|
Bloody posted:but what if i want to get data in/out faster?? how do i do like real usb or ethernet as painlessly as possible?? some sort of fpga would be my first choice. they can be inexpensive (for prototyping anyway) and have blocks for all sorts of things like USB, Ethernet, HDMI, etc.
|
# ¿ Nov 26, 2014 11:49 |
|
|
# ¿ Apr 29, 2024 12:39 |
|
I'm hoping my miniSpartan 6+ LX ships soon, I should start hacking together a VHDL description of NuBus for it I ordered it because it had both a ton of GPIO and HDMI out, my goal is to make a video card for my Mac IIci that can handle 1920x1080. I'll be able to connect it almost directly to the NuBus, just adding bidirectional level conversion, and everything else can be done in "software"
|
# ¿ Nov 26, 2014 22:22 |
|
how many gates would an FPGA reasonably need to support to emulate a Lisp Machine CPU? like first-gen, the CADR/Lambda/TI Explorer
|
# ¿ Dec 1, 2014 03:48 |
|
movax posted:the low-end parts from altera or xilinx should be able to do it easy (spartan-6, artix-7, cyclone iv or cyclone v) -- there was a recent kickstarter for a spartan6 devboard which would work, or check out digilent's de-0 or de-1 the Lisp Machines used custom CPUs specifically designed for ease of Lisp implementation. (which isn't the same as "running Lisp in hardware" as some people claim.) they also had writable microcode, so the lowest-level operations could be made as fast as possible the CADR emulator and Explorer emulator codebases provide C implementations that could probably be used to author some VHDL with effort, the main thing I'm wondering is if I'd get partway through and find out that lol, my Spartan 6 won't have anywhere near enough gates I suppose since the original systems were literally a few square feet of 7400-series chips I should be able to estimate the gate count from that (hell, even just based on physical density)
|
# ¿ Dec 1, 2014 07:16 |
|
movax posted:ahh, i see -- i still don't think you'd have a problem with the larger gate count spartan / cyclone devices, sounds like an interesting project though (especially to re from the existing emulators). wonder what the original were clocked at according to the Explorer docs at the Internet Archive, they had a 7 MHz microinstrunction clock. the processor board manual has a decent explanation of the internals including the microinstrunction format, hypothetically it should be enough to build an emulator for the cpu itself.
|
# ¿ Dec 1, 2014 07:57 |
|
movax posted:i'd be surprised if either the nexys-3 (spartan 6) or nexys-4 (artix 7) wouldn't be enough -- hopefully you can get academic pricing I ordered a Scarab miniSpartan 6+ LX a while back, hypothetically it ships soon, and I'll find out!
|
# ¿ Dec 1, 2014 08:06 |
|
Werthog 95 posted:hey so here's a hobbyist's question for the actual experts itt what sucks about it, is it something you could fix in the firmware? I have to say that this is in fact the very pro-est keyboard. at work I'm still using the first one my family bought with our Mac IIci in 1990, via the Griffin iMate adaptor. I have to plan for when the adaptor dies though, it has a button cell in it. so I'm keenly interested in whatever you come up with, pls subscribe me your newsletter.
|
# ¿ Dec 1, 2014 22:11 |
|
Poopernickel posted:I'd say one of the larger Spartan 6s would be plenty. Buy the biggest FPGA that the Webpack supports (probably either the LX45 or the LX25). the scarab miniSpartan 6+ I have on the way has an LX25 and 32MB SDRAM so I should be good. thanks!
|
# ¿ Dec 1, 2014 22:13 |
|
suffix posted:is codesourcery still a thing? how's LLVM/clang as an embedded compiler these days? should have pretty good ARM support.
|
# ¿ Dec 2, 2014 05:09 |
|
Jonny 290 posted:feels too Commodore-1541 to me how is this a bad thing? I mean, I've always been an Apple guy, and the IWM was an amazing hack, but giving the 1541 its own CPU made it awesome
|
# ¿ Dec 6, 2014 08:20 |
|
any suggestions on what level of scope I'd want for hacking on old-computer projects, with things like 25 MHz CPUs and 10 MHz busses? like, is a dirt-cheap modern scope fine? how about a used 100 MHz scope? is that overkill? would I be better off with a logic analyzer? saw an HP 1660 with kit for what looked like a decent price yesterday…
|
# ¿ Dec 14, 2014 22:40 |
|
check out this retro bit janitoring! a 68K single-board computer with ISA slots? boy howdy! I first read this when it was a series of articles in Radio-Electronics magazine in 1987, it was in large part how I learned to read schematics and how computer hardware all fit together at the low level does anyone even make cases for the PC-XT form factor any more?
|
# ¿ Dec 24, 2014 00:55 |
|
Mr. Despair posted:have a scope too Tek I was looking at those Saleae logic analyzers someone mentioned but it looks like, for the price of their 16ch model, I could get a used HP 1660CS with probes 136 channels! LAN interface with X11!
|
# ¿ Dec 24, 2014 01:54 |
|
just found out about OSH Park, which does small-scale and very inexpensive PCB fabrication with a quick turnaround. I'd been worried a bit about this thing I want to do because at 10 MHz I expect it'd be over what would normally work well with breadboard, wire wrap, and so on. but this place is dirt cheap, and a couple people I trust recommended it. I also found what looks to be a decent inexpensive layout app to use with them. so how cheap is it? per square inch per lot of three four-layer boards! (and half that for two-layer.)
|
# ¿ Dec 26, 2014 05:58 |
|
Jonny 290 posted:do we need a separate surplus thread, or should i just megapost surplus post would be a nice yosmas present any opinion on halted (HSC) vs the others? thinking of heading there tomorrow to buy poo poo, not just ogle it.
|
# ¿ Dec 26, 2014 06:13 |
|
Bloody posted:Also ten megs on a breadboard is going to be fine, your tenth order harmonics are still more than a meter and the parasitics will be irrelevant I was going by the last post in this thread on the 68 Katy and the other site linked from it which made me think running a >1 MHz bus onto a breadboard would be problematic.
|
# ¿ Dec 26, 2014 06:19 |
|
movax posted:All this talk of clock speed, and no mention of edge rate thanks! for me, I'm specifically talking about NuBus. I'm not sure how to figure its edge rate, but it's specified as 75ns unasserted, 25ns asserted, with signals asserted on the rising edge and sampled on the falling edge of its active-low bus clock. Edit: Found more detailed timing info towards the end of the pdf I linked. looks like it's not too bad... eschaton fucked around with this message at 22:03 on Dec 26, 2014 |
# ¿ Dec 26, 2014 21:40 |
|
I now have an HP 1660CS logic analyzer and scope! (pods and clips are on the way.) so far I've used the scope to view the action in a little 555 blinker. total overkill, sure, but it was fun to see the hardware I just picked up do something besides boot. also, it gave me the idea that I should also be able to diagnose and repair my digital tuner (a Korg DT-1) pretty easily now. I could also do that with a just multimeter, of course. I've also set it up on my network, and downloaded the user and programming guide PDFs from
|
# ¿ Dec 29, 2014 01:02 |
|
Phobeste posted:logic analyzers and scopes each have their own distinct uses unless you get the really really expensive ones that have both a) not poo poo analogue capabilities and b) a whole shitload of channels get one that used to be really expensive 136 channels of love
|
# ¿ Dec 31, 2014 21:26 |
|
Tin Gang posted:where do you even get probes for something like that? eBay has lots of people selling used test equipment, including Arcsech’s Tek TLA5204 ($10K!), and stuff like probes & pods too the latter seem to run $25-$100 or so for other logic analyzers, I don't know what the 5204 takes so I can't look up much in the way of details also lol windows-based test equipment
|
# ¿ Dec 31, 2014 21:37 |
|
stopped by Fry's and picked up an Intel Edison Arduino Kit, it's basically an Arduino but with two 32-bit x86 cores running at hundreds of MHz with gigabytes of RAM and flash and WiFi and Bluetooth instead of an 8-bit microcosm trooper with kilobytes of RAM and flash and a straw to the rest of the world downside: running Yocto Linux (or any Linux) on an embedded system wonder how hard it'd be to boot Darwin on it
|
# ¿ Jan 6, 2015 03:03 |
|
I'm the wearing your conference badge on stage
|
# ¿ Jan 6, 2015 03:10 |
|
kwinkles posted:i don't think the edison board ever did this. the first galileo board only had low-speed gpio on expanders, but the current galileo board has 12 high speed gpio pins right off of the SoC and more on an i2c expander. i am pretty sure the edison board has at least that many high speed gpios but i haven't worked with it so i'm not sure and i can't be bothered to look it up. the Edison can be configured for up to 44 GPIO pins, though as always the pins are shared by other interfaces.
|
# ¿ Jan 8, 2015 08:18 |
|
kwinkles posted:this sounds pretty rad. here is what i made with 12 full speed gpios: sweet! what kind of LED matrix is that? I assume it's i2c?
|
# ¿ Jan 9, 2015 06:21 |
|
so the Intel Edison stuff, with the Arduino board has an Arduino-compatible IDE. how should I go about figuring out what all of the Arduino stuff actually maps to in terms of memory-mapped I/O (or whatever) so I can wire it up to Lisp instead? oh, also, SBCL runs OK on the board. not great, at least in comparison to my MacBook Pro (i7), but acceptably well and probably a lot better than on Raspberry Pi.
|
# ¿ Jan 19, 2015 04:25 |
|
eschaton posted:so the Intel Edison stuff, with the Arduino board has an Arduino-compatible IDE. how should I go about figuring out what all of the Arduino stuff actually maps to in terms of memory-mapped I/O (or whatever) so I can wire it up to Lisp instead? figured this out! last week I found the source drop on Intel's site that includes the components implementing their Arduino support, which led me to their libmraa low-level C library for interfacing with I/O. just a quick compile against the already-installed headers and library, and I was able to use some C to blink the onboard LED. quote:oh, also, SBCL runs OK on the board. not great, at least in comparison to my MacBook Pro (i7), but acceptably well and probably a lot better than on Raspberry Pi. I also switched from SBCL to CCL for my Lisp, since it's the environment I'm more comfortable in on my Mac and it performs nicely on the Edison too. today I managed to figure out the madness that is CFFI and got a basic wrapper for libmraa's GPIO functionality written. now I can write code:
and blink the board's LED for n seconds! next I'm planning to wire up something to a breadboard and just try pushing some pattern of bytes to it as fast as I can and capture a trace with my analyzer, so I can see just what kind of bandwidth I have available (and maybe also check that Lisp isn't adding that much overhead)
|
# ¿ Jan 26, 2015 06:45 |
|
Bloody posted:bump tilt
|
# ¿ Feb 1, 2015 08:39 |
|
I made an LED blink on a breadboard via jumpers from the Edison Arduino board. and the LED didn't get hot or explode! I actually used a resistor, and the right resistor too! next I want to measure just how fast I can toggle that I/O from Lisp and C. I just didn't want to set up the logic analyzer before dinner. it was getting to be cheesesteak o'clock, you see.
|
# ¿ Feb 2, 2015 06:50 |
|
it's somewhat sarcastic as my teenage adventures in hardware were less than successful and half a lifetime ago now though I want it to not be a big deal. I want to make things that plug into other things, and I want to make things what do things. and well blink baby blink! and in C instead of Clozure Common Lisp: both are using libmraa via sysfs rather than memory-mapped I/O eschaton fucked around with this message at 07:39 on Feb 2, 2015 |
# ¿ Feb 2, 2015 07:31 |
|
Mido posted:with a bit of fiddling you can keep a 1wire device dead without power draw there's a lot you can kill with a bit of fiddling
|
# ¿ Feb 12, 2015 07:22 |
|
Blotto Skorzany posted:the necronomicon analog is magic already so this is absolutely right
|
# ¿ Feb 16, 2015 21:49 |
|
spankmeister posted:hello thread i like to build tube amplifiers have you funded your retirement yet by getting some wood & aluminum case made and setting up an online storefront to bilk audiophiles out of mad cash? you have a moral and ethical duty to separate them from as much as possible, you know.
|
# ¿ Feb 17, 2015 04:05 |
|
Bloody posted:also hdls are so goddamn tedious. yeah, lemme just declare my loving inputs and outputs in three places. boilerplate much??? in starting to look at Verilog and VHDL, I started seriously thinking about writing an HDL in Lisp that would generate one of them but let me write at a nicer level of abstraction.
|
# ¿ Feb 21, 2015 01:21 |
|
Bloody posted:seriously, do it. they are awful. i think verilog is slightly better, but just barely. are there any examples you can point to of good or bad Verilog, VHDL, and so on? also, do the Xilinx tools work in win8 yet? or are they "reasonable" on linux? I don't have a win7 system handy.
|
# ¿ Feb 21, 2015 03:53 |
|
someone else with a scarab miniSpartan FPGA board posted a simple text generation example to github, and just skimming it with little real understanding of Verilog I see several things I'd prefer to do with a "real" language even if it just winds up generating a similar mess if I can get the Xilinx tools up and running I'll probably try this anyway, just to see that board do something besides sit on my desk.
|
# ¿ Feb 21, 2015 04:10 |
|
also hey, a 6502 in Verilog, maybe I could make an Apple II with an HDMI port and an octuple-hi-res graphics mode (just gotta be sure to keep the strangeness of the graphics memory layout, otherwise it wouldn't be a real Apple II)
|
# ¿ Feb 21, 2015 04:17 |
|
Poopernickel posted:Killed self literally, I am a ghost now stay safe Xilinx ghost
|
# ¿ Feb 26, 2015 09:54 |
|
movax posted:fpga tooling on linux is only good for running headless builds imo; if you need to use the gui at all, stick with windows. linux is a second-class citizen (as it should be) if i were to install the Xilinx tools on Linux, would that be sufficient for new development if I just did everything through makefiles? or does the tooling really insist on an IDE-generated project file and so on, even for command-line builds? kind of like the IDE that I work on does
|
# ¿ Mar 2, 2015 00:55 |
|
Bloody posted:anybody write a tolerable to use shim for vhdl or verilog yet having to write c# programs to emit dozens of lines of boilerplate is getting really tedious I've been too busy with work stuff to do more than shoot the poo poo about writing a Lisp HDL that outputs Verilog. (plus you'd have to use Lisp to use it. and my employer would have to sign off for me to share it…) instead of writing c# to emit boilerplate though, could use use the T4 template system or something? seems like that might be a better fit for just expanding out boilerplate, and it's part of Visual Studio too.
|
# ¿ Mar 6, 2015 06:13 |
|
|
# ¿ Apr 29, 2024 12:39 |
|
Bloody posted:did not know that existed I've never used it, since I don't actually do Windows stuff, but I think it came out of shaggar's favorite framework (entity framework) for their codegen stuff.
|
# ¿ Mar 7, 2015 04:12 |