|
JawnV6 posted:no we've got a low powered arm chip that's already loaded to the gills with work and no hardware float, and it's coming down to 'deal with all the fixed point underflow and implement all these routines' versus 'send a couple hundred bytes up the pipe when needed and get the answer basically instantly'
|
# ? May 22, 2015 18:48 |
|
|
# ? Jun 3, 2024 23:59 |
|
c electron s: holy poo poo nightmare hellscape verilog mess is coming to life and like working in hardware and everything at the target clock rate haaaaaiiiil satannnn
|
# ? May 29, 2015 04:37 |
|
how fast is the target clock? how fast???? (fpga speed junkie) ps i have been janitoring bits in systemverilog a few weeks now and i'm really liking it. holy poo poo, multidimensional arrays and/or structs allowed in port lists? fuuuuuuuck yessssss
|
# ? May 29, 2015 05:56 |
|
BobHoward posted:how fast is the target clock? at work we (not me, the actually smart people) are doing a 250MHz design and it is like woah we're in the biggest part xilinx makes and its not enough
|
# ? May 29, 2015 06:52 |
|
250 is in the exciting zone for 7 series for sure. okay if you go nuts on the low level optimizations, but unless your design is a giant array of identical components you are going to be spending a lot hiring enough expert designers to fill a xc7v2000 with that much highly optimized rtl ive done some (very limited) poo poo in fuckoff nuts territory, aka 500 MHz. it was in virtex5 too. it was... entertaining. you spend tons of time knee deep in synthesis output figuring out how to write code to make it generate the luts and flipflops you want. this is sometimes v. painful when the synthesis tool absolutely.loving.refuses to extract the flipflop clock enable signal you coded in such a way that you could not possibly imagine the synthesis tool failing to recognize it as a CE
|
# ? May 29, 2015 07:10 |
|
BobHoward posted:ps i have been janitoring bits in systemverilog a few weeks now and i'm really liking it. holy poo poo, multidimensional arrays and/or structs allowed in port lists? fuuuuuuuck yessssss the main CλaSH developer asked if maybe he should deprecate the SystemVerilog back-end because he's brought up a Verilog back-end people replied saying essentially "oh god please don't" because SystemVerilog is so much nicer
|
# ? May 29, 2015 07:18 |
|
did anyone tell him that technically there is no such thing as plain vanilla verilog any more? true story, the standards committee decided that unlike c family languages they weren't going to bother maintaining the original language as a separate standard any more sv is mostly what verilog always should have been, except insofar as it has to preserve some unfortunate design mistakes to keep being a superset of verilog. that aside, they've fixed so many of the horrible glaring weaknesses
|
# ? May 29, 2015 09:11 |
|
52 MHz so not particularly blazingly fast and parts of it should actually be slower but we didn't feel like dealing with clock domain crossing just yet sysv would own if tools supported it but
|
# ? May 29, 2015 12:33 |
|
bump
|
# ? Jun 14, 2015 23:29 |
|
async clock crossing is sweet, one of those concepts that took 1.5 passes for me to understand gray codes, bubble gaps, its all p. neat
|
# ? Jun 15, 2015 00:36 |
|
JawnV6 posted:async clock crossing is sweet, one of those concepts that took 1.5 passes for me to understand tell me more about the clock crossing
|
# ? Jun 15, 2015 00:40 |
|
you see a clock pulse crossing zero. It's inverted. You don't throw an interrupt. Why is that?
|
# ? Jun 15, 2015 00:54 |
|
it's not that hard if you don't care about determinism, there's a head/tail pointer encoded w/ gray code so it doesn't matter if a signal from one domain doesn't settle by the time the other samples it you end up with bubbles and one of the interfaces probably has to be capable of backpressure omg flow control credits negotiated during link training
|
# ? Jun 15, 2015 01:05 |
|
stick an ipcore fifo on every boundary boom problems solved
|
# ? Jun 15, 2015 01:35 |
|
BobHoward posted:250 is in the exciting zone for 7 series for sure. okay if you go nuts on the low level optimizations, but unless your design is a giant array of identical components you are going to be spending a lot hiring enough expert designers to fill a xc7v2000 with that much highly optimized rtl same
|
# ? Jun 15, 2015 07:50 |
|
Bloody posted:stick an ipcore fifo on every boundary boom problems solved love me some fifos. currently writing a multi channel video framebuffer and I cant get enough of those. clock domain crossing, width adaption, flow control, burst smoothing, alll in one handy package.
|
# ? Jun 15, 2015 20:50 |
|
hey fellow bit janitors, im looking for a bluetooth soc that has a dac, any suggestions? it seems to be a surprisingly hard combo to find.
|
# ? Jun 15, 2015 20:54 |
|
JawnV6 posted:async clock crossing is sweet, one of those concepts that took 1.5 passes for me to understand once long ago i found a bug in a display controller that was causing a pixel to get dropped and i spent a whole day figuring out that it was an issue with a clock crossing fifo and i think that was the first time i really understood it
|
# ? Jun 15, 2015 23:55 |
|
Bloody posted:hey fellow bit janitors, im looking for a bluetooth soc that has a dac, any suggestions? it seems to be a surprisingly hard combo to find. anything from nordic?
|
# ? Jun 16, 2015 00:29 |
|
Bloody posted:hey fellow bit janitors, im looking for a bluetooth soc that has a dac, any suggestions? it seems to be a surprisingly hard combo to find. like this? http://m.ti.com/product/CC2540 ick 8051, etc.
|
# ? Jun 16, 2015 01:49 |
|
hobbesmaster posted:like this? http://m.ti.com/product/CC2540 i don't see a dac on that?
|
# ? Jun 16, 2015 04:02 |
|
movax posted:anything from nordic? nope. closest candidates so far are a silabs chip with a 4-bit current dac (current would be fine but 4 bits is wtf?) and a csr8670 which may or may not be usable but i will probably never know because they do not seem to be people who deal with low-volume people although the part is on digikey??
|
# ? Jun 16, 2015 04:05 |
|
Bloody posted:nope. closest candidates so far are a silabs chip with a 4-bit current dac (current would be fine but 4 bits is wtf?) and a csr8670 which may or may not be usable but i will probably never know because they do not seem to be people who deal with low-volume people although the part is on digikey?? http://www.mouser.com/new/panasonic/panasonic-pan1740-ble-module/ maybe?? edit: nope lol
|
# ? Jun 16, 2015 04:06 |
|
current best candidate seems to be slapping either a lovely standalone radio like one of the nordic chips with a micro that does everything else or one of these bluetooth soc chips with a lovely standalone dac but neither option is appealing
|
# ? Jun 16, 2015 04:06 |
|
ehh given the size i'd just as soon slap together a dedicated radio and a micro
|
# ? Jun 16, 2015 04:07 |
|
bonus points: wireless charging support. didn't even think might be a thing but i just saw one of the dialog bluetooth socs has it (but still no dang dac)
|
# ? Jun 16, 2015 04:09 |
|
http://www.reddit.com/r/electronics/comments/39zv50/recharging_a_lithiumion_battery_pack/ posted:Few days ago, I got lithium ion batteries from salvaged electronics. I hooked them up in the configuration that give more amps (P to P and N to N) and it gave out about 9000mA/H. I'm having trouble charging the battery pack. The battery pack gives about 4 volts and currently has 3.8 volts. I've charged it with a 500mA 5V charger with positive to positive and negative to negative. I've also charged it the opposite way positive to negative and vise versa. My voltage meter still kept saying that the voltage was the same even after one full hour of charging. I'm wondering what is going wrong and how I can fix it. FYI I've let this charge for about 6 hours straight and still no heat or voltage increase could be seen and felt. Also, the wires I used to charge them are connected to the wire that outputs the power. The diagram is here: https://imgur.com/dD7M9Vi
|
# ? Jun 16, 2015 05:28 |
|
is there anything like this test block but with a united states plug on it? it supplies line power to whatever's clamped into those terminals, but only when it's clamped shut. it would be just about the most useful thing for testing stereos that have had the cords chopped off, but i can only find the european versions
|
# ? Jun 16, 2015 05:55 |
|
Bloody posted:i don't see a dac on that? hurf I parsed DAC as ADC
|
# ? Jun 16, 2015 06:09 |
|
the best part of lithium ion batteries is that little bag that's in cells that contains the explosion
|
# ? Jun 16, 2015 06:10 |
|
quote:I know the dangers of working with Lithium batteries, that's why I never go above 500mA. Also, I'm majoring in Chemistry but I need this for electrolysis. I swear chemists are one of the biggest danger to electronics
|
# ? Jun 16, 2015 06:14 |
|
Bloody posted:current best candidate seems to be slapping either a lovely standalone radio like one of the nordic chips with a micro that does everything else or one of these bluetooth soc chips with a lovely standalone dac but neither option is appealing what's the use case? do you want something like a rn52?
|
# ? Jun 16, 2015 07:56 |
|
JawnV6 posted:what's the use case? do you want something like a rn52? basically a wireless arbitrary waveform generator with a few K of signal memory operating up to 20 kHz. size > power >> cost, so tightly integrated solutions are appealing
|
# ? Jun 16, 2015 14:24 |
|
does it have to be bt? there's zigbee stuff with DACs
|
# ? Jun 16, 2015 21:38 |
|
atomicthumbs posted:is there anything like this test block but with a united states plug on it? why dont you just get that and put a US plug on it?
|
# ? Jun 17, 2015 00:03 |
|
Sniep posted:why dont you just get that and put a US plug on it? because by the time it arrived from the uk my deadline would have passed
|
# ? Jun 17, 2015 01:06 |
|
I have no idea what those things would be called in the U.S. but we used a few my undergrad power lab. they were from the 60s, had less shielding, and supported 3 phase 208V incidentally the undergrad power was the most thing ever. so much cooler than the actually useful to my current career but full signals lab I was trying to take
|
# ? Jun 17, 2015 01:12 |
|
variac?
|
# ? Jun 17, 2015 01:13 |
|
just make one ffs
|
# ? Jun 17, 2015 01:14 |
|
|
# ? Jun 3, 2024 23:59 |
|
JawnV6 posted:variac? I think that's just a double (triple?) pole single throw knife switch with some extra protection the gently caress safety version is one of these and a cut off electrical cable: http://www.arborsci.com/knife-switch-double-pole-single-throw hobbesmaster fucked around with this message at 01:39 on Jun 17, 2015 |
# ? Jun 17, 2015 01:36 |