PM240 is no longer detecting attached sinks

[Sonartech]

My PM240 has started doing something strange recently - about 95% of the time, it "kinda" detects an attached sink, but won't let me do anything with it. This happens with or without the test software running. The voltage shows "5.06", but the Sink shows "Disconnected" and that's after I get the white full-screen pop-up showing "SINK CONNECTED". If I attempt to look at PD profiles for the attached power supply, the screen quickly flashes but does nothing (shows no profiles). About the only thing I've been able to do to get past this is to turn the whole unit off, connect the power supply, then turn it on - and about 10% of the time, the unit will see the attached power supply and show me its PD profiles.

When it gets "stuck" like this, it will reliably fail self-test #7 with error code 30-1. I'm running firmware v2.5 but I also did try to downgrade to v2.4 just to see if that made a difference but the behavior was the same. When it detects properly, it also tests properly, and the self-test does the whole cable-flip check and power testing - and everything passes. But that's becoming super rare, and I basically have to fiddle with the tester for 20 minutes before I get it finally see an attached supply.

Any ideas what's going on? I kind of worry that the FFT graph overheat issue might have made the electronics unhappy, even though this thing has worked fine up until about 1 week ago.

Thanks.



In the attached photo above, a PD-capable supply is attached but the PM240 won't acknowledge it's connected or show me any profiles.

[Nik (PassMark)]

Hi,

Thank you for the very detailed description of what is happening with this issue. Based on that information I have one guess at what might be occurring to cause this, which is that the communication link between the Protocol Controller Chip and the LCD Controller Chip is failing when the device boots up. I did some investigating into what could be causing this to happen and have made a small change in the firmware that will help prevent a possible rare edge case from occurring that can lead to this problem. I've created a new debug build v2.5.3 with this change now, would you be able to update the unit to v2.5.3 and see if it solves this problem please, you can get this updater tool from this link: https://www.passmark.com/downloads/U...ter_v2.5.3.zip

[Sonartech]

Hi, Nik -

I loaded 2.5.3 successfully but it doesn't appear to address the issue; the behavior is the same (but it's not worse, either).

[Nik (PassMark)]

Hi,

Thank you for testing that, we will investigate this further to work out what is causing this to happen and find a solution. Could you answer the following questions for me as well please to help me narrow down some more of the possibilities:

1. Have you tried using other chargers and get the same result?
2. When you connect the source to the sink port, does anything different happen?
3. Does the full screen pop up saying "sink attached" appear every time you connect the device under test?
4. When running the self test, do you start the self test with nothing connected, and then connect the source to sink port when stated in test 7?
5. Is there any difference if you power up the unit with a device under test connected or having no device under test connected until after power up?

[Sonartech]

1. Yes. I test PD chargers as a side-project, so I have a LOT of PD-chargers. I also have a PM125 that I use for 100W or less scenarios, just because I don't have to plug it in to AC and it does the same kind of testing, just less power. It works flawlessly. That said, I'd need to do some more analysis to see if this is something that one of the chargers I'm currently evaluating "initiates" or causes the PM240 to get into some kind of weird race condition [that persists through power-off resets?!?!]. If that's the case, then I would be happy to send you the suspect charger for analysis if there's a chance to improve the PM240 design or behavior. This is, after all, what these testers are designed to do, right? This would be much easier to sort out if I had two PM240's for side-by-side testing...
2. Source-to-sink, as in on the tester itself, like as part of the self-test? When the PM240 gets into this state, the self-test always fails step 7 with code 30-1. It never detects anything is connected and does the whole "if 10 seconds passes and it doesn't detect, press the fail button". When it is detecting as expected, the self-test passes.
3. YES, but only after I power-off reset and power back on. I can connect the power supply and I get the full screen white pop-up "SINK CONNECTED" message, then it goes away, and I'm left with the "Sink: Disconnected" message as shown in the picture I provided with my first message. So, it DOES see it, and it IS seeing the 5.0V from the attached supply, it's just saying "Sink: Disconnected" anyway.
4. I follow the self-test instructions as follows: No cables connected; go through the screen tests, then connect source-to-sink using the supplied 5A cable. From here, it'll either immediately tell me to flip one end and continues to eventually pass, or it doesn't even register that I plugged the cable in at all and I get the 10-second fail timeout option.
5. The ONLY way I've been able to get the PM240 to work when it gets in this state is to have the power supply (DUT) connected to the sink port when I turn on the PM240. About 25% of the time, I'll see the display show "5.0V" for an instant, then "11.0V" for an instant, then it pauses, and shows the power supply is detected and ready for testing. When it fails, however, I just get the white pop-up SINK ATTACHED and the same thing I mentioned in #3 above.

When it starts working, I can complete a range of testing for several hours - but if I physically unplug the DUT or turn the PM240 off, I have to dance with it some more to get it working again.

I'll do some testing this afternoon to see if I can nail down the behavior a little more. It's sitting in my lab ON but not doing anything right now (I don't use it unattended).

[Sonartech]

First off, thanks much for assisting me with this! Much appreciated.

Lots of testing this evening and good results to share. TL;DR - I think Passmark's going to want me to send them a power supply to them for analysis and testing.

Test methodology is based on the sequence of events leading to my first failure a few days ago: From an off state, I powered-on the PM240 and attached a new power supply for testing. I saw the "Sink Attached" pop-up, and saw 5.0V, but the display still showed "Sink: Disconnected" at the top, and I couldn't select any PD profiles - the screen just quickly blinks but never shows anything. After that, the PM240 is effectively "poisoned" until the next power-off reset and nothing else will detect, even if it's a known good supply. That's the basis for the following testing.

Requirements

1) Known good DUT USB-PD Supply, using an Anker 250W Prime desktop charger
2) Known good 240W cable; using a Thunderbolt 4-certified cable qualified with a total phase advanced cable tester
3) Suspect DUT USB-PD power supply

Initial Baseline: Test the Known Good equipment back-to-back 20 times from a clean POR (power off reset) state; insert & remove, no test software running. Expect to see PM240 detect the cable each time and settle on PD-FIX 5.0V, 3.00A profile.
Test results: PASS 20/20

Suspect DUT Testing Scenario 1: From a clean POR state, use known-good DUT Power supply and known-good cable 5 times first, then connect suspect DUT 5 times in succession and check for detection & contract negotiation.
Test results: PASS 5/5 using known good DUT, PASS 5/5 using Suspect DUT

Suspect DUT Testing Scenaro 2: From a clean POR state, attach suspect DUT power supply to Sink and check for detection & contract negotiation. If successful, record PASS, disconnect Suspect DUT and power-cycle PM240 for next test pass. Repeat testing cycle 10 times total.
Test results: FAIL 10/10. Upon connection, the pop-up message "SINK ATTACHED" appeared as expect, but "Sink: Disconnected" shows in display immediately after, and no profiles can be viewed, even though the display shows Voltage: 5.06V.

Analysis & Summary

The suspect DUT appears to be able to "poison" the detection capabilities of the PM240 on power up, if it's the first thing connected to the PM240. If a known good USB-PD supply is connected to the PM240 on power-up, subsequent detections using the suspect DUT power supply will also pass. If the suspect power supply gets connected first, however, any subsequent connection from any DUT, known good or bad, will fail. The PM240 must be power-off reset to restore proper detection capabilities.

This indicates two issues:

1. The PM240 cable detection mechanism can get "poisoned" by the suspect DUT if it's the first USB-PD power supply connected to the PM240 immediately following power-on. Since the suspect DUT can get detected properly if a known-good DUT is used first, this suggests that the PM240 somehow changes into a state that can allow it to properly detect the suspect DUT after it's seen a "good" DUT. That's possibly a PM240 firmware issue, possibly a wacky contract negotiation race condition scenario that's not covered by the state machine and causes subsequent detections to also fail. This also tells me that the detection mechanism in the PM240 isn't being "completely" reset on VOLTAGE=0.0 / cable removal like it would be with a power-off reset.
2. The Suspect DUT power supply is causing the PM240 to get into this state. I haven't investigated what it is about this supply that could be causing the PM240 to hang-up like this, but since the PM240 should be robust in all respects when it comes to testing "any" USB-PD supply, I would put need for corrective action more on the PM240 than the suspect DUT power supply - especially since the PM240 will properly detect the suspect DUT power supply if it's seen a "good" DUT power supply first.

So, would you like me to mail this supply to someone for a deep dive? Let me know

[David (PassMark)]

Thanks for the detailed investigation.

Yes, if you have one of these suspect supplies spare (or we can buy one off the shelf) it would good to get an example of the unit.

If you can Email us, we can sort something out.

[Sonartech]

Thanks for the detailed investigation.

Yes, if you have one of these suspect supplies spare (or we can buy one off the shelf) it would good to get an example of the unit.

If you can Email us, we can sort something out.

Howdy. I sent mail to info@passmark.com on the 10th and today. I haven't received a reply, but I'm ready when you are.