Petr Dvorak
@thatkicadguy.bsky.social
🚀 #ThatKiCADguy :: Content creator :: HW designer :: 🤝 (He/Him)
🤝 Electricity is all about connections. 𝐋𝐞𝐭'𝐬 𝐜𝐨𝐧𝐧𝐞𝐜𝐭!
👉 I am 𝙏𝙝𝙖𝙩 𝙆𝙞𝘾𝘼𝘿 𝙜𝙪𝙮
🚀 [email protected]
🤝 Electricity is all about connections. 𝐋𝐞𝐭'𝐬 𝐜𝐨𝐧𝐧𝐞𝐜𝐭!
👉 I am 𝙏𝙝𝙖𝙩 𝙆𝙞𝘾𝘼𝘿 𝙜𝙪𝙮
🚀 [email protected]
Sometimes the fastest way to correct a DRC violation is to override it. I'll tell you why and how.
When you route in crowded spaces, for instance, in between connector pad holes, as in the picture, it's challenging to fit in.
The solution can be to override the DRC for a while.
When you route in crowded spaces, for instance, in between connector pad holes, as in the picture, it's challenging to fit in.
The solution can be to override the DRC for a while.
February 13, 2026 at 6:06 AM
Sometimes the fastest way to correct a DRC violation is to override it. I'll tell you why and how.
When you route in crowded spaces, for instance, in between connector pad holes, as in the picture, it's challenging to fit in.
The solution can be to override the DRC for a while.
When you route in crowded spaces, for instance, in between connector pad holes, as in the picture, it's challenging to fit in.
The solution can be to override the DRC for a while.
This time, I wasn't an HW designer. I copied the example schematic from the datasheet, and it worked.
I almost feel embarrassed that I copied it 1 to 1.
But I couldn't resist checking the schematic. Plus, the circuit was bright and clean.
So I did it.
I almost feel embarrassed that I copied it 1 to 1.
But I couldn't resist checking the schematic. Plus, the circuit was bright and clean.
So I did it.
February 12, 2026 at 6:06 AM
This time, I wasn't an HW designer. I copied the example schematic from the datasheet, and it worked.
I almost feel embarrassed that I copied it 1 to 1.
But I couldn't resist checking the schematic. Plus, the circuit was bright and clean.
So I did it.
I almost feel embarrassed that I copied it 1 to 1.
But I couldn't resist checking the schematic. Plus, the circuit was bright and clean.
So I did it.
I felt like a nuclear plane manager the day of filming this demo.
Cables, DMMs, code, MCU, PCB...
It turned out to be the most satisfying video taking of the last 12 months.
The Texas Instruments INA2227 is genius.
Cables, DMMs, code, MCU, PCB...
It turned out to be the most satisfying video taking of the last 12 months.
The Texas Instruments INA2227 is genius.
February 11, 2026 at 6:06 AM
I felt like a nuclear plane manager the day of filming this demo.
Cables, DMMs, code, MCU, PCB...
It turned out to be the most satisfying video taking of the last 12 months.
The Texas Instruments INA2227 is genius.
Cables, DMMs, code, MCU, PCB...
It turned out to be the most satisfying video taking of the last 12 months.
The Texas Instruments INA2227 is genius.
A perfect docs
A worldwide accessible community
The smoothness of the ecosystem mesmerised me
Why the Zero?
I want to dig into Python embedded programming. With a big help of LLM, oc.
Sometimes, I deliver a device powered with a Raspberry Pi And I desperalely need to know how to test it, etc.
A worldwide accessible community
The smoothness of the ecosystem mesmerised me
Why the Zero?
I want to dig into Python embedded programming. With a big help of LLM, oc.
Sometimes, I deliver a device powered with a Raspberry Pi And I desperalely need to know how to test it, etc.
February 10, 2026 at 6:06 AM
A perfect docs
A worldwide accessible community
The smoothness of the ecosystem mesmerised me
Why the Zero?
I want to dig into Python embedded programming. With a big help of LLM, oc.
Sometimes, I deliver a device powered with a Raspberry Pi And I desperalely need to know how to test it, etc.
A worldwide accessible community
The smoothness of the ecosystem mesmerised me
Why the Zero?
I want to dig into Python embedded programming. With a big help of LLM, oc.
Sometimes, I deliver a device powered with a Raspberry Pi And I desperalely need to know how to test it, etc.
You might recall this TMF8806 EASY demo board.
I decided to dust it off and build a device around it.
TMF8806 facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
• Supply voltage 2.7 to 3.5V
I decided to dust it off and build a device around it.
TMF8806 facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
• Supply voltage 2.7 to 3.5V
February 9, 2026 at 6:06 AM
You might recall this TMF8806 EASY demo board.
I decided to dust it off and build a device around it.
TMF8806 facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
• Supply voltage 2.7 to 3.5V
I decided to dust it off and build a device around it.
TMF8806 facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
• Supply voltage 2.7 to 3.5V
I felt tired for a few hours last week. That's not a joke.
I dug out my old Beaglebone Black after years.
It was a big disappointment. The same as 6 years ago. I don't know why but the disaster started with the messy documentation.
No. 15 minutes and it travelled into the drawer again.
I dug out my old Beaglebone Black after years.
It was a big disappointment. The same as 6 years ago. I don't know why but the disaster started with the messy documentation.
No. 15 minutes and it travelled into the drawer again.
February 6, 2026 at 6:06 AM
I felt tired for a few hours last week. That's not a joke.
I dug out my old Beaglebone Black after years.
It was a big disappointment. The same as 6 years ago. I don't know why but the disaster started with the messy documentation.
No. 15 minutes and it travelled into the drawer again.
I dug out my old Beaglebone Black after years.
It was a big disappointment. The same as 6 years ago. I don't know why but the disaster started with the messy documentation.
No. 15 minutes and it travelled into the drawer again.
You might recall this TMF8806 EASY demo board.
I decided to dust it off and build a device around it
If you are not familiar with the ToF sensor TMF8806, here are some facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
I decided to dust it off and build a device around it
If you are not familiar with the ToF sensor TMF8806, here are some facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
February 5, 2026 at 6:06 AM
You might recall this TMF8806 EASY demo board.
I decided to dust it off and build a device around it
If you are not familiar with the ToF sensor TMF8806, here are some facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
I decided to dust it off and build a device around it
If you are not familiar with the ToF sensor TMF8806, here are some facts:
• Made by ams OSRAM
• 1 – 500cm distance sensing
• 1 – 1000cm after the new FW download
• OLGA 2.2 x 3.6 x 1.0 mm package
I finished testing on the new INA2227 Hardware Design Template board. It can't work better.
I'm impressed with
• how the IC is small and that I managed to reflow it flawlessly
• how well is the datasheet written
• how good did ChatGPT do when writing a simple Arduino driver for me
I'm impressed with
• how the IC is small and that I managed to reflow it flawlessly
• how well is the datasheet written
• how good did ChatGPT do when writing a simple Arduino driver for me
February 4, 2026 at 6:06 AM
I finished testing on the new INA2227 Hardware Design Template board. It can't work better.
I'm impressed with
• how the IC is small and that I managed to reflow it flawlessly
• how well is the datasheet written
• how good did ChatGPT do when writing a simple Arduino driver for me
I'm impressed with
• how the IC is small and that I managed to reflow it flawlessly
• how well is the datasheet written
• how good did ChatGPT do when writing a simple Arduino driver for me
When I studied a technical article by Texas Instruments called How the Dynamic Z-Track™ algorithm predicts unpredictable battery loads, I started seeking the smallest model.
In the ideal world, even the serial FETs would be integrated. I'm absolutely sure there are such ICs on the market.
In the ideal world, even the serial FETs would be integrated. I'm absolutely sure there are such ICs on the market.
February 3, 2026 at 6:06 AM
When I studied a technical article by Texas Instruments called How the Dynamic Z-Track™ algorithm predicts unpredictable battery loads, I started seeking the smallest model.
In the ideal world, even the serial FETs would be integrated. I'm absolutely sure there are such ICs on the market.
In the ideal world, even the serial FETs would be integrated. I'm absolutely sure there are such ICs on the market.
Start training your engineers.
It's not about the EDA tool.
Recently, I participated in a discussion about Xpedition vs the rest of the world. A few days (weeks, months...) ago, it was about Altium Designer vs the rest of the world.
It's running non-stop.
Choose wisely.
It's not about the EDA tool.
Recently, I participated in a discussion about Xpedition vs the rest of the world. A few days (weeks, months...) ago, it was about Altium Designer vs the rest of the world.
It's running non-stop.
Choose wisely.
February 2, 2026 at 6:06 AM
Start training your engineers.
It's not about the EDA tool.
Recently, I participated in a discussion about Xpedition vs the rest of the world. A few days (weeks, months...) ago, it was about Altium Designer vs the rest of the world.
It's running non-stop.
Choose wisely.
It's not about the EDA tool.
Recently, I participated in a discussion about Xpedition vs the rest of the world. A few days (weeks, months...) ago, it was about Altium Designer vs the rest of the world.
It's running non-stop.
Choose wisely.
I write about devices I love.
Help companies worldwide with hardware design.
From concept to manufacturing.
If you don't have a hardware design team, I am here for you.
Let's start slowly. I'll guide you through the entire process.
Step by step.
We'll grow together.
I'm one of yours.
Help companies worldwide with hardware design.
From concept to manufacturing.
If you don't have a hardware design team, I am here for you.
Let's start slowly. I'll guide you through the entire process.
Step by step.
We'll grow together.
I'm one of yours.
January 30, 2026 at 6:06 AM
I write about devices I love.
Help companies worldwide with hardware design.
From concept to manufacturing.
If you don't have a hardware design team, I am here for you.
Let's start slowly. I'll guide you through the entire process.
Step by step.
We'll grow together.
I'm one of yours.
Help companies worldwide with hardware design.
From concept to manufacturing.
If you don't have a hardware design team, I am here for you.
Let's start slowly. I'll guide you through the entire process.
Step by step.
We'll grow together.
I'm one of yours.
Early morning means a fresh mind. Today, I need only three things.
Focus, my AI companion and a new EASY device.
To keep myself focused, I sometimes use a simple 45-minute timer. I recommend trying it. When the timer goes off, take a short break or continue.
My AI companion today is ChatGPT.
Focus, my AI companion and a new EASY device.
To keep myself focused, I sometimes use a simple 45-minute timer. I recommend trying it. When the timer goes off, take a short break or continue.
My AI companion today is ChatGPT.
January 29, 2026 at 6:06 AM
Early morning means a fresh mind. Today, I need only three things.
Focus, my AI companion and a new EASY device.
To keep myself focused, I sometimes use a simple 45-minute timer. I recommend trying it. When the timer goes off, take a short break or continue.
My AI companion today is ChatGPT.
Focus, my AI companion and a new EASY device.
To keep myself focused, I sometimes use a simple 45-minute timer. I recommend trying it. When the timer goes off, take a short break or continue.
My AI companion today is ChatGPT.
When I tested it, it showed these flaws:
1. I should have used the transistor switch for the optocoupler input. The input impedance of the switch is too low for some generators.
2. I'm still unsure if the usage of the Schmitt stage buffer was correct or not.
1. I should have used the transistor switch for the optocoupler input. The input impedance of the switch is too low for some generators.
2. I'm still unsure if the usage of the Schmitt stage buffer was correct or not.
January 27, 2026 at 6:06 AM
When I tested it, it showed these flaws:
1. I should have used the transistor switch for the optocoupler input. The input impedance of the switch is too low for some generators.
2. I'm still unsure if the usage of the Schmitt stage buffer was correct or not.
1. I should have used the transistor switch for the optocoupler input. The input impedance of the switch is too low for some generators.
2. I'm still unsure if the usage of the Schmitt stage buffer was correct or not.
1 The tin ball, that is sitting between the R4 resistor Can I avoid it?
2 IC's pads aren't wetted completely. Is it an issue?
3 IC's pads are wetted but the shape isn't nice
4 The optocoupler's pads are OK
5 The big THT hole with ENIG looks beautiful when zoomed in
6 Diodes didn't move when reflown
2 IC's pads aren't wetted completely. Is it an issue?
3 IC's pads are wetted but the shape isn't nice
4 The optocoupler's pads are OK
5 The big THT hole with ENIG looks beautiful when zoomed in
6 Diodes didn't move when reflown
January 26, 2026 at 6:06 AM
1 The tin ball, that is sitting between the R4 resistor Can I avoid it?
2 IC's pads aren't wetted completely. Is it an issue?
3 IC's pads are wetted but the shape isn't nice
4 The optocoupler's pads are OK
5 The big THT hole with ENIG looks beautiful when zoomed in
6 Diodes didn't move when reflown
2 IC's pads aren't wetted completely. Is it an issue?
3 IC's pads are wetted but the shape isn't nice
4 The optocoupler's pads are OK
5 The big THT hole with ENIG looks beautiful when zoomed in
6 Diodes didn't move when reflown
Every ±20 dB, there is the x10 or ÷0.1 ratio
Every ±10 dB, there is the ×3.162 or ÷3.162
I learned this the hard way years ago, late at night, double-checking a gain budget that “felt wrong.”
The numbers didn’t lie. They just repeat the same logic again and again.
Every ±10 dB, there is the ×3.162 or ÷3.162
I learned this the hard way years ago, late at night, double-checking a gain budget that “felt wrong.”
The numbers didn’t lie. They just repeat the same logic again and again.
January 23, 2026 at 6:06 AM
Every ±20 dB, there is the x10 or ÷0.1 ratio
Every ±10 dB, there is the ×3.162 or ÷3.162
I learned this the hard way years ago, late at night, double-checking a gain budget that “felt wrong.”
The numbers didn’t lie. They just repeat the same logic again and again.
Every ±10 dB, there is the ×3.162 or ÷3.162
I learned this the hard way years ago, late at night, double-checking a gain budget that “felt wrong.”
The numbers didn’t lie. They just repeat the same logic again and again.
The MPS MP4056 documentation says mostly about the non-isolated solutions. I want it to be isolated.
I found a picture in the tail of the datasheet describing the Isolated Flyback Converter with Vo =24 V, Io=420 mA.
That's almost exactly what we need when targeting the output power of about 15 W.
I found a picture in the tail of the datasheet describing the Isolated Flyback Converter with Vo =24 V, Io=420 mA.
That's almost exactly what we need when targeting the output power of about 15 W.
January 22, 2026 at 6:06 AM
The MPS MP4056 documentation says mostly about the non-isolated solutions. I want it to be isolated.
I found a picture in the tail of the datasheet describing the Isolated Flyback Converter with Vo =24 V, Io=420 mA.
That's almost exactly what we need when targeting the output power of about 15 W.
I found a picture in the tail of the datasheet describing the Isolated Flyback Converter with Vo =24 V, Io=420 mA.
That's almost exactly what we need when targeting the output power of about 15 W.
The redesign took 5 times more than I expected.
I still don't know why. You're watching. Tell me.
The Active Scope Probe redesign project was an online thing.
Five KiCad Design Streams were needed before it's finished.
To be honest, in reality, it was 6 equivalent Design Streams.
I still don't know why. You're watching. Tell me.
The Active Scope Probe redesign project was an online thing.
Five KiCad Design Streams were needed before it's finished.
To be honest, in reality, it was 6 equivalent Design Streams.
January 21, 2026 at 6:06 AM
The redesign took 5 times more than I expected.
I still don't know why. You're watching. Tell me.
The Active Scope Probe redesign project was an online thing.
Five KiCad Design Streams were needed before it's finished.
To be honest, in reality, it was 6 equivalent Design Streams.
I still don't know why. You're watching. Tell me.
The Active Scope Probe redesign project was an online thing.
Five KiCad Design Streams were needed before it's finished.
To be honest, in reality, it was 6 equivalent Design Streams.
Integrated battery gas gauge and protector.
BQ27Z758 by Texas Instruments
Engineered for single‑cell packs. BQ27Z758 combines:
• a flash‑programmable RISC CPU,
8 16‑bit ADCs for simultaneous current/voltage sensing (down to 1 mΩ shunt), and
• SHA‑256 authentication.
BQ27Z758 by Texas Instruments
Engineered for single‑cell packs. BQ27Z758 combines:
• a flash‑programmable RISC CPU,
8 16‑bit ADCs for simultaneous current/voltage sensing (down to 1 mΩ shunt), and
• SHA‑256 authentication.
January 20, 2026 at 6:06 AM
Integrated battery gas gauge and protector.
BQ27Z758 by Texas Instruments
Engineered for single‑cell packs. BQ27Z758 combines:
• a flash‑programmable RISC CPU,
8 16‑bit ADCs for simultaneous current/voltage sensing (down to 1 mΩ shunt), and
• SHA‑256 authentication.
BQ27Z758 by Texas Instruments
Engineered for single‑cell packs. BQ27Z758 combines:
• a flash‑programmable RISC CPU,
8 16‑bit ADCs for simultaneous current/voltage sensing (down to 1 mΩ shunt), and
• SHA‑256 authentication.
Yes, it does. Let's calculate together.
7.5-digit Ω ≈ 10.000.000. log2(10^7) ≈ 24.
To be precise, it's 23.25 bits of effective resolution. At least a 24-bit signal chain is needed.
That's a lot.
I've never designed 24-bit signal chain devices.
7.5-digit Ω ≈ 10.000.000. log2(10^7) ≈ 24.
To be precise, it's 23.25 bits of effective resolution. At least a 24-bit signal chain is needed.
That's a lot.
I've never designed 24-bit signal chain devices.
January 19, 2026 at 6:06 AM
Yes, it does. Let's calculate together.
7.5-digit Ω ≈ 10.000.000. log2(10^7) ≈ 24.
To be precise, it's 23.25 bits of effective resolution. At least a 24-bit signal chain is needed.
That's a lot.
I've never designed 24-bit signal chain devices.
7.5-digit Ω ≈ 10.000.000. log2(10^7) ≈ 24.
To be precise, it's 23.25 bits of effective resolution. At least a 24-bit signal chain is needed.
That's a lot.
I've never designed 24-bit signal chain devices.
LED Shunt is a real device. Why didn't I know about it?
Do your LEDs ruin your strings? Read on.
Everywhere where preventive Maintenance is impractical or impossible.
Put the LED Shunt in parallel to each LED.
Sorted.
Done.
The NUD4700 is an electronic shunt which provides a current bypass.
Do your LEDs ruin your strings? Read on.
Everywhere where preventive Maintenance is impractical or impossible.
Put the LED Shunt in parallel to each LED.
Sorted.
Done.
The NUD4700 is an electronic shunt which provides a current bypass.
January 16, 2026 at 6:06 AM
LED Shunt is a real device. Why didn't I know about it?
Do your LEDs ruin your strings? Read on.
Everywhere where preventive Maintenance is impractical or impossible.
Put the LED Shunt in parallel to each LED.
Sorted.
Done.
The NUD4700 is an electronic shunt which provides a current bypass.
Do your LEDs ruin your strings? Read on.
Everywhere where preventive Maintenance is impractical or impossible.
Put the LED Shunt in parallel to each LED.
Sorted.
Done.
The NUD4700 is an electronic shunt which provides a current bypass.
PLNs inform us about changes.
End-of-life and NRND notifications should be tracked. IMHO.
• Edit the symbol description.
• Add a textual field to the symbol.
• Edit/Add a Status attribute.
The 3rd point is the best. But there is a catch. Check the image.
End-of-life and NRND notifications should be tracked. IMHO.
• Edit the symbol description.
• Add a textual field to the symbol.
• Edit/Add a Status attribute.
The 3rd point is the best. But there is a catch. Check the image.
January 15, 2026 at 6:06 AM
PLNs inform us about changes.
End-of-life and NRND notifications should be tracked. IMHO.
• Edit the symbol description.
• Add a textual field to the symbol.
• Edit/Add a Status attribute.
The 3rd point is the best. But there is a catch. Check the image.
End-of-life and NRND notifications should be tracked. IMHO.
• Edit the symbol description.
• Add a textual field to the symbol.
• Edit/Add a Status attribute.
The 3rd point is the best. But there is a catch. Check the image.
When I studied a technical article by TI called How the Dynamic Z-Track™ algorithm predicts unpredictable battery loads, I started seeking the smallest model.
I usually don't need multiple-cell solutions because my customers want small, usually wearable products using just a single cell battery.
I usually don't need multiple-cell solutions because my customers want small, usually wearable products using just a single cell battery.
January 14, 2026 at 6:06 AM
When I studied a technical article by TI called How the Dynamic Z-Track™ algorithm predicts unpredictable battery loads, I started seeking the smallest model.
I usually don't need multiple-cell solutions because my customers want small, usually wearable products using just a single cell battery.
I usually don't need multiple-cell solutions because my customers want small, usually wearable products using just a single cell battery.
.new thing is real. A quick and fun way to create new docs and files.
I accidentally found the .new thing in the release notes of my Dia browser.
Then, I asked myself. If Google have it, many services would do.
And yes, there are many.
Can you extend the list in the comments?
I accidentally found the .new thing in the release notes of my Dia browser.
Then, I asked myself. If Google have it, many services would do.
And yes, there are many.
Can you extend the list in the comments?
January 13, 2026 at 6:06 AM
.new thing is real. A quick and fun way to create new docs and files.
I accidentally found the .new thing in the release notes of my Dia browser.
Then, I asked myself. If Google have it, many services would do.
And yes, there are many.
Can you extend the list in the comments?
I accidentally found the .new thing in the release notes of my Dia browser.
Then, I asked myself. If Google have it, many services would do.
And yes, there are many.
Can you extend the list in the comments?
The idea was to ask you what your first KiCad version was.
Whether 6 or 7.
When I prepared this post, I visited archive.org and looked for kicad.org.
I got two screenshots, from 6.0.6 and 6.0.7.
So, the question had been killed even before I tried to.
Whether 6 or 7.
When I prepared this post, I visited archive.org and looked for kicad.org.
I got two screenshots, from 6.0.6 and 6.0.7.
So, the question had been killed even before I tried to.
January 12, 2026 at 6:06 AM
The idea was to ask you what your first KiCad version was.
Whether 6 or 7.
When I prepared this post, I visited archive.org and looked for kicad.org.
I got two screenshots, from 6.0.6 and 6.0.7.
So, the question had been killed even before I tried to.
Whether 6 or 7.
When I prepared this post, I visited archive.org and looked for kicad.org.
I got two screenshots, from 6.0.6 and 6.0.7.
So, the question had been killed even before I tried to.