Love you can now buy a esp32 display module, create a PCB for it. Then with 0 effort, download a 3d model (where you can enable/disable parts) and generate a same model for your PCB. Match everything up, and see if you made any errors in all the sizes, alignment, etc. It makes things so easy!

Watched this one this morning, www.imdb.com/title/tt0327.... Enjoyed it, wonder how I missed it up to now.

Played a bit with wifi settings and tweaking the queue'ing and got it to play without deadline miss. I wonder if I can push it to RGB666 instead of RGB565 (uses 3 bytes per pixel, so a big increase in space needed).

I hate shipping companies.. I was home, did not ring the doorbell, it was small, so it would fit in mailbox. No, just drop it in front of apartment complex.. Luckely I noticed it in time. and this is how it looked..

2nd version soldered.. Shacking hands is getting a bigger issue, needed two hands to place components, one to hold the tweezers, one to stabilize first hand. The hotplate way luckely helps straightening out the components. It all worked out, but I wonder at what point I cannot do this anymore :-/

Got my own si5351c board in from @aisler.net. The solder job is very bad (shaky hands today), but so far it is findable over i2c. And yes, I forgot to order the 25MHz crystal, but luckely I ordered this one because it can take in an external clock (all the hobby boards I found have the si5351a).

Another ali haul. Got a round 1.54" screen. I've been using 1.5" screens for years now, they are cheap ( ~3€), high res (240x240) and good viewing angles. This one is *slightly* larger, higher resolution 360x360 pixels, cheap (3.90 €) and is qspi. Hopefully this one is the new go to screen.

Made a small case for the Icepi Zero. Link here: makerworld.com/en/models/26...

The opensource tools for the zeropi/lattice are great. Got a platform with our riscv32i cpu up and running with little effort. A synthesis takes 10~14 seconds, even for a small design that is pretty amazing. Got PLL working, Tri-state input/outputs and the uart that is hooked up to the usb.

Got a icepi zero to play with. First step is to see if I can get my RiscV running on it and the sdram to work. Hopefully all with opensource tools!

*sigh* where ever you go, spam/e-begging will follow you around..

Now replaced the npn transistor in my flux 8120 by one a bit more appropriate. It still works, did a quick calibration check and it seems to be still pretty good, but not into spec. Still better then several cheap dmms I had lying around.

ESP32-P4 PPA weirdness. I wanted to make images fit the frame a bit better by cutting up a small slice at the top/bottom and scaling it a bit less to fit. However when I add that to the ppa_srm_oper_config_t, I get these weird little block pattern. As far as I can see, I do nothing wrong.

My ugliest fix ever.. and regrettable not without damage. (to desolder transistors, 2 pads (that did not connect) where lost, I managed to save a third. But fluke nixie dmm works again. The NPN died, PNP measured fine out of circuit. After replacing the NPN, the PNP failed. 8's and 9's all around!

Dove into my broken Fluke 8120A multimeter. A nice one with nixies. Found why it won't display 8/9 well. Most likely one (or two) transistors in the ADC 4th bit feedback line are broken. Now to figure out if i can still get these, or to find a suitable alternative.

The Reflective LCD waveshare board I've been using has some odd quirks: The RTC (with bat) seems to 'loose' state flashing the ESP, or pulling the main battery. Kinda defeats the purpose of the main RTC. I did manage to get the idle usage down (no deep sleep) to 2mA with some spikes on redraws.

A while back I got quiton jc-esp32p4-m3-dev and made an immich photo frame out of it. I saw a really cheap small esp32p4 board from dfrobot, the firebeetle 2 esp32-p4. At a local store ~15 euro, but has same spec (16mb flash, 32mb psram) and a DSI/CSI port, no eth. Drop in replacement for the frame

PCB came in today to duplicate my 10MHz signal. Now I have 6 50Ohm output on my reference clock. (forgot to take picture of soldered board before installing) So far it works fine. My Siglent AWG now uses this as reference clock (and according to my scope/freq counter) it increased it accuracy sign.

My desktop CPU is weird.. It has a large cpu cooler. Under full load the core sits at 78C. All cores have a 5.1GHz boost. Even then the cpu cooler is ~ 30C.. With heatcam it all looks cool, even at the cpu. I think to get cooler core temp I need to delid the cpu. Not sure I want to take the risk.

Finally got the 0.5x adapter for my microscope in. The different in image quality is night and day. (I still wonder if the 0.35 I had was broken.. ) No more purple glow around all bright stuff (I couldn't get rid off) and very limited focused area. Hopefully I can make better images of my work now.

Weird vintage find at work.. A BNC to banana plug without ground. Also allows a wire to be screwed in. For my next 200MHz measurement 😝

Had a nice example of how DDR memory keeps state for some time after powering off. Working on an Zynq setup, where FPGA writes to a FB in the shared DDR. I triggered an overcurrent causing it to power off. After remotely resetting the efuse, starting it up the test image was there, with some noise.

Wayland rendering is "fun".. You only know after the first 'frame' what exactly info is of the target monitor. This includes DPI and scaling factor.. This makes showing an UI at a certain physical size hard. Fractional scaling adds more fun, the application only gets an integer factor.

Working on running 80 fpga zynq boards over nfs, I wonder if I should keep using ubuntu.. 24.04 has many things broken (working in 18), only in the arm version. Got it working in the end, but required several 'firstboot' fix scripts to get everything going correctly. Maybe time to switch to debian.

@aisler.net oops it's broken.

Almost done, waiting for PCB to arrive from @aisler.net to add the finishing touch. (actually wire up the 6 10MHz outputs)

Took a picture of frontplate I was working on.. it got a whole black 'spot' after processing (it looked fine, then after a sec it showed this). Took the same picture in 'pro' mode and its fine.. Wonder WTF google camera app is doing..

Making a PCB to 'duplicate' my 10MHz reference signal. A quick test on perf board to test this out. Green below is the output. Not perfect, but good enough.

Ordered a 0x5 c-mount adapter on aliexpress. It listed 'DHL, Post-NL,..' as shipping. They used PDN. They made 1 attempt in the middle of a working day. Claimed they 'called' me. No more attempts, no way to contact them. The one number (mobile) I dug up, gives full voicemail box. The Worst...

Not wanting to drill into the case, I've 3d printed some caps for the holes that normally has the carrying handle. One has a nice 3d printed 3/8" thread that nicely takes my bnc connector.

My whole 'stable' clock started after rescuing a HP5315A. I wanted to calibrate it. Now it spiraled to give it an upgrade, a OCXO or external input. I saw people hacking it in using the OCXO option, but this was expensive (100 Euro). Going through the PCB I saw the extr. input header was there.

This is control that does 'rapid' control first if the offset is > 0ppm. Then, after 1 hour, it goes to slower control with a 1 hour window and updates every 20minutes. It looks like I need to switch to a longer window/interval for the control later. Currently on chip I keep around 2 hours of data.

Learned the hard way: If you want va_arg with 64bit arguments on 32bit hw you need to make sure the top of the stack is 8byte aligned. GCC generates an alignment ptr = ((ptr+7)&-8); and then uses that to find the 2 32bit values on the stack. It will put 64bit arg. on A2&A3, A4&A5 slot in stack.

On my OCXO I keep track of offsets in 30 seconds interval (PIO counts for 30 seconds). From this I can see that during an hour I have 12 blocks that have an 'offset' (6 +1, and 6 -1) . Not sure if this is instability in my measuring or the OCXO. Still very good and better then I have use for.

In my reference clock project we now can: * accurate count the 10MHz clock * Count against GPS PPS to get offset over longer period. Next step is to 'tune' the OCXO based on the calculated offset. The board uses a 204k pot-meter, allowing you to tune OCXO Vref between 1.96-2.32V.

Fun with SMD markings.. A divider, with 363 resistors, and 688 or is it 68B? 688 makes no sense.. but the last 8 looks more 8 then B, yet it was. Took me an (embarrassing) while that some where in 3-eia others in eia-96..

In latest ALI haul, got a few SI5351 clock generators in. Not for myself, but to help somebody out (results will be published). They are easy to use, not accurate, but good enough. Another thing in this haul was a bnc to banana plug cable, I bought one before locally, this was <1/3 the price.

hmm 'bug' in software, my opnsense router eaten through a large amount of data on the nvme... Spare: 100% Spare Th: 10% % Used: 26% Read: 140,971,432 [72.1 TB] Written: 56,692,629 [29.0 TB] In 23530 hours..

Another small alie haul.. For my gps conditioned 10MHz ref. clock, an active gps antenna. (I got two for 1/15th of the price of buying one local) I also got some SMA connectors, cables and a bunch of wire with silicone sleeve.

Center negative.. This must be done explicitly to screw your customers and sell your own power supply. Beside Brother my 30 year old Roland keyboard is the one other device with center negative...

Not sure if I messed up.. but 10Mhz ocxo, counting pulses over 1 hour. gives me this offset. So that is 3 pulses more over an hour.. now how to validate I did not screw up.. Specification is 0.05ppb, but it is clearly a used module (from ali).

Stupid connector gremlins.. NAS reported one disk dead.. (nvme).. so got it from my parents place (remote backups).. reseated the disk.. everything is just fine again... Happy the disk still worked, its now almost twice the price then when I bought it 1/2 year ago.

Thanks to @oldtaki.bsky.social who gave me this hint. I got some 'sensor casing' in, for my outside sensors. Now I need to remember how I did reprogram these sensors again to make the outside one use an sht31.

Why do they keep making pictures like this? That soldering iron is: 1.) Never used, the tip looks like it never been turned on. 2.) Held wrong, burning the finger of the user. 3.) Cheap model iron (not suited for this work) 4.) Not how you work on a GPU. www.maynoothuniversity.ie/news-events/...

First part of the prototype working. Now to design a PCB so I can actually adjust the Adjustment input of the oscillator. Not sure what the best way is to generate the reference voltage. I do have a 16bit DAC left from my sm7020 driver, those might work.

Using the PIO in a RP2040 I am trying to find the offset on my OCXO. The PPS signal of gps is on one pin, the OCXO output on the other. I could the pulses between 20 PPS signals. This works really well, I can see it has an offset of 0.13 ppm. Next step to write a loop to adjust the oscillator.

Why why did they thought a rubbery exterior that slowly degrades over years was a good idea.. Just spend 15 minutes removing it from my infrared thermometer that felt more like chewed gum.. I have 2 remotes from squeezeboxes that are unusable (pain to clean). I hate hate hate it and now avoid it.

Got an A.I. assisted patch for #rofi. Claiming 20,30 times and in some parts 15412x speedups. One huge patch touching removing 822 lines of code, adding 1336 files. Trying to test the claims (on multiple machines) I got a speedup on one*, equal on another and a huge slowdown on third*.

And I was wondering why I did not get any readings.....

Worst PC setup I've seen in while: PSU in top, CPU blows towards PSU. 1W fan in the back blowing out, one 1W fan in front intake. 2x1W fan inside the case. All fans (because of ipmi) running at 700rpm. Now the PSU is 1.5kW be quiet, CPU is a 64cryzen with 512gb ram. The PSU was to hot to touch.