Expansion board for the CD32

[alenppc]

Maybe UTSource is selling damaged chips.

[kipper2k]

the only difference i can see so far is i never used UTsource or am i wrong ?

[Amiga1992]

Maybe.
The weird thing is, as long as IDE is connected, the card works perfectly. If they were damaged, wouldn't they not work at all?
I mean, I made the effort to try stuff that filled the RAM, while IDE was plugged. No issues!

So maybe there is something else involved, that these chips may be more sensitive to, causing the issue.

I have no idea what I am saying anyway, I'm a moron. I can only post what I found, and test as thoroughly as I can.

I want to wait for Supaduper's word on what FW I am running, meanwhile I will go and buy that Xilinx programmer, so I can flash other firmwares.

[plasmab]

Originally Posted by Akira

Maybe.
The weird thing is, as long as IDE is connected, the card works perfectly. If they were damaged, wouldn't they not work at all?
I mean, I made the effort to try stuff that filled the RAM, while IDE was plugged. No issues!

So maybe there is something else involved, that these chips may be more sensitive to, causing the issue.

I have no idea what I am saying anyway, I'm a moron. I can only post what I found, and test as thoroughly as I can.

I want to wait for Supaduper's word on what FW I am running, meanwhile I will go and buy that Xilinx programmer, so I can flash other firmwares.

Its possible its timing related still but if thats the case the RAM is outside the spec.

Does anyone have an IDE terminator they can try?

[kipper2k]

Originally Posted by plasmab

Its possible its timing related still but if thats the case the RAM is outside the spec.

Does anyone have an IDE terminator they can try?

do you have a schematic, i may be able to whip one up

edit,

one more thing comes to mind, i did buy 12 of these chips from UTsource and i remember i could not get the original ram file to work so i wonder if it is an issue with the UTsource batch

[Amiga1992]

Hmm are there any open source IDE terminators out there? Seems lie an easy enough thing I can build at home.
Being that I am the only one reporting this problem, I ought to have one.

[edit] found this IDE terminator. Simple enough:
http://aminet.net/package/docs/hard/no4000ide

[plasmab]

Originally Posted by kipper2k

do you have a schematic, i may be able to whip one up

edit,

one more thing comes to mind, i did buy 12 of these chips from UTsource and i remember i could not get the original ram file to work so i wonder if it is an issue with the UTsource batch

I think we should try and get some from a different source.

I dont have a schematic for the ide terminator sorry.

[supaduper]

Originally Posted by plasmab

Very odd. I have those chips and they dont work for me either.

I bought 12 from Utsource like you steve, and I needed to flash some hynix with different firmware..remember when I could not get one Hynix set to work then i flashed it with a Slow.firmware you uploaded then it worked
Kipper bought his batch from a different seller..wonder if we just got a duff mixed batch

the other two boards out there are not a problem only Akira`s

Personnally think we should do as per the Github chart and leave them on the no list

and all Brian has to do is obviously test each board to see if they will still boot with the Board in place if yes then its good to go and he can do this with confidence that the board is good and do this to exhaust his hynix stock.
Its what I do now obviously as we were all concentrating on setting up the IDE to boot, but now we can do the cd boot check as well

[kipper2k]

Originally Posted by supaduper

I bought 12 from Utsource like you steve, and I needed to flash some hynix with different firmware..remember when I could not get one Hynix set to work then i flashed it with a Slow.firmware you uploaded then it worked
Kipper bought his batch from a different seller..wonder if we just got a duff mixed batch

i will make one more to see if it is just shthouse luck

[Amiga1992]

Originally Posted by supaduper

the other two boards out there are not a problem only Akira`s

Which other two boards?

[supaduper]

Originally Posted by Akira

Which other two boards?

Friend do not ask me about my personal business

[Amiga1992]

Originally Posted by supaduper

Friend do not ask me about my personal business

Haha, sorry, what I meant is, if you say they "are not a problem" does that mean there's two other boards with the exact same RAM, and whose users, whoever they are (Spiderman I bet one of them), have gone through the exact same steps I have and come to different conclusions?

If that's the case and mine is a confirmed one-off, I'll stop posting here and try other stuff.

But if they haven't done the exact same tests I did, one cannot say they are "not a problem". Let's remember I landed onto this issue by PURE CHANCE!

[solidcore]

Imagine it all comes down to soldering? Not that marks soldering is an issue as he's a top guy with excellent soldering skills - more over it's sometimes easy to overlook with the naked eye until you go right in there with a magnifying glass and bobs your uncle there's a short.

In my own experience when I first got a riser to test my board it wouldn't boot up, I took to resoldering the ram cpld TWICE until I saw a 'defective' message come up in red screen. After that I looked at the ram and what do I know a microbridge lol... I re-soldered the ram and it was working 100% .

@Akira based on the above info (mark's point that there are 2 out there working fine) and it could be an exclusive issue to your board (IDE having to be in to work?) it may be a solder fault OR worse case defective board, in any case can you go over each individual tiny little leg with a magnifying glass or microscope on all the chips and including all other components.

Of course I'm only speculating as we're going round in circles now lol.

[Amiga1992]

Yeah, we are. Unless those two other users did exactly what I did, I cannot still say it's only my board.
But going through the board with a glass is something I can do. I will inspect it later!

[kipper2k]

just built another using same ram and it works fine. (programmed using original RAM file)

[solidcore]

@Brian nice one mate, that's one step in the right direction for this ram.

[plasmab]

My HY board is unhappy and I have a microscope.


Sent from my iPhone using Tapatalk

[kipper2k]

i think best thing is to steer clear of hynix and go with samsung if that ticks all the boxes

[Stedy]

#2001

Oh I think we need as many opinions as possible. As I said, mainly to know if there's anything that could be avoided in the future, not because my particular setup isn't working all the time.

Stedy, what do you think mate?

Any ram that has the extra address pins and a 4k refresh will not work, also Oktoberfest beers are sweet .

The extra address pins will randomly change state if there is enough noise. This is equivalent to you reading a book then someone else changing the page you were reading when you looked away.

I have an idea about the IDE, can you run a memory test program? If it fails on D16-D23 might have an idea why.

Currently in Berlin enjoying some good beer so I'll pass on looking at datasheet s for a few days.

@kipper2k

Have you had any of your 4 layer TF328 PCBs back yet?

@thread

Don't forget the TF328 is an open source reference design, there will be bugs and you are beta testing. Be patient and have fun.

[alenppc]

Originally Posted by kipper2k

i think best thing is to steer clear of hynix and go with samsung if that ticks all the boxes

Or better yet, Alliance. At least it's 5V.