Recapping a Billabs BL27CB0P

[the_77x42]

I'm doing a capacitor replacement on my Billabs BL27CB0P. I'm going to document the process here for anyone else intending to do the same.

Ever since I got the monitor I noticed some squishing around the edges. In sidescrollers, it's like the image is being wrapped onto the screen as if it's being pulled off a roll of paper towel. Although I learned to just live with it over the last ten years, the last time I opened up the back of the cabinet I noticed a couple bulging capacitors. It's time for an overhaul. Hopefully it fixes the edge squishing, but if not, at least it will hopefully last another ten years.

The BL27CB0P is a 27" multisync monitor and there is little information about it online, but the original product page is archived here. Although it's a rebranded Wei-Ya, I can't find the original Wei-Ya model number. As such, I've had no luck finding schematics online.

I've setup a shared Google Drive folder with disassembly images and a capacitor inventory. The inventory sheet has tabs for the Mainboard, Neckboard, and VGAboard. I'm getting help from Peter at arcadepartsandrepair.com. He's going to be the one setting up a cap kit for me.

This will be my first cap kit installation, so I'm excited. Well, as excited as I can be for something that has 72 caps... Yikes. Stay tuned...

[the_77x42]

The first step was discharging the monitor and removing the board. . My tool of choice was a long screwdriver with alligator clips and solid core wire. One thing I did just to make sure everything was connected was do a continuity test from the tip of my screwdriver to the chassis. Everything came apart without issue. I took all the pictures in the post above so I had a record of how everything went back together. Since no cap kits for the monitor exist, I needed to take it apart and inventory everything. It will take a week or two before the kit arrives, so in the meantime I had to stock up on some tools...

I grabbed a Hakko FR-301 desoldering tool from tequipment.net. I figure I can do a cap kit install in some friends' machines as well. It's pricey, but it should last me a long time.

I also ran a voltmeter on every cap to make sure they were all discharged. I tested the capacitance of a few on the board and I noticed that one of the bulgers (C30) wasn't reading any capacitance... Maybe that one is the main culprit!

Stay tuned...

[grantspain]

surely this is a wei ya 3129 chassis or maybe a rodotron 666 chassis
in which case a massive amount if info is out there for these
in regards to your side compression this well documented on various forums, it is down to the value of the hv caps in the horizontal section

[the_77x42]

surely this is a wei ya 3129 chassis or maybe a rodotron 666 chassis
in which case a massive amount if info is out there for these
in regards to your side compression this well documented on various forums, it is down to the value of the hv caps in the horizontal section

Thanks for the leads! I was looking at the schematics for the 666 and unfortunately they are not the same. I had hoped that I could fix the side compression as documented here, but C86 on my chassis is a 1uF electrolytic radial.

I've also been looking at this thread, which describes my exact problem and attributes the horizontal compression to the VGA signal from my graphics card. I've setup my GroovyMAME profile based on a D9800 and have been meaning to experiment to get the exact monitor specs, but have not had the time. Regardless, since caps are failing on the chassis, I'll replace those then dive into the monitor profile.

[grantspain]

do you have a photo of the chassis, if its not wei ya then perhaps its a rebranded kortek

[the_77x42]

do you have a photo of the chassis, if its not wei ya then perhaps its a rebranded kortek

It's still mounted in the machine. All markings on it are Billabs. The photos of the disassembly are here, maybe those could help?

[grantspain]

pretty sure that is a wei ya m2929dc

[the_77x42]

pretty sure that is a wei ya m2929dc

That's what I thought as well, but I've been working with Peter at arcadepartsandrepair.com on the cap kit and after going over the M2929 schematic, they are very close (85%), but slightly different.

[the_77x42]

About a dozen capacitors in and I have a few general observations:

• Don't attempt to do this without a desoldering tool like the Hakko FR-301. It will be next to impossible for 72 caps.
• Add a small amount of solder to existing joints and the Hakko will perform even better.
• Tin the tip of the desoldering tool between each use to extend tip life.
• After removing the existing cap, apply real RA flux to the pads to make soldering in the new cap a breeze.
• If you're attempting to use desoldering braid on a difficult to reach component, soak the braid in flux first.

For this monitor specifically, the hardest thing to desolder so far was the RF shield on the neckboard. I removed as much solder from these small tabs as I could, but in the end I just heated the pad under the board released to get rid of the tiny amount left holding it together. By far the most difficult part to desolder was the tab on the bottom of the shield. I had to crank the desoldering tool up to 500 degrees with the largest tip and leave on the pad for almost ten seconds until the solder underneath the tab melted.

The VGA board was a cinch and the neckboard is almost done. I will post again when I begin the mainboard.

[the_77x42]

I decided against removing the flyback from the mainboard. It's a finicky operation since it's attached to a large heatsink. To support the board while it's upside down, I used an empty box and carboard shims to raise and level the heatsinks at either side of the board. The weight of the heatsinks is enough to snap the board if only one end is picked up, so I'm very careful to use my hand and forearm to support the whole board as I flip it over. My method is to desolder at least four of five caps before flipping the board over, minimizing handling time.

The large capacitors are actually easier to desolder than the small ones. I would recommend desoldering those your first time. I used a voltmeter to check the voltage of the caps and held a 10K resistor to them if any charge was left.

In some of the pictures you can see the nasty, yellow factory glue used to keep the caps in place. I had initially researched removing the glue using IPA (isopropyl alcohol), but after speaking with Peter, his suggestion was to just leave it. After desoldering the cap, it easily detaches from the glue just by carefully twisting and rocking it around.

Next up is the B+ filter and reassembly!

[the_77x42]

The B+ filter was super easy to install. I reassembled everything, triple checking all my pictures and...... a pure white screen. Panic. Then the realization that I must have bumped the screen pot on the flyback, so I adjusted it down and things are perfect. Well, the edge squeezing is gone, but the corners of the monitor are still very blurry.

At lower Screen pot settings, making the picture darker, the focus is much better, but adjusting the top two pots on the flyback can't perfect anything other than the very center of the monitor. Leaving the screen pot darker isn't viable because the brightness/contrast settings don't go high enough to compensate. It's been like this since I bought it, however. Hopefully I'll find an adjustment at some point.

On the board I have:

• B+ adjustment
• HV adjustment
• H center adjustment

[the_77x42]

I wound up fixing my focus issue by doing the following:

• Adjusted the convergence rings to perfect convergence
• Adjusted the three pots on the board above the yoke to align the corners better
• Increased the Screen/G2 pot on the flyback and lowered the brightness and contrast on the OSD
• Adjusted the focus pots on the flyback

The result is the best the picture has ever looked. There is maybe a one-inch square blurry patch in each of the corners, but it is still readable and not noticeable in game. I'd say this whole project has been a complete success.

[the_77x42]

Well, I had hoped my recapping would last ten years and it didn't even last eight. The horizontal linearity problem at 15kHz is back and worse than ever. I actually have a bit more experience with board rework now, so I'm hoping to fix this issue once and for all.

After some digging, I'm fairly certain this chassis is a Wei-Ya C2929D-6 (rev 6. that is). That should be the 15-50kHz model. I can't find any schematics for it though. With a similar chassis and problem, changing a capacitor value was able to fix the issue for these folks. Unfortunately, that caused other problems at other resolutions/frequencies.

Since I replaced the board with all Nichicon caps, I'm fairly certain they're still good. Thinking back, I'm not sure the cap kit fixed the right-side compression fully, just made it a lot better. I'm thinking now that it's possible it's the poly caps that are suspect.

I should note I was able to get into the advanced operator's menu by holding UP on the MENU button while the monitor powered on. That had an HL adjustment, which didn't do all that much, but the menu also wasn't available in 15KHz mode. Strange, just the standard menu was available. There's actually no way to H-POS the monitor all the way to the right. There's always about a 2-3" black bar and the image will wrap around itself.

[the_77x42]

I finally had some time to look at this.

R74A was measuring at 61Ohms and it should be 68 Ohms. I also noticed it had heat damage, so its replacement will be 5W instead of 2W. I also noticed a couple of the Nichicons that went in previously were a tad leaky, so they'll be replaced. I'm also replacing all the film caps with high quality Panasonic parts.

The main culprit looks to be a fried MOSFET Q31 in the horizontal section. All three pins were shorted and there was some heat damage. I tested all components in the region and they were fine, but I'm putting in some better ceramic caps (C73) and a higher quality MOSFET.

Fingers crossed!

[the_77x42]

Unfortunately, replacing the obviously bad MOSFET, the 61 Ohm with a 68 Ohm, most of the film caps, a few leaky electrolytics, and a couple bypass caps in the horizontal section didn't do much. I'm leaning towards a full recap. Again. I've tested the diode in circuit and the resistors close by the MOSFET. The only thing in the horizontal secion I can't test are the small transformers. They have a unique part number that I haven't been able to find:

LL-224-023-00
and
LK-228-050-00

Does anyone know what these are?
 
EDIT:
I did a bunch of more testing. This monitor is a strange beast. At any resolution other than 400x254, even 15KHz, I can get the pro menu (hold MENU UP when powering on the monitor) and those resolutions do not have any horizontal linearity issues. The screen compresses slightly in all four corners but it's not that noticeable. However, in 400x254, I can't get the pro menu, just the standard menu. I thought maybe the pro menu would be helpful here because it does have a horizontal linearity adjustment, but it never adjusts the right side of the screen, only the left (again though, this menu isn't available in 400x254). Also, in either menu, I can't seem to RECALL my settings. I think I was able to before.

Since I replaced those film caps and MOSFETs, I notice the picture no longer can wrap around itself when I do the horizontal positioning. I also actually think the linearity is slightly better.

Unfortunately, that MOSFET was tied into the proprietary IC on the board. If the MOSFET fried, I wonder if it took out the IC as well. That particular pin on the IC must have been for something like this 15KHz mode, because it's not used on the 30KHz variants that I have a schematic for. I almost get the feeling it has to do with turning on those S-correction caps.

I might give this thing one more go and replace the rest of the caps I found that were slightly leaking. I'll check the diodes and transistors as well. Maybe I'll try and identify the IC. Unfortunately, this is all close to exceeding my skillset. I don't have an oscilloscope and don't feel comfortable enough working on monitors while they're powered on.

EDIT 2 (FINAL, SUCCESS):
- I gave it another go and replaced 2/3 of the high voltage film caps (DigiKey didn't have stock).
- I also reflowed the solder on the small transformers and large coils.
- I also tested and replaced a bunch of the aluminum electrolytic caps, with a few in the horizontal section. I replaced them all with Panasonic FR caps, which are extremely low ESR.
  Many of the caps I tested were approaching or above 1 Ohm. They were all Nichicon caps from my first cap kit.
- I replaced C91 (2000uF, 35V). Despite it testing OK, it gave off the classic leaky fishy smell when I desoldered it.
- I tested all resistors, transistors, and diodes in the horizontal section. As far as I can tell, the blown MOSFET didn't affect anything else.

I wasn't too optimistic after doing all this work. And despite the Recall feature working again and the picture being absolutely amazing, at that one screen size, there is still significant right-side compression. I think this monitor might have been this way from the factory. One last ditch effort was to change my MAME configuration. Low and behold...

First I tried adjusting as much as I could in ArcadeOSD, but nothing seemed to help.

I'm running GroovyMAME 0.178. I do this because it has the best support for the old ATI card I am using. I turned off modeline_generation in mame.ini so GroovyMAME would use the modelines generated by VMMaker instead of generating them on the fly. Despite the two pieces of software having the same monitor configuration, the generated modelines for the Midway games were different... 400x256p @ 16.7KHz instead of 15ishKHz. I couldn't believe it too when the geometry at that modeline was ABSOLUTELY PERFECT. I mean, it's bang on now! I tried a bunch of different games and they all had perfect geometry as well.

So:

1. I have no idea what that MOSFET was doing.
2. All the components I replaced probably did nothing more than extend the life of my chassis.
3. My geometry issues appear to have been fixed in software.

In short: success?

[the_77x42]

After playing a few more games, I notice that I actually have VERTICAL linearity issues in 15KHz mode now. The centre of the screen is compressed where the top and bottom are fine. I had noticed this prior to the recent repairs as well.

I don't know why this only happens in 15KHz mode, but I'm thinking of just changing the monitor profile so that it only goes as low as 16KHz. That and above seems to have perfect geometry. Maybe there's something wrong with an IC? Maybe there's a coil or transformer that's out of spec just enough that it only affects that one frequency? I'm not sure. It seems like in 15KHz mode I either have horizontal or vertical linearity problems.

Without a schematic and without an oscilloscope, I'm happy to just Band-Aid it in software. I wish I went with a D9200 or similar all those years ago, at least I'd have better support resources.

EDIT: Wow, I fixed the vertical linearity problem with the PRO menu (hold MENU DOWN when powering up the monitor). It seems I am able to get that menu in 25KHz mode and that was the mode that had the worst linearity problems. What a ride.