Wilba

The only good (no noise) switchmode PSU I found was this one: http://www.jaycar.com.au/productView.asp?ID=MP3242&CATID=27&form=CAT&SUBCATID=381 All the other ones (i.e. regulated DC plugpack kind) I've used were noisy. I observed that if I "earthed" the negative output, the noise went away... but that's a bit of a hack and possibly dangerous or something. So I went looking for switchmode PSUs that had an earth connection (i.e. 3-pin power socket). I bought the one linked above and it's rock solid for sammichSID, not even a hint of noise, my transformer based regulated 12V DC plugpack is noisier than this one... but no good for sammichFM. Don't ask me why. Maybe because it is earthed instead of a basic AC/AC transformer.

What he said. Yellow wire is correct, pin is actually 11.

If you are resoldering R43, just use the excess lead from the resistor and solder to the pin of U22. D0 of JD1 connects via bottom track to the "via" below R31 between R31 pin 4 and 5, then via top track to pin 11 of U16 and then pin 5 of R30. So you can connect it to the "via", or to pin 11 of U16, or pin 5 of R30. If it were me, I'd use an insulated wire to pin 11 of U16.

Are you sure the noise is from the PSU and not from the SIDs? 6581 SIDs are pretty noisy anyway (i.e. very high noise floor).

I linked to one from Reichelt in this thread post: http://midibox.org/forums/index.php?showtopic=13161&view=findpost&p=121095 but link is broken now. Just wait a bit and my German slave nILS will assist you... :wink:

Do I flag this thread as **FIXED** ? :wink:

Very nice indeed! :thumbsup: However, it's a big tease if you don't show us how you made it... esp. parts that other MIDIboxers might be interested in... the case, how you made the panel, etc.

It sounds like there are multiple things wrong, which is hard to diagnose and fix remotely :wink: The easiest and quickest way to get this going again would be to send it back to me so I can swap out things like LCD, PIC, other ICs, etc. (and test your parts in a working sammichSID). Then anything not working I can replace straight away.

FWIW I used a 4x20 LCD on MB-6582 because it fit better... you can get components closer on the left/right sides, whereas with a 2x20 you can get components closer on the top/bottom sides. TK did use the extra two lines of the 4x20 but it's not essential, so you may not miss the extra bling. I was always thinking that a 3U rack design should try to use a 2x40 if possible, because the extra displayed items would be really useful, and you have plenty of width to work with. I was working on something like that a long time ago... MB-6582_rack.pdf

Nicely done, TAG! What tactile switch did you use for the bottom PCB? I considered doing this for my prototype but couldn't find a tactile switch with low enough activation force.... hence my labour-intensive extra contacts idea, which got very boring to make. One day I'll get around to finishing it so it wasn't a complete waste of time and money :wink: In the meantime, I bought an Axis-49 to satisfy my isomorphic keyboard cravings... there are some utilities to convert it to Wicki-Hayden (aka. jammer) layout. I saw this video and decided I needed a short-term fix: People watching this thread might also be interested in this too:

Awesome.

Be sure to check it is a "common pin" type resistor network... i.e. 5 resistors... and not the other kind with 3 separate, unconnected resistors.

I wouldn't say it's malfunctioning, it's just not very good tolerance. I would expect regulated supplies to have less than 5% tolerance. It should be OK for 8580 SIDs. If you don't get high enough voltage going into the 7809, you can try putting shunts in JBP to bypass the bridge rectifier. You must make sure the tip is positive (+) if you do that, but it will then put your 11.51V directly into the 7809 instead of losing ~1.2V from the two diodes in the bridge rectifier.

LOL Luckily, people who google image search for "SID synth" will find the results dominated by MIDIbox SID synths... my theory is it's because they are awesome.

I know creatorlars does silkscreen, and he's in the US.

If your aim is not to salvage a part (i.e. it's easily replaceable), then I advise sacrificing the part, chopping it into single pins which are then easy to desolder. You save a lot of time and it is less risk damaging the pads. After you get the part out, removing solder from small pads is tedious and annoying. I often use a vice to hold the board vertical, edge towards me, so I can heat the pad with a soldering iron from one side while sucking solder from the other. Sometimes adding more solder to the pad can help, so there's more of a blob to be sucked out, i.e. if it's only a tiny bit inside the hole, this can be hard to get into an easily suckable state. I have plenty spare resistor networks so just email me and I'll post you a new one.

I've been very slack and never got around to uploading the PCB layout PDFs to assist with troubleshooting. I guess 95% of the time, I diagnose what's wrong and people just fix their solder joints or replace parts put in the wrong way :wink: The troubleshooting section of the wiki now has PDFs: http://www.midibox.org/dokuwiki/doku.php?id=sammichsid#sammichsid_troubleshooting So now you can see which tracks lead between the PIC and the LCD header. There are 4 data lines and 3 control lines, all on the top layer (red).

If you are not getting text on the LCD (i.e. only "black bars") then that's the first problem you need to fix. The PIC has MIOS installed so you should get the MIOS boot screen. Check the connections between the PIC and the LCD header, solder joints on both ends of the LCD header, etc. It might help to remove the LCD from the CS PCB so you can resolder the joints without the CS PCB getting in the way. Perform continuity tests on all the pins of the LCD header, both ends, check there are no adjacent pins which are shorted, etc. Also, check that you are getting some MIDI output, i.e. the single "upload request" sysex message... or just use MIOS Studio 2 and query the sammichSID (you need both MIDI In and Out connected and MIOS Studio ports configured correctly)... it should report that MIOS is installed on the PIC. This will help validate that your PIC is running and should be sending data to the LCD.

A short beep when testing continuity between power rails is expected when there are capacitors between the rail. While the capacitor is charging up from the multimeter probe, it is fooled into thinking there is very low resistance. Once the capacitor is fully charged, there is no more current and the beep will stop.

It's a wiki. Login using your forum username and password. Edit the page. Preview before saving. Don't make a mess.

You may have shorts on the far left DOUT module (74HC595) causing the switch matrix to not work, buttons being pressed are thought to be other buttons being pressed. So check there first, then check solder joints on all the 74HC165 and check orientation of the resistor networks.

Nice! The list is not centrally maintained. Add yourself to the list! http://www.midibox.org/dokuwiki/doku.php?id=wilba_mb_6582_builders_list

I think the heat comes from both the 5V and the 9V/12V supply. They suck max. 100mA on the 5V, 40mA on the 9V/12V. So... P=IV... 0.5W + 0.36W (if 9V), or 0.5W + 0.48W (if 12V). I'm guessing... I don't really know the physics of this stuff :rolleyes: but that power has to go somewhere, doesn't it? So I guess it goes out as heat.

Some fans are noisier (on the audio) than others. From experience, fans with LEDs are often noiser. Powering from J25 is best because it's before the regulators of SID audio voltage (i.e. the 12V or 9V supply to SIDs). Also, dropping the voltage to the fan somehow will keep its RPM low and the fan noise down too, but I don't have advice on that... I suppose sticking an 7805 between J25 and the fan would work... it doesn't need smoothing caps IMHO. As for not powering SIDs... I can only guess that it's OK to leave out the 12V/9V supply to SIDs, since this is only powering the filter/amp inside the SID... let's say, I'd do it for my 6582A SIDs without worrying but for 6581 SIDs which spontaneously die on me even when packed in an antistatic tube, I would avoid even looking at them the wrong way, let alone rob them of power. :rolleyes:

They should work. For reference, I use Mouser part 511-M74HC595.