hi comrades update poooost

it's been a bit. three months or so since the last post. this project is deeply near and dear to my heart and it's not abandoned but there was a bit there where i fell off the face of the earth and i wanna fill in that gap.

i've been off and on for the past three months, but for the past two weeks in particular i've been on. i made a commitment to myself to spend time on this before work - this is my most valuable and reliable time slot in a day - and i've been reliably making updates every single day i haven't had a prior obligation. it's a habit now. i've been deeply personally turbulent lately, much more for better than for worse but in ways that really take most of my awake time - except this before work slot. wouldn't want to spend it on anything else.

the posts dried up for two reasons - one, the development recently has been feeling like the last leg of a marathon. nothing new is really happening, there's no news besides that i'm closer than i was last time, and there's nothing notable to report until i actually dot the final i and cross the final t and then i buy one and i get it in my hand. two, writing the posts actually took a lot of time - i was finding a few hours per week for sphynx time, and post-writing became one or two of them, because i care a lot about documentation and presentation and communications. that's kinda not sustainable. my fix for this has been taking things to matrix. sphynx has a matrix room now. i find it FAR more sustainable to bang out a couple quick 30 second messages when i finish up a work session than writing a post aggregating all of those every week or two. if you want know if/when/how i'm still working on the project, you can either check the git logs, which is where most of the work will live, or join the matrix room for more interactive and human readable updates (and even a place to hang out??). DM me your matrix handle for a link ✨

as for technicals, here's a brief summary of what i've done since the last post:

• implement some additional safety features, including but not limited to top-level current limiting diode to prevent overcurrent, inline resistor to also prevent overcurrent, and a parallel bleeder resistor to prevent the probe being weird inbetween pulses
• replacing parts that i spec'd from the JLCPCB extended library with basic library parts - each time i do this the sphynx gets $3 cheaper.
• in general, getting all parts at least in the extended library so that you can order a sphynx, not even own a soldering iron, and get to using it. this has involved a lot of part swaps, some which are moving targets - spec a part, find it on JLCPCB, replace it, do other stuff, come back, it's out of stock and you have to replace it again. very frustrating.
• updating PCB graphics - better knob indicators, better labeling, paring down unnecessary test points and jumpers and other junk, etc.
• probably more but i'm forgetting

realistically, this is the last post until i have the sphynx lite RC2 in my hand. as it stands this is likely my whole work sesh for the morning but i dearly love you all and you deserve to know what i'm up to and this was a good use of my morning. thank you for everyone who's been following. let's talk soon.

heyyy friends

It's been longer than I wanted it to be, I've had a hard time lately, everything from work stress to coming out to my highly conservative family to some particularly nasty dysphoria, had me cruise right into a nice little light to moderate menty b - I'm fine, it was a growth opportunity, sometimes you just indulge these things and get in your car and drive with the music up real loud and hop out somewhere you can watch the sunset and you cry alone for a while and you savor it and you come out of it cleansed. I'm good. Being a girl is hard. Still figuring it out. <3

This is a light update in terms of work I've accomplished but it's a big update in terms of how this project becomes something more community related. the title of this post is actually a double entendre, you see - not only do I have issues, but the sphynx project also has an issue tracker.

Issue tracker for sphynx (the electronics): https://todo.sr.ht/~_410bdf/sphynx?search=status%3Aany

Issue tracker for sphynx-site (the website, both content and style): https://todo.sr.ht/~_410bdf/sphynx-site?search=status%3Aany

This changes a lot. I've been stressing with each post about ways I can ask y'all for help, because I 1) want to take some of the load off of myself, and more importantly, 2) turn this into a truly community project, where people all across the world leave their mark and have this be something we do as together as we can. Now that I have an issue tracker, I can start unloading my brain asynchronously from these posts, providing all the information in a neat compartmentalized fashion per issue, and teeing things up for community members to take on. It also serves as a great work journal, I still make, comment on, and close issues that I work alone, so it's a good place to check in to see that I'm still doing things (even though it's been a minute )

I've done some electrical work since last post too, pretty much just bug fixing and improvements from RC1, and if you'd like to read about it, it's all (albeit very tersely) captured in the issues! Go dig around and have fun!

The site is a little fresher too! Just a bit, it's still not inhabitable yet, but we're getting there. https://sphynx.diy Huge shoutout to @Edie@hexbear.net, it put in a really a lot of work and sent me more than a few patches, including fixing some bugs I had spent kind of not a little time on, plus, of course, doing the entire original Jekyll port and the entire original CI pipeline. You're real for that, thank you

OH! And message me on matrix??? @410bdf:matrix.org say hi! I love hearing from y'all! Or say hi here! Huge shoutout again to @Edie@hexbear.net for reaching out on Matrix and talking with me and hyping me up and talking about the project together with me!

Not much else, this is a quick one to let y'all know I'm still alive and to share out the issue tracker so I'm more confident to ask for help and have it be effective. As usual love y'all. I've been using RC1 and it still works but it's only getting better from here. Byeee 🫶

Hi comrades! I'm back! Sorry for the delay! I've been keeping busy, with this project, with work, and with getting familiar with a new town. This update is fun because it's actually entirely devoid of electrical engineering nerd shit! Since last update, I've focused on two things - getting the framework of a website together, and actually getting usage hours on the device so I can know how it handles, what it's missing, what it does well, etc. I'll break this update into those two sections.

Device usage

I think I have about 5 hours in the device so far, with I'm guessing a couple thousand or so pulses. I've tested my face, eyebrows and beard, I've tested my hands, I've tested my arm, I've tested my leg, and I've tested my chest. The general routine is as follows:

• Get clean and get comfy.
• Get all the peripherals in order and get the board set down somewhere stable. For me, this is getting good direct bright lighting, getting the foot pedal in a good spot, getting a clean (or at least clean_ed_) probe in the probe holder, getting the return electrode connected [1], and for some work areas, particularly with vellus hairs, getting a magnifying headset on.
• Dial in estimated power levels. The thickest and darkest hairs I've tested on like about 2 mA at whatever voltage it takes to feel the burn, usually >= 10V, and pulsing for the whole 10 second duration. I've gotten facial hair kills at 1 mA/6 V/6 seconds, and vellus hairs on my hands can typically take 0.5 mA/6 V/5 seconds. These are determined by trial and error and I almost always fiddle with them multiple times in the middle of a session; there's no rigorously defined science besides the Units of Lye calculations, which are very loose. More current leads to faster kills but it hurts more and it's easier to accidentally overdose and get a hyperpigmentation scar. I've never managed to do this accidentally, only on purpose, see below.
• Put in the probe and hit the pedal. Wait out the indicator light [2], feel for the burn, and at the end of the pulse, pull out the hair.
• Repeat!

Some areas are clearly not regrowing, others, it's too early to make a call, but I'm pretty confident that it'll be minimal, possibly even better than salon electrolysis, due to a number of advantages to self-work that I'm finding that I'd love to detail here. There are some cons too.

• pro: you can feel the power. You very quickly get a feel for what a good follicle kill feels like. A well killed hair will cause a burning sensation strong enough that it lasts for a few seconds after you stop applying current. You can also feel overkill - I did this a couple dozen times to experiment, the follicle isn't any deader than any of its neighbors, but one or two of the deliberate overkills has what appears to be a hyperpigmentation scar that looks like a freckle. They're already fading, but they might be permanent and might not. I have never accidentally overkilled but it's definitely possible to do so, especially if you have a healthy appetite for pain and/or are working impatiently. An overkill probably takes at least 3 times the current*time that the minimum safe kill does, so it's not super close.
• pro: you can feel the pluck from both sides. I do have minimal testing on another person and feeling the hair pull on one side is helpful (a dead hair slides out with a pretty constant slide, a live hair tends to hold on and hold and hold and then break loose all at once). However, when you're both feeling and plucking, you can feel for the signs of a killed hair more effectively - killed hairs tend not to hurt on the way out and you can give them a light test tug and not feel it. A hair that hasn't been killed will hold on and cause a little bit of pain with a test tug, allowing one further way to check your kills. I strongly recommend both parties having experience (just a couple hairs) on both sides when doing co-work to foster this kind of mutual understanding that allows for better communication.
• con: obviously you're not getting a lot of your body. You need a friend to get everywhere, particularly a lot of the spots that are really important. My biggest issue has been trying to self-work my neck. It just doesn't wanna go, the angles do not hit. I believe it's an area that will be possible with more practice and dexterity, plus a smarter setup - I'm going to try a mirror setup to see while lying on my back, but it's frustrating. :angery:
• con: fatigue. When you're working on someone else, you can get comfortable. When you're working on yourself, you're very often uncomfortably contorted. This poor ergonomics greatly shortens session time and poses an issue for our comrades with disabilities. My DIY sessions usually last an hour tops; I have some pain issues in certain positions and this does NOT help.

I also have some notes on equipment:

• [1] - I had a good idea that I'm a thousand percent recommending for self work. Salons typically use a holdable piece of metal for the return electrode. I am using an ECG electrode. For partnered work, this is nice because it allows your recipient to get more comfy and do things with both hands, but for self work, this is almost indispensible - it lets you keep both your hands free. I'll put instructions for making and working with both on the website.
• [2] - Visual cues suck. You do not take your eyes off of the hair for the whole cycle. I initially had the lights hard-wired and the buzzer switched - the lights absolutely should be switched, they're not useful most of the time. I'm working without a buzzer due to the soldering mishaps in the previous post and it sucks, the beep cue is almost certainly a thousand times more useful.
• Pedals are great for a salon-like context of a laying recipient, a seated operator, and personal space. For every other configuration - standing in front of a mirror, laying on the couch next to your bestie, etc., you probably don't want a pedal as much as you want a hand switch or a bite switch.

This isn't all inclusive, but I'm definitely happy to get thoughts down prior to starting work on the manual! which leads us to...

The Website

It's on jekyll now! We're finally ready to start hosting actual, real content there! So, so many thanks to @Edie@hexbear.net , who actually single-handedly ported the old single page splash screen to Jekyll and on top of that pushed out a bunch of fixes for my sloppy half-assed jekyll code. I really, really appreciate you, you're genuinely helping both to motivate me by both making indispensable contributions, and also just by being a person out there in the world doing this with me.

The appearance and organization are both super preliminary, and there's not any actual content, and the design is also very much a prototype. The important thing is that now I (or anyone else!) can just write markdown and have it reflected online in an easily readable and shareable format. Email patches are a great way to get both code and content on the page, I try to review them at least twice a week or so. Now that things are set up, it's extremely convenient to put content up. I think to pilot the website, I'm going to make a more formal tutorial for the pencil based probe to start. The one I made per @YearOfTheCommieDesktop@hexbear.net's instructions is still doing phenomenally, thank you for your research and development. Similar thanks for being a part of the project in a meaningful way and making this a team effort.

Come take a look! https://sphynx.diy

What's Next

So the project is hitting a bit of a fork, where we finally have multiple parallel work streams at a time - the online manual needs to be designed/written, and the RC2 version of the PCB needs to be designed and ordered. I have a list of changes from working with the RC1 that I'll be rolling in, that'll be the next post. I'm also going to start using https://todo.sr.ht to track issues, both to keep myself organized and to publicly advertise what we need to get done in a neat encapsulated way. I'll have details for that up on an #8.5 post in a couple days. For now, honestly I feel bad for drastically overrunning my two week timeline and leaving y'all in the dark for so long, so this post is going up ASAP. I've been busier than usual and probably busier than I plan to be in the future, so I'm more optimistic for a timely #9 post, although that may be in three weeks and not two due to some plans of mine.

I love y'all. RC2 might be the release. We're just weeks away from other people benefiting from this project materially, it's just refinements from here. Thanks for the support, and you know the drill, stop by, say hi, ask questions if you want to understand things better, make suggestions, all of it. See ya next time.

it's in and it's built and it FUCKING WORKS, at least like 80% of the way - I haven't fully put it through a full round of testing yet but every single functionality I've tested - including things that were first-time builds that I hadn't prototyped yet, like the adjustable LDO on the output and the double schmitt trigger falling edge detector, all seem to work! I even splurged a little bit and got some trans knob caps .

What I've been up to

I don't know what to tell you, read 'em and weep. The board is built, you see it, it goes hard. Pretend my soldering isn't shitty and I cleaned the flux off.

Full list of things that work:

• Output voltage control knob works, tested open loop (top end is supposed to be 16 and it's closer to 15.5, I might just not worry about it.)
• Output current control knob works, tested through a 1K resistor.
• Timer and knob work from 2 to 10 seconds.
• Lights work*, although I think I killed the 555 timer by soldering near it with the board powered . It worked great for a few minutes and then I soldered in S1 and it stopped working.
• Foot pedal and probe work.

I don't know if it has any new bugs when specifically connected to a human being. I suspect that if I were to try to use the current sink on the low side of the probe, which I'm bypassing right now, it'd be whacky - I never made an effort to resolve that bug, and I think I might just descope that to get this thing done at all. I'm already longing for a fresh start on some things that I think I can get by launching into the microcontroller-based version of this. There are already enough redundant safety measures in place for me to feel comfortable, although I'd really like to test the JFET current limiter, which I didn't populate because I couldn't source a good JFET and I also couldn't be assed to determine the correct resistor value. I'd really like either that or a current limiting diode on the high side of the amplifier before I call this done. That's going into the next rev. There are also some non-breaking bugs that I still need to design out, like for example I designed in 20K potentiometers for the current and time control, but Alps Alpine only makes 20K potentiometers in audio trim, meaning the knob angle isn't really one to one with the output. I need to replace those with different values. Other little things like that too.

Next up

I'm going to work with the board for a bit and see how it handles! I need to get some hands-on time with it to truly learn how it works, what it's missing, and what I need to change. I'm also kind of just looking to reap the benefits of this thing personally to be honest, I'm getting really tired of shaving

(plus i'm meeting some girls who are kinda into this thing and might wanna go hour for hour on some mutual aid electrolysis time over comfort shows and snacks on the couch )

I think the "alpha" model of releases was a mistake. Because of the nature of hardware development, I think I'm going to change to a "release candidate" model - it makes it clear that any RC board isn't ready to be used, and it allows me to promote any one at any time once I've designated it good enough. So, I'll begin work on RC2 once I'm deeply familiar with this board!

I'm also deeply neglecting the site, mainly @Edie's jekyll port that I really deeply truly appreciate and I never figured out how to apply the patch for. I'm sorry friend 💔 Can you work with me to get that patch applied? I tried to tackle it on a super low executive function day and I just made so little progress applying your patchset and I wanna have it up so bad but it was fighting me and I needed rot time and I quit trying after like twenty minutes

Any ways to help?

There are kind of a few things actually! In no order of priority, with loose guesses at difficulty:

• Getting the library I'm using updated with 3D models. It's not super necessary but having cute renders is always cool, and it'd be handy for anyone who wants to design an enclosure. Not hard, not easy.
• Fixing some of the footprints, particularly making sure all the knobs look the same, making the font nicer, etc. Not super easy.
• Doing a JLCPCB cart catalog audit. I think I'm close to everything being in JLCPCB's catalog? I'm not sure though! If there's anything missing I'd love to know so I can try to design it out in RC2! Kinda boring but not too difficult.
• some more?? i'll edit them in in the morning i'm up LATE

If any of these things sound like a thing you want to make an attempt at, let me know! I'll work with you to get you started.

sloppy post today, this is deeply not accessible for non-technical audiences and leaves a lot out, so please let me know if you want deeper explanation on anything and i'll add detail! I just wanted to make sure i got my post up to let you know I'M STILL FIGHTING BABY

i'm still very not settled with the move, and burnout is closing in with my job and my new trans social life, but I'm at equilibrium, I'm okay to keep spending the amount of time I am on this, but I wish I could be spending just a little more. Life is okay though. Great, even.

As always, stop by, hang out, say hi, ask questions, tell me what you've been up to, design review me, however you'd like to be involved is good by me! I'll see you in the comments 🥰

cross-posted from: https://hexbear.net/post/2679948

things are HAPPENING besties

hello comrades! It's been a minute! I'm back with my first tech update since upending my entire life and getting re-settled! Let's talk

What I've been up to

See for yourself, pictured is the first alpha of the Sphynx Lite. Schematics are on git, go check it out (and if you want to set up KiCAD to view or edit but aren't quite sure how, say so in the comments and me or someone else will help you out!)

Issuing guidance on the usage of these things is complicated. I'm always going to be a little apprehensive recommending that people use this, because I'm a very cautious person. For now, I think my official guidance is:

• you should understand the circuit
• you should have basic debugging equipment
• you should have the time to work with it extensively
• if you meet all of the above, I provide this with no warranty and no guarantee, knowing you full well have ideas as to what you might want to use it for
• If you have a decent working knowledge of electrical engineering and want a walkthrough, I might make a Matrix so I can actually communicate with people looking to get it going??

As it gets more mature and I revise and revise until I drop the alpha, I'll trust it more.

This was a push across the finish line. I optimized for getting something testable done with the correct topology, layout is a little dicey (cut off silk screen in the bottom right ) and subject to change, I just needed something to test on. I was gone for over two weeks from my last post here, mainly because it was just a lot of work with nothing intermediate to report. But now it's here! This has all the working parts of the last boards merged into one, with the new additions of:

• power conditioning
• 9V battery holders
• additional safety measures, like the redundant current limiting JFET
• configuration jumpers! and test points!! !!!

Oh, and I whipped up a cute lil logo. :3

Schematics

schematics in here

Board Layout

board layout in here

Board Renders

board renders in here

Next up

I have a lot of really good work from y'all that I need to capture and incorporate. Particularly, my immediate plans while this board is fabricated and shipped are to:

• Incorporate @ComradeEd@lemmygrad.ml's patchset porting the site to Jekyll - really thank you so much for this, I've been extremely focused on the board, and now that I have a minute, I can get the site going so I can use it to show off, post guide, aggregate educational resources, etc. This is going to be absolutely necessary for the project to have the reach and accessiblilty I'm hoping it gets. Thank you.
• Build a probe per @YearOfTheCommieDesktop@hexbear.net's work into coming up with a probe holder. I have a handful of the Yasutomo's to experiment with. This is going to be a game changer, currently I'm doing self-work with an alligator clip and it's extremely irritating. An ergonomic and reproducible probe with an extremely cheap parts list is critical for the project to function. Thank you.
• More of you provided meaningful help than just these two, these are the two who's work I'm directly interacting with this minute. Thank you to everyone who has stopped by and made suggestions, I've read them all.

The next post will be me building this out and reporting on how well (or whether) it works, and documenting changes for the Lite Alpha 2.

I also badly need to make a BOM. With some effort, anyone who's built a board together can buy all the parts for this and make one, but it'd be a thousand times easier if I just made a single cart that you can buy that includes every single component. If you want to make an alpha board for fun or debugging, maybe hold off until I have that out.

Any ways to help?

Honestly, those of you who are following this pretty closely have a good read on what's going on, what's needed, and what's upcoming - keep being interested! If anyone wants to make one, I do recommend waiting until I release a BOM, but very soon we miiight be at the point where other people besides me get one of these in their hands.

I could also use some polish on the logo. It's fine, but it could be cuter, and the lines are a bit funky. I'm going to put SVGs on git soon and if anyone feels like cleaning them up, rearrange the kitty cat face so it's cuter, fixing my wonky paths, let me know!

Also, and I simply cannot stress this enough - when I've been exhausted, when I've been deep in executive dysfunction, when I've been not feeling up for it, I've read through all your encouragement and support throughout the duration of this project and it's helped me to keep pushing. I would unquestionably not be this far without y'all. Thank you all so much.

As always, stop by, hang out, say hi, ask questions, tell me what you've been up to, design review me, however you'd like to be involved is good by me! I'm thinking I want to make the expected cadence of posts once every two weeks, just because I'm busy and I don't want to cause alarm when I miss a week. I can always surprise y'all with more frequent posts too.

P.S. - I have no hair regrowth in test areas.