Kickstart 3.1/Amiga 3000 weirdness

[icbrkr]

I've been upgrading the A3000 as many of you might have read.  Now I have a weirdness question that's been affected by 3000.

Ever since I put in Kickstart 3.1, it takes a long (long) time to start to boot.  The screen flashes red (ie, like a bad KS ROM) but after about 2 minutes it will kick on and the drives will spin up.  If I remove the SCSI HD from the chain, it pops right up and gives me a "insert disk" screen.  I haven't a battery installed in my 3000.  Is it possibly scanning for HDs now since I don't have a battery installed?

[melott]

I've been running my A3k for 3 or more years without
a battery with no problems. (ie long startup delays)
I play around on my A2k alot with different hardware
and I've found some harddrives are alot slower when
booting than others. My personal choice are Quantum
drives, they seem the fastest and very Amiga friendly.
I've had the most problems with Seagate drives.

Also you may have a faulty set of roms ..you are using
the rom tower..correct ?
I've read that if you don't use the rom tower things
don't run correctly (don't know for sure since I haven't
tried it myself)

[Zac67]

My 6.1 board got Kick 3.1 around a year ago - no problems ever of that kind. Maybe 3.1 doesn't like the battery missing, I've replaced mine.
Does this happen only on cold boot or also on reset?

[odin]

Doesn't the A3k save SCSI settings and such in NVRAM or something? Perhaps that needs the battery not to suffer from amnesia?

[vic20owner]

Right, and I think *some* A3000's react differently than others when the battery is missing. If you do a warm reset from the keyboard during the delay, will it reboot and come up quickly? If so it could just be that the drive isn't initializing fast enough on power up... but I'd lean towards the NVRAM.

[icbrkr]

vic20owner wrote:

Right, and I think *some* A3000's react differently than others when the battery is missing. If you do a warm reset from the keyboard during the delay, will it reboot and come up quickly? If so it could just be that the drive isn't initializing fast enough on power up... but I'd lean towards the NVRAM.

If I do a warm reboot, there's no issues at all which is why I was leaning towards the battery.  I didn't have a ROM tower at all in this model, just standard socketed Kickstarts.  I think I'll be making a trip to Radio Shack today and building the lithium battery mod.

[Zac67]

Sounds pretty much like the battery - make sure the voltage is high enough when using a Lithium/diode mod. The voltage drop from the diode might make the battery seem empty rather quickly.

I replaced the battery with a fresh new NiCd one - and will replace it every couple of years.

Another choice is to replace the battery with a NiCd or NiMH one, but rather then soldering it directly to the board you use a pair of wires to put the battery where it can't do much harm - preferrably in a closed box. Be aware though that your Miggy might not be stored horizontally all the time in the future.

[icbrkr]

What's the minimum voltage you can use?  Would it be possible to put a few NiMH AA's in a battery holder and attach it to the board?

[Zac67]

Definitely - you'd need three cells for the required 3.6V.

[melott]

icbrkr wrote:
What's the minimum voltage you can use?  Would it be possible to put a few NiMH AA's in a battery holder and attach it to the board?

You could take a look on EBay ...
I have both the Coin Battery (all required parts) and
the Barrel Battey listed in the Amiga section at a
reasonable price .....

[patrik]

@icbrkr:

Is the scsi-bus terminated correctly and is does your A3000 feed termination-power to it (see the "Specific Troublesome Devices" part on this page)?

The reason I am asking this is that in 3.1 the scsi.device driver is newer/different and might react differently to scsi-bus issues.


/Patrik

[icbrkr]

patrik wrote:
@icbrkr:

Is the scsi-bus terminated correctly and is does your A3000 feed termination-power to it (see the "Specific Troublesome Devices" part on this page)?

The reason I am asking this is that in 3.1 the scsi.device driver is newer/different and might react differently to scsi-bus issues.


/Patrik

This is a good answer here because more or less, this is roughly what the issue was.  I slapped in a Quantum this morning and the problem seems to be fixed.  It now blinks red for about 1/2 sec and boots off the drive.   I'm going to try the "battery holder" idea and slap in a couple of rechargables later this weekend.

[patrik]

@icbrkr:

Something like this maybe? :=).


/Patrik

[icbrkr]

patrik wrote:
@icbrkr:

Something like this maybe? :=).


/Patrik

Yep!  I used an old cordless phone I had lying around for its battery pack and it's in there now  Finally, I have a clock that works!

[icbrkr]

Okay, I had some issues with the new HD (I didn't know if it worked correctly.. it loses data constantly).  I was about to rip out the new KS 3.1 chips and drop back to the original 1.4 ones and softkick 3.1 until a new HD arrives when I noticed something odd - my SCSI controller chip in the 3000 is marked as "Proto" !  Is this normal??  What's the revisions of this chip?