The BurgerDisk gets two little siblings.

Colin Leroy-Mira's BurgerDisk and DominoDisk just received a firmware update that lets Apple II users load DOS3.3-ordered .dsk images directly.

#### General availability Since the end of the Kickstarter, a few people who missed it contacted me to get a device. A few people also expressed frustration at wanting to get one, but finding the shipping cost to USA off-putting. The first thing is solved: I have opened a small web shop on my site. The second thing is being solved: I have contacted Joe of Joe’s Computer Museum, and he’s interested in distributing variations of the BurgerDisk in the USA. Variations, because building a BurgerDisk is kind of a lot of manual work. I designed two devices that are easier to assemble (their PCBs are less DIY-friendly, both go hand in hand – they’re still open-source) for him. Keep reading for more details. #### “Fat fingers” option I’ve designed a variation of the microSD module for full-size SD cards, and a variation of the enclosure to accomodate for it. The SD module I have no idea how to ship just the module internationally without losing money on it, so for now it is only an option when buying a complete device. It’s also open-source, so anyone can have them made easily. A “fat, old fingers” version of the BurgerDisk #### Different form factors The first device I designed for Joe is the Mini BurgerDisk. It is 10.5×7.5cm. The second one is the DominoDisk, which I am really proud of. I think it’s the smallest Smartport hard-drive, chainable or not. It’s small enough that you can still put your Apple II beneath the Monitor //c’s stand. The DominoDisk plugged at the back of my Apple //c, with a 5.25″ drive connected to it Setting up and booting from the DominoDisk Both of these devices share the same features and firmware as the BurgerDisk. #### Last bit of news: the price of the device Some of you may be aware that the D-SUB 19 connectors that the Apple II uses are not manufactured anymore, and are hard to find. There are ways to find some on eBay, and that’s what I have been doing so far. The prices on the eBay listings, however, are… something, and the availability uncertain. I already ended two seller’s supplies during the Kickstarter, and this was not sustainable. This is very much still in the air, but I’ve been trying to find a way to source those connectors in larger numbers and at lower prices. For the male connectors, I bit the bullet, and bought 500 from Big Mess O’Wires. Steve had ten thousand of them custom-made a few years ago and shares :-) For the female DB-19 connectors, there was no such easy way. I called every electronic component shop in my city, just in case, without a lot of success. The conversation with most of them was short after a “Sorry, we don’t have those since years”. One of them embarked me on a short emotional rollercoaster, with the conversation going like this: * “I’d like to buy D-SUB 19 connectors, female. Would you happen to still have any?” * “Let me see…” <noises of rummaging in drawers> “Yes, I do! How many would you want?” * “How much would it cost for 50-100 of them?” * <more rummaging> “I got… three of them, so 9€” (I bought the three of them). I have then scoured the internet, using different search engines, a lot of search queries in quotes, a lot of Google Translate, and I have managed to find some leftover stocks of what probably is what I want in various Eastern European countries. Probably, because some of the pages only mentioned a reference, “no picture, no description, let’s hope”. Some of them agreed to ship internationally. Some did not… I reached out to people who seemed nice on the Fediverse, and got help from two exceptional persons, Alexey in Serbia and Kenneth in Estonia, who agreed to take time to help a random nerd, receive my orders, and send them back to me! Packages have not arrived yet, but if everything goes well, I should have secured about 300 female connectors – from North Macedonia, Lithunia, Serbia, Estonia and Hungary. If this works, this should be more than enough to meet the demand from the whole Apple II community, and it should also allow me to make the devices 15-20€ cheaper, even with double-shipping costs and customs duties. #### Sharing and caring Also about price, 10% of each sale’s benefits will go to our local transgender rights association, Clar-T.

Following the release of BurgerDisk, I have been asked what it would take to handle daisy-chaining in Fujinet. I am going to explain what I’ve done and what I’ve found in this post. ## The basics – being a Smartport device Implementing a Smartport device and having it work alone on the bus is rather easy. There a few lines from the Smartport connector that we’re going to use. **Line**| **Use** ---|--- +5V| Power for your device GND| Ground for your device PH0 aka REQ| The four stepper motor phases, and REQ PH1| PH2| PH3| RDDATA| Data read WRDATA| Data write WRPROT aka ACK| Write protect signal, and ACK Apple has repurposed signals in a smart manner, so that non-smartport drives would not react to Smartport packets. ## Smartport bus basics The Smartport bus has three states: * RESET: PH0 and PH2 are high. PH1 and PH3 are down. * ENABLED: PH1 and PH3 are high. PH0 and PH2 are ignored. * DISABLED: any other state. When the Apple II boots, it resets the Smartport for a short time. Devices notice that only PH0+PH2 are high and act accordingly. Afterwards, each Smartport packet is exchanged with the bus ENABLED, PH1+PH3 high. When the packet has been exchanged, the bus goes back to DISABLED. When the bus enables, the computer wants to send a packet to a device. A device informs the computer that it is ready to receive the packet by raising ACK high. The computer replies by raising REQ (aka PH0). Data is exchanged, then the device lowers ACK to indicate reception, the computer lowers REQ, and the bus disables. This is simple. The necessary connections look like this (**Note:** the pinout is **not** the one used in BurgerDisk. The pinout shown here was cleaner to draw the schematic): ## Chaining with other devices Of course if multiple Smartport devices are on the bus, things get a bit more complicated. First, a device must know whether a given packet if for itself or another device. Devices have IDs, which are assigned at boot, right after the Smartport RESET: the computer sends INIT packets, with incrementing device IDs, until everything got an ID. In order to avoid every device getting IDs at once, when the Smartport bus RESETS, every device in the chain must pull PH3 low on its daisy chain port. This has the effect that every device after the first one sees the Smartport bus as DISABLED, and will not be able to answer INIT packets. Our device handles its INIT packet(s) and registers its ID (or IDs, if multiple volumes are present). When it is done, it stops pulling PH3 low on the daisy port, and goes back to mirroring it. The next device sees the bus ENABLED, and is able to answer the next INIT packet(s) and register its own ID(s). **This means that our device has to control PH3 on the daisy port.** How does the computer know that everything has an ID and the INIT is over ? In each INIT packet response, the device sets a flag to inform the computer “there are more volumes after this one” or “we’re done”. To determinate whether there is another Smartport device on the Daisy port, a device has to check whether the HDSEL signal is grounded. **This means our device has to ground its input HDSEL (in order to be recognized as a Smartport device), and check the daisy out HDSEL (in order to recognize the presence of a next Smartport device).** In addition to that, if the device behind is a “dumb” floppy disk drive, we have to disable it while the Smartport bus is enabled. Otherwise, even if the phases lines don’t make sense to the stepper motor, the drive will start running. **This means that our device has to control DRV1 and DRV2 on the daisy port.** Basically, we have to: * Disable daisy PH3 before we have our ID. Mirror PH3 as soon as we do. * Mirror DRV1 and DRV2 when the bus is disabled. Disable them when it is enabled. The rest of the lines can be shared. Our schematic now looks like this: Notably, we added diodes to the WRPROT (ACK) and RDDATA lines to avoid undesired interactions with other devices driving them. Some input lines that we don’t have to filter are connected directly from IN to both the microcontroller and OUT (PH0, PH1, PH2, WRDATA). Some input lines we don’t use at all are just connected from IN to OUT (EN3.5, WREQ, +12V, -12V, and of course +5V and GND). Inputs we have to filter are connected from IN to the microcontroller, and from the microcontroller to OUT (PH3, DRV1, DRV2). ## The case of /DRV2 DRV1 and DRV2 lines control enabling dumb 5.25″ drives behind the Smartport devices. LOW means enabled, HIGH means disabled. DRV2 is only used on computers that can have two 5.25″ drives on the external bus: the IIgs and the IIc+. But this line is, once again, repurposed by Apple, and on old 5.25″ disk controller cards, this pin (17) is a +12V line, which would fry the microcontroller if the user was to connect it directly to the Arduino. To be safe, it has to be decoupled. We will use an optocoupler: (**Note:** I didn’t yet receive the PCB batch with the optocoupler setup and it might be wrong). ## The firmware First we have to setup pin modes: * WRDATA (D9 here) is input pullup * DRV1 IN and DRV2 IN are input (D7 and D8) * PH3 OUT (A5) is output. We pull it low at boot as we expect a RESET. * DRV1 OUT and DRV2 OUT (A3 and A2) are outputs * HDSEL OUT (A4) is input pullup * By default, WRPROT (ACK) is input, RDDATA is input pullup. This allows the device to be silent on the bus when required. The precise algorithm of the firmware is best explained in code, but here is a bit more detail about the important question of “how to not interfere with another device when a packet is not for us?” When the Smartport bus enables, we pull ACK high and so do every device in the chain. We receive a command packet, which contains the destination device ID. If the packet is not for us, we have to let the other device answer, without interfering. This is done by: * not acknowledging the packet (not pulling ACK back low) * muting ACK and RD lines * waiting until REQ is back to low (the computer has acked the answer from the other device). On the other hand, if the packet was for us, we: * acknowledge it by pulling ACK low * wait for the computer to pull REQ low * answer the computer with our response packet Afterwards, we always re-mute ACK and RD. **Watching out for infinite loops** As soon as we need to ignore packets, it starts being dangerous to receive packets when the bus is enabled: when REQ is back to low, we’re going to notice that the bus is still enabled, and start trying to receive another packet, pulling ACK high and waiting for REQ to be high – which may never come. So, our ReceivePacket function must switch from a simple /* tell computer we're ready */ set_ack_high(); set_ack_output(); while (req_is_low()) { /* wait for computer ack */} while (wrdata_is_low()) { /* wait for start of transmission */} // receive packet to verifying that the bus is still enabled: /* tell computer we're ready */ set_ack_high(); set_ack_output(); /* wait for computer ack as long as bus is enabled */ while (req_is_low()) { if (smartport_get_state() != SP_BUS_ENABLED) { return ERROR; } } /* wait for start of transmission as long as req is high */ while (wrdata_is_low()) { if (req_is_low()) { return ERROR; } // receive packet ## More? I think going into more details in this page would make it harder to digest, and most questions/doubts should be answered by studying the schematics and firmware code. However, if I’ve left out important details here, feel free to ask questions in the comments. Note that I don’t claim to be an expert of the Smartbus, and am also far from an expert in the hardware side of things. What I made here makes sense to me, seems logical and safe, and does work on different computers with different daisy-chaining setups, so I think it must be a fairly good implementation.

(For those who missed it, BurgerDisk is a device I am making, a daisy-chainable Smartport hard drive for the Apple II. Kind of like the SPIISD, but chainable, both ways). ## Two DB19 options The first difficulty with this device is that DB19 connectors are hard to find. They’re not manufactured anymore, and stocks have dried up, leading to annoying prices and uncertainty. This is especially true for the female DB19 connector, the one at the back of the device, that other devices will plug onto. Female DB19 connectors come in two flavours, “Solder type” and “Angle mount”. “Solder type” means the connector is equipped with solder cups at the back, and they were primarily used to make cables. “Angle mount” means the connector comes with pins, angled 90° from the connector, and they were primarily used to solder on PCBs. In order to ease procurement as much as possible, I first made a PCB where you could use either one or the other. The first PCB’s double connector The first PCB with a SolderType connector soldered The first PCB with the Angled connector soldered This was OK, but not great: when using the Angled connector, one would have to cover the edge connectors with tape, or risk shorts with the DB19 body. The options cardinality starts here, with two versions of the PCB. the v2 PCB, with edge connectors for the SolderType DB19 the v2 PCB, with pin connectors for the Angled DB19 Both connector style must be at the same place in the enclosure (right at the back), so the PCBs are not exactly the same size, and the constant is the relative positive between the connector’s metal body and the mounting holes. Of course, each DB19 connector will not be at the same height relative to the PCB: the Solder Type one is aligned vertically with it, while the Angled one is on top of it. The connector’s height difference (angled one is upside down as its pins are bent and don’t fit anymore) I wanted to be able to have a single enclosure, so I needed to be able to modify the PCB’s height inside it. I did that with washers of different heights: 7mm and 2mm. If using a SolderType connector, you can put 7+2mm under the PCB. If using Angled, put 2mm under the PCB, 7mm over the PCB. In both cases, the total height is the same, so the mounting posts counterparts in the top section of the enclosure are the right length and maintain the PCB correctly. One of the multiple test prints trying to line up everything This project taught me that iterating hardware is much more time consuming and expensive than iterating on software. For the enclosure, I print small parts instead of the full ones until I’m satisfied with the result. This spares both time, money, and ecological footprint. A test with the SolderType connector Result is starting to be satisfying ## The Apple logo I have made a small Apple-shaped notch in the front of the enclosure, so that people who happen to have an Apple logo from a vintage, broken device they don’t care about can put it in the front and have their enclosure look Real Good. At first, I thought I’d need a version of the enclosure without that notch – that would be better for people who don’t have that logo – but I found that a 3D-printed blank logo looked cool. So, we are still at a single version of the bottom part of the enclosure, usable with both type of connectors and with/without the Apple logo. The placeholder logo The real logo ## The SD card form factor So far, the BurgerDisk’s storage has been a microSD. But in this community, we’re starting to get old, and these things are so small. So, I wanted the option to store data on a full-size SD card. Of course, both modules have different footprints, and I still did not want two versions of my enclosure. So, I sized my module support for the full-size SD module, and added screw mounting holes for both of them. the MicroSD module footprint the full-size SD module The mounting posts for both modules As you can see from the images, these two modules have different pinouts! The full-size one has two grounds, can be powered by 3.3V, and also… they’re in a different order. I handled that on the PCB, providing both connectors, so it’s easy to connect to one or the other without having to fiddle with pin order. The mission to have a single version of the enclosure for any option was complete ! But… ## The (micro)SD slot’s witdh and height The microSD’s slot is approximately 17mm wide, while the full-size SD is 29mm wide. The full-size SD’s slot is also 0.4mm higher. 3D-printed things tend to not like “overhangs”, where matter is deposited over nothing. This is why things are often printed face-down. Given the lip join between top and bottom parts of the enclosure, I couldn’t do anything but make two versions. If I only did one, with the larger slot, it would be possible and easy to insert the microSD in the enclosure instead of its holder, which would be very annoying. So, there are two versions of the enclosure’s top. Choose the one you need. If that proves to be a bother, I could also decide to drop the MicroSD version entirely, and just have people use an micro-to-full-size adapter if needed. ## In conclusion While I am not super happy that I have to maintain two versions of the enclosure’s top, I’m still quite happy with the fact that there are only two variants of the enclosure, while it still manages to handle 8 differents possible combinations of (connector / logo / SD size).

While I’m still waiting on the next step of the Kickstarter (finalizing the pledges and addresses should take them another week), I’ve kept myself busy and imagined variants on the BurgerDisk device. As I said elsewhere, I would love that everyone who wants to be able to use modern storage **and their floppy drives** at the same time, to be able to do so. The Kickstarter proved two things: * I’m far from the only one wanting to do that * Exporting from France to the USA is expensive, and off-putting some enthusiasts At least it’s extremely expensive at my scale. For me, shipping 400g of things outside of the EU is 30 to 40€, and to the USA? Add the orange fascist clown’s tariffs and it’s all the way up to 65€. Luckily, Joe from JCM-1 is interested in providing USA people with my device. What he’s not very interested in, however, is spending one hour assembling every one of them, which I can understand. The BurgerDisk requires quite a bit of work to gather parts and assemble, as can be seen from the Bill Of Materials: main PCB, adapter PCB, MicroSD PCB, Arduino, 139 solder points, LED cabling, box printing (5.5 hours). It is possible to order assembled PCBs for the main PCB, sparing the need to solder components and headers, but all the rest will still be needed. So Joe suggested to me that I look into options that would make for easier assembly. I have looked into that and am experimenting with two smaller designs: the “Mini BurgerDisk” (which does not have a special, cool name for now), and the DominoDisk. Both of these get rid of the Arduino Nano and the MicroSD module, integrating both into the main PCB, leaving only the need to solder the DB19 connectors. A family of daisy-chainable Apple II hard drives Peeking inside the enclosures, it’s easy to see how simpler the assembly of the two small siblings is: The three models, opened. The two small versions have exactly the same features as their big brother. They’re all cabled the same so share the same firmware. They have an ICSP header for updating it. The ATMega328P-AU on them is configured the same, so uploading firmware is the same process. They both have an LED, and of course a daisy-chaining port. So far, both of them are prototypes and are not ideal. The MiniBurgerDisk you see here has the decoupling capacitors layout all wrong, and the DominoDisk is much too deep at 54mm including connectors. I’m working on a 4-layers version of the PCB that will allow it to be more compact. Still, in its current version, it’s compact enough that it doesn’t interfere with the IIc monitor stand when that sturdy Disk //c is plugged into it. Viewed from the side, there is ample room to put the Apple //c under the monitor stand Of course, these devices will also be open hardware. I hope that both of these versions will be ready in April or May!