My new baby!

[Moggy]

And I love how you have them on your wall - I need to do that to win back some space on my workbench.

Yup that MIDI unit won't build itself.

[nollkolltroll]

Just read again and saw:
"exposing serial and jtag lines"
Wow! Maybe it will be possible to download code directly to the zeddy and/or to the sd-card? It sure would speed up my turn around time when developing...

[sirmorris]

I have to admit I've been considering that option myself for a long time now... It's just been a matter of priorities.

[Paul]

It's just been a matter of priorities.

MIDI MIDI MIDI MIDI MIDI MIDI MIDI MIDI !!!

[sirmorris]

[Moggy]

You laugh when paul says it yet you took out a restraining order against ME when I GENTLY SUGGESTED MIDI!


Makes 10 x 8 picture of Charlie and spends rest of afternoon throwing darts at it.

[sirmorris]

So I was finally moved to attend to the issue that I was up till now pretending didn't exist which was that zxpand+ would reset every time an sd card was inserted. I knew this was related to the fact that sd cards can flatline a power supply when they're hot inserted. They draw a large current and that in turn makes the available amount of voltses dip. Usually I'd just throw capacitors on the board until the errant behaviour stopped.

However this time it didn't. So I sat down and thought a bit and realised that the classic 'spand design doesn't do this. You can insert cards to your heart's content with nary a flicker. This time I thought I'd be clever and hook the SD card to the new 3.3v regulator that powers the CPLD. It turns out that CPLDs don't like spikes or dips in their power supply and as far as I can tell the device was browning out and resetting. This isn't an issue on my other devices where the CPLD isn't in charge of decoding memory accesses as well. There the cpld might be resetting but it doesn't matter so much.

So after another few thinks I changed the board to use the classic design with the meatier 5v supply on sd-card feeding duties, via a neat little voltage limiting arrangement. I was about to send off for more prototypes when I thought of a different way of going about it.

IMG_3482.jpg

(3.2 MiB) Downloaded 243 times

Here you see the result of me going about it differently. I quickly designed and etched a replacement SD card pcb which has the new power arrangement 'pon it. Jammed it onto my working board, and hey presto! validated the design without another £100 layout and 2 week delay. Woohoo!

You can also see the work I've done trying out different mixers for the AY. I hacked together a board which has the AY outputs and audio out taken off to jumpers. I can replace the resistor network just by popping one off and slipping another on. Sweet! Here's another view.

IMG_3486.jpg

(3.73 MiB) Downloaded 243 times

So I have no idea if anyone is interested in this waffle at all but perhaps it shows that I'm not just twiddling my thumbs ignoring the elemidiphant in the room

[sirmorris]

I've also decided against having 2 reset buttons. While more is definitely more it wasn't in this case better. I'm going to stick with the 90 degree switch on the edge of the board. Having used both for a while now it doesn't seem odd anymore to reach down to the side of the machine. It also removes a lot of wobble potential. Not that you're generally bothered about wobble when your finger's on the reset button but it satisfies some part of me that wants to see the elimination of as much wobble as possible.

It will also make case designs simpler. Existing cases can have their end caps modified, a much easier proposition than poking another hole in the back.

[nollkolltroll]

Looking very nice! Can't wait to buy one.

[siggi]

So I was finally moved to attend to the issue that I was up till now pretending didn't exist which was that zxpand+ would reset every time an sd card was inserted. I knew this was related to the fact that sd cards can flatline a power supply when they're hot inserted. They draw a large current and that in turn makes the available amount of voltses dip. Usually I'd just throw capacitors on the board until the errant behaviour stopped.

This kind of problems can also be caused by too thin +5V and GND lines. High currents (even for a short time) can cause voltage drops AND level shifting of the GND reference potential along the lines, caused by resistance AND the inductance(!) of the lines. So normally thick +5V lines and thick GND lines or (even better) GND planes(!) are used to avoid this problems (in combination with capacitors).

For test puposes you could add thick wires for +V5 and GND, leading from the ZX81 connector directly to the SD-card, and some capacitors (10uF + 100 nF) next to the SD card.

Regards
Siggi

PS: http://www.zx81.de/cooltour/hymne.htm