c0nsumer

Tonight I finished all the soldering and then fully tested and assembled one unit to be sure the end panels and such are right. Here it is. Now I just have to finish up the remaining PCBs, test them out, assemble them, and package them along with the PCB/end panel sets.

now you can sit back and watch the mayhem happen... sloopy. That's the point. Or more specifically, that's the middle of my workday. I don't want to take any time off from work for doing actual SDrive NUXX stuff.

[email protected]:~> sudo at -l Date Owner Queue Job# Thu Aug 4 14:00:00 EDT 2011 XXX c 10 [email protected]:~> Woo! Everything's set to automatically open ordering at the listed time.

It's all done via Google Checkout which takes credit cards.

I'm pleased to announce that orders for the second run of SDrive NUXXs, both assembled devices and/or PCB/End Panel/Microcontroller, will be accepted beginning on Thursday, August 4th, 2011 beginning at 2:00pm Eastern Daylight Time (2011-08-04 @ 14:00 UTC-4) via this page: SDrive NUXX Ordering Page Note that this page is up, but ordering is not live. At the date/time stated above the page will change and it will become possible to place orders. Pricing is as follows: Assembled SDrive NUXX: US$150 SDrive NUXX PCB, End Panel, and Microcontroller Set: US$40 These prices include USPS Priority Mail postage within the United States. For non-US destinations there is an US$10 shipping surcharge to cover the additional international postage costs for each SDrive NUXX or every five PCB, End Panel, and uC sets. When placing your order, if you are not in the United States please add an appropriate number of International Shipping options to your order. If you are an non-US customer and do not add this surcharge I will have to cancel your order and ask you to resubmit it. There are only 30 Assembled SDrive NUXXs and 20 PCB / End Panel / Microcontroller Sets available. After these are sold out the ordering page will be disabled. At this time I have no intention of a third run, so be sure to get your order in early. All items will ship within a few days of orders being placed.

I'm pleased to announce that orders for the second run of SDrive NUXXs, both assembled devices and/or PCB/End Panel/Microcontroller, will be accepted beginning on Thursday, August 4th, 2011 beginning at 2:00pm Eastern Daylight Time (2011-08-04 @ 14:00 UTC-4) via this page: SDrive NUXX Ordering Page Note that this page is up, but ordering is not live. At the date/time stated above the page will change and it will become possible to place orders. Pricing is as follows: Assembled SDrive NUXX: US$150 SDrive NUXX PCB, End Panel, and Microcontroller Set: US$40 These prices include USPS Priority Mail postage within the United States. For non-US destinations there is an US$10 shipping surcharge to cover the additional international postage costs for each SDrive NUXX or every five PCB, End Panel, and uC sets. When placing your order, if you are not in the United States please add an appropriate number of International Shipping options to your order. If you are an non-US customer and do not add this surcharge I will have to cancel your order and ask you to resubmit it. There are only 30 Assembled SDrive NUXXs and 20 PCB / End Panel / Microcontroller Sets available. After these are sold out the ordering page will be disabled. At this time I have no intention of a third run, so be sure to get your order in early. All items will ship within a few days of orders being placed. -- Since I've also been talking a bunch about production status here, I'm quite close to finishing up the PCBs, with probably one more evening of work to go before assembly will be complete. After this comes basic electrical testing, insertion of uCs, initial power-up testing, final assembly into the enclosure, functionality testing, and then packaging. I expect to have this complete by August 4th so that shipping can happen as soon as orders are processed, but there's a slight possibility (depending on how late I end up working some evenings at my day job) that they won't be ready to ship until a few days afterward. And yes, I'll probably have some more photos to post tonight. (After this post I'll go back to this thread being used solely for production status updates and whatnot. The thread about orders is here.)

I've got 30 being assembled, 29 will be available, and I'll post the date/time ordering info once they are all available. I intend to give everyone at least 24 hours notice of when orders will be accepted, and then on that date/time the order page will go live. It'll stay up until all complete devices and pcb/panel/uC sets are sold.

Here's some more teaser / production progress images. There's a bunch of PCBs with all washable parts fitted, flux residue on a board, a clean board after washing, and the boards set out under a fan to finish drying. Tonight I finished soldering all the washable parts, washed all the (water soluble) flux off with hot water and a soft bristled brush, then blew most of the water off using compressed air. The boards were then propped up to dry under a fan in an air conditioned basement (read: dry air) overnight. After they have ~24 hours to sit I'll solder on the non-washable parts (switches, SIO connector, DB25) with no-clean flux. Then comes testing, fitting into enclosures, more testing, and then packaging.

...And made-in-USA, for the most part. F. I could say Designed and Assembled in Detroit if I wanted to stretch things a wee bit.

The SD slots are a lot easier than it looks, I managed to do 2, and repair a Dingoo with a similar (but smaller) slot @Consumer, do you have any prices? I am thinking maybe building a very barebones version using one of your PCB's and hide it in my 1200xl Agreed. These SD connectors are quite easy. There's a bunch of space between each pin and the only thing to worry about is not using too much solder. For pricing I'm looking at US$150 for an assembled device, and US$40 for a PCB/Panel/uC/screw set. All prices include shipping in the US. For international shipping it'll be an extra $10 for each assembled device and for every five PCB / end panel sets. The work-in-progress ordering page is here if you want to see it, but orders there just go to the sandbox right now so they can't be completed. As mentioned earlier once the devices are complete and packaged I'll set a date/time when I'll start accepting orders, announce that here, and the order page will go live then. This should give everyone a good chance to order without a surprise start time, since once the parts are sold out they will be gone. For the time being I'm not planning on a third run.

No. The kit will only contain the PCB, end panels, and preprogrammed microcontroller. I don't currently have any extra SD connectors to offer this as an option, either, although it does seem like a good idea. I wonder how I could do something like that... Hmm... Hmm. I'm thinking I could do pre-soldered SD card connectors on the PCBs, but that'd push the cost up $10. Those connectors are almost $5/ea, plus shipping for me to do another order just for them, plus my time. I'm not sure it'd be worth it.

No. The kit will only contain the PCB, end panels, and preprogrammed microcontroller. I don't currently have any extra SD connectors to offer this as an option, either, although it does seem like a good idea. I wonder how I could do something like that... Hmm...

Tonight was capacitor soldering time, which means more progress is being made. Next up? The last couple capacitors, the LDO voltage regulator, then the LEDs.

Hehe, thanks. I'm working through them. I just have to find some time to finish them off, but I figure that the PCBs will be assembled in a week or so. (I've got a bike race on Saturday and loads of other personal plans that have to come first.) I'm going to be taking orders for and shipping it all at the same time, just to make the post office trips easier. I'm thinking end of the month is not an unreasonable target.

Indeed! You can't say c0nsumer doesn't document thoroughly! I find that photos tend to be super-convenient documentation a few years down the line. Looking back through them gives a bunch of hints for how things were done. It's also kinda nifty to look at old projects and remember doing bits of them.

Here's another. The box of these was sitting on my front porch and I didn't find it until after posting the previous image. I now have all the requisite parts. Now I can kit up the PCB+Panel+uC sets and finish the rest. Yay! (Note that there is protective plastic on all the panels; that's not wear or scuffing.)

Here's another teaser image. All resistors, the diodes, and the DIP socket are now fitted. After this it's on to the pin headers, crystal, capacitors, voltage regulator, LEDs, and DIP switches. Then it's time to wash and dry the boards.

Any of you work anywhere with larger HP or Xerox copier/printer units? These commonly scan to network as PDF via the automatic document feeder. 15 minutes standing over it letting the feeder do its business, a bit of reassembly in Acrobat, and voila!

Sure thing. The ATmega8 is super-common so it's probably supported. One thing to remember is that AVRs need a clock present for serial programming. So, if you set a AVR to use an external oscillator (done via the fuse bits, and the SDrive uses an external oscillator) and one isn't connected the chip will simply appear dead. You have two ways around that: Use HVPP (high voltage parallel programming) which doesn't require the clock, or program it in the SDrive NUXX board where the oscillator is attached. When I first started using AVRs I found this out the hard way, as I'd thought I was repeatedly killing chips when programming them. Turned out that as soon as I set the fuses for an external oscillator serial programming would simply stop working. To get around this for the batch programming of the SDrive chips I use an AVR Dragon fitted with a ZIF socket in HVPP mode. This, along with avrdude allows me to flash them from command line on my Mac. For what it's worth, the command line is as follows: avrdude -q -p m8 -P usb -c dragon_pp -U lfuse:w:0x3f:m -U hfuse:w:0xdf:m -U flash:w:SDrive.hex:i -U eeprom:w:SDrive.eep:i (This is the setup that was shown a few posts back.)

30 assembled, 20 PCB / end panel / microcontroller sets. To be honest, since I'm doing this from home and fronting all the costs I'd rather make enough to ensure that I sell out reasonably quickly. I'm not a shop (just a guy who likes making things), so while I want to make these available to people I'm not that interested in sitting on a back stock or doing a massive production run. Also, hand-assembling these is a bit tedious. If I were to do a larger run I'd probably redesign it as all surface mount, microSD card, make it much smaller, and have them assembled for me. But, if I did that I'd be cutting out the DIYers (of which I'm one) who want to build SDrives themselves; something that I don't want to do.

Sure thing. And yes, it should be pretty easy. I very intentionally used the stock SDrive firmware and made the device completely pin compatible so that other firmware (or future versions from C.P.U.) could be used on it without issue. Specifically, maintaining that compatibility is the reason for the mess of vias / signals that need both sides of the board (as seen here) between the AVR and the LEDs/switches. I wanted to rearrange the pins to make the layout simpler, but it wouldn't have been possible to do that without slightly changing then recompiling the firmware. I didn't want to add that layer of hassle for users who may wish to change the firmware somewhere down the line, so I stuck with the same pins that C.P.U. used.

Here's thirty (30) SDrive NUXX PCBs with SD card connectors fitted. Yes, this means assembly is under way. The first (and only non-through hole) part is now fitted, so I can now move on to everything else.

As long as the Willem programmer will do ATmega8 chips then it'll be fine. The AVRs are socketed, so you'd just pull it, reprogram it in the Willem, then reinsert it. Or you can just use the onboard programmer; your choice. Er, also, you could use the on-board programming header to plug in an external programmer. The MISO, MOSI, SCK, Reset, and GND lines are exposed on a .100" header. Just power the SDrive NUXX externally, remove the SD card, and use the header to program with your existing programmer, if you want. Just be sure grounds are connected together.

SteveS didn't make the SDrive (he's just the GURU) [edit] to add that SteveS DID make his own Ghetto SDRive (or whatever he is calling it this week) Some guru! I still haven't figured out how to put the hodge podge of parts of my "Barebones SDrive" (that's the name du Jour) into the case I chose for it. I have no excuse for failing to complete the build since this thing doesn't have any buttons or LEDs but I'm talking about v2, a standalone version. The original project was just an idea I had to turn a 600XL into a game console. It needed to have mass storage and it needed to be cheap. Unfortunately the 600XL rebelled before I had the chance to install the Barebones SDrive and it's now the 600XL is DOA.. (I plea the 5th). This is the Barebones SDrive as I actually intended to use it (minus the temporary cable). Disclaimer: Don't try this at home! I mean it! -Steve Sheppard Thinking about this, you could probably do a bare bones SDrive on a PCB not much larger than the SD connector itself. Or go microSD and the whole thing would be no larger than a stick of gum. The problem would be soldering, but if you have 30 or 50 or so made and outsource the assembly to a shop with proper assembly line stuff it probably wouldn't be very expensive. You could then just blow the firmware on there via a header.

SteveS didn't make the SDrive (he's just the GURU) [edit] to add that SteveS DID make his own Ghetto SDRive (or whatever he is calling it this week) Sorry, I should have said guys who make SDrives. I very much know it's the C.P.U. guys who did the original SDrive firmware and schematic.