Original PC Engine (White) RGB Amp Installation & Overhaul Guide

[blueraven]

Hello again!

Over the past few days I've been working on installing an RGB amp into an original (white) PC Engine #PITG001, and have run into several difficulties and differences from the other PC Engine/TurboGrafx 16 systems. There is also no repair thread associated with the Original PCE, So I figured I would use this opportunity to explain some of the differences between the Original PCE and the Coregrafx/Coregrafx II. Credit is due to thesteve for constant hacking/guru help in chat literally every step of the way and telling me what to do, chop5 for counterpoint, and BlueBMW for the Coregrafx chart that I will reference later.

Day #1:



While digging around looking for a file to clean up a casemod, I came across this Original White PCE that I had forgotten about in a box with some PSP stuff. I plugged it into my IFU, and I had no power, so I pulled it apart and began the journey of tracking down the problem... The PCE would pulse in the IFU like crazy.



Opening Up the case, everything looked DECEPTIVELY clean, with the exception of the clunky old RF port, (top right on the chipboard) that looks like a piece of 8-bit tech from a William Gibson novel. I then decided to check the underside of the unit for any obvious problems.

I found the grounding plate soldered on two points that lifted clean off with the touch of an iron. I then was looking at the bottom of the board:


 
I ran the wires for the RGB amp off of the Expansion Bus, following the RGB Amp guide on TG16PCEMODS.com, and powered the system on, to check the voltage to the BUS. I then did the Jailbar Fix, installing the two blue caps on C131 and C135. Everything read correctly.



The Next step was to remove the cyberpunk RF port...



...3 solder points on the bottom that go a heck of a lot quicker with a fine-tipped iron and some flux. I don't have a de-soldering station and didn't have a torch handy, so i just took my time with a wire braid and took it all out. It is important to note that the three holes that hold the RF port in are ground, so you should re-fill them with solder when the port is removed for easy access to ground when installing the RGB chip or Region Mod. The back of the port is connected to three tiny pins, which are video, sound, and ground. desolder them from the bottom of the board after you get the 3 main points off.



This felt like a good logical time to stop for the night.

Day #2:

Aargh. That was fun. But at least that damn RF port is gone now and you can finally open up the systems true potential! The RF signal is nothing short of terrible, and while I have been playing games on composite for all these years, I wanted to get the eye-popping colors i used to see in arcades when I was younger... so on to building the 8pin DIN connector! When you assemble it, make sure that the R, G, and B signal are separate, using a thin-gauge wire (30AWG Kynar is recommended, but I didn't have any).



It looks like Spiderman! Now that the switch is assembled, I Remembered the mounting issue, and opened up the PCE and stared at it for awhile. The gaping hole leaves a lot of mounting questions (get your mind out of the gutter ), but the real trick here is that the LEFT front ground pin on the DIN8 centers the plug perfectly in the system.



I assembled the completed switch and cut ALL of the pins except the two grounds: The left front pin, and the center pin for the DIN8. I then soldered the left front pin to the front RF hole (above), and hot-glued the switch into place once resistance was confirmed at the plug... and then came the wierdness. Checking the plug, I realized that It was putting out 9V instead of 5, and that there was absolutely no power past the power switch. It just...stopped at a point under the heatsynch... It was the fuse!! Without the parts, I put the project away for the next night.

Day #3:

...I thought it was over, and I went to test, and..... NOTHING. Then I realized I had the grounds incorrect, so I ran them correctly, then made a quick trip to the electronics store to get the 1uf 50v caps and the fuse that blew the circuit out. I disassembled everything the 7805, fuse, heatsynch, and caps, and replaced the fuse. An open circuit yielded R, G, and B flashing in repetition, so I knew the DIN was correct! I then installed the caps.



...And then something went horribly wrong. I was trying for three days to get this thing working with steve and chop's help, and i didn't even notice that the grey CABLE WAS COMPLETELY DISCONNECTED. I made 20 cuts and stripped the wire, and desoldered JP101.



Then I reattached all 21 points on the card-side of JP102.



Almost immediately after I did this, the other three cable connections snapped off in the same way  . This put me at 21/84 connections on the cable, so I did the most important thing I could do, and unplugged the iron for the night and poured myself a stiff drink.

Day 4:

Gearing up for what is one of the most annoying repairs I have had to do, I got plenty of coffee and and prepared for soldering about 100 connections. There were at least 76 more to be done on the cable side of things so I prepped the main circuitboard, and removed all of the soldering slag and bits of wire. I then cleaned the area with alcohol to make sure there was a clean surface...



...and then followed suit with the second cable on the Hu-Card slot. It was already disconnected so it was pretty darn easy to completely desolder JP102. JP 101 was attached from yesterday and holding on pretty well.



I cleaned the boards again, and positioned the unit in the casing of the PCE, and changed the batteries in my headlamp to get the best view of the repairs. With a needlenose pliers I bent each of the pins at a 45-degree angle and they went into the main board clean, save for some stuck stands that took time to work through. I then bent the pins back and soldered the wires from the bottom of the board. Cable JP102 went on clean.



Instead of immediately reattaching the bottom of the cable, it was time to do the oBey mod to save an extra disassembly in the long run.

I had the Hu-Card slot apart, and after looking at it I thought "I could finish a region a region mod on this f*cking thing in an hour or so, everything is already apart, and for the first time I don't have to dig through tons of wires to cut the pins and solder through the board!" I grabbed the remaining oBey chip I had from turbokon, and did a quick install that lasted just over an hour, the chip went in clean, and I used purple wire for PCB side, and orange wire for the Hu-Side. Everything was flat as the slot was removed completely from the base of the system



It was then time to reattach the 42 pins on the base of the system, and now that the top-ed was completed, it was much easier to put together the two assembly's. I used the method that Ergot talked about when he posted earlier today, and soldered the two ends of the cable to keep the ribbon's geometry intact, and not tug on any extra wires. The next photo doesn't have the end soldered to show the length of wire left to allow the bend and not do any further damage a the solder points. You can see the oBey chip fuzzy in the background.



The most annoying part of the repair came next, where I had soldered two of the wires on JP101 too close to the board, and they suffered the same fate as the pre-repaired cable, and snapped right off. I cut the ends, pulled the wire through the bottom of the board with the iron at the top, and cut back the two wires (one of which, wire #3 facing the camera is 5v power) and did a spot-repair with the leftover purple wire. The shrink-tubing is there to remind me that the wire is +5v.



Going through all of these and adding the surprise region mod was tedious and I played with the length of the exposed wires again to make sure that I got a clean bend and that there were no stress points. I repositioned the main board with the soldered top assembly into the case to hold everything steady and got ready to gently bend back the cable assemblies and hope that it tests correctly.



...I then noticed that the 1A car fuse had popped again, and that I had no power to the main board. I took the extra fuse that I bought (just in case) and reinstalled, and tested with the meter. My suspicion before even testing the unit was that the pulsating power issue remained, so as soon as I had power to the main board, I hooked up the oBey chip to power, and tested. It registered across the meter on the pins and with 5V power. I unplugged the console, and slowly and carefully bent the ribbon cable back into place. The whole time, I'm thinking "No Whammies, No Whammies".



...Man it looks slick, but I powered it on in the IFU and STILL, the repetitive power surging that must have blown the fuse. On the RGB PVM, a green screen that fades to turquoise and then deep blue (not the game). So it's still dead, with the original issue. Everything has power, and the chips ARE working confirmed with the meter, and the game is playing on the board, but it not being exported. All pins on DIN8 are putting out proper voltage, all caps reading on the meter... but still this wierd short. Steve said that it was a power overload.

If I can't get this thing working I'm going to make it into a necklace and wear it like Flava Flav.

Day #5:...

So after hacking this unit up I got to the point where I couldn't figure out what was wrong, and I sent it to thesteve for testing and repair. The following post will show what steve did to correct the work and get everything working perfectly!

--

I will keep updating this until I get the project finished.

Goonies Never Say Die

le (hard)caw

[blueraven]

...And here are the results of thesteve's amazing repair on this unit:

He did one wire bridge which was causing the unit to have the solid color when running standalone, and found A PIECE OF TIN FOIL in the AC socket which is what was causing the power surge!  He also completely sliced off the ribbon cable (where I made it look like an octopus) with a craft knife, and cleanly re-soldered the ribbon cable, and remounted the chips so that they were not flying around the case. My hat really is off to steve, I really learned a lot about this unit and how to do things right after the initial mod, and what needed to be changed so I wouldn't make the same mistakes again.

Picture 1 (from thesteve): DIN remount, RGB amp mount, using existing wires. SO much cleaner than how I had it set up, and the DIN is functioning correctly.
 


Picture 2: Mounting of the region chip, cap added to switch, 30AWG kynar wire. Absolutely beautiful work. he ran the wires thru a slice in the ribbon cable to completely bypass my clearance issue, and the wires were so thin they didn't even get in the way. 




So there you have it!! The finished product. And I wouldn't have opened it up in the last photo but I accidentally plugged the wrong power supply into it and blew the fuse... and make sure that ground plate is installed correctly on these units! that's kind of a prerequisite to get them to work right

Ninja Warriors (PCE)                   ......And TaleSpin (TG16)


If you try one of these, Good luck! 

[MNKyDeth]

That's pretty awesome man. Seems you are making progress and that's always a lot of the fun. Moving from one problem to another until the system works like new or better, with RGB.

Keep it up, revive the beast!!

[Necromancer]

Why'd the ribbon cable break off?  Brittle with age?

[blueraven]

That's pretty awesome man. Seems you are making progress and that's always a lot of the fun. Moving from one problem to another until the system works like new or better, with RGB.

Keep it up, revive the beast!!

Cheers, thanks MNKyDeth!

The pleasure lies in the journey, not the destination. This lil' turbob box will be a beast when its done!

Why'd the ribbon cable break off?  Brittle with age?

Yeah, the ends are stranded wire, and just moving the Hu-Slot back and forth a few times snapped the bottom clean off. It was pretty damn shocking to find that after all of the work on the board had been done, and the most obvious (and large) part in the system was the culprit. I ended up completely desoldering the bottom of the board and stripping the other three sides of the cable. I think that the cable is the weak spot on all of these; thesteve has told me that he's had to do a few of these in the same manor, and I know that chop5 has completed this repair on several occasions.

That said, I'm not sure if the Coregrafx/CoregrafxII setups experienced the same cable problem. It seems that they corrected quite a few of the design flaws on the board with the CG photo that BMW put up as the cap replacement guide.

...Anyway, more soon! - caw

[Ergot_Cholera]

An easier way of stripping back the plastic of the ribbon without separating each wire is just to score across the plastic with a craft knife and metal ruler on each side, you should then be able to grab the plastic at one side and just tear it off revealing the newly exposed wires ready to solder back into the board.
Straighten each exposed wire and tin if necessary, then position all back into the holes of the board at the same time (a bit fiddly but possible), solder the two end ones to hold the cable in place then solder the rest.
I have had to do this a few times and find this the easiest method

[blueraven]

An easier way of stripping back the plastic of the ribbon without separating each wire is just to score across the plastic with a craft knife and metal ruler on each side, you should then be able to grab the plastic at one side and just tear it off revealing the newly exposed wires ready to solder back into the board.
Straighten each exposed wire and tin if necessary, then position all back into the holes of the board at the same time (a bit fiddly but possible), solder the two end ones to hold the cable in place then solder the rest.
I have had to do this a few times and find this the easiest method

Thanks, Ergot. That's much simpler than the method I was using. If I have to do this fix again, I will remember to use a craft knife. I did solder the end wires, though and that was crucial in keeping the system geometry stable as I had to do 2 spot repairs (re-repairs) on the cable after I bent it over the first time. That was the best way to keep the cable in place. I was also glad that I left a bit of room between the plastic and the board so that it won't stress the wires as the original install did. I have a lot more play in the cable to wrap JP101 and JP102 inside the case.

I'll post some updates here in a bit, I decided to go a step further than I originally expected.

[blueraven]

Updated the thread with Day 4's progress. Finished the ribbon cable repairs, and randomly installed an oBey region mod chip. I know you may think that the latter was completely unnecessary, but everything was completely apart and it took 1/3 of the time that it normally would have.

Unfortunately, the original problem in the first post is still persisting . Has anyone run into the IFU-surging/overload issue before? IFU is confirmed working (green light on) perfectly with another PCE, but when this PCE is plugged in to IFU with only IFU power the light doesn't come on, and the power keeps pulsing. 

  Any ideas? Thanks in advance. 

[mickcris]

probably doesn't have anything to do with your problem, but some of them need that bottom ground shield installed to connect the analog and digital grounds together at the expansion port.  not sure if yours is one of them as there seem to be a couple different versions.

also, you are supposed to use ceramic caps not electrolytic to do viletim's jailbar fix.  from his documentation, the electrolytics aren't going to do anything to help.

[blueraven]

probably doesn't have anything to do with your problem, but some of them need that bottom ground shield installed to connect the analog and digital grounds together at the expansion port.  not sure if yours is one of them as there seem to be a couple different versions.

it DOES have the ground plate. And I have been testing the unit with it off  . Thanks for this observation, when I re-test the unit later this afternoon I will make sure to install the ground plate. One quick question about that: One of the 2 ground traces (the smaller of the two that you just mentioned) for the grounding plate lifted the copper trace under the solder, and now that you mention it it was never connected  but I can still get the solder to stick if I fold the pad over. I'll post a photo later but is there another place that I can pull that ground from?

also, you are supposed to use ceramic caps not electrolytic to do viletim's jailbar fix.  from his documentation, the electrolytics aren't going to do anything to help.

I will replace with ceramics and re-test later, hopefully. Thank you for pointing these two things out!

[mickcris]

you can just connect the 2 with a wire if yours is one that it uses the shield to connect.  check on the expansion port if GNDa and GNDd are connected without the shield
http://www.gamesx.com/misctech/pcebp.php
then if they are not, determine a point where the shield connected that go to each and then run a thick wire to connect them both.  they may already be connected on yours as I have seen shields that dont connect them also.

I think it should have been connected in more than 2 places originally.  i'm sure you can find a couple spots to connect both grounds.  there should be lots of places to connect them together.

edit:
looks like yours might have them connected already, but wont hurt for you to check yours anyways. I pulled mine apart to look at it again:


this was one that I got from gamedrhk a long time ago and they had left that shield off.  when i tried to use it on a av booster i modded for rgb, it didnt work.  after some troubleshooting, thats when i figured out the shield on that board connected the grounds together.  the black wire (on the left side) is connecting the analog and digital grounds.

[ApolloBoy]

Your wire is way too thick, and I'm also not sure why you used a full-size DIN when a mini-DIN is a much better choice for the white PCE.

[esteban]

Your wire is way too thick, and I'm also not sure why you used a full-size DIN when a mini-DIN is a much better choice for the white PCE.

He loves in Siberia (or the equivalent thereof). Supplies are limited.

[blueraven]

Hello again,

There was a lot of progress on this unit, and I ended up sending it off to thesteve to correct some of my repairs. I felt kind of irresponsible leaving this "done wrong" for so long, but life gets in the way, etc. It was a great learning experience learning on this unit, and I ended up making so major adjustments and modifications to my own repair work, made some cool tools, got a lot out of the comments everyone made here. Thanks it helped me step up my game.

As you can see from the above posts, its a spiders-next of wires inside, and while I tried my best to make the photos look good, this work could have been a lot better. it was about learning how to hack the unit instead of being a "Guide". but I'm glad that Necro sticky'd it nonetheless.

Your wire is way too thick, and I'm also not sure why you used a full-size DIN when a mini-DIN is a much better choice for the white PCE.

You are completely right. The case was drilled too large to accept the plug and I tried to fix it with some epoxy. A mni-DIN would have been a much better fit and I wouldn't have had to the case repair if I had done this.

Also you are 100% right about the wire size. After your comment and seeing steve's repair photos, and again first-person when I blew the fuse on the unit, thesteve switched me over to 30AWG kynar wire; and I then went and bought a few rolls myself for future repairs. Way more low-profile and a lot easier to fit in the case.

-- edited the first two posts with updates. Will have pictures from the PVM soon and will update the thread again.

[HuMan]

Congrats on that insane repair procedure, that PCE deserves to live after all that. This makes me grateful all I've had to do recently with my white PCE is replace all the capacitors. Strangely, it doesn't have any vertical lines.