EEVblog #1107 - Shocking 4K BENQ Monitor Problem!

[EEVblog]

Why does this 4K BENQ monitor mysteriously switch off on it's own?
Inadvertently re-visiting an old video.
The chair strikes back!

[wraper]

Why does this 4K BENQ monitor mysteriously switch off on it's own?

Most likely because of the crappy video cable. It's not switching off, it just blacks out because of the received data corruption. Seen this crap myself with 3rd party cable which had the shield but it wasn't connected to anything.

[elmo]

It just dawned on me that this may be the problem with a Eizo monitor I have at my workplace. It switches off and on seemingly random, worse on some days and other days not at all.
I resolved the issue by switching to display port instead of DVI.

[MaxSimmonds]

I actually had this problem at work, I've just finished a graduate traineeship with ESA (I was making a digitally controlled Buck + Boost Converter). I was having some serious issues with it, working one minute and stopping the next. The main issue was that each time this (at that point) unknown issue occurred, it would blow the arse out of my high side and low side FETs on the buck side - direct short between G-D-S, all three!

Of course, after about 5 days of no luck, testing everything I could think of, I was beginning to get quite disappointed. So, staying late one Friday night, I thought I had solved the issue - it had been running fine (in open loop conditions) and no failures. A failure was quite obvious to see since the power supplies would go in to over current protection.

"Excellent! Time to grab a coffee and leave it for a while to see if the problems solved". So, I stood up, and heard the devastating click of the relays in the PSU going into short circuit protection. At this point, it seemed far to coincidental.

I remembered, from a couple of years ago, watching the video Dave made on this phenomenon back when I was at university (probably in my first year!). I didn't have (and still don't unfortunately!) a scope of my own, so didn't get around to testing it. However, that Friday night I realised what a pain in the arse ESD was. I re-soldered two new FETs (DPAKS, so a total pain to unsolder and resolder 15+ times) and stood up from my chair again, this time probing the gates of the MOSFETs (possible making the injection worse?). And sure enough, it failed again, with some horrific interference (if anyone's interested I have some scope shots somewhere)



So, thanks a lot Dave! No one believed me, even engineers with 30+ years couldn't see how my chair was causing my converter to fail. I explained I watch your video a few years ago and tried it out, showed them, and after they checked all my grounding/PCB layout/etc, they agreed it was the chair!

In the next few says I brought in one of those lights with the piezo crystal starters that was empty of gas. It was a lot more convenient to troubleshoot the issue than standing up and down from my chair all day! Interestingly, this problem only showed itself when the weather changed and it got more dry. (BTW: Sitting on an ESD mat was my solution for a while, but I like the cardboard box idea too!)

[joseph nicholas]

I took apart a bad quality vga cable from China and cut the chock apart to find it was , well there was no chock at all in the plastic cast, on both ends.

[aqarwaen]

 few years ago i had similar problem with brand new monitor.took it back same day to shop back and they replaced it with new one..new one didnot have that issue..

[BravoV]

I took apart a bad quality vga cable from China and cut the chock apart to find it was , well there was no chock at all in the plastic cast, on both ends.

Use magnet instead, to detect the ferrite.

[joseph nicholas]

Great advice never though of that.

[Artlav]

Huh.
I have a similar problem on my DIY laptop.
It's a metal case grounded to everything inside of it, it got a system management board that have an HDMI->LCD converter on it and a little PC box acting as the actual computer.
If i scratch the case with my fingernails the screen starts flickering and might turn off for a moment.

Haven't actually bothered tracing it down, but now i'm curious about just what kind of shielding issues on HDMI could cause this kind of problem.

[DanaParsons]

An easy way to reproduce this is with a discharge spark from a Van Degraaff generator. These will shut down monitors and projectors within 10 feet or more. I had always assumed that the monitors themselves were not shielded properly but I suppose it is possible that the HDMI is the culprit. Interestingly, iMacs do not seem to be susceptible to this phenomenon. It would be pretty interesting to track down what kind of shielding would solve this problem.

[hamdi.tn]

My dell monitor black out too everytime i plug anything in main or powerup the lab powersupply.

[Ranayna]

12 years back, just when i started my IT apprenticeship, we had something similar happen.
We use USB tokens for two factor login to our computers. When the token is unplugged, the workstation immediately locks.
At that time, we were all sitting in a row. Each time a certain colleague got up from his chair, the computers of his two neigbors immediately locked themselves. Somehow the static discharge would interrupt the USB ports and cause the token to lose it's connection for just a moment, enough to lock their workstations.

The phenomenon only stopped after the offending colleague got a new chair. And we never saw that issue anywhere else in the company since then.

[wraper]

Haven't actually bothered tracing it down, but now i'm curious about just what kind of shielding issues on HDMI could cause this kind of problem.

No wonder it  has problems. Besides no proper differential pairs and impedance, HDMI runs over several ground splits. On top of that connecting several ground pins (separate shield for each diff pair) through one common via is ridiculous.

[T3sl4co1l]

Ya know, I wonder how pervasive it is to see badly grounded shields on HDMI.

USB is almost universally bad, so there's that.  HDMI is a very similar interface, it could be the same idiots are making the same mistakes there.

For those not in the know:
1. Q: When is differential not differential?  A: Always.  A pair is two wires, and each wire has a voltage with respect to ground, period.  When the common mode voltage goes beyond the input range of the receiver, poof, no more signal!  Typically, high speed interfaces (CML, LVDS, etc.) only receive within the supply range (often 0-3.3V).  (Industrial interfaces, like RS-485 and CAN, have a wide range of +/-10V or more, but aren't that fast.  Ethernet is fast AND has a whopping +/-1500V range, because it's transformer isolated.)
2. There is only one uncommon special case, where a connector's housing/shield shouldn't be connected to ground.  For >98% of cases, the shield must be grounded, at least through multiple capacitors, if not directly.  (What's critical is RF grounding, not DC galvanic grounding.)

Tim

[Artlav]

No wonder it  has problems.

Well, it works. :/

My excuse is that it's a hobby project with no third people use potential, but i can totally see how someone would design a product with the same mindset.

[isopropilick]

Back in January i brought an BenQ BL2710 monitor and i had the same problem! (I work with high power RF transmitters) It results in to that the only thing that i had to do is to disable the eco sensor (located in the little IR window in the front).

[seb30]

Hi,
the best ESD tool is the gaz lighter ! you can trigger device reset at 1 to 2 meter.
The manufacturer probably tested product for compliance with a lot of ESD protection diodes and
removed them to reduce cost on production.
By the way, on compliance test, this is the manufacturer who declare if the behaviour under imunity test is acceptable or not.
SeB.

[Fungus]

It just dawned on me that this may be the problem with a Eizo monitor I have at my workplace. It switches off and on seemingly random, worse on some days and other days not at all.
I resolved the issue by switching to display port instead of DVI.

One of my monitors blanks when I plug/unplug other devices in the room, maybe it's the same thing.

[BrianHG]

If it's a data error on the HDMI lines, usually you should get a glitch, not a screen on-off cycle.
if it is a glitch on the +5v sense line, HDMI pin 18, making the monitor think you disconnected/reconnected the PC, or, the PC power has cycled, this might make some sense.  Some HDMI cables might not shield this signal, or, some video cards feed 3.3v on this line making for a easily triggered low signal with any interference.

If you are using HDMI, try a low resolution like 1280x720 (if your screen actually will allow a low res mode...) and see if the screen still blanks out.  If it still does, the problem could be on the +5v sense line.  If so, you might need a thick HDMI cable which has an outer frame ground shield for all the conductors in the cable.

Personally, I still prefer the old DVI connectors with bolts.  I know HDMI is smaller, but, I have had less issues using the DVI plugs and cables, even when feeding HDIM compliant signals through the DVI connector.  I feel there is even a difference even if one side of the cable is on a DVI connector with the other side using a HDMI connector.

[wraper]

If it's a data error on the HDMI lines, usually you should get a glitch, not a screen on-off cycle.

With HDMI usually it will go blank before syncs with signal again. Sometimes image corruption without going blank is possible. Displayport always becomes blank.

not a screen on-off cycle.

It's not on/off cycle.

Personally, I still prefer the old DVI connectors with bolts.  I know HDMI is smaller, but, I have had less issues using the DVI plugs and cables, even when feeding HDIM compliant signals through the DVI connector.  I feel there is even a difference even if one side of the cable is on a DVI connector with the other side using a HDMI connector.

DVI does not work at 4k even at 30Hz.

[BrianHG]

If it's a data error on the HDMI lines, usually you should get a glitch, not a screen on-off cycle.

With HDMI usually it will go blank before syncs with signal again. Sometimes image corruption without going blank is possible. Displayport always becomes blank.

not a screen on-off cycle.

It's not on/off cycle.

When I say on-off, I basically meant the blackened pause.
If Dave was using HDMI, some of time when he got up from his chair, if there was data corruption, there should have been a few times when a bunch of pixels glitched up, along the exhibited occasional display going blank.  This is why I suspect the +5v sense line.  The syncs are embedded in the HDMI video data lines and if the data errors went through at that point in time, and a cheap monitor HDMI decoder cant deal with an error there, the screen will probably go temporarily blank.

One thing I haven't taken into account is that the monitor's HDMI receiver's cable length equalizer may respond to low frequencies and seeing an electrostatic pulse would cause them to mute down the receiver's amp for a long enough period to mute out enough lines of video where the monitor cant cope with so many lost h-syncs.  However, with a problem like this, almost every single time got up from his chair, the monitor should have been affected.

I don't have experience with display port, so I can only assume you are right about it's data corruption error artifacts.

[wraper]

there should have been a few times when a bunch of pixels glitched up, along the exhibited occasional display going blank.

Pixels glitching is hard to achieve even if you intentionally try to do this. Data corruption needs to be very very tiny for this to happen. It may happen with low quality or long cable at high resolution but not when transient happens.

[Red Squirrel]

We have this same issue at work!  I work in a NOC, we had 6+ monitors at one point, but got upgraded to 4k so we don't need as many.  Been through many PC upgrades and changes (normal PC life cycle) and even moved desks to a different part of the office.  No matter what, each and everyone of us has 1 monitor that will do this.  Even before the 4k ones.     It's always only 1 monitor and always the same one.  It seems to be triggered by any sudden movement, not just chair.  Ex: someone walking close, for me it's if I move my car keys.  It's very random.  We kind of gave up trying to figure out the cause because it seems to be so sporadic.  Every now and then it goes black and stays black and has to be turned off/on.   What's worse is if you turn off a monitor, it causes Windows to go haywire and shuffle all our applications all over the place.  My work place is the only place I've ever seen this happen. 

[nctnico]

Why does this 4K BENQ monitor mysteriously switch off on it's own?

Most likely because of the crappy video cable. It's not switching off, it just blacks out because of the received data corruption. Seen this crap myself with 3rd party cable which had the shield but it wasn't connected to anything.

Same here with a long video cable. I lost track of the standards but it isn't VGA. Switching on/off something which produces nasty spikes causes the display to blank.

[BrianHG]

  Ex: someone walking close, for me it's if I move my car keys.  It's very random.  We kind of gave up trying to figure out the cause because it seems to be so sporadic.

Hey, your monitors may have those old fashioned Ultrasonic Remote controls from the late 60s and early 70s.  I used to dangle my keys and the chiming sound would trigger a channel change, or, volume/power on and off....