After another night of working on synths, I have something to show for it. Tonight I got the PAiA FatMan I am rebuilding (photo gallery retired) working, and as a test I had my x0xb0x triggering it, and both of their audio mixed together.
Listen here: http://www.nuxx.net/files/x0xb0x_and_fatman.mp3
Also, flying wires to a front panel are a pain in the ass. Photos later of why.
Well, this is good. I’ve been in touch with MaverickLabels.com about the error in the printing of the x0xb0x panel overlays. I’ve been told that the reprinted panels will either ship today or tomorrow. This makes me quite happy, because hopefully I’ll have the synth finished by the end of the week. So, while there was a problem with the printing, that company is definitely doing a great job getting it sorted out. I feel a lot better about this all now.
Now, just two more synths to finish… Well, that’s before the 9090 project (and this will be a HUGE project) begins. More on these other two later.
[This is a cross-post from here in the ladyada.net Forums, redone in HTML instead of BBCode.]
Well, that should whet people’s appetites.
That photo above means that I’ve received the panels from MaverickLabels.com. There are both good and bad things about this…
First, the good:
– The panels look excellent. The colors are exactly what I wanted, and the print quality is outstanding.
– Everything appears to align nicely to my x0xb0x panel.
Now, the bad:
– See the image above? See how it is missing the lines from the bank indicator switch through hole to the numbers? They aren’t right.
This means that I’m not going to sell / ship them yet… MaverickLabels.com is closed until Monday, at which time I’ll be calling them and seeing what they will do to sort out the problem. I find it interesting that these lines are missing, as the digital proof which MaverickLabel.com sent has them in there.
I am hoping that they will sort out the problem and get corrected labels shipped to me quickly.
In the mean time, here’s a few more photos involving the finishing of my x0xb0x and the panel overlays:
– High res scan of one of the panel overlays I received tonight to show the detail of the printing and the error. (Please note that the colors aren’t accurate, and the sparkly appearance is because of the texture of the Lexan.)
– Stack of all 12 panels, not long after they were removed from the shipping package.
– The original proof sent by MaverickLabels.com.
– Glare-filled comparison of an x0xb0x panel overlay (with error) and the original x0xb0x panel.
– A smaller version of the scan of the panel, with the missing lines pointed out. This was made so I could send it to MaverickLabels.com.
So, that’s that. Everything is on hold until Monday when I can talk to the Customer Care people and see what they are going to do to sort out the problem. I don’t see a problem with things getting sorted out, especially if they are willing to put a guarantee like this on their site. But, we’ll see…
Per usual, I’ll keep posting updates here.
-Steve
So, yeah. Tonight I tried to make the printed circuit board detailed in this PDF on this page for making it possible to trigger my Casio SK-1 via MIDI.
Well, I learned two lessons in doing this…
First, my technique for registering the two sides on a double-sided PCB appears to be successful.
Second, my method of laser printing the mask on vellum and exposing to fluorescent light for 10 minutes isn’t good enough for fine lines. As you can see above, something went awry and the photo resist between the lines wasn’t exposed to enough light.
I think I’m going to try an experiment with various levels of detail and a smaller single-sided PCB with a piece of laser printer transparency, just to see how well this method will work. I don’t think it’ll work out so well, but it can’t hurt to try…
If this doesn’t work, I guess I’ll have to order some special paper just for doing this kind of thing. That won’t be cheap…
Well, at least I’ve got a few more unexposed boards to play with.
Ah. I just did it. I ordered a set of 9090 panels, ROMs, and the PIC.
Not sure how I’ll trigger it yet… That’s an exercise for another day. Maybe a MIDIbox SEQ or something.
Hmm… Should I build a 9090?
In short, it is the synth portion of a TR-909, but without a sequencer…
Well, last night I finally received my Mouser order with the parts that I was waiting for.
Hmm. Perhaps I should backtrack a bit… After finishing the electrical assembly of my x0xb0x, I found that the LEDs were way too dim. After building the Elenco XP-720K power supply (photo gallery retired), I was able to test out the LEDs with different resistors and find some appropriate ones. It turns out that a 820Ω resistor in conjunction with the LEDs worked out to be exactly what I wanted, bright enough to be seen under a bunch of lighting, but not glaring when in the dark. So, last week I desoldered all 40 resistors which tie to the LEDs, and last night I hooked in the 820Ω resistors, put the x0xb0x back together, cleaned off the flux, then got it all working.
After that installing the new resistors I went through and set everything up with the new Fluke 179 and got it all set up right. The frequency counter in it was especially handy, as it made setting the offset between octives (this synth is 1V/oct). Not long after the synth was all done being tuned, and it sounds great. Then I broke it… More on that later.
Yesterday, as a prerequisite to tuning the synth, I also finished up some work on the Elenco XP-720K (photo gallery retired) by adding heatshrink tubing to the worryingly unprotected IC leads, installing new Bourns 10-turn wirewound pots in place of the original 300° carbon ones, and new knobs from EPD (450-2016). The new knobs were needed as the shafts of the new pots are of a different design, and I also wanted some which allow for a bit finer control. While I like how they look I’m still not sure if they allow for fine enough control. We’ll see…
With the new pots, I am able to easily set the supply voltage to within a few thousandsths of a volt, and with some gentle nudging of the knob I can generally hit the exact voltage I want. Because the pots are wirewound there seems to be a limit to how fine the control I have actually is, but things are still fairly accurate. Adding a second set of wirewound pots (200Ω) as a ‘fine tuning’ control in parallel with the existing ones would help give more control, but it would require drilling more holes and… I don’t know if it is needed. Regardless, I am quite happy with the finished power supply.
So… About how I broke my x0xb0x. Well, to make a long story short, I was playing around with writing my own firmware for it, and after flashing it one of the times, it no longer worked. Details of all what happened can be read here, but the solution ended up being flashing it with alternate firmware, then going back to the original firmware, then wiping the patterns in the EEPROM. I think I had corrupted part of the EEPROM, and as the EEPROM is read as part of setting the tempo as the x0xb0x boots, I think I was making it crash on boot.
Also, I’ve found my firmware with the blinky startup sequence causes my x0xb0x to crash randomly, so that needs work. That’ll come later. Thusly, I’m just using the normal, and originally shipping, v1.04 firmware.
I think I’ll also tag some specific photos of interest here:
· The pile of LED-driving 10KΩ resistors pulled from my x0xb0x.
· Calibrating the x0xb0x’s power supply to 5.333VDC.
· Playing the x0xb0x via MIDI. As a nice surprise, a velocity over ~100 seems to trigger accent.
· The newly heatshrink tube’d IC leads in the Elenco XP-720K.
· Rear view of the Bourns 3504S-1-202 pots in the Elenco XP-720K.
· Two front views of the front of the Elenco XP-720K with the new pots and EPD 450-2016 knobs: 1 · 2
Well, that went well. What you see above is my first hand-etched PCB, and is also a layout I did myself. That right there is the PCB I’m going to add to the PAiA FatMan which I am rebuilding, and provided I interpreted the schematic for the part correctly, that should add two extra octaves of harmony to this synth, giving it some nice, thundering bass.
I should mention that the traces are a shiny silver in person, although I’m not completely sure that I tinned them enough. Also, yes, one of the pads on the left side isn’t quite right. I’m not sure what is with it, as it appears to just be matte in person. It doesn’t matter too much, though, as the reason I made two is to both use up space on the board (it’s 2″ x 4″ and my design was nowhere near that large) and so that if one got screwed up, there would be a second.
So, I guess at this point I’m ready to try etching the double sided board for adding MIDI to my Casio SK-1. I would have done it tonight, but it is getting a bit late to start everything over from scratch…
Anyway, yeah. This makes me happy. Hopefully tomorrow I’ll be getting the order from Mouser which will contain the new resistors for my x0xb0x (photo gallery retired), parts for the FatMan rebuild, and the new pots for the Elenco XP-720K (photo gallery retired).
Mmm… Projects…
UPDATE: Damnit. I don’t know how I didn’t notice it until just now, but there is a 90º bend in one of the traces, when I tried to do everything in… Well… Not 90º joints. Guh. I feel like an idiot, because i know this is one thing that I was trying to avoid… And I somehow missed it in all the times I stared at the printout.
Ah well, It’s only a power line, so it shouldn’t hurt anything… I still feel dumb about it, though.
On the left is my brand new Fluke 179 which just showed up today (thanks, Chad!). In the middle is the Radio Shack DMM I had received from my parents a number of years ago as a gift. My first DMM, if you could say that… On the right is a crappy ‘Cen-Tech’ brand DMM which I purchased from Harbor Freight solely to compare a batch of transistors.
While I didn’t expect the Harbor Freight meter to be even remotely accurate (after all, it cost $3 or $4), when hooking all three DMMs to the same power supply, it is surprisingly close to the Fluke. Wow.
This isn’t to say the Fluke wasn’t worth it, as it is amazingly responsive, easy to use, and does all sorts of things well… And being properly calibrated and all, I can trust its results. I’m just amazed that the general piece-of-crap from HF was more accurate than the one I’ve been using for years. I guess I should have replaced it sooner.
Hey, look. It’s the first PCB layout I’ve ever done… Seems to be okay, too. The autorouter didn’t do as good of a job as it could have, but that’s okay. This isn’t what I’ll actually be producing the board from, I just wanted to be sure I had a good understanding of how to do it.
Oh, and that board up above there will be for this subharmonic generator for the PAiA FatMan, the two octive version. It’ll be connected to the board and front panel via flying wires. (Yes, there will be photos once I actually do it… The parts are on order.)
And yes, that should say “Subharmonic Generator” in there. I’ll fix it later.
