@GTV reviews the Cosmic Fantasy 1-2 Switch collection by Edia, provides examples of the poor English editing/localization work. It's much worse for CF1. Rated "D" for disappointment, finding that TurboGrafx CF2 is better & while CF1's the real draw, Edia screwed it up...
Main Menu
Menu

Show posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Show posts Menu

Messages - thesteve

#1
Yes they are connected together
1081 is the source of the signal and 1082 listens for it
#2
Your system is failing between position check and focus lock
#3
Ok for starters the issue was unlikely motor related
Order of operations
Start sent from software
Check door switch
Check head position by bumping the limit switch
Mirror check (laser bobs)
At this point the laser head will trigger the F ok pin on the CXA1081 to the CXA1082
If the 1082 sees the pulse it moves to focus check (spools spindle and moves lens up into focus lock)
If focus lock is achieved it adjusts spindle speed to match data clock and begins read
#4
odds are its actually open connections between the HuC6270 and its ram chips
the traces open up where they pass through the board
#5
1 pole of battery is routed through the Jack as NightWolve just said
The indicators are fragile and often get decapitated when someone digs in the sealant
#6
the battery power not working is normally from tarnish in the external power jack
#7
odds are its just a dirty power jack, but could be something more obvious
#8
air stations are great for some applications, but risky for others
some of the plastic parts melt way too quick and these old caps will burst before the solder melts
#9
honestly your blog seemed about right (didnt really promote the mistakes)
as for the old solder, it doesnt all have to come off as long as its clean (a touch of flux restores it as long as its not soaked in cap fluid), but the old cap leads must
the fluid corrodes under the solder sometimes so the pads dont bond anymore, removing the old legs exposes that
solder should always cover the pad thin, because a thick blob can mask poor adhesion

The way i do it is I prep the pads with a puddle of solder on them, then add a drop of flux to the top
next i hold the cap in place and reflow the solder melting the cap leg into it
the flux on the pad preps the surface of the wire and helps the solder bond to it
#10
what KC said was correct
there are 2 huge traces that run across the front of the unit sending 5V to the CD system
the VIA near the 47uf caps by the LED gets eaten away and you lose 5V to the chips causing this issue
#11
im sure your iron was a big part of the issue
i used a 40W temp controlled one @700F
liquid flux helps allot as well
simple rule (less is more) you want the solder to be smooth and follow the surface of what your soldering
any lumpyness to the solder indicates low temp or contamination and should be cleaned and redone
#12
Final diagnosis
this was entirely solder and prep issues
several caps were not soldered down on both sides
some of the old cap legs hadnt been removed from the pads, and the solder hadnt been heated enough to flow
the 3.3uf cap was overlooked
the new cap legs were excessively long and some were touching other traces
removed all caps and reinstalled completing the cap change, unit started right up loading fine with some audio drops
slight adjustment with scope and its running well
#13
most shooters on the system are actually good, but not many have complex scoring or a way to grind
#14
worst case i designed and turbokon sells complete replacement controller PCBs
#15
these things are still quite reliable, we just deal with so many we are bound to see the bad ones
also the age means a few more have failed

to the OP issue, getting nothing on CDR is generally a bad laser
some lasers just dont work at all on CDR and poorly on pressed (or visa-versa)
#16
a warm up could be a few things, but caps are the most common
all the adjustments effect the CXA1081 and CXA1082 function as those 2 chips work in tandem to lock the laser onto the track and generate a pulse train from the CD
the CXD1167 reads the pulse train and adjusts its speed based on buffering
the NEC chips deal with translation between the core system and the sony chips
any of the 3 sony chips can cause issues and become intermittent (seen many bad)
also some switching supplies can cause read issues (try an old style one)
#17
ok just fixed it
found Q304 failed open (very common sound issue)
switch had low on one side and floating on the other (not right)
found no pulse from IC104 (replacement solved switch issue)
laser was also bad (no light)
confirmed working with new laser
#18
you have an intermittent connection somewhere
i have see ones that wont boot till they have been on a while
#19
im not saying that doesnt happen, but my impression was it was image related and the OP did specify it was an RGB modded system
this can easily be checked by using composite instead
#20
on the express its power hum, but thats not the OP issue
he is having SCART hum which has to do with the current running through the grounds on the SCART cable
there are a few ways to reduce it (mostly shorter cables or larger wires) but you wont completely eliminate it especially on bright screens.
a dark screen will have little hum due to the low image current
#21
im not opening your links
for IC520 you can use the HC or LS series chips, doesnt matter to it
only thing to watch is you need the narrow version  otherwise install gets a bit tricky
also i dont think IC520 has failed
dont recall which pin gets the pulse from the lid switch, but start by probing the switch itself (should be pulse 1 side open, both sides closed)
either way trace it back to the chip, just if pulse confirmed you only need to trace the switched side
#22
ok lost of the same issue here sounds like
look for missing address pulses to the D91317GD (almost every time)
IC520 could be bad, but will often also keep cards from working (check pulses at slot with cart out)
#23
before you get too far try Rock On
the game music doesnt wait for the video chips to start so often if the RAM and 6280 work, you will get sound
without a probe or scope it gets a bit tougher from there
full stop system check the CE pins on the 6280 to see what chip is currently active
there are 3 CER for the ram chip CE7 for the 6270 and CEK for the 6260
the 6270 also has a RDY pin which tells the 6280 to wait if low (often with a bad 6260 6270 or connection to 6270 this pin stays low)
also seen several 6280 with dead CE pins that simply dont ever go low (mostly CER)
#24
the quickest way to find this (for me) is to probe the data pins at the slot and the address pins at IC501
with cart out
the IC501 pinout is posted D91317GD
#25
could also be an open connection to one of the center references
check C322 and C323 voltages
should be around 4V
#26
may need a separate power supply connected, but it does work
all the pinouts are the same RGB ect
#27
ZD302 is an 8V zener diode used to feed Q304
D306 is part of the anti thump mute circuit
#28
IC520 is a high failure part that is used to switch between card and cd system
IC104 handles the cd system I/O for switches and track counter ect
the lid switch gets a pulse from IC104 and sends it back to IC104
D102 is a simple diode used as ESD protection for IC104 (lid switch passes through it)
cd bios calls cd system and says please wait, cd system responds with lid open or active..........lid open is shown if thats the response
please wait is shown until the cd drive responds
#29
just reposted all the images from the original post
they had gone missing
#30
/10202674019810402.jpg
/10202674018850378.jpg
/10202674018130360.jpg
/10202674017170336.jpg
/10202674016250313.jpg
/10202674015650298.jpg
/10202674015010282.jpg
/10202674014210262.jpg
/10202674013690249.jpg
/10202674013010232.jpg
the missing images from my original post


EDIT 2022: Images restored at top.
#31
well this thread sure ran off topic
#32
Quote from: guestSchrödinger's switches can't be stacked and can.
By the way, I like this project.
I drew this up when I started my cordless project
#33
just a note, the stock supply will get worm with a blown fuse as the fuse is on the output
#34
well i generally avoid ebay, as i want to keep in touch with the customers in case anything does go wrong or they have something else they want done
#35
a short in IC520 will also do it as thats the interface chip between the systems
#36
actually that is normal
white screen normally is in the core system, not the cd circuit so start there
#37
snes region lock is based on a code lock (much like the nes)
there are ways to bypass it but some games double check it
#38
/20989219_10207624584051414_1999850446767350690_o.jpg

that's the 2-port, but the 5 should be easy to extrapolate from it
note that those switch chips can't be stacked but i found some that can
#39
/20989219_10207624584051414_1999850446767350690_o.jpg
#40
Don't see anything related to fault indication, so I second check your voltages
#41
So it does
Never hit power on the remote before
#42
Stand by light?
Never knew it had one
#43
Quote from: xantoz on 08/05/2017, 12:44 PMHi, lurker posting for the first time

I currently have a project making myself some wireless controllers for various retro consoles (yes, yet another one of those). Thus far I've gotten a PoC SNES sender (I just use the extant shift register in the controller, so it would be possible to just make a plug-in box to make un-modified controllers "wireless") and receiver working. In addition I intend to use a unified over-the-air protocol making controllers interchangeable between different consoles.

I'm currently looking to implement the receiver side for PCE, and as I would like the ability to support several controllers to one receiver at once (the nRF24L01 I am using natively supports 5 RX pipes, how convenient!), it would be nice to implement a native multitap in firmware at the receiver end. This way I do not need to have 5 plugs coming out of a box into a multitap into the PCE. I am trying to implement as much as possible in firmware, to keep the count of external components low.

However, while I have been able to find documentation and schematics on the PCE pads (including ones with turbo functionality and the 6-button pad) I have found none on the multitap.

If anybody has any clues it would be helpful. Right now my only idea is to hook up a logic analyzer to the multitap and observe it in action.
OK I recently actually did this
The pce has 2 outputs and 4 inputs on the port
On 1 controller the line simply enables the controller or swapped but in the tap one causes a counter to select the next controller, and the other resets the count
The counter needs to loop after last controller
#44
Oh, missed that
Laser defect or adjustment then are your only options

Sent from my HTC One_M8 using Tapatalk

#45
As I said before, the connections are the most likely fault

Sent from my HTC One_M8 using Tapatalk

#46
It now is either the 20256 ram, the HuC6270 or the connections between

Sent from my HTC One_M8 using Tapatalk

#47
Not offhand on the R
Hacked up card slot could be the only issue, so start there

Sent from my HTC One_M8 using Tapatalk

#48
Quote from: LowRes on 07/30/2017, 05:25 PM
Quote from: thesteve on 07/30/2017, 04:36 PM
Quote from: LowRes on 07/30/2017, 03:15 PMI have a Turbo Duo-R that has graphic glitches once you start playing a game.  The loading screen is fine, you can load a game no problem and even the title screen of the games seem to be fine, but once you try to start playing, the graphics are all messed up.  Should I start with swapping out the ram chips?  does this sound like a ram issue?
It depends actually
Likely a ram issue, but in most cases it's the traces not the chip
Which chip can often be determined by the specific image that's corrupted on certain games

Sent from my HTC One_M8 using Tapatalk
am I correct to be looking at the 2 HSRM20256LM12 ram chips? or are there other ram chips on the board?  Prince of Persia for example works perfectly except the sprite for the character is missing, Dracula X has a lot of random glitches, the Title screen looks good except the red writing is missing

I would link to some pics but for some reason photobucket is asking to upgrade my account to do so.  I'll have to find another hosting site

thanks for the help
If it works correctly on cards, the hsrm20256 are fine
It actually has 4 CD related ram banks
And 2 core system banks
Each bank is 1 or 2 chips

Sent from my HTC One_M8 using Tapatalk

#49
Except pin8 isn't ground

Sent from my HTC One_M8 using Tapatalk

#50
The controller board has the same pinout as a standard PCE controller
Not sure about switching between boards
Several of the NES games are ported if you have an everdrive

Sent from my HTC One_M8 using Tapatalk