Another IIci ROM hack

[tt]

Well, that's the part I'm still trying to figure out. I haven't routed anything manually yet.

Got it...if I got to Design, "Make Rat's Nest" it shows the connections.

I'm looking at my Radius "SE30 Bus Adapter" and it has the 1-40 row numbers going the other direction, basically A1 in the top-right corner for the orientation of the board you posted. Not sure if this is an issue...

 

[bbraun]

Yeah, that's what I'm concerned about. Maybe I made the part backwards?

 

[tt]

I think so...it the notch/naming on the Radius board seems to agree with the GttMFH. Trash can check too, I think he has the same board.

 

[techknight]

I've got what I think is the connector, the chips and the connections all into Osmond. I'm not sure I've got the orientation of the board correct. As in, the chips on the right side of the board when the card is installed into a machine.Anyway, my files are all here. The 030Card.osm file is the project file and should be all that's needed. The others define parts, list of parts, and connections. Osmond is a free download, and works on OSX, so if anyone wants to take a look, feedback is appreciated.

U1-U4 are 29F040's

U5 is 7404

U6 and U7 are 7408's

U8 is 74125

I can/have made a simple app and an AVR that reads an SD card, prints its contents in the app, and gives me 512bytes in which sector/LBA that i request. The problem lies in the ability to take that, and make it into a driver that allows finder to use it. Something that I cannot do.

Hook me up! I'll take a stab at it. Or if you'd like to do it yourself, the ROMDisk driver should be a fairly simple C based example using CodeWarrior 10. It's really just a lot of glue that boils down to a memory copy. The complications usually come in when the memory copy becomes something more complicated, like calling another driver, or something along those lines. I'd be glad to help out in any way.

I haven't figured out how (or if it's reasonable) to make something appear to the SCSIMgr. Everything I've done is more along the lines of the floppy & EDisk driver, that just present a "drive" to the OS, which is a linear array of 512byte blocks. When the "drive" structure is given to the OS, your driver is associated with it and the Prime routine will be called whenever a read or write operation is performed. The drive will show up in the Finder when a diskEvt event is posted using PostEvent. If the disk needs formatting, you'll get the "Uninitialized disk has been inserted, do you want to eject or initialize" dialog.

Icon retrieval and information about the size of the disk, whether it can be ejected, is read-only, etc. is handled by the driver's Control call.

Ill go and dig through my files and try and find it. its been long ago, and i made it as a proof of concept precursor to making an MP3 player. (SD card, not USB).

But about the mac and C, Thats the whole thing. I dont program in C, and i know very little about it. Everything programming related that I have always done was in BASIC/Visual Basic.

I compile AVR Code in BASIC, the few mac apps i have written was in RealBASIC. I cant code C. hehe. Also the libraries I used for the SD card were part of the compiler I use, and uses the SPI of the SD card which is about as slow if not slower than the MAC Serial. LOL. But it was a proof of concept, and it did work.

 

[uniserver]

i agree with Trash, not to say you would even use DDR2 ram, I have literally a bag, 66 sticks to be exact, of 512meg DDR2 533mhz ram.

my point is, ram disks are shot down because of cost, yeah a TB of ram is still rather pricy!, but were only talking about maybe 256mb to 512mb's would be plenty!

and sick fast! throw a 3.3v lithium battery on there just incase there is a power spike. USB wold be cool, popping in a 8gb sd card that costs only a few bucks, that can mirror ram by clicking on an icon, and then pooping that into any modern computer of your choice.

 

[techknight]

To do routing, in most, if not all CAD software you need whats called a Netlist. This is the network persay, that ties the entire system together.

So you cant really draw a line from pin 1 of chip 1 to pin 2 of chip 2 unless you have a netlist that ties the 2 together to whats called a "ratline"

Every cad software is different, but the one I use is like this. Called FreePCB (which is free).

Where does the netlist come from? Well theres 2 ways. 1 you can build it manually. but, 2, it comes from the schematic. For example EAGLE. If you draw up a schematic and it has no errors, it will build a netlist and import it into the CAD router. This way you CANT make mistakes.

 

[techknight]

The easiest way for me to do a serial based mass-storage device with an SD card for starters, is having a MAC driver as you say to present a disk to finder with 512byte sectors.

But say if we can draft up a protocol quick-like, we can make this work.

For example a protocol that would do like this:

(Serial port)

Send a byte out the serial port to tell the external device that I request a read, or a write. So, $0A for read, $0B for Write (hypothetically).

So in this example, $0B for Write.

Receive ACK.

Send LBA Addy.

Recieve ACK.

Send 512bytes data + CRC.

Recieve Ack once complete. ACK here serves 2 purposes. 1, to let the driver know the data was received properly. 2, the external device is no longer busy processing the write. Driver wont get the Ack until the write is complete, 2 birds with 1 stone.

a NACK would indicate bad checksum/data and require driver to resend.

Then a check somewhere that if it fails too many times, report write error to OS.

Same thing comes with a read.

$0A for read

ACK

Send LBA

ACK + Data+ CRC

Driver checks validity and replies ACK/NACK to external device. NACK means the device will resend the data. Failure here too many times, report read error to OS.

Something like this, I can code microcontroller software to do. But i need the MAC software/driver/control panel to handle this. Now of course, this is dirt simple. about as simple as it gets. No provisiing for control such as maximum LBA size, etc... isnt provisioned here. this is simple read and write.

LBA size could be another command, like $0C.

$0C = Get Disk Capacity.

ACK + total blocks LBA + CRC

 

[techknight]

Oh, and not to mention if we code this up, we could put it in ROM, no? then we could boot old machines off a serial based SD card reader. SLOW... YES... but it would work!. Besides im thinking Serial-SD card for sneakernet anyway. not for booting, or huge file transfers.

And once the protocol is finalized and in place, we could make the PDS card work, only difference being is that will run a parallel data operation instead of serial. Or you can just use a PIO chip that can interface to 030 and add a "third arbitary serial port" that only the driver knows about

My goal personal wise is to replace the modem card in my mac portable with an SD card reader and sneakernet things back and forth without floppies. Hell i could even throw a mux in there and toss a serial based WiFi card in there and have network at the same time as SD storage.

We kinda need a new thread for this as this is a whole different tangent. Oh well at least i might be on to something? But i could engineer the hardware and firmware to make this happen. But i would need someone such as bbraun and/or doug3 to write the mac driver to interface.

I laid out a simple protocol as a starting base, maybe someone can refine it or change it. either way. That could be the starting tree if we can iron out a solid protocol then driver to hardware interfacing would be simple. Just follow the protocol and everything would be ok :scrambled:

Edit: P.S. The device could be wireless too. In theory. Serial SPP modules are dirt cheap, and with that in mind if they are paired together, can create a serial wireless link. So in theory it could be wireless. This could be handy for localtalk. I have done this once with xBee but i dont like those modules that well, but with xBee modules they can work in broadcast mode so multiple machines can be on the same wireless channel and be on the same network. BT is point-to-point so that isnt an option.

 

[bbraun]

tt: Thanks for catching that, I think I've corrected the orientation. I've updated the project file with the renumbering change.

techknight: sounds good. I've sent you some protocol feedback, and for a serial device I can get started by doing some serial comms against a modern machine just to test it out and get a jump start before having hardware.

 

[trag]

Sooo cool. I wish I had time to participate.

Remember, when using Osmond (or any PCB software) to pay attention to whether you're seeing a view of the layer or a view through the board to the "back" of the layer. I laid out an 8 pin SOIC mirrored on my first board because of that.

Here's a little old discussion about SCSI Mgr, which might be interesting. It's too bad that "Driverguru" just dropped in for a few days and then went away again. Still, if you get to where you need some specific questions addressed, it might be worth trying to track him down:

http://68kmla.org/forums/viewtopic.php?p=61401#p61401

I had always thought it terms of interfacing a USB chip directly to the Mac's bus, but the idea of using a microcontroller as a DMA/driver engine for the USB controller is great.

Still, I think it might be worth taking a look at the USB chip I referenced up thread. I didn't read the whole datasheet, but it said something about an on-board microcontroller in the introductory blurb. With the configurable I/O pins, one might not even need any programmable logic to interface it. And it says it has 5V safe inputs. If that really means what it says, level shifters should be unnecessary.

http://www.ftdichip.com/Support/Documents/DataSheets/ICs/DS_Vinculum-II.pdf

This one is available at DigiKey in the 64 pin QFP for ~$6.

Edit: Bah! Never mind on the above chip. It has a really nifty 40 pin GPIO mode which could probably be made to interface with the 68030 bus without any glue, but while it is "USB 2.0 compliant" it does not support 480 Mbps (High Speed) because "the power requirements are deemed excessive for VNC2 target applications."

Double Bah!

 

[tt]

tt: Thanks for catching that, I think I've corrected the orientation. I've updated the project file with the renumbering change.

Cool...It looks consistent with my artifacts.

But....you forgot to include the right-angle pass-through connector on the other side. ) But now I'm reminded of a question I've had, is the PDS slot completely parallel if you want to add more slots say for a dummy adapter that Trash has brought up?

 

[bbraun]

Thanks. Passthrough is beyond my routing abilities at the moment. In fact, I've routed 7 traces in 3 hours and needed 4 layers to do it. How many layers can I have? I'm thinking I'll probably need 40-60 layers.

 

[tt]

Yeah, I was J/K, that would add many more lines to route. I would try to get it in 2-layers otherwise the fab cost is much higher. Then in general the options seem to jump from 2 to 4 layers for protos.

Maybe the traces you are using are too wide? This shot shows you can fit two traces in between each pin: viewtopic.php?f=29&t=19369#p183281 Also, you can use vias to jump sides of the board if you need to cross-over a trace.

 

[Trash80toHP_Mini]

I think so...it the notch/naming on the Radius board seems to agree with the GttMFH. Trash can check too, I think he has the same board.

Actually, mine's the DayStar Adapter with provision for adding Zeners to damp over-voltages on the signals.

Come to think of it, these cards would be perfect as a Prototyping Riser/Chassis for playing with the PDS in your open topped IIsi, bb, or on your bare MoBo SE/30 testing config, tt. I'll bet macmetex still has at least half a bucket full and the price is right.

Just de-solder the RA PassThru Connector and solder headers onto the back side of the board, then solder a prototyping board directly onto the headers and you're good to go. Add the Zener Diodes if noise becomes problematic.

You can now wire wrap directly from the headers to wire wrap sockets for the ICs. Heck, the first Mac Mobo was made in wire wrap . . .

. . . as were the first two PC Prototypes. PaleoTech ROCKS! :approve:

 

[bbraun]

I would try to get it in 2-layers otherwise the fab cost is much higher. Then in general the options seem to jump from 2 to 4 layers for protos.

Haha. You're funny. No seriously, if all I get are 4 layers, I may as well stop now.

 

[tt]

Well, let's see what you have... I am playing around a little bit with it. BTW, do you know how to swap routed traces to another layer in Osmond?

 

[bbraun]

I've got the current state here. I've got vcc, ground, and 6 data lines, down to 3 layers.

 

[Trash80toHP_Mini]

How many layers can I have? I'm thinking I'll probably need 40-60 layers.

You must be looking at PCB layout the way I do, like a graphic that makes sense in your head. Doin' it for real requires calming your mind, letting go of your beliefs and expectations . . . trust the force.

IOW, do the rat's nest thing and let the software worry about the routing, that's where application of the force comes into play. Wielding the force like a Paduan Learner ends up with too many required layers as well. By using just Apprentice level Jedi finesse earlier in the thread, dougg3 managed to pull out of an autoroutation death spin by manually placing a few, key, guide path inducing traces before using the force.

A real PCB guy is a Jedi in his application of the force. A Master PCB guy plays Jedi Mind games likely to make Master Yoda's head spin. )

edit: we're back up to p.40 again after being there temporarily . . . what's up with that? :?:

 

[trag]

I don't know what's going on with needing so many layers, but I have a couple of suggestions.

First, use the tutorial. I think you did already, but just in case you skipped the bits on routing....

Second, read the specifications for the PCB service you hope to use and set your trace width and trace spacing to match. I believe that Seed Studio allows down 6 mil (6/1000") traces. Signal traces do not need to be any wider than that.

Third, keep the annular rings around your vias small. You can probably make them about 10 mils, (20 mils wider total than the via drill hole) and be okay, although I remember reading some handy guidelines somewhere once.

Fourth, this is more a matter of style, but if you're not using a netlist, first choose your PCB size. Then put your components on the board, generally where you expect them to land. Then use the (connection?) tool to connect the various pins which will connect together. Now you have a rats nest.

Now start routing the traces. When you need to add a via, you'll want to create the via in place, then use one of the tool (forget which one) to connect the via to the rats nest trace. Then cut the trace between one of the pins and the via. The pin should change colors because the netlist still says it should be connected, but it isn't any more. Then switch active layers of the board, and connect the pin and via back up on the other layer. If the pin doesn't exist on both layers, you'll need a second via, of course.

At least that's the way I remember doing it.

Fifth, try to use one layer for horizontal traces and the other layer for vertical traces. Switch between them with vias. This is not a firm rule -- it's down right floppy, but it's a useful state of mind.

Here's a two layer board I laid out. There are something like 300 pins on both connectors and I put it on 2 layers:

Both:

Front:

Back:

And here's the resulting board:

Didn't work though. That's a two layer board trying to act as a 40 - 50 MHz CPU adapter for the ANS. One of these days, I'll get back to it...

The JPGs are kind of low res. If you like I can upload the Osmond file when I get home tonight. It's about twelve years old, so I'm not certain that the file format is still compatible.

Actually, come to think of it, I have 50 or 60 saved files for this one design, if it hasn't been lost in the hard drive catastrophes over the years. Fifty or 60 saved files, because when I'm laying out circuit board I save often, and always do Save As, so I can revert to an earlier design.

So, my collection of saved files really is a record of the progression of the layout. Would that be helpful?

 

[James1095]

I never had much luck with autorouting, usually it makes a big mess of things and needs a lot of manual cleanup and optimization so you don't end up with 5-10 vias on a single trace, noisy traces running through sensitive analog areas, etc. I manually route all my PCBs, it's an art, it takes time and patience, but the result IMHO is worth it.