Wilba

Same pinout? It's completely different! It's a single row of 16 pads at the top edge, instead of 2x8 on the left side. The ultimate white/black LCDs are the Optrex STEP displays: http://www.optrex.com/products/partdetail.asp?PartNumber=C-51505NFQJ-LW-ALN http://www.optrex.com/products/partdetail.asp?PartNumber=C-51505NFJ-SLW-AIN I have the white on black one. It will work, but the pinout is slightly different, the backlight pins are at the opposite end of the header, which means you need to cut a couple of pins on the header and connect them to the pads on the LCD with two wires. It's an easy job to do, if you're interested, I'll post instructions and a photo of how I did it. You'll probably need a new male header for the LCD, I have plenty of spares.

It sounds like this display is like the ones I have, they only have a single LED on the side, and so they don't need the maximum current supplied by the standard MIDIbox Core backlight circuit (and thus also MB-6582 base PCB), which is around 280mA or more. I recommend inserting an 82K resistor in series with R4 (the 1K) or replacing it with machine pins (like you probably used for the SID filter capacitors as recommended by the build guide) so you can try different values. This will reduce the maximum current supplied to the LCD backlight. I'm surprised the LED hasn't burned out already. Perhaps I'm totally wrong. What colour is this display? Does it look like it is uniformly illuminated or is it brighter on one side? If it's a negative white display (or anything other than yellow-green), it's probably edge-lit (single LED). The typical current (25mA) suggests only a single LED. This doesn't quite explain the flickering though... if it was a current supply issue, i.e. too much current draw by SIDs + LCD, then it would stop flickering when there was only one SID. What's the J3 measurement with only one SID? 4.5V DC is too low, it should be much closer to 5V (like 4.9V or more), so there might be a current supply issue at play here too, so you can confirm that by measuring J3 voltage with different loads, like one to eight SIDs installed. It can't hurt to try replacing T1 and checking the solder joints in that area, in case a poor solder joint causes this intermittent flicker... but also, if it's only happening on power on and not later, that's suggesting the load is too much on power on and 5V supply is unstable but then becomes stable later. So you'll have to consider both possibilities (bad solder joints/T1 vs. current supply).

Good to hear it works now, but I don't understand what you mean... can you draw it on the image I uploaded in the previous post?

POIDH! Good to hear it all works now. :D

IC2 is the optocoupler, it's only used for MIDI In. The two wires of a MIDI cable are "looped back" via a LED inside the optocoupler, so it is electrically isolated from the device. The phototransistor inside the optocoupler is what generates the signals to the PIC. It should have no effect on MIDI Out (i.e. if you remove IC2, MIDI Out should still work). The LCD datasheet suggests you could run the backlight directly with a 5V supply, it has its own current limiting resistors, however, an additional 128mA on the 5V supply might make the 7805 run too hot. So in this context, you could probably run it with JBL=5V and JR4=closed (i.e. normal MIDIbox Core backlight circuit for high-power LCDs, supplied with 5V).... turn the brightness pot down as much as possible to keep current low (avoid heat in heatsinks). The resistors on the LCD should keep the current under 128mA if the supply is 5V (JBL=5V). FWIW you can test actual current draw of the LCD by connecting your multimeter to read current (amps) passing through the two pins of JBL - instead of a shunt in JBL, you connect the multimeter to JBL instead, so the current goes through the multimeter. That way, you could try the "high power LCD" jumper settings and tweak the brightness pot to get the exact current you want (i.e. <128mA)

Are you measuring this "no resistance" with the IC installed? If yes, take out IC3 and try again. See if there are any shorts between adjacent pins of IC3 socket (and pin 7, 9, 14, I don't know why you checked these before but keep checking things again to find anything that isn't right). Try also manually lighting the LEDs via the pins of IC3 and IC4 sockets (remove the ICs). You can connect one of pins 2,3,4,5,6,7 of IC3 to ground, then connect one of pins 15,1,2,3,4,5,6,7 to 5V. This should power one LED of the matrix. You can take 5V from pin 16 and ground from pin 8 of either IC3 or IC4 sockets. Do not put too thick a wire into the IC socket pins or they will become looser and won't make good contact with IC pins afterwards. Something like a cut resistor lead is OK, or the solid core wires used in prototyping "breadboards".

Reinstalling IC2 might have fixed something, if the contacts were not good before. Can you show me a link to the LCD datasheet? AFAIK there is no requirement for LCD "LED array" backlights to have a specific voltage, like 5V, maybe this is a minimum voltage because of forward voltage drop of two LEDs in series. If it says max current of 128mA (a strange value, actually) then you should configure the base PCB headers JBL and JR4 for "high-power LCD backlight" and set brightness pot less than half-way (probably 1/3 is best).

http://www.midibox.org/dokuwiki/sammichsid

Maybe this picture can help. You can see where the LED cathodes for the top right row are connected to the 74HC595. It's possible the solder joint there is bad, or there's a short somewhere to another pin. Test connectivity between the LED leads and the pin of the IC from the top (don't assume the IC socket pin and the IC pin must be connected).

Let's assume that since you did upload successfully, the sammichSID MIDI Out should be working and unlikely to have suddenly stopped. Therefore it's probably a setup issue. You should be able to do a loopback (PC MIDI In connected to PC MIDI Out), play some notes on the keyboard and see events in the MIDI In window.

Ho ho ho. Ya koo tocha ka poonoo nee sok nyee.

The white/blue LCDs just have one LED on the right side. I got a bunch of these LCDs cheap, but the LEDs were green. Don't ask me why they made such a yucky combo. Anyway, I just bought some white LEDs, took out the green LED and put in a white one. So what you can do is, test if the LED is working by itself, and also test if there is power going to the backlight pins of the LCD header (on the base). Testing LED is easy... just use a 5V supply through a resistor, anything between 100 ohm and 1K will be OK for this test. Disconnect the LCD from the base PCB (it can stay screwed into the CS PCB). Touch the wires onto the pins of the LED, where they are soldered to the LCD. You can also check it at the top two pins of the header, these pins would connect into the B+ and B- pins of the J16 header on the base PCB. You can test if the backlight pins on the base is working too. Make sure you have JBL set to 5V, and JR4 doesn't have a shunt. Stick a LED into the B+ and B- pins of the header. A spare LED that came with the kit is fine. Cathode (short lead) goes to B- pin. Most likely you have a bad solder joint somewhere, as I test each LCD before packing... but in the worst case scenario that the LED isn't working, I can send you a new white LED, it's pretty easy to replace... this is the main reason I chose to "recycle" these LCDs because they have easily fixable backlight LEDs and the replacement LED is a common 3mm LED.

sammichSID is an implementation of MIDIbox SID V2 Synth, which uses two SIDs in a stereo pair. Go to www.ucapps.de and click on the links in the left navigation bar under MIDIbox SID V2 Synth Such as: http://www.ucapps.de/midibox_sid.html http://www.ucapps.de/midibox_sid_manual.html MIDIbox SID V2 Synth works like a SidStation - it's a MIDI controlled synth, with internally stored patches, loads of parameters for tweaking, etc. No disrespect towards SidStation, but MIDIbox SID has a lot more features in it's synth engine, and two SIDs sounds fatter in lead patches (Lead engine) and can be used for polyphonic patches (Multi engine) or two simultaneous bassline sequences (Bassline engine).

Are you receiving the upload request in the MIDI In window?

MIOS Studio only uses MIDI ports. TM-1 should work, but not in any turbo mode, only standard MIDI baud rate.

OK something is wrong. It only takes 82 seconds for me. For the upload to work, both the PC and the PIC must be running their serial comms at the same speed, so it's not an issue with that... i.e. each BYTE is being sent and received OK, but there must be massive delays between bytes. The problem would therefore have to be at the PC end, since there's nothing at the PIC end that could slow down an upload. What kind if MIDI interface are you using? Perhaps there's a conflict with other USB devices or it expects to be in a USB 2.0 port or some other issue.

grizz: that's the one I normally buy for kits. I've had to buy Mouser 512-MM74HC165N for the current batch of sammichSID. I think Digikey had some also but they were more expensive.

"forever" because of errors or just slow to upload? Post screenshot after the upload, it should show the speed at the end. It couldn't possibly be slower than my M-Audio Delta 1010LT.

I forgot to mention another option... the lead or wire that's supposed to be soldered to a missing pad can instead be soldered to a track that connects to the missing pad. Just scrape the solder mask off the track. However, for JD6 and JD7, it doesn't matter what the pad is like on the bottom, it's only connected on the top layer to the resistors.

12V AC 500mA works. If using unregulated, any higher than 1000mA (1A) is too much, input voltage would be too high because the current load is much lower than the rating. If it's regulated, it doesn't matter if the current rating is high, as voltage will be constant.

I don't think they're reattachable. If a pad is broken, you would attach a wire to the lead or wire in the hole, and then connect the other end to another pad which would be connected to the broken pad by a track. In this context, with a pad on the other side still good, perhaps you can solder the new wire at the same time as soldering the remaining good pad - i.e. poke it through if the wire and "part" are on opposite sides.

Possibly you have a short between the encoder pins/switches and the heatsink. It could also be a bad connection between the PCBs, or a bad solder joint on the 74HC165, i.e. if pins are floating around then you will get the effect of random button presses and encoder turns. I don't think it's related to power supply. 12V regulated 1.5A is good, you have plenty of power and it won't run too hot.

"switchmode" and "transformer" are the two types of plugpack internals. Switchmode will be noisy for sammichSID - you'll hear it in the audio. Most of the new plugpacks are switchmode, especially if they're light and high current... I'm pretty sure practically every mobile phone charger is now switchmode. Electronics stores should still stock the old "transformer" kind.

Yes... I really like what you did with the extra buttons.

Livid's boards don't support RGB SMD LEDs as far as I can tell, only monochrome. Maybe you're thinking of SparkFun? https://www.sparkfun.com/commerce/product_info.php?products_id=7835 The problem with those is the cost. SparkFun's ones are cheaper and are known-good to support even fat 5mm RGB through-hole LEDs. 10 USD for 16 rubber buttons, so 160 USD just for 16x16 rubber caps! Oops I forgot the other 33 buttons on the proto... so it's more like 200 USD just for the rubber caps! Not many people would want to spend that.... VERY TEMPTING THOUGH! Looks great TK... can't wait to see it with the panels and button caps!