00:02 <dbj314> The spec seems to let you get away with as low as 44 mA when pulling a line high, and as low as 95 mA when pulling a line low.

00:02 <whitequark[cis]> I only have a cursory understanding of the PCI electrical spec

00:05 <dbj314> Same, honestly.

00:07 <dbj314> Which is why I mainly came here to ask about the ATF15xx series, because the people who spent a ton of time reverse-engineering it might be able to answer weird questions about it that aren't in the datasheet. Though the cap charging timing was a good idea. I might try that.

00:08 <whitequark[cis]> I mostly reverse-engineered the toolchain

00:10 <dbj314> For toolchain related questions, when looking through the datasheet and prjbureau, I got the vague hint that "TTL" mode might be push-pull, and "CMOS" mode might be open-collector. Is that true?

00:10 <Wanda[cis]> that sounds wrong

00:10 <whitequark[cis]> that sounds the exact opposite of true

00:11 <Wanda[cis]> in general, CMOS is the "symmetric" mode

00:11 <Wanda[cis]> input threshold at VCC/2, equal output drive strengh in both directions

00:11 <Wanda[cis]> while TTL has lower threshold, and also stronger N drive than P drive

00:12 <whitequark[cis]> "classic" TTL does not have a P drive, it has a pull-up

00:12 <Wanda[cis]> (FPGAs/CPLDs don't necessarily implement both; sometimes they let you change only the input levels, sometimes only the output strengths)

00:12 <dbj314> The datasheet lists "Output Low Voltage (TTL)", "Output Low Voltage (CMOS)", "Output High Voltage (TTL)", but no "Output High Voltage (CMOS)". Is that just a datasheet mixup?

00:12 <whitequark[cis]> like... you have a small value resistor and a 40 mA capable driver, that's TTL

00:12 <whitequark[cis]> I honestly don't know what the datasheet is on about

00:12 <Wanda[cis]> XC4000 has a funny feature where "TTL output" uses an N-channel transistor for high output instead of P-channel, making the output level lower

00:13 <Wanda[cis]> vendors do weird shit.

00:13 <whitequark[cis]> these are the two main fuses relating to the output driver

00:13 <whitequark[cis]> there isn't a CMOS-or-TTL switch

00:14 <Wanda[cis]> they could also be talking about TTL and CMOS in the datasheet implying their hardware is compatible with both while not, in fact, actually having two distinct modes

00:14 <whitequark[cis]> yes, that's what I would gues

00:14 <whitequark[cis]> s/gues/guess/

00:15 <Wanda[cis]> (some Xilinx devices are like that. not all. you're not going to figure out the difference from looking at datasheets.)

00:16 <dbj314> I saw those 2 fuses, I just couldn't tell which what corresponded with what. Also, the IO Standards table doesn't talk about which fuses ead to what. https://whitequark.github.io/prjbureau/options/buffer.html#i-o-standards

00:16 <whitequark[cis]> oh wait

00:16 <dbj314> *lead*

00:18 <whitequark[cis]> yeah, so there is an IO standard fuse, for ATF1504BE and larger

00:18 <whitequark[cis]> this is not relevant for AS

00:25 <dbj314> I have an (analogue) oscillioscope, a digital multimeter, and a few microcontrollers. Between them I might be able to figure out how much current the pins can output. And maybe even figure out what the i/v curve looks like, though that is getting really far into the "I have no idea how to do that" territory.

00:26 <whitequark[cis]> what kind of scope?

00:26 <dbj314> Tektronix 2465

00:28 <whitequark[cis]> oh that's a good one

00:30 <dbj314> It doesn't really matter right now, because I don't have any ATF15xxAS chips at the moment. I was (and still might be?) in the stage of deciding if the project has enough of a chance of success to take any action at all regarding it.

00:32 <dbj314> Though at this point I am convinced enough that these will be useful for something

00:40 <dbj314> Hah. I should not order them from microchip directly, because I do not want them in multiples of 90. Why do they even list prices for quantities 1-24?

00:41 <dbj314> Mouser works. Bye.