A300 + A604n

[jarp]

Hi all,

Just got A604n and RapidRoad USB I ordered. I have Amiga 600 rev.1 a.k.a Amiga 300. When A604n is installed, computer won't boot. Power led is lit and my screen regognizes signal but that's it. Black screen. Nothing more. I actually have two A604ns and neither of them works.

So I am just wondering that is there any known compatibility issues between A604N and Amiga 300? Anything I could try? I even removed my accelerator (Furia) and everything, but no luck.

Visual inspection of the PCB shows no problems with PCB traces or anything. I am quite afraid that this is unfixable problem... If anybody has good ideas then please share them, million thanks!

Ps. Mobo is recapped yes.

Edit: I also cleaned the connector multiple times with IPA. No luck.

[kipper2k]

Hi,

pull the a604n back a bit, do not push it in all the way, see if that fixes it

[bubbob42]

You could also clean the connector with a soft pencil rubber. IPA does not remove layers of oxide reliably.

[jarp]

Originally Posted by kipper2k

pull the a604n back a bit, do not push it in all the way, see if that fixes it

Originally Posted by bubbob42

You could also clean the connector with a soft pencil rubber. IPA does not remove layers of oxide reliably.

Both were good ideas but unfortunaelly no, they did not help. I even tries switching back to kickstart 2.05 from 3.x just in case, but no, unfortunately no help.

[jarp]

Sigh... Nothing I tried worked. So I just printed out schematics and I am going to measure every pin of the connector and see that there is proper connection between them and corresponding chips.

[jarp]

I measured every single pin of the memory expansion connector and every pin is connected to correct chip pin (sometimes via resistor). I was hoping for simple broken trace or something, but no.

However I also found out that while A604n is connected then _RESET is being held to ground all the time. Explains why I get black screen only and no leds. But why that happens? Is A604n keeping _RESET grounded or is it some Amigas own circuitry?

Need to study schematics more...

[jarp]

Probably need to admit defeat and start to hunt A600 rev 1.5 mobo. I've been studying schematics and seems that _RST signal (and _HLT for the CPU) is produced by the Gayle and distributed to pretty much every other major chip in the system including the CPU. When _RST is low then all the chips are doing nothing.

_RST signal is then routed via 74HCT244 buffer to expansion connectors and name of the buffered signal is _RESET. Since _RESET is connected to the output of the buffer then I assume expansion boards cannot actually reset Amiga? Thus A604n cannot deliberately keep _RESET and _RST low.

Now I did not check datasheets of every single chip but what I've read it's Gayle and Gayle only who decides about resets. As far as I know there's two conditions under which Gayle keeps _RST low; _KB_RESET is low (Amiga+Amiga+Control is being held down or machine has just booted and 555 timer is erm... charging?) and other one is +5V being too low.

Now on my Amiga with a604n connected _KB_RESET is high and and +5V is well in spec. Thus I am at dead end, there's no obvious mechanical faults nor there is any obvious reason Gayle is keeping Amiga in reset with a604n connected. It just must be some kind of compatibility problem, Gayle just thinks that machine cannot be allowed to boot? Why is that, I do not know.

[DopPie]

Hi,

Did you get any further with this?

I have the exact same issue. Also with a rev 1.1 board with the label A300. I was thinking about recapping the board and see if that does any good.

The strange thing is, with a different memory expansion the board works fine. Just doesn't take the A604n. Could still be the A604n is dodgy.

[jarp]

Unfortunately no, I found no way to make it work. I contacted Jens and asked if there's any known issues with A300 and he said he is not aware of any. He asked to send boards back so he could test them but I haven't done it yet since I find it hard to believe I got two broken boards and I am hunting rev1.5 mobo now.

Unless, of course, whole A604n batch is faulty... How old is your A604n? Mine two are from the latest batch which were made available just recently.

Ps. Let me know if you find a way to make it work. I am considering re-capping my re-capping my A300 with tantalum caps since I've had some very slight video problems after re-capping (some caps were near impossible to solder, there could be cold joint or something).

[DopPie]

It's brand new latest batch. Jens said there was one other report which must be you. Hopefully they test some of them before shipping. I doubt the whole batch is dodged.

Monday I will see if the card works with a rev 1.3 board. It it doesn't it's going back to amigakit.

[jarp]

Cool, let us know how it works with rev 1.3. If it does not then most likely there are problems with the batch and I will send my boards back to Jens. Don't want to do that if they really are not broken, probably can sell them locally without any problems.

Then the problem is that which chip memory expansion will work? Perhaps kipper2k's board would work...

[amigoun]

I just installed my newly bought A604n in my rev1.3 motherboard (equipped also with Furia accelerator) and it did not boot, just hangs. I bought 2 A604n but none works. But it worked fine with A604n from the older batch. Seems the redesigned a604n from the latest batch has issues.
I see C9 is not soldered on my A604n boards, maybe that could be a problem?
I will soon test in rev1.5 motheboard.
edit: tested and on rev1.5 mobo it works.

[DopPie]

Jens can expect alot of returned material when the expansion has issues with two board revisions. For me the expansion is only required for the indivision. Maybe I can wire wrap it to my working expansion. It does not seem to use an awful lot of signals.

I'll keep you guys posted.

Edit: please contact Jens on issues. Only with plentiful information he can find a fix.

Edit2: Just got word the guy with the 1.3 board cant get it to boot at all. So, this week my board just goes back to amigakit. Too bad.

[jarp]

So so far we've seen problems with rev 1, rev 1.1 and rev 1.3 boards with 5 different a604n boards.

Uh oh, for me this sounds like faulty batch or then design fault but I would not believe Jens would do such mistake so it must be faulty batch...? Looks like I need to send my boards back, sigh, I wanted to avoid it due to extraordinary high prices Finnish postal service is charging from registered packages...

Would be very interesting to know what is the problem, just for academic interest.

[Schoenfeld]

Hi,

I've been made aware of this thread by a customer, and I have to make a test with an A300 board. Yes, Gayle decides about resets, and A300 boards have Gayle revision -01, whereas A600 boards have Gayle revision -02. At this point, I have no idea about what could go wrong, as the majority of signals the new A604n design is using are inputs to the CPLD. Never even thought about making a separate test with the (pretty rare to me!) A300 board. I do have one, but need to dig it out. I have scheduled this for this coming week.

Will keep you posted here. This must be a tiny issue, easy to fix.

Jens

[DopPie]

Originally Posted by Schoenfeld

Hi,

I've been made aware of this thread by a customer, and I have to make a test with an A300 board. Yes, Gayle decides about resets, and A300 boards have Gayle revision -01, whereas A600 boards have Gayle revision -02. At this point, I have no idea about what could go wrong, as the majority of signals the new A604n design is using are inputs to the CPLD. Never even thought about making a separate test with the (pretty rare to me!) A300 board. I do have one, but need to dig it out. I have scheduled this for this coming week.

Will keep you posted here. This must be a tiny issue, easy to fix.

Jens

Good to hear you are on the case. I'll hang on to my board a little longer then. If you need me to measure something on my board let me know.

[DopPie]

All well, all I wanted was an indivision. First I had to wait for the A604n to be in stoack and now this.....I lost patience so last evening I hacked the indivision in wirewrap style



This probably voids my warrenty but I don't really care.
It is actually working quite nicely.


So, once the A604n gets a patch I'll install it the proper way.

[Schoenfeld]

Found the cause of the problem. The short version is: Remove capacitor C4 from the new A604n if you experience startup problems - see first picture. Warranty will not be affected by this! Please only solder on the card if you really need to. Mind the prime directive: Never change a running system.
Long version:
When doing some research on the Amiga chipsets, I found that Agnus and Alice tend to get blitter problems if they have been in use for a very long time (>15 years in heavy use). It appears to be an ageing problem, and I have no idea what the underlying chemical theory is - it's just an observation so far.

As I described here on the German A1k forum, I have been able to revive an A1200 board where I thought that Alice is defective. In reality, only the negative bias voltage was flaky, and I have been able to fix that with an added capacitor on Vbb.

Later on, I checked an A600 that displayed a similar behaviour: Blitter problems during a "Sanity" demo (Arte is one of my favourites!). I was able to fix that with a small capacitor on Vbb - without the need to exchange the Agnus chip.

This has led me to the decision to add a Vbb capacitor to the new A604n design. However, not all A300/A600 boards actually have a Vbb-capable Agnus! My A300 is equipped with an Agnus type that outputs a 14MHz clock on the Vbb pin, which of course doesnt like to be loaded with a 100nF capacitor. Once the cap is removed, it works fine.

I have added a few pics of Agnus chips - just for you to see the difference between Vbb-type Agnus chips and non-Vbb types.

Jens

[jarp]

Awesome Jens, it worked! I removed C4 and my Amiga 300 now boots with 2MB chip ram. Thanks, I didn't believe this was fixable without sending boards back to you but wrong I was!

Edit: Thanks for the explanation also, very interesting at least for me! I studied schematics and found note about Vbb and 14MHz clock but was not smart enough to connect the dots. And thanks for "letting" fix this by ourself, would've hated if I would had to send boards back!

[amigoun]

Good to hear the fix is that simple, I am going to fix my A604n's soon!