ok, that helps... Can you find the hidden data?

I'm trying to figure out these chips. a lock, of some sort, seems to have been put on the contents, and the TL866 'cant' clear it. The datasheet is a bit awkward to follow, they do some strange remapping stuff.

morning random image: really flat circuit for capacitor dropper mains indicator lamp

almost hit my weekend goals (only missed by about 2 miles) I managed to do they clay, and get the shifter off the parts truck, which wasn't even on my list. I think I know how to do the belt, I should try now. I have to think on that state machine, and the FFT thing is gonna keep me up all week

Today I also managed to remove the shifter from the parts truck and plate it off. the goal is to pull the engine and have it reworked for next summer. This is that plate I made yesterday.

After waiting about 10 minutes, I poured off the top, there wasn't actually much in the bottom, a few lumps. A pouring bucket with an inner lip would be awesome. Now I let it settle, and gradually pour the water off the top.

ok, I gave the mud a good stir. The stirring thing actually helped get out a bunch of the bits of roots and stuff, I didn't know there was that much in there. Gave it about 10 mins to settle.

FFT: If I optimize for inputs where the imaginary component is always 0 going in, I could leverage that part of the array to do the butterfly in the first pass of the calculations....

dear god, I have to tear down this FFT thing WAY more than I thought, this isn't just silly on the top, its silly ALL THE WAY DOWN. I'm starting to question if the original authors knew how to code.

hahaoh no, its 1am already, how did that happen?

oh god..... what if I told you that you dont need to do the butterfly re-sort on your FFT data, by modifying the code that does the loop.... OMG, its all been a lie.... I have yet to test these results.... FFT...

ok, back to the FFT....

I also have a bin filling up with proprietary / OTP / DRAM chips that I dont want, but would be into donating to someone doing die shots (this could last a while eh?) there might be interesting things to learn about what chips lie in these carriers...

Some sort of custom relabeled atmel chip, I'm curious as to what it might be. I think it would be an arm controller of some sort, as its a BGA I cant really R/E from the pcb that well.

WOULD ANYONE LIKE to have... I have a set of (maybe 3) BGA TMS320 processors, I'm not into trying to use them, but if you have a project they would be perfect for, please contact me.

heh, no, its ok, that'll buff out just fine... I hand install these chips, so all the bent pins from the scavenge process really isn't an issue.

OK!!! RETRO peoples (please boost) These are mac serial ports for printers. If anyone needs these for a repair or whatever pls ask, I specifically pulled them on the grounds I suspect their hard to find and a retro restore might be in need of them.

I'm gonna call that a night, its a descent haul.

Here I cracked open a relay that had case damage during removal, turns out the relay was ok, but I'm not going to use it like this. A cut-away view of how it works, the solenoid on the right pulls in a lever that pushes the contacts on the right.

The parts are not all the glory, sometimes the disassembly reveals an idea that is worth more than the parts. In this case its a way to do a toroid base, something I have challenges with.

I spent some time stripping more pcbs, making good progress clearing stuff that only has a few parts I'm instereted in.

Meaningless numbers: Twitter: 1 follower / 5 posts Mastadon : 1 follower / 14 posts Bluesky: 1 follower / 3 posts

"You look depressed, want a cookie?" "How will that help?" "It wont help with the depression, but you will then also have a cookie." Depression cookies.

When all you want is a picture, but you have to suffer thru a youtube video with ads, introductions, lead-ins, ads, and a table of contents. I think this will solve the DIY belt I'm trying to make AND put the DIY-sandals back on track.

I have also been doing some PCB stripping. The scrap out bin if filling up nicely :]

How to not CNC a sheet metal part. By Rue Mohr. Step 1, CAD it up, print 1:1. Be sure to mark hole centers!

Things to do this weekend: - design hardware FSM that uses the new ROM adapter PCB - dig and clean some clay - totally rewrite the FFT code to use completely pre-calculated indexes. well, for now I got a symbol for the ROM adapter made....

@mle.bsky.social tasked me with attempting to produce some 'wild clay' after I asked for more details on how it was done. - fair - This is all going down on mastadon, but I thought I'd share. Instructions were as follows:

Things to do this weekend: - design hardware FSM that uses the new ROM adapter PCB - dig and clean some clay - totally rewrite the FFT code to use completely pre-calculated indexes. well, for now I got a symbol for the ROM adapter made....

This morning epiphany, that will eat my weekend, is that almost all the code in an FFT just calculates 3 indexes. (highlighted here is the code that doesn't calculate those indexes) I can pre-compute this and just run a single for loop. This will be a radical simplification.

Epiphany last night, one of those "oh, well duh" ones, most of the FFT alg is just building 4 indexes, which is really only 2, but they could be a precomputed array.... FFT FSM...

at 60ns the limit of the ROM FSMs is about 16MHz, but realize thats faster than a 16MHz cpu, becasue every cycle can do IO.

The next step with these ROM adapters is to make hardware state machines. I have since noticed I can make faster machines with cache SRAM, but I want to finish this path. hehe, I have to reverse engineer the pinout on these too... mmmm 19 input FSM LUT...

😅 heh, problem... You may notice "This is not the 28F008 you were looking for", as it turns out, my programmer cant unlock(?) a sector of the 28F008 chips I have (a lot of). The adapter does work tho!, its much better than the assembly on the left otherwise required to program these chips.

My @aisler.net PCBs arrived!!!!! They sponsored these ones, I'm super excited to get one assembled and tested. The primary board here is a TL866 flash ROM adapter, BUT, if I screwed it up, I can cut it up into 3 other alternate-population PCBs, two oscillator modules and a LED test

And so the goal is to build something that allows for disconnecting a RAM from a state machine to load/reload it with a config via either a microcontroller or a slower ROM state machine... I didn't get any sleep last night...

The thought occurred to me last night that the state machines I want to make would run faster from cache SRAM than the flash roms I been collecting. but of course they would have to be loaded first, which is a horrid design decision.... which all FPGAs use... huh....

I did some sorting too.... excited to see what fet drivers are in here.

After all that I stripped down more circuit boards. hah, another 6800 cpu. Lots of cool memory chips.

pump impeller!!!! Well most of it, this is my buddies chuck, he's going to make an insert for the middle along with parting it off and stuff.

CNC machine go BRRRRRRR....

I'm helping a buddy machine a new pump impeller, the plate at the bottom was machined with a tight pocket for the chuck to register a zero location. The bolts and link nuts serve as jack posts for the clamps.

A short session of parts stripping tonight, what do you see thats cool? I know what I like!

"Working wuth a shape shifter and it got weird cause over the course of the day he would gradually distort and every so often you would see him do a readjustment."

Hmmmm 6800, must be precursor to the 6803 so many wires. ;]

One of the cool chips from todays pull. MX635A, RMS to DC converter.

stripped a few more pcbs this evening.I got a good mix of the hotplate and heatgun working. wondering if I should modify the heatgun with a temp adjust....

The parts haul is getting really juicy. its still tiny tho, I did sort some of it. The most popular 74xx series you find on modern boards is 74244, followed by 7404, then 74245 and 74374

I have quite a cache of stuff to strip down later.

I filled up a box and a bit today, which is great.