A4000D boots with A3640 but not BFG or TF

[Tumbleweed]

Hi all,

I have ome weird issues with my A4000D when I upgraded the A3640 to the TF4060.

Installed the TF4060. Switched on. No boot off HD. Tried booting off floppy - boots off WB3.1. Thought it was a HD installation issue, but have MMULibs.

A4000D is KS 40.68 and OS 3.2.2.1. Swapped out the TF4060 with the A3640 - all good. Boots fine.

Thinking the issue was with the TF4060 I installed it in my A3000D. Swapping out the BFG9060. Works fine - no issues (The TF4060 onboard IDE not tested - 3000D doesnt have the INT2 mod).

I then tried the BFG9060 in the 4000D. No boot. So thinking the issue is with the 4000D.

What was common with the TF4060 and the BFG in thre 4000D was I coudl not warm reboot the A4000D. Power light flashed but it did not complete the warm reset. This does not happen with the A3640 - warm reset works as normal.

On tha bsis Im thinking a problem with the reset circuit - but why wouldnt it impact the A3640 only the BFG9060 and TF4060? The BFG uses internal clocks; the A3640 external as does the TF4060.

How can I test the reset circuitry is working ok?

Weed

[Castellen]

How can I test the reset circuitry is working ok?

I'm not familiar with the BFG9060 or TF4060, but are they supposed to work with the A4000 onboard IDE?  There could be some kind of limitation with the design, especially at faster clock speeds.

A quick look at the sales brief suggests it has a programmable phased lock loop for 68060 clock generation.  Have you tried configuring this for 50MHz operation and checking if the onboard IDE interface then works as expected?


Regarding your question of testing the reset operation, simply look at what _RESET is doing.  The easiest place to access it is pin 1 of the IDE connector.  When you first power up the system, _RESET should stay low initially, then the power on reset generator (part of U150/Gary) releases the line and it goes high (approx. 5V) and the system enters run state.

The keyboard 'warm' reset works in a similar way; U150/Gary looks for the keyboard holding the keyboard serial clock line low for an extended time to indicate a reset request, and U150 responds by holding the system in reset state (IDE connector pin 1 low) for approx. 250msec, before releasing it (IDE connector pin 1 high) and the system enters run state.  If it does that, then the logic on the main board is working as expected.

[Tumbleweed]

Thanks Castellen.

I measure voltage = 4.85v on pin 1 of the IDE from a cold power on.

Warm reset: ctrl left amiga right amiga power light dims Voltage drops to 0.14v; release power light brightens voltage jumps to 5.77v; then falls back to settle at 4.85v.

The TF defaults to 50 mhz; you can change the clock frequency through software to set it at say 75mhz but it always starts at 50 mhz from what I've read. The clock jumpers Ive set as:

J100: 2-3
J104: 2-3

I have a CSMKII I can also try but suspect the results will be the same.

With the TF I can boot to workbench from floppy and then access the HD - all partitions show. I can also run Amiga Test Kit (ATK) and Versatile Amiga Test (VAT) from floppy.

The 4000D did have battery damage which I repaired. As part of that repair I replaced the original v09 Buster with v11. It wouldnt boot with the original Buster which I concluded was defective after running diagrom. Also at the time it wouldn't boot with the original A3630 but did with the A3640. At the time I thought I had a problem with the A3630 but maybe the fault was/is somewhere else and switching to the A3640 masked it. I'll look at it again tomorrow.

Much appreciated.

Weed.

[Jeff]

Is the Buster socketed? Paul "Acill" has a Youtube video out about the same type of issues with an A4000.  Getting to the bottom of the problem took some hours of advanced troubleshooting. The end result was to surface mount the Super Buster IC - all problems cured.

[Tumbleweed]

Hi Jeff,

Buster is socketed. Desolidering the socket and surface mounting the chip is probably beyond my soldering skills though. When I repaired rhe mother board it was a few track repairs and resoldering the smaller chips U975, U976 and U177. I'll have a proper look at it this weekend.

Thanks for the suggestion.

Weed

[Castellen]

I measure voltage = 4.85v on pin 1 of the IDE from a cold power on.

Warm reset: ctrl left amiga right amiga power light dims Voltage drops to 0.14v; release power light brightens voltage jumps to 5.77v; then falls back to settle at 4.85v.

With the TF I can boot to workbench from floppy and then access the HD - all partitions show.

It's unlikely that reset will be high immediately at power on as you'd be seeing a lot more issues.  To measure this correctly you need to use a dual trace scope, trigger when the 5V supply rises, and use the second channel to measure the time until _RESET transitions from a low to high state.  Which will typically be something like 200mSec, you won't see that with a multimeter.

In any case, the reset hardware sounds as though it's working normally, that's not the cause of your problems.  It's going into run state after a warm boot, but there's somewhere else that it's hanging.  You'd need to look further to see what's going on, which would be to see if the system is attempting to boot from ROM (_ROMEN going active), see if _OVERLAY (U350 pin 2) is getting set low, etc.

So it sounds as though the onboard IDE interface is actually working with the TF4060, but the issue is that it's either not accessing the IDE interface initially, or isn't correctly interpreting what it sees on the interface.  Nothing in the U975, U976, U177 area is going to cause the problem you're describing.  You can see if the IDE interface is being accessed by measuring if _IDE_CS1 (pin 37 of the IDE connector) becomes active (low) soon after power on.  As with most signal measurements, you need to do this this with an oscilloscope as it'll only be a few milliseconds in duration.

It would also be worth sanity checking the operation of both those CPU boards in another A4000D just to be certain they do behave as expected.

[Tumbleweed]

Thanks Castellen. Sounds like I need more advanced kit to test properly. My multimeter isnt sufficient.

The BGF and TF4060 both work fine in my 3000D. No issues booting and stable.

The TF is really odd - boot from floppy and seems to be ok ish. Can see hard drive and all looks ok. Just wont boot from HD.

What I did see is when I warm reboot (with the TF) the power LED dims and then blinks continually looks to be stuck in a reboot cycle but never actually reboots. it doesnt do that with the A3640. Reboots as it should. Same happens with the BFG, LED blinks continually and then is stuck in a reboot cycle but doesnt complete.

Weed