[X] My Assistant

After 17 years knowing and using Linux, today I finally learned how to use rsync. holy shit.

Tried to get in touch with nvidia regarding how to make a desktop shortcut (clickable icon) to rotate my display.

CODE

C:\Windows\System32\rundll32.exe "NvCpl.dll,dtcfg rotate 1 90"

is something I found somewhere, but it was very old information, and not working.

But they could not help me.
Saying you could not control the nvidia control panel features through command line.
I'm not entirely convinced, but I will give up.

Meanwhile, on better OS'es…

CODE

xrandr --output DVI-D-0 --rotate left

Also nvidia-settings works on CLI. At least on better OS'es.

After 17 years knowing and using Linux, today I finally learned how to use rsync. holy shit.

I ostensibly know how to use rsync. I have used it in the past. But honestly I just tend to make tarballs or cpio archives and push them over scp because rsync is annoying and the documentation is huge and almost but not quite incomprehensible.

I wonder if the BSD's have better documentation for it.

Edit: [man.openbsd.org] Yes, yes they do. Mainly because of a smaller feature set.

This post has been edited by dragontamer8740: Apr 3 2022, 20:05

with rsync I feel like everyone has their favorite flags that they then just always use for every command, without really knowing what it all did. and rsync has those flags that turn on a bunch of options
on that topic my goto (bless the huge command history) is rsync -avP

This post has been edited by cate_chan: Apr 3 2022, 21:27

Meanwhile, on better OS'es…

CODE

xrandr --output DVI-D-0 --rotate left

Also nvidia-settings works on CLI. At least on better OS'es.
I ostensibly know how to use rsync. I have used it in the past. But honestly I just tend to make tarballs or cpio archives and push them over scp because rsync is annoying and the documentation is huge and almost but not quite incomprehensible.

I wonder if the BSD's have better documentation for it.

Edit: [man.openbsd.org] Yes, yes they do. Mainly because of a smaller feature set.

Yeah all this while I'm the tar guy too and it worked perfectly for my workflow. J
Just recently I have some tasks that involve moving thousands of ever-changing files every few hours, so the tar way gets tedious quick.

with rsync I feel like everyone has their favorite flags that they then just always use for every command, without really knowing what it all did. and rsync has those flags that turn on a bunch of options
on that topic my goto (bless the huge command history) is rsync -avP

Huh. I thought -a is the answer for everything.

This post has been edited by rozos: Apr 3 2022, 22:34

Just recently I have some tasks that involve moving thousands of ever-changing files every few hours, so the tar way gets tedious quick.

Yeah, I recently backed up my pr0n directory to another computer on my lan and used rsync to try to update it less wastefully.

absolutely BOTW2

New gainclone module showed up. Hopefully soldering that in fixes the amp, then I can give the pair a proper test. It'll be a bit annoying to setup, but I may try an A/B test of the Pyle class-D and Tandberg gainglone class-AB.
Not convinced there's any point to it though.

My preamp seems just fine after some work- put a less cheesy 1/4" headphone jack in it, and reflowed all the pads on one end of a PCB. I felt little pops when doing the two dual pots, so I think they had cold joints.
The AC adaptor was lol-bad. It looked really cheesy, so I decided to take it apart. No screw under the label, so I gave it a solid whack to loosen it up. The side shattered and the end shot off. The circuit was sensible enough, but I probally could have opened it with my bate hands if I tried. Line power jumped to the board with 24awg wire and iffy soldering on the prongs. Of note, it had no effective isolation. The class Y capacitor to reference the DC side was a basic 1KV ceramic cap in lieu of an RU listed class Y. The transformer was where it got real bad. No glue on the core, only poleyester tape. Not safety criticial, but it's gonna whine. Took the core out, and the outer layer of tape off. Secondary was on top, and primary in the center. The two ends of the primary crossed over each other tightly, and a secondary lead crossed over a few mm away... So the only isolation was two layers of enamel, with the bonus of the first spot to fail will flash over to the other spot before it burns clear. Oh, and no fuse either. The only good thing I have to say about it is that it didn't have any fake approval/listing markings.
Replaced that with a much higher quality unit.

Looks like I finally have something to motivate me to buy an arduino USB host shield.

There's an AVR-based implementation of 'checkm8' (bootloader exploit for iphone X and earlier), which would mean I no longer have to boot an OS X VM with USB passthrough and re-sign a jailbreak app every damn time I reboot or run out of battery on an old ipad I was given running iOS 9.3.5 (which I mainly use for looking at schematics while I'm working on electronics). Xcode is confusing as hell.

[github.com] https://github.com/a1exdandy/checkm8-a5

Damn my relatives for religiously updating their old iShit before they give it to me. The newest device I have an untethered jailbreak for is an iPhone 4S running iOS 8 since it was my primary phone until around 2015 (when it was still getting updates).

The AC adaptor was lol-bad. It looked really cheesy, so I decided to take it apart. No screw under the label, so I gave it a solid whack to loosen it up. The side shattered and the end shot off. The circuit was sensible enough, but I probally could have opened it with my bate hands if I tried. Line power jumped to the board with 24awg wire and iffy soldering on the prongs. Of note, it had no effective isolation. The class Y capacitor to reference the DC side was a basic 1KV ceramic cap in lieu of an RU listed class Y. The transformer was where it got real bad. No glue on the core, only poleyester tape. Not safety criticial, but it's gonna whine. Took the core out, and the outer layer of tape off. Secondary was on top, and primary in the center. The two ends of the primary crossed over each other tightly, and a secondary lead crossed over a few mm away... So the only isolation was two layers of enamel, with the bonus of the first spot to fail will flash over to the other spot before it burns clear. Oh, and no fuse either. The only good thing I have to say about it is that it didn't have any fake approval/listing markings.
Replaced that with a much higher quality unit.

I've taken to buying all my ac adapters on digi-key/mouser in recent years. When I'm not using some old radioshack adjustable-voltage brick, that is.

Update:
(IMG:[i.imgur.com] https://i.imgur.com/4m9eWEa.png)
A bit late to be sorry now... but I guess it's better than him bragging about it.

This post has been edited by dragontamer8740: Apr 5 2022, 22:10

Jack Dorsey is always like 'Oh lawl, I regret that we didn't know what we were doing could be construed as lying at time from your perspective, we will try to do better and by that I mean to the minimum required by the law and actually probably just doing it again"

maybe someone needs to go back in time and prevent Tumblr from being started, it seems that was where the cancer first grew.

Got the chip soldered in, wired the preamp up to the amps, and amps to the speakers. Listeed and it sounded pretty good. After some 3 minutes, it made a loud harsh buzzing that rapidly ramped up to very high volume, and I yanked the plug. Looked over my setup and the same thing happened again- roughly 3 or 4 minutes from cold start, it did that with a couple second time delay in either channel. This time an amp blew one of it's slighty undersized fuses, and both my speakers weren't happy. Both amps and both speakers still work fine for short tests with aux in directly to my phone, and phome not charging.

I need more fuses to investigate, some people on IRC were stumped by the time delay. My best guess is either the local ground or the input line to the amp chips is charging up with some DC voltage. The audio and dc grounds of all 3 devices are commoned, but otherwise floating. The only tie to anything is likely the class Y cap in the preamp's wall wart. As for the input I bodged into the amp, there's a 101 poly cap to ground, a 225 ceramic cap in series, and a 221 ceramic cap to ground at the module- I can't say if it has a DC patch to anything. Same for the local ground, I don't think it has any DC leakage path.

This is all purely speculation, mind you. I need to setup something of a dummy load instead of my 301 IV speakers and try to probe around when it's going on. With my luck, it'll dissapear on detailed bench testing.

Get some bookcase speakers I said. I already have a monoblock, just get another I said. I was kind of looking into just getting an integrated amplifier/receiver/??? last night, and was failing to find stuff with bluetooth input and low-level outputs.

This post has been edited by Wayward_Vagabond: Apr 5 2022, 15:20

The rear speakers in my setup are too soft. They are supposed to be atmos but they are always pretty quiet.

Got the chip soldered in, wired the preamp up to the amps, and amps to the speakers. Listeed and it sounded pretty good. After some 3 minutes, it made a loud harsh buzzing that rapidly ramped up to very high volume, and I yanked the plug. Looked over my setup and the same thing happened again- roughly 3 or 4 minutes from cold start, it did that with a couple second time delay in either channel. This time an amp blew one of it's slighty undersized fuses, and both my speakers weren't happy. Both amps and both speakers still work fine for short tests with aux in directly to my phone, and phome not charging.

No feedback loop anywhere?

What frequency was the buzzing (roughly)?

I finally did something I should have done YEARS AGO:
I tried copying the winhlp32.exe file from Windows XP to my Win10, then drag&dropping a .hlp file requiring it on the exe, and IT WORKED, JUST LIKE THAT.
I had seen lots of complicated things on Internet, and it gave me the impression it could never work, but it did.

Of course if a program tried to summon its help itself, it would not work, but after adding that exe to the sendto menu, now I can save time without having to fear decreased performances or an update deleting my stuff.

Uhh... It's been available for newer OSes for ages. There's an official release of it for Windows 8.1 you could have extracted.
Windows 8.1: [www.microsoft.com] https://www.microsoft.com/en-us/download/de...s.aspx?id=40899
Windows 8: [www.microsoft.com] https://www.microsoft.com/en-us/download/de...s.aspx?id=35449

Windows 7:
* (32 bit) [web.archive.org] http://web.archive.org/web/20200803205226i...B917607-x86.msu
* (64 bit) [web.archive.org] http://web.archive.org/web/20200803205226i...B917607-x64.msu

Windows Vista:
*(32 bit) [web.archive.org] http://web.archive.org/web/20160229194024i...B917607-x86.msu
*(64 bit) [web.archive.org] http://web.archive.org/web/20160229194024i...B917607-x64.msu

Thought: Fuck microsoft for removing downloads for old OSes. There is no reason these links should not be permanent.

This post has been edited by dragontamer8740: Apr 6 2022, 06:21

A loop or feedback issue seems like it'd start immediately. There's some potential the amp is going into oscillations. I'm gonna try to reprpduce it in a way that doesn't threaten to blow up my speakers. I got just a hint of 'warm' smell from them on the second go.
The sound was very harsh, even at low level. When it got loud, there was a 60Hz tone coming through, but that may have been the power supply overloading, and the clipping following the ripple, as I think the amp was deep into clipping when it got really loud. A possible idea is to use a buck transformer on the line voltage- I can knock 40 or 50V off the line voltage. That'd bring the power rails down to +/-20V instead of 35V.

Entire getup was the two amps sharing a C13 Y cord, then an RCA cable from either into the preamp. Preamp plugged into a nornal double insulated switching wallwart. The outlet it was all plugged into has no ground at it. I could individually ground lift the two amps to remove any loop, but I don't think such is to blame. Preamp fed via bluetooth and speakers not referenced to anything. The speakers have a cap on either tweeter, but the woofer is directly across the terminals.

There's a chance it may be a thermal thing, but the chassis didn't feel that warm. I'll check that as well when trying to recreate it.

I just taped a piece of paper to my drawing tablet. I remembered other people did it, but it's my first time doing so. I have slightly shaky hands sometimes; it worked wonders to stop that, due to the extra friction.

A loop or feedback issue seems like it'd start immediately.

Might start at supersonic or subsonic frequencies, though. It ramps up in volume, right? It isn't just going zero to 100 instantly?

There's some potential the amp is going into oscillations.

I consider that to be a form of feedback loop.

Alternately, if the power supply is going crazy when this happens, might be a bad ground.

Or maybe it's a failed decoupling capacitor that saturates and then shorts? (depends on the failure mode of whatever kind of cap they use though)

Could it be thermal expansion causing some un-isolated winding, solder pad, or something of that sort to shift and make/break contact with something? Or arc?

Just throwing ideas off the wall, never had the problem you describe.

This post has been edited by dragontamer8740: Apr 5 2022, 22:10

Too long since I've worked on amps.



It is hilarious that Elon Musk is on the board of Twitter in return for not taking over the company.

-Cynical; hopefully Twitter will fail

-Positive: Hopefully Twitter will fix some of their practices

He's not allowed to take over the company before the 2024 election. At least they have their priorities straight.

This post has been edited by EsotericSatire: Apr 6 2022, 00:11

Uhh... It's been available for newer OSes for ages. There's an official release of it for Windows 8.1 you could have extracted.

Nah, those are among the "complicated things" that I tried in the past, that additionally didn't work properly, making me think an exe from Win XP would definitely not work because using deprecated OS features or something.
Also copying the XP exe means no install, no registry filled with more crap, more simple happiness like in MS-DOS era.

Nah, those are among the "complicated things" that I tried in the past, that additionally didn't work properly, making me think an exe from Win XP would definitely not work because using deprecated OS features or something.
Also copying the XP exe means no install, no registry filled with more crap, more simple happiness like in MS-DOS era.

You chose this by continuing to support microsoft's increasingly bad products. I have near zero sympathy.
Of course, DOS was also awful if you ever had to touch config.sys or autoexec.bat.
Anyway.

CODE

wusa file.msu

or

CODE

7z x file.msu

…done

And no, they haven't changed that program in years and years. I wouldn't be surprised if it were bit-identical to in later versions.
I don't see a point in getting it from an XP install when it's a download away though.

If you actually want DOS, go use DOS. And IIRC you can still use the XP SP1a progman in windows 8 (maybe 10). I know you could in 7 for a fact. Guess how progman saves information, though? (registry)

Could be that those MSU's are just "update clients" or something and the program itself isn't inside them, though. That sounds very microsoft-y.

Finally, I will once again mention [github.com] https://github.com/rel-eng/QWinHelp

This post has been edited by dragontamer8740: Apr 6 2022, 17:08

Correction: Input path is a 93pF decoupling cap, a 2.2uF series cap, then another 220pF decoupling cap at the chip, all caps ceramic. There's no DC leakage path present I can see with my meter either polarity at the chip. The datasheet of the chip shows the input directly feeding the base of a transistor. It also shows a 1K resistor in series with the pin fed by a 10K pot to ground. Somebody on IRC last night mentioned a 1K series resistor without consulting the datasheet or knowing the exact part..

For the test rig, I used an 18R 10W resistor in series with a trash 4Z speaker for either amp. It went about 5 minutes, then repeated going insane. With the smaller speaker and more gain used, I could hear it breaking up more clearly, and there was no 60Hz note when it finally lost it. Of note, the little speaker let out some smoke (Which is fine, I was literally going to bin them; 3" full range ~20W speakers that just needed a refoam or the OEM foam reglued- Instead some dipshit tried fixing them CA and made them a write-off.) So, it was well deep into clipping when it broke up. Heavy clipping tends to make square waves, which deliver roughly double the power of a clean sine wave. Also telling, the case of the amp stayed cool, so I can rule out thermal effects.

So yeah, amps are going unstable, because I didn't RTFM the datasheet. I'll add the series 1K resistor after my blocking cap and a 10K or 100K to ground.

This post has been edited by Wayward_Vagabond: Apr 6 2022, 15:29