Amiga 3000 conversion to PAL not working
 

My Amiga 3000 has arrived from the USA and I went about changing Jumper J200 to PAL and I swapped the oscillator with a 28.37516 MHz PAL version.

I finished the 110V -> 230V conversion of the PSU. I followed this manual and the mod went fine: http://amiga.serveftp.net/A3000_HardwareGuide/psu.html
I cut the wire and used heat-shrink to make sure those (now) two wires can never touch.

The system now runs directly of 230V and is perfectly stable.
I measured the voltages on the 15pin motherboard connector before and after and the voltages stayed the same. So that's good. I hooked up the PSU to the A3000 and it booted fine but...
... it still seems to run in NTSC mode. PAL Video look like crap, NTSC Video is fine. Lotus II music still too fast.

Unfortunately, one leg of the old oscillator broke off and was stuck in it's tiny little socket. I would not come out, not even with mild controlled violence and verbal encouragement.
A stump stayed behind in the socket. So what I did is cut the same pin on the new (PAL) osc. to length, then gently pressed the other 3 legs of the osc. in their sockets until the shortened pin touched the stump of the old pin. I then carefully soldered the stump (by lack of a better word) and the shortened pin of the new oscillator together.

What if the solderjoint is not good. Cold joint, not connected properly. It looks good but what if it isn't? Could that cause this effect? The pin that broke off is the one directly next to the pin marked with a dot in the corner (see the 3rd foto).

With my trusty multimeter, I tested continuity between each of the 4 legs of the new PAL oscillator with their solder-points at the underside of the motherboard. I get solid, stable readings for all four.

SysSpeed 2.6 shows "Power" to be 60Hz. In both PAL and NTSC Highres Laced workbench 3.1 screenmodes. See attached screenshots. We don't have a 60Hz grid here (it's 50) and the PSU is now plugged in directly into a 50hz 230V wall outlet so i'm a bit confused.

Broken pin of old oscillator
https://eab.abime.net/attachment.php...1&d=1675494851

SysSpeed power-info after PSU mod.
https://eab.abime.net/attachment.php...1&d=1675494851
https://eab.abime.net/attachment.php...1&d=1675494851

I found this (a comment from user Davideo back in 2008)

Could this be it? I converted my PSU to 230 successfully. I measured all the voltages on all the pins of the 15-pin Motherboard connector and they are all the same before and aft. When I measured that the "TIC" pin still measured 2.665 Volt after the modification, I got suspicious.
Should that voltage not have changed? That must be why SysSpeed reads the power as 60Hz.

In the several Blogs and forum posts concerning the A3000 PSU 110V to 240V mod (or the other way around), nobody ever spoke about the TIC signal.

If I want to avoid having to change J350 from "tic" to "vSync", my PSU must start sending the 50Hz tic signal and not 60Hz. Does anyone know how to do this?

Edit: Found this on Ian Stedman's Amiga Powersupply (convert PC PSU to Amiga 3000, 4000 etc):
I guess what Mr. Stedman means is that PC powersupplies don't generate a TICK signal (or the correct one) and that one must start using vSync instead.

The PSU should already be sending a 50Hz tick if that's your local AC line frequency?

That's what I thought too. And in my mind, that must be the reason why the powersupply's TIC pin voltage stayed the same before and after the 230V modification.
Before the mod the USA spec 3000 was powered by a professional 230V -> 110 converter. On the label on the converter it says "Input 230V 50Hz - Output 110V 50Hz".

And I sound like a broken record but the oscillator is PAL, J200 is set to PAL. And still, SysSpeed says Power is 60Hz and the fact that games like Lotus 2 runs too fast, music plays too fast, that displaying PAL graphics is total crap while NTSC graphics are normal, must mean that the system is 60Hz.

I just don't understand how, other than that there must be a 60Hz TICK generator in the PSU, which does not care about the AC line frequency.

Does your multimeter has an option to measure frequencies? If so, hook up the black cable to one of the ground pins and the red cable to the brown pin of the A3000D PSU. Compare with this image: https://ibb.co/BVqYcfc

I have a Voltcraft VC170. When I set it to "Hz" the reading on the lower middle pin (brown cable) the reading is 149.9 Hz
When I press the "Select" button, it changes to "Duty Cycle". This reads 49.4 %

This is definitely some magnitudes too high. This should read +/- 50 Hz in your case.

Have you tried setting the J350 jumper to VSYNC to verify that the psu tick is definitely the cause?

This schematic for the power supply shows that the PSU definitely generates the tick signal from the AC input
http://amiga.serveftp.net/Schematics..._schematic.pdf

That is the next step. But as we need to understand why my PSU is on 150hz, and fix that, we (well, my repair super-guy) will open the PSU up to see what is going on in there.

I've seen tick go haywire or go missing on some a2000/a3000 power supplies. But the solution is J350, PSU tick is not very relevant for most use cases.

I'd find it more likely that your multimeter is misreading for some reason than tick actually going at 150Hz.

The TICK circuit in the PSU is quite simple for one, and for another if it was running at 150Hz I'd not expect the system to work properly at all in either NTSC or in PAL modes

I am not 100% familiar with this process but I believed that the 8372ab 2MB AGNUS found on the A3000 has a pin (41) which is (unfortunately) named TEST but (for 8372A and 8372AB) actually selects between NTSC (VSS) or PAL (OPEN) on reset. Maybe you need to change on the motherboard connected to this pin? Check the schematics of the A3000?

We used 2 different makes and models of Multimeter and both read 149.9Hz.
I don't know why the Amiga is not going bananas either. It seems to ignore it. But I'm not knowledgable in this field

We also checked J350 and it was, and still is, set to "Tick". We will change it though if we (the repair shop) can't fix the PSU to act normally and start sending a 50Hz tick.
In the shop they have an original 230V A3000D PSU and that reads 50Hz.

I have 8372B as my MoBo is a Rev. 9.x
From what i've read, Revisions 6.1 to 8.9 mostly had AB's.

8372AB - Amiga 3000 from Rev 6.1 to Rev 8.9 (NTSC/PAL)
8372B - Amiga 3000 Rev 9 (NTSC/PAL)

The enhanced (ECS) 8372B Agnus (model 318069-03). Essentially an 8372A with 2 MB RAM support, all addressable as chip RAM (20-bit pointers). Some batches of the 8372B even use the 8372A packaging, with the B distinctly added after the 8372A part. These read like 8372AB, but internally they are 8372B parts.
Unofficially called the “obese” Agnus to signify the yet increased 2 MB chip RAM support.
Typically found on: Amiga 3000

Source: http://theamigamuseum.com/?s=agnus


Aren't you confusing the Pin41 thing with the A500 (you know, masking it makes PAL pre-selection permanent). Jumper J200 on the A3000 basically acts like the Pin41 trick on the A500.

B and AB are identical. Pin 41 switches between PAL and NTSC, which is directly connected to J200.

And I have said elsewhere, unless you do Genlocking (ie need absolute sync with an external PAL/NTSC video source), changing the crystal isn't required.

FYI it was me who first collected the information about all the different Agnuses (Agnii?) that is used everywhere on the internet today. At the same time dispelling lots of myths and rumours not to mention incorrect information.

e.g. Their product numbers, their markings (e.g. dates, version numbers, what Vbb means), the systems they were in, their capabilities and their pinout.

It was for BBoAH when I was a contributor (albeit the version of BBoAH Agnus page available today isn't the final one I wrote but an earlier version). The information was collected here on this forum and Amiga.org with the help of lots of members.

No, I just didn't remember/know that Agnus pin 41 was connected to J200 (I could have looked it up).

Back to my Powersupply. Salocinx (my :great repair shop) here in Switzerland, who chimed in a few posts above, has hooked up his Digital oscilloscope to the Tick Signal. Not only is the signal doing the mamba (fluffing about between 141 and 149.9Hz at the moment), the signal is also of pretty poor quality. Not a nice steady pattern at al.

He suspects Varistor RV21. There are (many) A3000 110V PSU's out there where RV21 is not rated for voltages above 120V (with some margin above). He will open up my PSU soon to see inside. If his suspicion is correct, that could be cause of the trouble.

Salocinx is not the only one mentioning this. Anthony Hoffman from New Zealand recommends changing this varistor as well.
Source: http://amiga.serveftp.net/A3000_HardwareGuide/psu.html

Mr. Hoffman has been the only person to ever mention this. All the others that i've found never spoke a word about a varistor being changed. I was hoping that, because my PSU was produced relatively late in the A3000's lifecycle, the only difference would be the "110V bridge-wire". On the other hand, if Commodore could save a few pennies by using a cheaper, less capable varistor in 110V PSU's, that would not suprise me at all either...

Obviously, someone going from 230V to 110V does not need to worry. Only those who are going the other way (like me) might run into this.

Anyway, my PSU will get completely rebuilt. Out with the old, in with new, fresh components. I just don't trust that PSU anymore and now that it has it's guts hanging out, might as wel replace everything to put my mind at ease.

My Amiga3000, was originally a NTSC machine as I imported the MB from the USA.
I don't have the original PSU, but use a AT PSU.

I just set the Jumper to PAL, and the Tick to onboard.

sysspeed reports that it's running with power at 50 Hz, and in PAL.