CCFL Tube on Oscilloscope

[bostonman]

This is somewhat carrying over from another thread I had regarding a BNC connector question on a scope I'm repairing and kind of deviated with a dim screen.

It's a LeCroy scope with a very dim screen (screen brightness raised to the max setting). From my understanding, it's a cold cathode light (the screen has a model number: Sharp-LQ084V1DG21) and I'm getting some knowledge on this along with having disassembling it to figure out how it works.

It's an L shape bulb that sits on the edge of two sides of the screen. From reading and my observation, it shines line across the screen to light it. The CCL has a two-pin connector (I assume high voltage - more on the voltage measurements I took will follow below), and the screen has a ribbon cable (I assume all the screen data).

The two-pin connector plugs into a driver board (?) that's fed by a five-pin connector. From reading and brief discussions, the CCL can be replaced with LEDs which are cheap, but the driver board is expensive. My goal was to take some voltage measurements to figure out what feeds the driver board, and then research what is needed to drive a LED replacement.

The five-pin connector that feeds the driver board has two pins that are chassis ground, one pin is +12v DC, and two pins that I can't figure out. They don't have DC or AC voltage on them, I didn't see any waveforms when I scoped them, and nothing changes as I sweep the screen brightness.

The two-pin connector that feeds the CCL seems to have around 300V AC on it. First I used a high voltage probe because I expected the voltage to exceed my multimeter, but I saw 0v with the voltage probe. I resorted to using the multi meter directly and saw about 300V AC, however, the interesting part is that connecting probes to measure the voltage resulted in an extremely dim (almost blank) screen.

Even more interesting is that leaving only one probe connected (i.e. dangling in the air) resulted in a slightly brighter screen, but not as bright without any probes connected. My guess is the lead length is causing too much inductance and messing up the transformer impedance.

In any case, my questions are:

How does this screen work? Does anyone know what those two pins do? I expected to see a positive voltage on one pin (which I did - the +12v DC) and other pins varying as I change the brightness level (which I didn't).

My goal is to find out what load this driver board has and then find out if I can use a basic circuit to drive a replacement LED strip.

[moffy]

Jim Williams wrote the book on CCFL power supplies. Two of his app notes are:
1. https://www.analog.com/media/en/technical-documentation/application-notes/an65fa.pdf
2. https://www.analog.com/media/en/technical-documentation/application-notes/an49.pdf
you might find what you need there.

[moffy]

If you haven't seen it this is a service manual with schematics for a Lecroy Waverunner LTxxx scope: http://www.ko4bb.com/manuals/49.195.59.33/LeCroy_LTXXX_WaveRunner1_Digital_Storage_Oscilloscope_Service_Manual.pdf
page 8-55 shows a 5 pin connector that goes to the backlight they are as follows:
1. +12V
2. GND
3. Open collector enable signal
4. GND
5. Connected to a HC4051 and looks like it might control brightness.

[bostonman]

I haven't read the app notes, but did skim the schematic - both look like they'll be a great help.

If I had to guess, that open collector is for the screen saver to turn off the screen. Pin 5 is connected to R149 which is 0 ohms, so that's a bit baffling; and the fact I didn't see any signals on this pin while adjusting the screen brightness.

Maybe things will make more sense once I read the app notes (I'll do that later when I have more time).

[Ice-Tea]

Fun fact: I 'repaired' a few of these without any LEDs. Sometime, the reflector foil behind the CCFLs coils up, obstructing the light from the tubes rather than reflect them. Even removing the foil improves brightness considerably.

On a side note: I don't know which lecroy you got exactly but there's quite a few that have fragile front bezels. Be carefull with those.

[BrokenYugo]

The CCFL driver typically just spits out straight high frequency AC (20+kHz) from the transformer more or less directly into the tube with some really basic regulation of the primary current/power, voltage sits wherever it sits, unit will likely draw arcs unloaded, not a suitable LED driver, similar concept in that they're both current regulated but with very different voltage/current levels. Depending on the DMM this may be past the high frequency roll off point. I wouldn't recommend poking around the secondary side if you don't know what you're looking for, especially using anything of value.

[bostonman]

It's an LT344 - I got it for $100

What exactly is fragile about the front bezel? If you're referring to standoffs breaking off, you're a bit too late on that helpful tip.

I have read about the foil being an issue and good to know it can be removed without any issues.

not a suitable LED driver

My plan (assuming removing the foil and/or the foil isn't blocking anything) is to replace the driver board with a custom board. At the moment I'm uncertain how much current is being drawn, so I'm uncertain if a replacement board will just be simply an adjustable regulator that maybe I can use a pot to set the brightness rather than an elaborate board that is software controlled.

[Ice-Tea]

Just about everything? If you apply to much force, the entire thing breaks. Especially likely if you forgot a screw and figure "surely it will click out when I try a bit harder".

Ask me how I know.

Not sure about this exact model but I think the bezel is secured with screws on the bottom of the bezel? Screwing them in too hard sometimes also breaks the bezel, especially if something is not lined up 100%

100$ is pretty sweet, though!

[bostonman]

Disassembling this wasn't easy. The front panel (thankfully?) doesn't screw in from the bottom, but from the rear. If I remember correctly, a screw in each corner and one or two in the middle.

At one point I was trying to remove one of the corner screws and the screw kept turning and turning. Either I cracked off the standoff or it was broken which I discovered after removing the front.

This model LeCroy is much older than I initially thought. The guy was selling it for $100 and claimed the "only" issue was channel 1 BNC was broken (more like bent). I suspected other stuff was bad, but thought the $100 gamble was well worth it. Funny enough, he claimed to have others interested, and, not wanting to wait for a more convenient time and risk losing, drove about 40min in snow. Being four channels is a nice addition to my equipment, but thought being a LeCroy, it was new enough to have a USB so I can save waveform data.

Unfortunately it has a 3.5" which I guess can work if I buy a USB to 3.5" and transfer saved images from a 3.5".

I will be careful of the front bezel - thanks for the heads up.

[nctnico]

The two-pin connector plugs into a driver board (?) that's fed by a five-pin connector. From reading and brief discussions, the CCL can be replaced with LEDs which are cheap, but the driver board is expensive. My goal was to take some voltage measurements to figure out what feeds the driver board, and then research what is needed to drive a LED replacement.

No, the LED strips come with the driver board. Just take the CCFL driver board out and replace it with the LED driver board. The LED strips can typically be cut to length but this needs adjusting the current on the driver board by changing some resistors.

But simply changing the CCFL tube might be an easier option. These are mounted in a cartridge that can be pulled out of the TFT screen. These could be locked in place with a screw or a metal slip that is bent to retain the CCFL cartridge.

[bostonman]

No, the LED strips come with the driver board.

Unless I missed something or I didn't look hard enough, the LED driver board I found was in a kit that cost $150.

[bostonman]

If you haven't seen it this is a service manual with schematics for a Lecroy Waverunner LTxxx scope:

I'm not understanding the mapping on the schematics. Earlier I measured all the pins on the HC4051.

The three address lines (A, B, and C) are all stuck at 5v. X0 through X7 are all stuck at 0v, and, as previously mentioned, the I/O (pin 3) is 0v.

Working backwards to IC85, the three Q outputs sit at 5v, the D inputs are clocking, and the CLK and CLR sit at 5v.
 
Although HC4051 doesn't seem to be working correctly since X0 - X7 are all stuck at 0v, Edit: I may have thought about this wrong. The common I/O (pin 3) I assume is sitting at 0v because all X0 - X7 are on and they are just resistors without any voltages.

I figure I'd trace the CLK and CLR on IC85, but I'm not understanding the mapping. Using the easier one (CLK), the offsheet connector has a '3' next to it along with a node name of 'NPORT5W'.

I'm assuming this goes to the sheet with the '(3)' - page 8-7, however, I'm not finding NPORT5W.

Am I interpreting the mapping wrong or missing the node name on page 8-7?

[floobydust]

Looked good until the price...
I would hit up CCFL Warehouse and see what they can offer.

[Ice-Tea]

No, the LED strips come with the driver board.

Unless I missed something or I didn't look hard enough, the LED driver board I found was in a kit that cost $150.

They're dirt cheap:

https://www.ebay.de/itm/354236208521?itmmeta=01HRVCTJRBF4DVDSB0ZMRP54Z4&hash=item527a1f9589:g:kuMAAOSw33Vi6R8N&itmprp=enc%3AAQAIAAAA4LYcjMTzaj7mpB04M9RIg6ann62hn1XhB31GUo3JVjQKI65wwa%2BOsYK32tKY2SnpK%2B8TlUqSDjFUgxLs0xRbDyrWji%2BKubb3X%2BceZr7rDhOGlKiVGHX%2FuclFXIcgqxAr9TRriYJ%2FugSDuwuburH6R80ZYrLv%2F73M6vMlcW%2Fc03DBXG%2Ff6iDQVWpOigZS5BonlbhbUCcfnY4wV%2FS%2FIhzdLAS14I9w8ZWmY3h4Nv3JOE2Axpqw26L1tvcrP8TUhCr02xgPtWRX9dHdxYFLnNbTWgp2iz3RGIb9kUvxdZXGziN%2B%7Ctkp%3ABFBM1qzq7MZj

[bostonman]

Looked good until the price...
I would hit up CCFL Warehouse and see what they can offer.

Hahhaha that was my statement as well. I'll look at CCFL Warehouse and the eBay stuff looks good too.

I haven't had a chance to read through the links above (app notes and the write up on CCFLs), and maybe the answer is in there. At the moment I'm wondering why I'm not measuring activity on CLK and CLR which is indicating the logic isn't clocking the chip that controls the brightness, however, the brightness knob does work.

Since the eBay (and maybe CCFL Warehouse) has cheap prices, maybe I'll just buy one, but I'd like to understand the logic better before I just swap.

[bostonman]

CCFL Warehouse had the strip, but not the inverter board (which seems to be the expensive part).

I ordered some from eBay, but they are shipping from China and will probably be a few weeks or so.

Referring to the schematics in the service manual, does anyone understand what the numbers mean? I know I asked this already, but I spent time tonight reading them and trying to locate the connections on other pages. Page 8-55 shows NPORT5W with a 3 next to it, and I found NPORT5W on 8-53; not 8-3 as I expected.

So I'm just trying to figure out the nomenclature.

[moffy]

Referring to the schematics in the service manual, does anyone understand what the numbers mean? I know I asked this already, but I spent time tonight reading them and trying to locate the connections on other pages. Page 8-55 shows NPORT5W with a 3 next to it, and I found NPORT5W on 8-53; not 8-3 as I expected.

So I'm just trying to figure out the nomenclature.

Take NRESET on 8-55 it has 1,2,3,... The first drawing in this set of drawings is 8-49 which occurs after the MainBoard PCB layout. 2 is 8-51, 3 is 8-53 and 4 is 8-55 etc. I think the schematics are grouped by PCB.

[bostonman]

I think the schematics are grouped by PCB.

This would make sense.

When I finally found the other node, I thought it was odd it was on 8-53 when it had a 3 showing on the schematic, so I thought maybe they are just eliminating the 8-5, but then I questioned how would someone know it's on 8-53 and not 8-43 or whatever.

[bostonman]

If you haven't seen it this is a service manual with schematics for a Lecroy Waverunner LTxxx scope: http://www.ko4bb.com/manuals/49.195.59.33/LeCroy_LTXXX_WaveRunner1_Digital_Storage_Oscilloscope_Service_Manual.pdf
page 8-55 shows a 5 pin connector that goes to the backlight they are as follows:
1. +12V
2. GND
3. Open collector enable signal
4. GND
5. Connected to a HC4051 and looks like it might control brightness.

The LED strips arrived (I bought four) including the converters. Using a bench power supply, I tinkered with the LEDs to figure out what needed to be adapted into the LeCroy scope to incorporate them.

On a side note, knowing the converter board needs to be adapted to the exiting connector in the scope, I cut the wires to perform some bench testing. I didn't realize I flipped the connector ends (the end that was connected to the board was now the open end that I cut) and resulted in the red wire being connected to ground and the black to Vcc. It took several minutes (and two boards) before I discovered the input voltage was reversed due to plugging the opposite end.

I'm extremely confused over this scope. As listed above, the pins on the existing board are correct and pin 3 needs to be pulled low for the screen to turn on (I assume this is something to do with screen saver or whatever - but I verified that open and the screen is off, pulled low and screen is on).

Pin 5 has a few milli volts on it regardless if it's open or connected, the voltage doesn't change with adjusting the screen brightness (although the screen brightness works because I can dim the screen to off all the way to 100% brightness), Using a current meter, I tried measuring current and didn't see any fluctuation (although maybe I should have used a Fluke in series), etc...

After trying to measure voltage, current, and changes in both while adjusting brightness, I removed the CCFL so I could look at the lamp while adjusting the brightness, took the LED strip and let it shine down the screen (although it wasn't perfect because I rested it on top, the image was more than visible enough to see the screen), and adjusted the brightness. The CCFL didn't change brightness at all.

My confusion: I thought screen brightness was caused by the CCFL changing intensity and this doesn't seem to be the case. My other confusion is what the heck is pin 5 doing? The screen remains off if pin 5 is blank, but I don't understand the need for this circuit.

If I had to guess, I'd assume a brightness pot exists somewhere, or a menu setting exists somewhere that I'm not locating, however, I haven't see anything that points me to a pot or a menu setting.

The replacement LED strip needs some work to fit into the scope; nothing too difficult. It needs +12v (12v to 30v) which the scope provides, ground, 0-5v for brightness, and >2.7v to 5v for enable.

From current measurements I took, the 12v needs almost 1A (scope should provide that) but brightness and enable take very little current which makes things much easier.

[bostonman]

I haven't located any other brightness settings, so I'm uncertain how brightness is being controlled or the purpose of that circuit.