ShutDown Pin does not go HI - v2.8

[paulkosm]

Finally, I got a time to continue on my  project - Internet radio player: PI 3, 7 inch DIS, p-amp, amp, and power distributor for the all components. Anyway, I soldered that jumper on back cs I want to have it work in latch mode; the momentary button mode is not acceptable in my case unless I want to build some gizmo that I can shutdown by depressing the button for 0.5 sec... no, wait a minute, more then 2 sec and less then... I am lost here. I would have to hang an instruction next to my radio for the people in my household - "How to turn on and off this radio. A stopwatch will help". Weirdo. I want to have a standard behavior power switch so the latch mode is the closest: press once = ON, press second time = OFF. For the reset function I can utilize the run port on PI.
So, I have soldered pin terminals onto this switch and I tested it without connecting to RPI. I wanted to see how the outputs and input behaving. I simulated the connection with RPI GPIOs. testing that I have found that the ShutDown pin does not go HI.

POWER ON stage:

1. Power ON (latching push button): LED status is on; all power outputs are HI -5V - NORMALL
2. After about 20 sec 5V is applied to the BootOK pin (to simulate the BootOK): the green LED next to the pin is ON - NORMALL

--->The ShutDown pin is 0V at this point

SHUTDOWN stage:

1. Power OFF (latching push button): LED status start blinking and after  1 sec the red LED next to this pin goes ON and the ShutDown pin... goes LOW - 2.3V

2. After 15 sec the 5V is removed from the BootOK pin. After additional few seconds the switch shuts down the power - NORMAL.

The ShutDown pin stays at 2.3V until the power is off.

Conclusion:

Everything behaves normal except the ShutDown pin does not go HI to 5V - stays at 2.3V . What I was reading from your description that this pin should go HI to trigger the shutdown process on PI - which makes sense. But for some reason it stays LOW at 2.3V. Am I missing something here, or the switch is defective?

P.S. I have attached JPG picturing data log . The orange line is the BootOK, and the blue is the Shutdown pin. As you can see it went to 2.3 v.

Best regards

[paulkosm]

This is my reply to my own post. Actually a question.
Usually the HI signal should be over 2.5 v to be considered HI. I started digging online how this applies to RPI 3. Actually there are sort of mixed information not really consistent in regards to the exact voltage threshold on RPI 3 as there is no firm info on this issue. I have read few articles about this and on YouTube one dude did real measurement of voltage threshold on RPI gpios. He confirmed that LOW is 0 to 1.3v and HI is 1.3v to 3.3v. Therefore, if the shutdown pin on ATXRasp produces 2.3v, would this be enough for RPI to recognize the input as HI?
Can anyone confirm this from own experience? Thank you.

[Felix]

Hi Paul,
Thanks for the explanation and details, the SHUTDOWN signal pin should go higher than 2.3V when it's HIGH. Sounds like you already tried debugging/testing the ATXRaspi while not attached to Rpi.

Is it possible to desolder the jumper on the back (to revert it to momentary mode) and try it that way and see if the SHUTDOWN signal is higher than 2.3v?
It's remotely possible you have a faulty board even though the boards are tested for all functions to ensure they would assert the signals properly. But let's try to get yours figured out and we can even do a return and repair/replace to make sure you are satisfied with your ATXRaspi.
Let me know what is easiest for you .. try to debug some more or just return it for me to resolve the problem?

the momentary button mode is not acceptable in my case unless I want to build some gizmo that I can shutdown by depressing the button for 0.5 sec... no, wait a minute, more then 2 sec and less then... I am lost here. I would have to hang an instruction next to my radio for the people in my household - "How to turn on and off this radio. A stopwatch will help". Weirdo.

Most people seem to understand it once they see or try it once, but I see your concern with other users from the household (although I can't remember anyone really complaining about it being too complex). Even so you can actually request it without the reset or delay if that is really confusing. The "stock" firmware is more "complicated" to offer more functionality and most people seem to appreciate that. Then some customers order the boards with custom firmware or even custom hardware functionality after they try and like the board. I try to make everyone happy.

[paulkosm]

Hi Felix,
thank you for your response.
Of course, you are right, most people will do, but that's not the point; if I build something that I will be only the one who will use it I can do however I want, but if I build a device designated for a public use - people in my house are considered as "public" in such case - as an engineer I tend to bond to rule "keep it standard". As when comes to hardware as software engineering. Otherwise the user will turn a way from your design.
Do not take me wrong; I am not trying to be rude. I just said: "...in my case". I understand that other people might like it the other way and they have a right to like it the other way. I am not trying to dictate the way you supposed to run your business. Anyway, this a nice piece of hardware and very affordable.

Back to the main topic here.

So, I decided to connect it to my RPI regardless the 2,3 V and see. As I said before: I found mixed info regarding the voltage on RPI GPIO pins. Basically,  3.3V it is the max the should be connected those pins, and 1.3V to 2.8V RPI considers as HI. I gave a shot and it worked. As I mentioned I am using in latch mode. BTW: in momentary mode it also produced 2.3V.
Since I do not know RPI and I never programmed in Python I was curious how the script works so I added logging points to the script - debugging - with an output to a text file. I was able to see what was happening to pins 7 ans 8 during boot up and shutdown. Python seems a very sweet language to program with.

NOW THE PROBLEM
It worked for several days; I was able to shut down gracefully my RPI and I was a happy man. Before yesterday, I noticed that when I connected power to my radio the bootok green led went ON before pressing the power button. Next, I removed the switch from the radio and just connected RPI power supply to the USB power IN with the same result: the bootok led is ON before turning the ATX switch on. I measured the power out ports and they have 5V. It seems that the power leaks to out put pins. of course I tested my power switch and it is fine. I do not want to troubleshoot this circuit because I believe it is not my task to do at this point. Felix, can you tell me what would be your idea here? Thank you and best regards. 

[Felix]

I removed the switch from the radio and just connected RPI power supply to the USB power IN with the same result: the bootok led is ON before turning the ATX switch on

Did you completely disconnect ATXRaspi?
If not, and you left the BOOTOK/SHUTDOWN signals connected to your Pi which is powered separately, then of course they will "leak" and could light up.

[paulkosm]

As I said "...removed the switch from the radio".
More precisely: let's say, the switch is connected to RPI and everything is powered off. Now, if the 5v from power supply is applied to the power input, the power out pins go high. BEFORE PRESSING THE POWER SWITCH TO ON. The bootok led goes hi because the power is applied to RPI which should not be at this point. Anyway, the power switch is useless- because the power out is already on.

If the switch is COMPLETELY disconnected from RPI it behaves the same except the bootok led is off obviously since there is no signal from RPI.
Like I said:  the power "leaks" through the output pins. The voltage appears on them when it should not. I don't know whether I can explain it in a simpler way.
To cut the, I mean to save you the time on your  posts I can assure you that if I solder something it is soldered as it said in books. Connections are per instructions. Circuits are handled using electrostatic prevention. What else I can add - I don't know. Can you post the schematic diagram and your source code, or it is a top secret technology of yours? LoL.

[Felix]

I'm completely confused. A picture would be worth 10 posts to figure out what you're doing.
Anyway I will gladly debug the board for you if you want to return it. I will repair or replace any proven faulty board.

Removed to me means you taking it somewhere else from the circuit. Away. Like .. not connected to the Pi AT ALL, not a single wire. Then how is it possible the BOOTOK/SHUTDOWN signals come on, unless it IS connected to it, makes no sense. I don't see the problem though.

I am only concerned about ATXRaspi, not how the Pi behaves with it. Only with the ATXRaspi FULLY and COMPLETELY connected (not partially, its not meant to be used partially connected, and I can't guarantee doing partial connections won't damage it).

Also call it ATXRaspi please so we both know what we refer to, switch can mean many things.

[paulkosm]

OK, my radio is a system that is made of a number of components. Like any radio. in this case these components are: RPI, touch screen, amplifier, pre-amplifier, power supply, speakers, and ATXRaspi that works as POWER SWITCH that fulfills omitted component on RPI like the one found on full fledged motherboards where the role of such switch is to be a part of power management, so the system can be shutdown gracefully.
Next, "...removed the switch from the radio", or "...the switch is COMPLETELY disconnected from RPI" means that switch (ATXRaspi ) is disconnected from the system which means there are no wires from ATXRaspi  to RPI or any component in ma radio and ATXRaspi was  tested as as a standalone device. For the test/troubleshooting, the power is taken from separate power supply - RPI power supply  that I have bought for the RPI. This power supply is not used to power my radio.
Next, "...If the switch is COMPLETELY disconnected from RPI it behaves the same except the bootok led is off obviously since there is no signal from RPI" . This quote is out of context now, but if you read my whole post were I refer to ATXRaspi  as a SWITCH - sorry for the confusion, maybe I should be more specific with naming so you would not be more confused -  I stated there that:

"... the switch is connected to RPI and everything is powered off. Now, if the 5v from power supply is applied to the power input, the power out pins go high. BEFORE PRESSING THE POWER SWITCH TO ON. The bootok led goes hi because the power is applied to RPI which should not be at this point"
- it means this: when power cord is connected to power in port on ATXRaspi  the 5V appears on power out PINS WITHOUT switching ATXRaspi  ON. 5V should appear only after the latching switch is pressed so is set to ON position. Here is the malfunction; there is power output on regardless the latching button is used, or even is not connected to the ATXRaspi  as a matter of fact. The bootok led goes on before the RPI boots up when ATXRaspi  is connected to RPI. If ATXRaspi  is tested standalone obviously bootok does not go ON because ATXRaspi  is REMOVED or DISCONNECTED from RPI.
Again, the point is: it is not important whether ATXRaspi  is connected to RPI or disconnected. 5V appears on power out ports right after it gets power from power supply.
You know, I did not want to be annoying that's why I asked you for schematic diagram and source code but you omitted this point in your reply by being conveniently silent about it. Maybe it is a business secret. You know, this size of circuit could be easily reverse engineered including hex code from the chip but of course I AM NOT going to do this because I respect someones's intellectual property, I am not your competitor - I live of programming, I am an end user of ATXRaspi , this radio is my one time shut project. I wanted just safe your time, that's all. My point is that when I buy something that becomes defective right after few days of using I turn to the re seller for exchange or refund and then I get shipping label so I can ship it back to the reseller.
Finally, It is not about the $15 or so, it is about my time, and I have to admit that I am being a little tired of the replies back and forth so I have proposition:

I will send my ATXRaspi  back to you at my cost and you will test it - I believe in your honesty because I do not have a reason not to.  If you will say that is not defective and I did something wrong I will pay for the additional shipping cost back to me. If you admit that it is defective and you do not have a reason to believe that I did something wrong, you will replace it, reimburse my shipping cost, and you ship it back to me at your cost. What you say, is it fair enough? Well, after what I saw in this forum I understand that you may suspect that some users are not competent. Maybe I am not too, I leave it to your judgment.

Here is short video clip of my ATXRaspi

[Felix]

The return policy is fair and I try to give a guarantee that you get a working product. The easier you make it for me to see the problem, the faster we can work out the cause and fix it, with or without a return. I only have so much time to read posts, and I like quick sweet facts, photos and visuals help a lot, long descriptions make me tired. Sometimes users do damage boards, and I do not evaluate their competency since I try to make it easy for the whole spectrum of users from beginner to advanced. Some soldering skill is required to make good clean solder joints free of shorts. On the software side I provide a written and video tutorial on how the board is supposed to be used and installed etc, but some skill is also required there, but quite minimal.

I believe your unit might be damaged. Please return it for a 1 time inspection/repair/replacement. The return address is the same as the one on the package you received. Otherwise you can email me for details.
Thanks

PS. One last thing. I've seen cases where an obscure solder short created such confusion. I mean solder shorts under the header's plastic, i'm not kidding, just look at this post for an example. Please give it a close check and maybe something as simple as that could explain this 3 mile long thread.