ReAmiga1200 _KB_RESET/HALT question

[imqqmi]

I'm in the process of building the ReAmiga 1200 board from scratch. I've got all the parts I need except the reset generator, but that's on its way. In the meantime I thought I'd check out how the reset circuitry works and maybe get the Amiga run code without the reset generator. The Diag roms are placed in the rom sockets, even without roms the CPU doesn't appear to show any signs of life. The 14MHz clock is there though and the CPU is warming up (about 30 degrees C).

The _RESET signal is pulled high with a 10K resistor (ReAmiga PCB doesn't have a cap of 330nF to ground, so no delay during power up).
_KB_RESET is also going high, which means that the Amiga is getting out of the reset state correct?
The CPU is held in HALT state though. I then reset manually by shorting pins 9 and 10 on the TP1 connector but still nothing happens, the A0 of the CPU doesn't appear to start counting up and the CPU is still held in HALT state (_HLT is 0V at R951A).
Would this indicate an issue with Gayle, which is keeping the CPU in HALT or with the reset generator?
I've checked, double checked and triple checked for shorts, found 2 shorts on the CPU and removed them. Budgie and all other chipset ICs have no shorts. Still this behaviour persists. Either I don't understand it yet (most likely) or something is off.

I do get a signal on memory address pins (not yet installed, I'm at the LED blink pins stage with the MC68EC020 soldered in, part 2 of the guide), I assume this is the video part reading the memory for the video DAC? So some part seems to be working.

Does anyone have any pointers? Should I wait for the reset generator to arrive or should I look for problems with or around the budgie?


Thanks,
imqqmi

edit: Budgie should be Gayle

 

[imqqmi]

After reflowing all the ICs, caps and resistors, replacing the CPU, removing the CIA nothing made a difference. I've checked for shorts 100 times but couldn't find any.

But I've found the problem. I didn't follow the build guide to the letter. Instead of going through all caps and resistors using the Locator I went ahead and used the silkscreen on the PCB to find the value for the particular component and pick it from my sorting box and solder it on. This worked well for me to get most of the 'bird seeds' on the PCB. This way I don't have to hunt for each same value location all over the place and can just place components close to the previous components.

Some of the silkscreen is very tiny, though I did use magnification, part of the value like pF and nF was very similar and the lower part of the p was cut off and the circle blended together due to the small size and the silkscreen not being able to resolve such small details. This made me place 22nF on some spots instead of 22pF. Luckily the colour of 22pF is different to the 22nF one so I could spot it easily once I was on to the right track.

Normally with decoupling caps this shouldn't matter but these were impedance matching caps (with matching resistance) on digital signals and the factor 1000 higher capacitance on a logic signal has the effect of logarithmic slopes of charging and discharging capacitors in the logic signal waveform which should have much steeper slopes. Some pulses didn't quite make it to full 5V or 0V and looked like spikes. This will certainly cause things to go wrong, in my case the processor didn't even start running code. Apparently Gayle has some functionality that does some checks and until some conditions are met the processor is kept in reset.

Once I replaced the caps with the proper value (all 68 ohm + 22pF networks) the processor sprang to life and started executing code.

I can finally continue the build