Kiwi - a 68k Homebrew Computer

Hello,

as I told there are some issues with the current design I want specify here.

• My symbol for the WD177x had a double counted pin. This resulted in an short between D3 and SYS_RESET. The symbol and the schematics which can be downloaded are corrected. But my PCB layout data still have this short (I had to cut some traces with a knife).
• The PS/2 interfaces need pull-up resistors. I wanted to use the internal pullups of the ATMega8, but this is not enough. For PS/2 devices it is ok, but USB devices which come with an adapter sample these signals at powerup when the ATMega hasnt finished setting up everything.
• The distance between each ferrite at the joystick connectors has to be increased. Or they should be dropped at all...
• Power jumpers for both SIDs should be changed as well as jumpers for different filter capacitors should be added. At the moment it is possible to set jumpers for both 9V and 12V. This is not good. And I would add two more filter capacitors (switchable via jumpers) to have have the correct values for MOS6581 and MOS8580.
• The negation of the /MO signal (floppy) is wrong and the logic of the two /DS signals differ. Don't know, how I did these bugs.
• Maybe change the circuit to use an overclocked WD1772 (switched by pin 2 of the floppy connector) for HD floppies. But HD is out of the WD1772 specs...
• Add a voltage monitor for CS1 of the real time clock as claimed by the datasheet.
• Add a LC(?) filter near the main power plug to get rid of the ripple noise and/or improve the overall PCB layout.

So these are the issues I would have to address anyway before I consider the PCB ready for production. Maybe there are more which I didn't notice yet.

Simon

Will examine this during lunch with some friends

Great! Now we hopefully see a progress here.

Today I started to rework the schematics as promised.

simon wrote:My symbol for the WD177x had a double counted pin. This resulted in an short between D3 and SYS_RESET. The symbol and the schematics which can be downloaded are corrected. But my PCB layout data still have this short (I had to cut some traces with a knife).

Done!

simon wrote:The PS/2 interfaces need pull-up resistors. I wanted to use the internal pullups of the ATMega8, but this is not enough. For PS/2 devices it is ok, but USB devices which come with an adapter sample these signals at powerup when the ATMega hasnt finished setting up everything.

Done!

simon wrote:The negation of the /MO signal (floppy) is wrong and the logic of the two /DS signals differ. Don't know, how I did these bugs.

Done!

Simon

simon wrote:

• Add a voltage monitor for CS1 of the real time clock as claimed by the datasheet.

I am seeking for a nice device to monitor the voltage. It should be small (8 pin DIP prefereable). I found the Dallas/Maxim DS1210 (which seem to be obsolete) and some devices from Maxim (MAX807, ...). I will have to compare the different devices.

simon wrote:[list]
[*]Power jumpers for both SIDs should be changed as well as jumpers for different filter capacitors should be added. At the moment it is possible to set jumpers for both 9V and 12V. This is not good. And I would add two more filter capacitors (switchable via jumpers) to have have the correct values for MOS6581 and MOS8580.

This is straight forward and my task for tonight.

simon wrote:[list]
[*]Add a LC(?) filter near the main power plug to get rid of the ripple noise and/or improve the overall PCB layout.

Still not sure if this would address the ripple noise. Lee mentioned another power routing on the PCB. Anyway, as for the schematic adding this is straight forward as well. This should be another task for this weekend.

Simon

simon wrote:I found the Dallas/Maxim DS1210 (which seem to be obsolete) and some devices from Maxim (MAX807, ...).

Well, the DS1210 is obsolete as it is not RoHS compliant. In contrast the DS1210+ is RoHS compliant. Probably I will use it for the RTC.

Bogo

simon wrote:

simon wrote:I found the Dallas/Maxim DS1210 (which seem to be obsolete) and some devices from Maxim (MAX807, ...).

Well, the DS1210 is obsolete as it is not RoHS compliant. In contrast the DS1210+ is RoHS compliant. Probably I will use it for the RTC.

Bogo

I've had the flu the last couple days. Well, I ordered a few DS1210 and created the schematic which should prevent the RTC from loosing it's time and date. I will build a small prototype PCB to test this solution before I merge it to the rev. 2.

BTW, are people reading this forum and are following my progress?

Simon

simon wrote:

• Power jumpers for both SIDs should be changed as well as jumpers for different filter capacitors should be added. At the moment it is possible to set jumpers for both 9V and 12V. This is not good. And I would add two more filter capacitors (switchable via jumpers) to have have the correct values for MOS6581 and MOS8580.

Done!

Simon

simon wrote:

• Add a LC(?) filter near the main power plug to get rid of the ripple noise and/or improve the overall PCB layout.

Done!

Well, I have finished reworking the schematics. When I got the DS1210 I will build a small drop-in test PCB for the RTC. If the test is ok I will proceed creating a new PCB layout.
If you want to contribute something to the schematics you will have only about two or three weeks left from now on...

Simon

people are reading this atlest me