logic state inputs

1.) multiple inputs 1,2,3,4,5,etc
2.) Reset input
3.) Enable input
4.) Clock input

What can i use please?
 

Wendy

Joined Mar 24, 2008
23,421
Generally you don't need to know exactly what a chippie is doing, most failure modes are pretty obvious. This assumes, of course, you have a working design to begin with. If one input to an AND gate is high, and the other pulsing, then you should see pulsing on the output. No pun intended, it is logical.

A high speed oscilloscope or a decent logic probe (which is very inexpensive, BTW, especially if homebrewed) is all that is needed for 95% of most failures. Most failures are straight forward.
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
how do u use a logic probe or oscilloscope on a encoder or multiplexer?

What are most of the failures that u have learn or seen?
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
An encoder is a circuit that changes a set of signals into a "code".
A decoder is a circuit that changes a code into a set of signals.
An encoder is a device used to change a signal (such as a bitstream) or data into a code

demultiplexer is normally required to break single data stream back down into the original streams.

The encoder has the limitation that "only one input" can be active at any given time. If two inputs are simultaneously active, the output produces an undefined combination

A priority encoder is such that if two or more inputs is given at the same time, the input having the highest priority will take precedence
 

Papabravo

Joined Feb 24, 2006
21,225
There are many techniques used by experienced engineers and technicians to evaluate and troubleshoot circuits. They mostly involve some kind of repetitive action. Most of the boards that I have designed over my 45 year career had a processor in them. This allows the construction of test and diagnostic code that was part of the product. In the days of analog oscilliscopes, that was about the only way to get a trace you could see and interpret. Sometimes we spent more time designing test fixtures than designing products. It all depends on the time and resources you have available.

From what you're are asking I'm assuming that you are trying to do things on a limited budget with the electronic equivalent of knives and stone axes. You need to realize that there are few if any cookbook methods that can be applied to solving your problem. Each design is different, and not everybody is suited for this kind of work.
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
ya i know

since i don't have a logic analyser the only other way is to use switches and lamps

switches to +5 volts for the inputs and lamps for the outputs
 

beenthere

Joined Apr 20, 2004
15,819
Do you realize that troubleshooting a failure is not a matter of dilligently testing and verifying each and every input and output on the board? You do need to be able to evaluate what is not working and make up a test methodology based on that.

If one of several outputs is not right, the the testing may be confined to the elements that produce that output. Testing all the other stuff on the board is not likely to do anything but slow you down.

If nothing works, then start looking for a general failure, like the power supply.

Nothing helps as much as understanding the principle behind the circuit's functioning, so symptoms can be related to possible failure mechanisms. That understanding will also allow you to use the oscilloscope more effectively.
 

Papabravo

Joined Feb 24, 2006
21,225
ya i know

since i don't have a logic analyser the only other way is to use switches and lamps

switches to +5 volts for the inputs and lamps for the outputs
I recommend inputs be connected to a pullup resistor and a switch to ground. That way you can set the input to either state. It is a rare circuit that does not have its inputs connected to something. You can't just go adding switchs all over the place since they will be fighting with other outputs that want to control those inputs. In order to help you evaluate different troubleshooting stategies it would help to have a concrete example.

BTW, "lamps" are more than just a little bit obsolete. An LED(Light Emitting Diode) is far more useful and is infact a key component of the aforementioned logic probe.

There are some basic things you can do.
1. Verify that each chip has power and ground.
2. Verify that there are no floating inputs.
3. Verify that periodic signals like clocks are beating up and down.
4. Put your finger on each chip and make sure you can keep it there.

After that you really have to understand what is going on.

If the circuit has never worked because it is a brand new design then that is different from a circuit that has worked, but for some reason has failed.
 
Last edited:

Thread Starter

relicmarks

Joined Oct 13, 2006
355
how do u use your oscilloscopes Trigger mode, when testing digital circuits?

What are some "test methodology" for digital electronics?

1.) Check the supply voltage on IC chip pin to make sure its getting power
2.) Check the clock pulse
3.) Check the ground to the IC chip
4.) Check the inputs and outputs

The hard part for me is that my Signal Generator only has ONE TTL output , and most of the logic IC chips have like multiple inputs , so i need a multiple TTL output Signal generator. So its hard for me to test the inputs and outputs at once in a Cycle

So i guess its best to use Switches to +5 volts to test the inputs since i don't have multiple signal generators
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
I recommend inputs be connected to a pullup resistor and a switch to ground.

1.) What resistance value should the pullup resistor be to not LOAD the circuit?


Encoder- turns Highs and lows into CODE
Multiplexer- turns Highs and Lows into DateSTREAM

A DVM,oscilloscope,LED or LAMP can't indicate or monitor these, so how can i please?
 

Papabravo

Joined Feb 24, 2006
21,225
how do u use your oscilloscopes Trigger mode, when testing digital circuits?

What are some "test methodology" for digital electronics?

1.) Check the supply voltage on IC chip pin to make sure its getting power
2.) Check the clock pulse
3.) Check the ground to the IC chip
4.) Check the inputs and outputs

The hard part for me is that my Signal Generator only has ONE TTL output , and most of the logic IC chips have like multiple inputs , so i need a multiple TTL output Signal generator. So its hard for me to test the inputs and outputs at once in a Cycle

So i guess its best to use Switches to +5 volts to test the inputs since i don't have multiple signal generators
I use auto triggering on the channel of interest to see if "anything" is happening. Then I switch to "normal" triggering and select either positive or negative edge ass appropriate. Then you adjust the trigger level to see more or less of the edge.

You seem to be focused on verifying that the chip manufacturers have done their job. Trust me on this, that it is a colossal waste of time on your part. Focus on understanding how the collection of chips should be working and you won't have to spend so much time with "switches and lamps"
 

Thread Starter

relicmarks

Joined Oct 13, 2006
355
Then I switch to "normal" triggering and select either positive or negative edge as appropriate.

1.) Thats another of my problems, how do i know if its positive or negative edge triggering? how can i tell looking at the schematic? any hints that would tell me?

2.) Is this mostly the CLOCK signal that chooses the circuit to be positive or negative triggering?

Focus on understanding how the collection of chips should be working and you wont have to spend so much time with "switches and lamps"

2.) well the switches and LED's will tell me if the chip is faulty or not

Yes how the "Collection of chips" is working together is the hard part
 

Papabravo

Joined Feb 24, 2006
21,225
I recommend inputs be connected to a pullup resistor and a switch to ground.

1.) What resistance value should the pullup resistor be to not LOAD the circuit?


...
What kind of chips are we talking about. TTL is different that CMOS is different than ECL.
Generally 1K-4.7K for TTL
22K to 47K for CMOS and low power applications
For the 55th time will you get off this kick that a single answer applies to all situations. And let's start getting a bit more specific about your actual problem. Are you trying to test individual chips or are you trying to test a complete board?
 
Top