Cable Tester Automatic Power On

Thread Starter

dcbingaman

Joined Jun 30, 2021
1,065
I am designing a cable tester for scope probes, DMM leads etc. It will basically be a small box with a green 'cable good' indicator and a 'red' cable bad indicator (LEDs) and will run off of a 9V battery. A small PIC microcontroller will handle sensing when a cable is attached and then performing a pass/fail test.

The power source has the following requirements:
1. No on/off switch, the cable being plugged in will automatically turn on power to the microcontroller via a 3.3V LDO.
2. When the battery is replaced, the unit will also power on automatically and perform a 'battery good' test.
3. After the microcontroller performs the cable test, it will remain on for about 30 seconds after the user removes the cable being tested. The microcontroller will turn itself off.

I have the following sub-circuit designed to handle the above requirements. Note: I am not using the shown 3.3V LDO, but it will be similar with an 'Enable' input like this one:

1663451466102.png

Circuit notes:
- When the battery is connected, C1 starts to charge, turning on M1 which then enables the 3.3V LDO that provides power to the microcontroller. (This is a sub circuit only, the microcontroller is not shown).
- D3 then latches power on, by placing the 3.3V output to M1 holding it on.
- Cable_In1 and Cable_In2 are typically the 'ground lead' of the cable being tested. These are shorted together when the cable is connected turning on M1 which then latches power on to the microcontroller. This turns off M2 which provides a 'Cable_Connected' digital input to the microcontroller.
- R6 and C4 are not part of the circuit but here are simply emulating the microcontroller 'MICRO_SUICIDE' digital output.
- M1 and M2 are not the final parts. I simply selected them being they are part of LTSpice at the moment.
- When the microcontroller sets 'MICRO_SUICIDE' high, Q1 turns on via D4, which in turn turns on Q2. This latches Q1 and Q2 on which turns off power to the LDO. Which then allows, C1 to charge and turns off M1 resetting the circuit.
- D1 and D2 protect the circuit from a battery being plugged in 'backwards'
- C2 just does some filtering
- Filtering for the LDO is not included here but would be on the final circuit.

I am considering:
- Replacing Q1, Q2, D4, R9, R8, and R7 with a small SCR to reduce parts count.

I have also included the LTSpice circuit.

Here is my question:
This seems rather complicated to accomplish my requirements. Does anyone know of a simpler way to accomplish this without so many parts?
I also just noticed that M2's gate is connected to one of the outside connections. This should probably be protected from ESD damage from an ESD source, what is a good device to use for that?

Thanks!
 

Attachments

Last edited:

MrSalts

Joined Apr 2, 2020
2,767
Why don't you just have the microcontroller control the EN (enable) pin on the second chip? Once disabled, the current draw should be minimal and should be on the datasheet. Removes the bulk of your bulk.
 

Thread Starter

dcbingaman

Joined Jun 30, 2021
1,065
Why don't you just have the microcontroller control the EN (enable) pin on the second chip? Once disabled, the current draw should be minimal and should be on the datasheet. Removes the bulk of your bulk.
Because it cannot perform the following requirements:

1. No on/off switch, the cable being plugged in will automatically turn on power to the microcontroller via a 3.3V LDO.
2. When the battery is replaced, the unit will also power on automatically and perform a 'battery good' test.
3. After the microcontroller performs the cable test, it will remain on for about 30 seconds after the user removes the cable being tested. The microcontroller will turn itself off.

The microcontroller turning itself off can be done that way (i.e. using the enable pin), but how do you only turn on power when the battery is replaced or when a cable being tested is plugged in? The other issue has to do with turning off the LDO: It could easily turn right back on when the voltage to the microcontroller shuts down because Q1 will turn back off when power is removed. Thus I had to 'latch' the setting to guarantee it does not turn back on.
 

Sensacell

Joined Jun 19, 2012
3,432
Really flakey cables will pass your tester.
You need to validate a threshold of resistance below 5 ot 10 ohms to really test a cable.


This might not be what you want but it's simple and works well.
 

MisterBill2

Joined Jan 23, 2018
18,167
The very first thing I ask is just exactly is the device supposed to detect? What conditions indicate a satisfactory cable? Reliably verifying whatever cable properties constitute a "pass" is vastly more important than automatic switching on.
So the very first thing is what is considered an "accept"able?
For verifying the 10:1 attenuator cables the check will not be so very simple.
And, in addition, the probe calibration check is already present on many scopes, and will verify calibration as well as continuity.
 

Thread Starter

dcbingaman

Joined Jun 30, 2021
1,065
The very first thing I ask is just exactly is the device supposed to detect? What conditions indicate a satisfactory cable? Reliably verifying whatever cable properties constitute a "pass" is vastly more important than automatic switching on.
So the very first thing is what is considered an "accept"able?
For verifying the 10:1 attenuator cables the check will not be so very simple.
And, in addition, the probe calibration check is already present on many scopes, and will verify calibration as well as continuity.
The most frequent problem I have ran into is bad ground leads on scope probes and occasionally a bad probe (no signal at all). I have rarely found a situation where DMM probes and BNC cables are 'partially' working, that is a resistance that is 'higher' than normal. They typically (at least for me) end up simply being open or the contact inside a BNC for example is intermittent.
For the scope probes I was planning on sending a square wave into them and looking at the output after amplification using an ADC when in the x10 position. BNC cables would be similar but with no amplification. I may also generate some circuitry for verifying the probes have the 10Meg (x10) and 1Meg (x1) input impedances.
The other property for the tester is 'drop out' detection. That is if the cable is intermittent (very common) the tester will 'latch' onto that. That is the microcontroller will be looking for any intermittent drop out and latching onto that. This allows the user to attach the probe and then 'bend it around' in an attempt to create the failure. Intermittent probes seem to be the thing that most drives me crazy. I have been troubleshooting issues with circuits only to find out the probe is 'intermittently' not working correctly.
 
Last edited:

Thread Starter

dcbingaman

Joined Jun 30, 2021
1,065
The very first thing I ask is just exactly is the device supposed to detect? What conditions indicate a satisfactory cable? Reliably verifying whatever cable properties constitute a "pass" is vastly more important than automatic switching on.
So the very first thing is what is considered an "accept"able?
For verifying the 10:1 attenuator cables the check will not be so very simple.
And, in addition, the probe calibration check is already present on many scopes, and will verify calibration as well as continuity.
Good point on the scopes having that capability. Believe it or not, for some reason I always seem to forget about that. It is like a 'special feature' to the scope, that you don't normally think about using when problems show up. I have used that feature before, especially when adjusting the capacitance of the probe in order to get a good flat square wave out of them (calibration). Probably for scopes that don't have that output, a really simple tester could just be a box with a battery that puts out a square wave that you could then look at with the scope.
 

MrSalts

Joined Apr 2, 2020
2,767
First, I always store my probe by clipping it onto the test point (calibration point) on the scope. After the cope runs through it s startup routing, I make sure I get my 1V x 1kHz signal before I do anything.

Personally, I would design it super simply to make sure I can instantly know the problem is my probes and not some issue with my complicated circuit.
I would add a momentary button to check if the battery works at before I test a probe, then I'd test the probes. Then push the buttons again to make sure the battery works if any probe fails.

No programming to worry about and it will only fire if the momentary is pressed or a probe is attached.

81D64F02-30A1-4787-A088-3320C51262BE.jpeg
 

MisterBill2

Joined Jan 23, 2018
18,167
How about the first step being with the probe attached to the scope, and connecting the ground lead to the tester. IF there is a good connection the green LED lights and the scope power gets switched on. Then with the probe on that test check point, observe the square wave. At that point the probe is veridied and checked and ready to be used. No complex turn on circuit and a complete check.
 

Thread Starter

dcbingaman

Joined Jun 30, 2021
1,065
First, I always store my probe by clipping it onto the test point (calibration point) on the scope. After the cope runs through it s startup routing, I make sure I get my 1V x 1kHz signal before I do anything.

Personally, I would design it super simply to make sure I can instantly know the problem is my probes and not some issue with my complicated circuit.
I would add a momentary button to check if the battery works at before I test a probe, then I'd test the probes. Then push the buttons again to make sure the battery works if any probe fails.

No programming to worry about and it will only fire if the momentary is pressed or a probe is attached.

View attachment 276520
"First, I always store my probe by clipping it onto the test point (calibration point) on the scope. After the cope runs through it s startup routing, I make sure I get my 1V x 1kHz signal before I do anything."

Simple and effective. Thanks! I will try to get into the habit of doing that.
I also like your simplified cable tester.
 
Top