exxos blog - random goings on

[exxos]

So this is what the next-gen Falcon ST-RAM board would basically look like.

NG.PNG (75.35 KiB) Viewed 1271 times

[Badwolf]

Typique!

IMG_5971.jpeg (207.5 KiB) Viewed 1269 times

BW

[Badwolf]

So this is what the next-gen Falcon ST-RAM board would basically look like.

Inline termination resistors?

Not going to use 4-way packs?

BW

[exxos]

Inline termination resistors?

It's a "T" type filter. 2 resistors in series with a center tap to a cap to 0v. It's the only way to squash it all. It does help decouple noise going back and forth between the videl and RAM as well.

There is still a small amount of noise on the RAM side but it's below 200mV and not ideal probing .. Well, don't care I've beefed up the power rails a bit to see if that helps as well.

Been trying to figure out the impedance of it all. I think 1mhz into 100pf was about 1.5k. So the videl shouldn't see to much abuse from the filtering. The rise time goes from about 6ns to 12ns. But unavoidable really. Overall even if there is more of a load on the signals, it can't be as bad as having all the abuse from all the oscillations and spikes anyway.

Edit:

RC filter basically. Filters out crap over about 30mhz.

[Badwolf]

Inline termination resistors?

Its a "T" type filter...

Aha, proper EE stuff.

I know you're not hand assembling them, but that many passives makes me itch.

BW

[exxos]

Aha, proper EE stuff.

I know you're not hand assembling them, but that many passives makes me itch.

You can have a PCB if you want

Getting them assembled other than the RAM. I'll have to try and transpose RAM from a already built board to keep it a fair test. Not looking forward to going through all that again.

[Badwolf]

You can have a PCB if you want

Ah, that's very kind but it turns out I have a surfeit of RAM boards now!

BTW, I've been working on resuscitating my last DFB1r5 board today (the one I earmarked for you, but melted a socket on).

Progress! Two bodge wires so far...

IMG_5976.jpeg (213.04 KiB) Viewed 1213 times

Just a problem with DMA, I think. Hopefully just one more bodge.

BW

[Badwolf]

Yep, BGK was broken. Another DFB1 enters the world.

IMG_5977.jpeg (164.29 KiB) Viewed 1208 times

Only a 36MHz oscillator in there, so 224% is about right.

Now, three bodge wires in the top half of the chip is disappointing. I'd better check the lower half of the data bus before I even think about soldering on any SDRAM. Also: not sure I have any, actually

BW

[Badwolf]

TT-RAM all good (40MHz oscillator this time):

IMG_5978.jpeg (186.37 KiB) Viewed 1193 times

The only issue is that the FPU doesn't work. I seem to remember I had a problem with this socket when I first fitted it. It may just need replacing, but this board is so beaten up and the FPU is such a fickle beast anyway, I'm not going to attempt it myself.

So, @exxos, this one's for you to play with. I'll chuck it in the post on Monday.

IMG_5979.jpeg (251.24 KiB) Viewed 1193 times

BW

[exxos]

The only issue is that the FPU doesn't work. I seem to remember I had a problem with this socket when I first fitted it. It may just need replacing, but this board is so beaten up and the FPU is such a fickle beast anyway, I'm not going to attempt it myself.

So, @exxos, this one's for you to play with. I'll chuck it in the post on Monday.

OH nice one

I will probably break it before getting to the FPU tests anyway, so all good If something lasts more than 10mins in my hands, then you can considering it a more final design