Plasma Panel experiments?

[electroyeti]

Hi everyone,

This is my first public posting to this forum so "Hi" to everyone and sorry if some of the following questions seem a little "dumb" on my part.

I am a semi retired ex-tv service chap here in the UK who has spent many years in electronics both as a hobby and as a paying job. I enjoy "tinkering" and recently had cause to scrap an old LG 50PX4D plasma tv but as part of the strip-down and recovery of components I found myself curious about the general operation of plasma displays.

Following some research on the Internet I stumbled upon a posting from back in 2013 entitled "How a plasma display panel (PDP) works" on this forum by tom66 and was very impressed by the detail. I also read an application note from the International Rectifier company (an-1088) both of which have gone some way towards improving my understanding of plasma display technology.

As part of my tinkering and to satisfy curiosity I should like to light-up a few pixels on the old panel. I could just use "brute force" and connect a few electrodes at random to a tesla coil but aside from a few coloured "flashes" from the display I dont think I would learn very much :-)

So my plan is to apply a few static DC Voltages to selected panel pins via current limiting resistors to light-up a few pixels but a few things still "bug" me regarding the operation of plasma panels. Reading the IR application note the wording implies that the panel pixels need to be driven with an AC waveform in order to remain illuminated, is that correct or can a pixel be turned-on and remain lit with only DC Voltages applied?

The application note also describes a sort of generic plasma display cell with five electrodes per pixel, x,y and three "address" electrodes which appear to be for the R,G,B primary colours. Looking at my panel I am a little unsure as to which connections peform these functions. At first sight I would have said the panel does not appear to have enough electrodes to support the number of pixels required for a HD display but I am obviously missing something. The only connections I am certain of are those to the "Z-Sustain" pcb which appear to be three broad copper strips. Would these be the R,G,B "common" electrodes perhaps?

In short, I supose I am asking what would be the theoretical minimum connections required in order to light-up a selected pixel or group of pixels and do plasma display panels contain any "in-built" electronics for addressing the pixels or are they just gas-filled cells addressed by a matrix as the application note implies? Would it be safe for the PDP to simply connect a few DC Voltages between pins (with current limiting) in order to de-mystify the connections and light-up a few pixels?


Regards,

Des (electroyeti)

[tom66]

Re: Plasma Panel experiments?

The first thing to understand about PDPs is that they are bloody complicated beasts. In fact I still find it amazing that of all the technologies that were competing in the flat-panel display market these were somehow the least complex or easiest to implement (until LCDs came about.)

You can consider the AC-PDP, which is used in all plasma TVs (no plasma TVs use DC-PDP; some early plasma monitors do, but only very small sizes), to be a type of memory device. Data is programmed into the panel by means of the address lines (red, green and blue along the bottom of the panel: for a 50" 720p (~1365x768) display, there are over 4,000 address electrodes) and lines are selected using the Y-buffer. Data is programmed onto each line during the ADDRESS (or SCAN) phase of the cycle.

When data is programmed into the panel, it is a state of either being "on", which means the pixel will discharge when sustained, or "off", which means the pixel will not light. There is no inherent greyscale capability. Each line is programmed into the panel sequentially so for a 768 line display, all 768 lines need to be programmed with 1365 horizontal pixels of data. This write happens within 0.1ms, which is to say, damn fast.

The purpose of the Y-sustain is to provide one half of the sustain waveform, and also provide the -Vy erase/initialise voltage (-200V) and the Vsc scan/write voltage (150V, relative to -Vy). The Y-buffer provides a separate scan signal to each line of the panel - hence the complexity of the ribbons connecting it. The purpose of the Z-sustain is to provide the common sustain signal, which is the other half of the sustain waveform. It is applied to the "Z" electrodes of the panel (sometimes called X), which is common for the whole panel. The different Z-sustain ribbons are not responsible for R, G and B and in fact the set will usually run just fine with one Z ribbon connected although this may lead to the FPC cable overheating. The Y- and Z- operate as a pair of boards that fires the sustain voltage at 180 degree phase shift. So when the Y is driving it up to Vs (~200V), the Z is driving it to ground (0V), and vice versa. The cycle is generally repeated at around 100~200kHz. The Z-board also provides a voltage called Vzbias which provides some bias on the panel during erase and write.

After the pixels are written a high-voltage alternating current is applied to the panel. This alternating current causes cells with wall charge programmed into them to light up. To get greyscale levels, varying on-time widths of sustain are used, and different cells are programmed on each cycle, to create exactly the right amount of light from each pixel. Dithering and error diffusion is also often applied to increase the effective bit depth. A common, budget PDP will be written at 600Hz, for a 60Hz input image. So that's 630 million pixels per second. Pioneer 9G plasmas use 28 subfields and a 100Hz refresh rate, so the 1920x1080p panel is written 2800 times a second: 5.8 billion pixels per second! Now you can get some appreciation for the complexity of the hardware to control all of this!

The panel must also be reset and initialised on each cycle, which is performed by means of ramp waveforms. These ramp waveforms can cause a discharge due to the constant-current behaviour of the gas in the panel. They are responsible for the light background glow of the panel when running, and have been a target for manufacturers to reduce as much as possible. Pioneer and Panasonic have achieved excellent, low background levels of light by precisely controlling the waveform applied to the panel, and getting the gas mix correct. Whereas LG doesn't usually bother so their panels have black levels that are pretty high. (LG, since 2009 or so, had always seen PDPs as budget large screens, rather than the high-performance displays they were. Correspondingly they put little R&D into them past this date.)

The way an AC plasma panel works is that typically a DC voltage of even a few hundred volts will not cause a discharge. The reason is pretty simple. It is essential to the write process that cells that have not been written do not discharge, or the intensity of light across the panel will be affected. A design flaw in the reset waveforms of some LG plasma display panels causes an effect known as "maldischarge" where pixels will spontaneously discharge under operation, or will fail to be completely erased on each cycle. This is usually fixed by upgrading the TV's firmware which will change the drive patterns used.

And... now you understand that...

Unfortunately it is quite hard to get a PDP to light without supplying high frequency AC to it. If you applied around 300-400V DC to the Y and Z electrodes you could probably get some glow discharge across the whole panel. You would need to find a way to write to the cells, via the address buffers, to get actual colourful light out of certain pixels. This would be difficult. I can't think of an easy way to do it. You could try treating the panel as a DC-PDP and apply a strong voltage across the Y and A (address) electrodes, however, trying to connect to just one of ~4000 lines along the bottom will be quite hard. If you apply a strong discharge to the address ICs, you would probably be able to get some of its outputs to stick to the high Va voltage (~60V) leading to "stuck lines". It would then be a matter of discharging voltage along the Y electrodes and you might be able to see some light from the panel.

[electroyeti]

Re: Plasma Panel experiments?

Hi tom66,

Thankyou for taking the time to present such an in-depth and educational response. From the texts I had allready read it seemed that plasma display technology was indeed complex but after reading your fascinating response the word "complex" seems like an understatement :-)

I found your response very interesting and will have to read and re-read it a number of times for it to fully sink in. Having said that, just reading your post and looking again at the pcb's and PDP ribbons it all makes much more sense now. All understood regarding the reason for AC operation of the cells, I see now why its done the way it is. It also makes clear the reason for PWM control of pixcel brigtness.

Even if I never get a flicker out of the panel its a fascinating excercise just studying how it works. Sadly, here in the UK plasma displays are viewed as not being very "green" due to the high power consumption of early models and EU directives which have effectively killed the import and sale of plasma displays.

BTW your explanation goes some way to explaining the faint glow visible even when the picture was meant to be dark. It aslo explains why (when the Z-Sus panel failed a few years ago) a faint image was still visible on the screen.

Thanks again for taking the time to reply and in such comprehensive detail. Much appreciated.

Kind regards,

Des (electroyeti)

[tom66]

Re: Plasma Panel experiments?

Plasma TVs were killed off by a few factors.

LED/LCD TVs had dropped in price by too much. Manufacturers like LG and Samsung had relegated most PDPs to the budget end of the spectrum and they just couldn't compete. Panasonic was being outpriced by these manufacturers when they were trying to sell their premium, high-quality offerings. They were the first to drop out in 2013, and LG/Samsung killed it off in 2014.

Contrary to popular belief there was no EU regulation prohibiting plasma TVs from being sold - they just had to be marked with a lower energy rating. Panasonic's 1080p plasmas got only an energy rating of "C", compared to most LCDs achieving "A" and "A+".

Another factor was how expensive it would have been to develop proper 4K panels, which was reasonably easy to do with LCDs.

Finally their brightness wasn't great in a brightly lit store, and you can't see black levels well. They were for the discerning customer - the average shopper went for price.

OLEDs are the true successor but have a few issues like lifespan to improve upon before they become a product I can fully recommend. For now I'm quite happy with my 2011 and 2012 Panasonic PDPs which have excellent image quality, low black level, and relatively low power consumption.

[electroyeti]

Re: Plasma Panel experiments?

I had no idea the EU regulation was just a "myth", at one point it was even mentioned on the news warning that plasma tv,s would vanish from showrooms within a year.

Just out of interest, we replaced the BER plasma with another LG but this time an LCD 60"
Not perfect but certainly the blacks are darker than the old plasma set and it uses a fraction of the power. But to be fair it is almost 12 years newer :-)

electroyeti