Digital circuit inputs turn on thresholds, turn on voltage?

[Billy Mayo]

So what continuous voltage can I apply on the output of a TTL or CMOS chip? 2 or 3 volts is that safe?

Is that why they use the logic pulsar because its not a continuous voltage and its a short duration to cause damage to the output of a TTL or CMOS chip?

What voltage can I apply to the output that is continuous DC voltage that is safe?

 

[Billy Mayo]

applying continuous +5 volts to a LOW signal will cause a SHORT?
Or applying a low signal to a +5 HIGH signal will cause a short?

 

[KeepItSimpleStupid]

yep, just change your sentences to "LOW OUTPUT signal" and "HIGH OUTPUT signal". The paths are different, but it's still a short.

 

[Billy Mayo]

Oh ok thanks sorry about that

When there is no load on a TTL or CMOS chip , what output signal is?

Is a pull up or pull down resistor a Load? what is a load? what can i use as a load on the output of a TTL or CMOS chip?

what continuous voltage can I apply on the input or output of a TTL or CMOS chip that is safe? Can I use an external power supply with a 1K or 2.2K resistor in series to limit the current So I can use a contiinuous +5 volts to inject on the inputs or outputs when I'm troubleshooting Logic circuits?

 

[audioguru]

I told you the output voltages of TTL when it has a load. I have not iused TTL for about 35 years so I do not have any to measure without a load.
The output of Cmos with no load is 0V for a low and its supply voltage for a high.

You can use 0V for a low and use the supply voltage for a high as inputs to TTL or Cmos. TTL MUST have a 5V regulated supply but ordinary Cmos can have a supply from 3V to 18V.

 

[MikeMl]

...
what continuous voltage can I apply on the input or output of a TTL or CMOS chip that is safe? Can I use an external power supply with a 1K or 2.2K resistor in series to limit the current So I can use a contiinuous +5 volts to inject on the inputs or outputs when I'm troubleshooting Logic circuits?

You should never try to force a TTL Output high by connecting it to +5V. That will blow up the TTL output driver.

You can get by temporarily grounding a TTL output to force it low for a short time.

 

[Billy Mayo]

Oh ok I never knew this stuff sorry

So how can I "force" an input or output HIGH or LOW , without damaging the TTL or CMOS chip?

I am trying to Force it since I'm troubleshooting Logic circuits with the TTL and CMOS chips in circuit

I need to know how to "Force" a HIGH or LOW on the inputs and outputs

I don't want to use a Logic Pulsar because the Pulsar toggles from high and low , its a squarewave output , so it's hard to troubleshoot using it

 

[Billy Mayo]

I don't know what the signs of a bad TTL Or CMOS multiplexers, adders, counters, gates, flip flops, do when they are bad or failing ? what have you seen or come across please?

 

[KeepItSimpleStupid]

They get hot. They don;t change state. TTL low is lower than 0.6 V. You have to look and understand the input and output stages for the logic family. They can also oscillate. A dead chip that takes out a supply is pretty nasty to find without the right equipment like a tone milliohmmeter.

Remember injection, or signal tracing are the usual troubleshooting techniques. When you do either, you verify that it doesn't work and then start in the MIDDLE. If your tracing and a signal is not there, then it's in the 1st half. Then you bisect the first half and figure what half it's in and then bisect that half.

Linearly tracing stuff is nuts.

You have to use your judgement. #1: Is always power supply and ripple. Can't stress that enough. Make sure the inputs are present and the controls set properly.

 

[Billy Mayo]

A dead chip that takes out a supply is pretty nasty to find without the right equipment like a tone milliohmmeter.

What u mean by a dead chip that takes out a supply? like in the power supply or how does it take the supply out?

When a Test technician is testing 100's of the same boards, they do find some boards have higher Supply buss or lower supply buss than it should because some component or IC chip is cause that problem. It is hard to find those. The Power buss will be good at 100VDC, VCC will be +5 , VCC +12 , the power buss will be 150 volts or 170 volts DC when BAD , the VCC +5 , VCC+12 will either be the same or no voltage at all. The main problem is the power DC buss is HIGHER or LOWER

Remember injection, or signal tracing are the usual troubleshooting techniques

I do injection using a function generator if a stage is not working when its an audio circuit. You can inject the function generator at any stage and signal trace it.

But you guys have made me learn that Injecting or Injection Techniques with Logic TTL or CMOS circuit is bad and can damage the chips which I didn't know before.

Linearly tracing stuff is nuts.

Do you mean linear tracing? is signal tracing one stage after another in a row?


When there is a short VCC or there is a component shorted to VCC or shorted to ground, how would you troubleshoot this?

do you use an ohm meter/ continuity checker and put the RED probe on VCC and use the black probe and go to each component on the circuit until you get a beeping sound making continuity, then you know you found a short?

My DVM meter doesn't measure milli-ohms to check for shorts

Is there a way to use an oscilloscope or a DVM meter to check for shorts? short to VCC or short to ground?

Any Techniques you use or know

 

[KeepItSimpleStupid]

Just suppose one TTL IC is shorted in a mess of 35 or so. When you turn on power the 5V supply goes into current limit. How would you find the bad chiP.

I told you about binary searching without telling you the term. Linear means you start at stage one and continue incrementing by a "stage" until you reach 10. Bisecting or binary searching is all about bisecting the faulty sections.

If you guys are troubleshooting a lot of similar boards, you should have some of these on the bench: https://www.huntron.com/products/2800.htm
With a good and bad board side by side, it should be easy to find the faults, unpowered even.

Multi-voltage supplies typically have dependencies. You have to be aware of them.

You guys should also invest in an IR camera that can be shared. A very expensive, but useful piece of equipment.

If your troubleshooting the same stuff and see lots of the same stuff, then you should be doing some sort of failure analysis with those boards. The symptoms should suggest possible failures even before you start.

You might want to take a look at some older Keithley, Tektronix or HP/Agilent service manuals to get an idea as to how they should be written.

When I was troubleshooting an early Scanning Electron Microscope, the most useful thing was the card extender. It was mostly 741 OP amps, but it was really easy to fix.

When I was dealing with a Kevex Auger mass Spectrometer, that was WAY different. Only schematics.

When i was dealing with DC power supplies and vacuum gauges, nearly everything could be fixed by knowing the symptoms. Relamping was done last. If the guage didn't turn on, there were 1 of 2 problems: ripple on the -15 V supply or a bad relay for the filament. It's like, those were the only issues unless someone spilled acetone on the meter face. Then there was calibration.

A DC power switching supply would basically get very noisy. Replacing about 8 or 9 capacitors would fix it every time.

A lot of other recurring problems were due to surges that were either power line related or loose internal connections. One had to do with non-isolation of a 100 kV power supply. Those I had to engineer fixes.

Many of the maintenance issues I had to deal with were also mechanical. Disk drives, fans, slidewires, potentiometers etc.

I did a fair number of repairs to high powered audio amplifiers. That takes some skill, because not catching a bad part takes out all of the new parts.

One of the most stupidest problems I had to deal with was a SBC (Single Board Computer) that would not reset when using a linear power supply. It worked fine with a switching power supply for power. Annoying, for sure. I just replaced an IC with a Schmidt Trigger version and fixed.

 

[Billy Mayo]

Thanks for the info

When you turn on power the 5V supply goes into current limit. How would you find the bad chiP.

Never came across a power supply going into current limit mode? what is that, do you mean the output of the power supply is only output is current not voltage?

Multi-voltage supplies typically have dependencies

What do you mean by multi voltage supplys have dependencies? what issues u mean?

You guys should also invest in an IR camera that can be shared

What would an IR camera do for troubleshooting PCB circuits?

 

[KeepItSimpleStupid]

A regulator such as an LM317, is 1.5 A regulator. When 1.5 A is exceeded it tries to deliver 1.5 A at a lower voltage. That's current limit. It would happen if an IC was shorted.

Dependencies
A 150 VDC supply may be dependent, ie. present only when the 5 V supply is present, or something liken if the 5V supply is 4.5 V, the DC supply would be 140 V. There could also be power supply sequencing.

IR Camera:
If can find hot and cold spots

So let's say you had a PCB and another both on for 10 minutes and took an IR photo of both. It's very possible that a non-working IC could show up too cold or too hot. See: https://www.vision-systems.com/arti...meras-tackle-pcb-inspection-applications.html

 

[KeepItSimpleStupid]

A shorted IC usually takes out the power supply or at least the PS goes into current limit. So picture a long trace of +5 and a ground plane for ground with a bunch of IC's hanging off of it. The "center IC is shorted" . A tone milliohmeter connected between ground and along the +5 trace will "show" a smaller resistance as the probe nears the short. They can be made rather cheaply.

 

[Billy Mayo]

PS goes into current limit

So Current limit means the voltage is lower but the PS current output stays the same?

Thanks for the info.

A tone milliohmeter connected between ground and along the +5 trace will "show" a smaller resistance as the probe nears the short

It's hard to find a DVM meter that has a milli-ohm setting, i don't know why they don't put them in DVM meter to test for shorts

My last job, the tech use to use one that a DVM that measured milli ohms

He also tested for shorts using a method of injecting 0.5A with an external power supply set at 0.5Amps and would measure the voltage until he would see where the current was going and had the biggest voltage reading, I'm really not sure how he did this but it is a common method used in the 50's.

 

[KeepItSimpleStupid]

A milliohmmeter is an entirely different animal and so is a meter that can measure gigohms. Usually milliohmeters require 4 or possibly 5 wires. One Fluke Ohmmeter that I used had a zero to 2 ohm compensation adjust. "Zero" resistance ammeters are nice animals too.

This is a design I like, but not the simplest that I have seen on the web: http://cappels.org/dproj/dlmom/dlmom.html

The injection technique could work too.

---

Current Limit comes in a couple of flavors. One is Constant Voltage/Constant current and the other current limit, or foldback current limiting. Some supplies have a Save Operating area meaning a range of voltages and currents that the supply will operate at.

 

[Billy Mayo]

Testing for shorts using a method of injecting 0.5A with an external power supply set at 0.5Amps and would measure the voltage until he would see where the current was going and had the biggest voltage reading, I'm really not sure how he did this but it is a common method used in the 50's.

Do you know anything about this method and how do you use it?

You have to use this method when the the PCB circuit is not powered on, the power is OFF and you inject it with a very small voltage

I have tried this method and im not sure what i'm doing really because it doesn't work for me to find out where the current is going and the biggest voltage drop is where the short is at

I was never able to track down a short using this method, but it does work if you know what you're doing

The Tech would always say when there was a short on the PCB ** where is the current all going**

Do you know how to use this method please?

 

[Billy Mayo]

A milliohmmeter is an entirely different animal and so is a meter that can measure gigohms. Usually milliohmeters require 4 or possibly 5 wires. One Fluke Ohmmeter that I used had a zero to 2 ohm compensation adjust. "Zero" resistance ammeters are nice animals too.

Yes you're right about the 4 to 5 wires , but that is different, you don't have to use the 4 or 5 wires , the DVM will still measure milli ohms with 2 wires, The HP Rackmount DVM meter did that I used at my old job.

Yes with the top line flukes, you zero out the probe leads, but it still doesn't measure milli ohms 0.0034 ohms 0.0068 ohms , this is way below 0.2 ohms which most DVM meters do