Lol. These videos are great bedtime stories for Devs too. Literally watch Andreas Spies great content with the intent to fall asleep, helps that he releases ~12am in my time zone.
I'd love to see a video on pi filters and ferrite beads. I see them in every one of your newer design. I've read application notes and articles on them, but figuring out the right impedance value to use is very tricky.
That's definitely on my list of videos to make. You're right that there's a lot one needs to take into consideration when choosing an appropriate value. I hope a video (in the near future) will clear that up!
I simulated a pi filter with ferrite like that (using full manufacturer spice models) and saw that it definitely can resonate, so there is a specific frequency that is actually amplified rather than filtered. I don't know if that is only an issue if there is stimulus at that frequency range though? Or perhaps the source impedance combined with the first cap means it's actually not providing gain? Or maybe this is the reason Phil had to add that big RC filter, since the pi filter wasn't doing the filtering that was thought?
Very nicely done and big thanks for walking us through your choices. I’ve been wanting to build this very same product for my own use. As you have paid such close attention to the headphone amp, I would just purchase the PCB from you. Thanks again.
If anyone is curious what the audio sounds like my audiophile friend and whisky enthusiast described it as "full-bodied and intimidatingly full on the nose, shrouded with the scent of lapsang souchong, tobacco, iodine, vanilla and sherry spice"
Thanks Phil! This thing has kept me busy for a few months already.. Finally made a working version myself! Although a little unstable and noisy.. Probably due to poor component selection (already found out I should use thin film resistors and S0G caps in the signal path, I missed that the first 100 times I watched this video) After sorting those out I'm planning to make a version with a Baxandall tone control adding to it. Cheers!
5532 makes and incredible power amp, 16W only, but its incredibly quiet and precise, designed and built one, took Selfs design as a starting point. 5532 is an amazing chip to this day.
Thanks for this video Phil. I've been looking at your videos in this series for any tips on how to do the opposite: drive USB headphones from stereo analogue audio, haven't found anything yet that doesn't use some complex microcontroller with full USB host.
I read that book too. I continued trying to design a 4 stereo channel mixer with EQ. I gave up when I got to about 24 5532s and about 200 components. I could layout one channel, but the PCB design software gave me no way to duplicate them on the PCB. They coudn't even be copy pasted. For drive power I went for the OPA551s which gives 70V 200mA+ to place with per channel.
I thought about designing a mixer at some point as well. It would make a great project, but there's so much detail to be sorted out.. Maybe when I'm retired :D
where can i go learn to catch up to the "basics" mentioned in this video? Stuff like LDOs, pi filter, bi-directional filter, differential pair and signal routing, DNPs, floating shield, ferric bead?
Hi Phil! I am currently taking your mixed signal design course and I love it. It is one of the things I look forward to, making progress and taking notes each day. I am also excited for your next course and was wondering when we can expect it?
Thank you very much, Andrew - I'm glad you're liking the course. I'm intensely working on the next course, which packs a lot of content. Hopefully within the next few months, sorry for the long wait!
I am a software engineer who wanted to make a small PCB, and did, but now I'm obsessively watching all your videos 0_0, my brain is sore but now I have a new obsession, I made a Filament drier, next maybe I'll make my musician friend an AMP? Tho it might suck!
Great video Phil. Quick question about the bias generation: does this circuit sacrifice power supply rejection? Since the bias voltage comes from a voltage divider instead of a reference, it seems like any PSU ripple will end up in your signal chain. Some of this may be cancelled by the concatenation of inverting and non-inverting stages, though group delay will eat away at this.
Cool. Building some like this has been on my backlog for ages. I still have some PGA2311 for volume control somewhere because apparently I like to make things complicated for myself :)
Hi Phil, thanks for the great video as always! Quick question - do you have any videos that give a bit more detail regarding how the 30pF capacitors provide EMI protection? Are they just forming very high frequency low pass RC filters with the series resistances?
Hi Louis, Yes, that's basically it! Even if they're not populated, I like to place series + parallel element at every I/O in case they need to be fitted for compliance reasons.
Regarding using Class I (C0G) capacitors instead of Class II (X5R, X7R, etc.), the reason for distortion in Class II MLCCs is largely the hysteresis introduced by the spontaneous polarisation reversal process in the magnetic domains of the barium titanate dielectric. Class I dielectrics are not spontaneously polarised, so when no electric field is applied the crystal lattice does not have any polarisation. Class II dielectrics, on the other hand, form spontaneously polarised magnetic domains of random orientation when there is no electric field applied, and the application of an electric field causes reversal of that spontaneous polarisation to align the domains with the field. This leads to hysteresis and effects like domain wall heating, hence the distortion.
11:06 I believe the TPS61040DBVR boost converter is configured wrong. When having a look at the datasheet, the formula for maximum load current is given. With this configuration of L1, R4, R5, D3 and an assumed efficiency of 80% (reasonable mentioned in the datasheet) the maximum load current is only 56.9mA. When having a look at the NE5532 datasheet the total supply current without the drive current is 8 to 16mA. In the design we have 9 NE5532. So just in idle without accounting for the load currents and taking the best case supply current, the devices are drawing 72mA from the 15V supply. Hence, I believe the TPS boost converter is not sufficient for this design. Please correct me if I am wrong. But I believe a fix would be to increase the Inductor of the boost converter to 15uH.
I have tested the design at home and the TPS provides a voltage of only 14.56V instead of the expected 15.8V. This voltage also aligns with the maximum switching frequency of 1MHz as stated in the datasheet. @PhilsLab
As far as my research went, another alternative would be to use the LM27313 as a drop in replacement for the TPS. Pins are the same and capabilities are similar. Furthermore, not a single component surrounding the TPS must be changed.
At around 22mins in what's a "Pot an shunter" My favourite op-amp for headphones was the good old LM759 easy to use and enough power for most uses. Just goes to show how old i am.
Hi Phil, really enjoyed this walk through and even though I've completed your mixed signal design course, I've made some extra notes based on this video as well regarding the signal chain. I'm interested to know though what exact parts you selected for some parts of this circuit, specifically around the power supplies and filtering. Is there a BOM available that would match the schematic you presented in this video? (I couldn't see anything on github.) My first board (since taking your course) is taking 5V from a Raspberry Pi header (so noisey) and boosting and filtering it, much like this board does. I've selected different components and its about to arrive from PCBWay this coming week, so I'm excited to do some noise tests once it arrives, but I can see there are other options you've selected here I'd be interested to know about. Thanks again.
Hi Phil, Your content is great and has helped me in my day to day work in my career. I was wondering if you're ever planning on looking at Bluetooth audio, specifically 2 seperate speakers with one common Bluetooth connection. Airpods would be an example of something with this connection, but something bigger for speakers. Much appreciated and keep up the great work! Adam.
Thanks for the video. Out of curiosity, have you finally decided what video will you do in a live video? You suggested some videos ago and I am eager for the day that launches. Seeing a profesional Hardware designer doing the layout of a PCB is such a treasure for a beginner.
Thanks for watching, Alejandro! I'm planning that for the (hopefully) 100k sub video, but it won't be "live" but rather full-length following the design steps (~3hrs). It'll be using KiCad 7 but I haven't quite locked in what I want to make yet..
@@PhilsLab Hi Phil's, it seems good anyway. I have gain a lot of expertise and tips just by watching your videos, specially on those videos where you make a PCB Layout review. 🧐
@@PhilsLab That's what I thought. For hobby, the quality of your boards are amazing but what amazes me more is the frequency with which you make these PCBs.
This is a pretty nice design, I'm curious as to how the analog side stacks up with something like the Neve headphone amp. Great video as always, keep it up Phil!
@@PhilsLab Honestly it probably performs very similarly. The THX AAA-888 (which is beloved by many in the audiophile community, at least from my experience) is just a power opamp with some standard auxiliary circuitry.
The Neve is based around a TPA6120 amplifier, which was originally an ADSL line driver. It's a great amplifier though, if somewhat cranky stability-wise. There are a number of DIY TPA6120 designs out there that would allow you to achieve Neve levels of performance without the silly price.
Are the ground planes different between each layer at all or do all the vias simply cut through all the layers? I thought there's some benefit to having separate analog and digital ground, but it looks like you're only separating power signals via a series resistor.
I'm aware that this is now a past project and your store is a low priority at the moment, but it would be amazing to see another round of assembled NE-XT USB boards. I've been running mine for months and it's been flawless, superior to everything else I've got both in sound quality and reliability. Heck I'd even get a few manufactured myself but I'm not sure if you intended to release the repository for it.
Thank you very much for your kind comment - I'm very glad to hear that you're liking the amp! I actually have a couple assembled boards left. Feel free to contact me via phil@phils-lab.net, should you be interested in them - thanks!
Next time try the capacitance multiplier after a boost converter. With a single transistor you will get a real bang for the buck. Capacitance multipliers are very good at clearing noise in the audio spectrum. Also you can use any generic buck converter in an inverting buck-boost configuration to get a negative rail. With +/- rails life is much easier in an audio realm, especially if you plan to use op-amp based analog filters. Xonar Essence sound card used 90 cents buck converter designated in a datasheet for a LED flashlight to get a negative rail. 😂 And they achive -114 db of SNR.
Very cool, thank you, Phil! About boosting 5V - not the first project, when I'm a bit sad, that PC (not all?) can't give you more than 5V via USB PD protocol. With all this pi-filters and very nice design, it's enough to be called "audiophile" :) BB's are very cool - I've worked with PCM1792a.
Usb-pd can deliver up to 20 volt for 100 watt total iirc no? This is of course optional. Of course this is for usb 3.2 or 4.0. (Maybe 3.1 too but that has up to 20 watts total I think, would have to look that up.)
Hi Phil, Big fan of your channel... Just wondering... since it's a 4 layer board and you have dedicated 2 of the plane layers to GND.... wouldn't it be been better to have a split on one of the plane layers, and have power delivered via plane layer or better dedicate a complete layer for power delivery and split for different voltage rails ?
Check out the video on my channel on stackups and buildups for a detailed explanation why I choose this stackup (e.g. return paths, field control, etc.). Power planes are not needed for this type of design and only become relevant for high-speed digital.
Kickass! If my time weren't taken up with full-time engineering (mechanical) it'd be really interesting to dive into a serious project like this. Any restrictions on international shipping (Australia here)? It's not like a US-only thing?
Thanks! I ship pretty much anywhere in the world (incl. Australia). Just have it set-up as a flat-rate 15 USD (as this covers pretty much all cases, although sometimes it does cost more).
what program did you use for simulating those paralleled op amps ? Asking since i am not sure how to pick that balanceing resistor that is 1 ohm in your case i want to make a charge pump that gives a few mA of negative voltage and i am thinking of useing 2-3 op amps for buffering my clock signal .
Thank you for your awesome detailed videos, I have learned a lot. Please make a playlist of videos for firmware development for different microcnotroller. Specially I want on chips by WCH, they are very cheap.
@@PhilsLab thanks for reply! So what should i do if i want just pad on my pcb? Do i send them a note along side gerber files? Can i still use your pad pcb file?
Brilliant Phil, I just ordered my first 4-layer board today and I needed an ADC for the next project so --- that's a bit handy! :) Wish I could justify Altium. I'm stuck with KiCad (which is great for what it costs) or TINA which I haven't gotten to grips with for PCB design - and I'd better be quick as demo will expire!
@@PhilsLab Thanks Phil, I can't tell you how much I've learned from you, Eric Bogatan and Rick Hartley over the last year or two. I've been out of the game for decades and getting back in there is like visiting a whole new planet with SMT being all the rage... it wasn't even "out' when I last designed a board.
Thank you very much, Marc - I'm glad to hear that the content has been helpful. Yes, certainly a lot of things have changed throughout the last few decades. For me the coolest thing is how much the price of manufacturer and assembly has come down in the last few years!
@@PhilsLab It would be easier if KiCAD (in my case) had better footprint libraries and could send the BOM direct to PCBWay etc. but it's free so I shouldn't really complain. It's a shame Altium don't do a version for us retired hicks. :)
Would you like to make a video on software as you did for the filters and other argument?! I would like to propuse you the OTA, Over The Air firmware update. Would be very interesting, both for the ESP32, or some wireless capable STM32 MCUs.
I am curious what's your take on using capacitance multiplier circuits (eevblog #1116) instead of using pi filters and ldo. I also did a karaoke board and I ended up using them for filtering the noise from the switching regulators.
Loss of headroom is the issue with cap multipliers (on a 5V rail that's about 15%). Purely passive filters can nail HF noise coming off the USB rails (which are notoriously noisy) whereas "multipliers" seem more attuned to L/F ripple.
@@PhilsLab Another suggestion that I had in mind was for PCM2704 DAC chip. Looks like it has better THD+N performance. Too bad it is not drop-in replacement as pinout is different since it's only a DAC.🤷♂
@@pavlobilous4030 PCM2706 + e.g. PCM5102A is a proven combo. But, I'd rather see Phil design around CS4398 with differential outputs. Cirrus Logic made a great DAC right there.
Hi Phil, have you ever looked into how soundcards work when it comes to sending audio data from analogue inputs towards the computer and at the same time being able to receive them? Are there specific ICs that are able to do this? edit: btw, I really enjoy these audio related videos and thank you for mentioning Dougles' book, I was looking for such a book!
@@austayo okay, so I am not so sure you were talking about the Vin pins of the PCM2900 but it seems like it can do exactly what I was asking about :) Either way, thanks for the comment! :D
This is pretty awesome, I'd love to see a design walk through for a microphone amp + USB converter, basically going in the opposite direction for those small few of us looking at the custom microphone space (inspired by this video from DIY perks: ruclips.net/video/LoQu3XXIayc/видео.html&ab_channel=DIYPerks) where the circuitry left a lot to be desired.
![NLE Choppa - Or What ft. 41 (Orchestra Mix) [Official Music Video]](http://i.ytimg.com/vi/biOFASEgSOM/mqdefault.jpg)
These videos are analogous bedtime stories for EEs.
Lol. These videos are great bedtime stories for Devs too. Literally watch Andreas Spies great content with the intent to fall asleep, helps that he releases ~12am in my time zone.
I'd love to see a video on pi filters and ferrite beads. I see them in every one of your newer design. I've read application notes and articles on them, but figuring out the right impedance value to use is very tricky.
Yes indeed a video for this topic would be awesome
That's definitely on my list of videos to make. You're right that there's a lot one needs to take into consideration when choosing an appropriate value. I hope a video (in the near future) will clear that up!
I simulated a pi filter with ferrite like that (using full manufacturer spice models) and saw that it definitely can resonate, so there is a specific frequency that is actually amplified rather than filtered. I don't know if that is only an issue if there is stimulus at that frequency range though? Or perhaps the source impedance combined with the first cap means it's actually not providing gain? Or maybe this is the reason Phil had to add that big RC filter, since the pi filter wasn't doing the filtering that was thought?
Don't need this device. Yet I need to watch Phil's video guaranteed to be sprinkled with golden nuggets of info. Cheers!
Thanks! Hope there are some of those in this video :)
Best electronics channel on RUclips
Thank you!
Very nicely done and big thanks for walking us through your choices. I’ve been wanting to build this very same product for my own use. As you have paid such close attention to the headphone amp, I would just purchase the PCB from you. Thanks again.
Would love it if you release the layout routing as a class! Definitely would sign up for it!
Quackphones
I wanted to do the same EXACT project when I finish my grad proj.... thank you so much, phil, for the great videos and knowledge!
Very glad to hear this will be useful! Good luck with your grad project :)
If anyone is curious what the audio sounds like my audiophile friend and whisky enthusiast described it as "full-bodied and intimidatingly full on the nose, shrouded with the scent of lapsang souchong, tobacco, iodine, vanilla and sherry spice"
+1 It's all about that woody herbaceousness
Thanks Phil! This thing has kept me busy for a few months already.. Finally made a working version myself! Although a little unstable and noisy.. Probably due to poor component selection (already found out I should use thin film resistors and S0G caps in the signal path, I missed that the first 100 times I watched this video) After sorting those out I'm planning to make a version with a Baxandall tone control adding to it.
Cheers!
Awesome, glad you got your design working, Benjamin! Good luck with your revisions!
5532 makes and incredible power amp, 16W only, but its incredibly quiet and precise, designed and built one, took Selfs design as a starting point.
5532 is an amazing chip to this day.
Thanks for this video Phil. I've been looking at your videos in this series for any tips on how to do the opposite: drive USB headphones from stereo analogue audio, haven't found anything yet that doesn't use some complex microcontroller with full USB host.
I read that book too. I continued trying to design a 4 stereo channel mixer with EQ. I gave up when I got to about 24 5532s and about 200 components. I could layout one channel, but the PCB design software gave me no way to duplicate them on the PCB. They coudn't even be copy pasted.
For drive power I went for the OPA551s which gives 70V 200mA+ to place with per channel.
I thought about designing a mixer at some point as well. It would make a great project, but there's so much detail to be sorted out.. Maybe when I'm retired :D
where can i go learn to catch up to the "basics" mentioned in this video? Stuff like LDOs, pi filter, bi-directional filter, differential pair and signal routing, DNPs, floating shield, ferric bead?
Hi Phil! I am currently taking your mixed signal design course and I love it. It is one of the things I look forward to, making progress and taking notes each day. I am also excited for your next course and was wondering when we can expect it?
Thank you very much, Andrew - I'm glad you're liking the course. I'm intensely working on the next course, which packs a lot of content. Hopefully within the next few months, sorry for the long wait!
Every electrical engineer in a nutshell. Why buy a solution when I can create a better solution.
I am a software engineer who wanted to make a small PCB, and did, but now I'm obsessively watching all your videos 0_0, my brain is sore but now I have a new obsession, I made a Filament drier, next maybe I'll make my musician friend an AMP? Tho it might suck!
Why are you using bidirectional diodes for the USB data lines?
Send it to Amir/ASR and julian krause
I think one of these on my desk at work would radically improve my efficiency
Very nice. So it's basically an handmade USB audio card.
Why are the first two versions removed from your channel? As analog amplifiers they are still very useful for viewers
Great video Phil. Quick question about the bias generation: does this circuit sacrifice power supply rejection? Since the bias voltage comes from a voltage divider instead of a reference, it seems like any PSU ripple will end up in your signal chain. Some of this may be cancelled by the concatenation of inverting and non-inverting stages, though group delay will eat away at this.
Cool. Building some like this has been on my backlog for ages. I still have some PGA2311 for volume control somewhere because apparently I like to make things complicated for myself :)
Hope you get around to building your own soon then! :)
Hi Phil, thanks for the great video as always! Quick question - do you have any videos that give a bit more detail regarding how the 30pF capacitors provide EMI protection? Are they just forming very high frequency low pass RC filters with the series resistances?
Hi Louis, Yes, that's basically it! Even if they're not populated, I like to place series + parallel element at every I/O in case they need to be fitted for compliance reasons.
Regarding using Class I (C0G) capacitors instead of Class II (X5R, X7R, etc.), the reason for distortion in Class II MLCCs is largely the hysteresis introduced by the spontaneous polarisation reversal process in the magnetic domains of the barium titanate dielectric. Class I dielectrics are not spontaneously polarised, so when no electric field is applied the crystal lattice does not have any polarisation. Class II dielectrics, on the other hand, form spontaneously polarised magnetic domains of random orientation when there is no electric field applied, and the application of an electric field causes reversal of that spontaneous polarisation to align the domains with the field. This leads to hysteresis and effects like domain wall heating, hence the distortion.
I did not see the #24 lab referenced in the video. Did it get removed?
Hi, Phil! Is it even possible to compete with off-the-shelf products in terms of sound quality? Can DIY beat Chi-Fi budget devices? How do you think?
What determines the ideal output impedance? I have 250Ohm headphones for example.
11:06 I believe the TPS61040DBVR boost converter is configured wrong.
When having a look at the datasheet, the formula for maximum load current is given. With this configuration of L1, R4, R5, D3 and an assumed efficiency of 80% (reasonable mentioned in the datasheet) the maximum load current is only 56.9mA.
When having a look at the NE5532 datasheet the total supply current without the drive current is 8 to 16mA. In the design we have 9 NE5532. So just in idle without accounting for the load currents and taking the best case supply current, the devices are drawing 72mA from the 15V supply.
Hence, I believe the TPS boost converter is not sufficient for this design.
Please correct me if I am wrong.
But I believe a fix would be to increase the Inductor of the boost converter to 15uH.
I have tested the design at home and the TPS provides a voltage of only 14.56V instead of the expected 15.8V.
This voltage also aligns with the maximum switching frequency of 1MHz as stated in the datasheet.
@PhilsLab
As far as my research went, another alternative would be to use the LM27313 as a drop in replacement for the TPS. Pins are the same and capabilities are similar. Furthermore, not a single component surrounding the TPS must be changed.
can the baxandall volume control part implement before a bass treble board of a power amplifier? Will it be worth the extra work?
At around 22mins in what's a "Pot an shunter"
My favourite op-amp for headphones was the good old LM759 easy to use and enough power for most uses.
Just goes to show how old i am.
Hi Phil, really enjoyed this walk through and even though I've completed your mixed signal design course, I've made some extra notes based on this video as well regarding the signal chain. I'm interested to know though what exact parts you selected for some parts of this circuit, specifically around the power supplies and filtering. Is there a BOM available that would match the schematic you presented in this video? (I couldn't see anything on github.) My first board (since taking your course) is taking 5V from a Raspberry Pi header (so noisey) and boosting and filtering it, much like this board does. I've selected different components and its about to arrive from PCBWay this coming week, so I'm excited to do some noise tests once it arrives, but I can see there are other options you've selected here I'd be interested to know about. Thanks again.
Hi Phil,
Your content is great and has helped me in my day to day work in my career.
I was wondering if you're ever planning on looking at Bluetooth audio, specifically 2 seperate speakers with one common Bluetooth connection. Airpods would be an example of something with this connection, but something bigger for speakers.
Much appreciated and keep up the great work!
Adam.
Thanks for the video. Out of curiosity, have you finally decided what video will you do in a live video? You suggested some videos ago and I am eager for the day that launches. Seeing a profesional Hardware designer doing the layout of a PCB is such a treasure for a beginner.
Thanks for watching, Alejandro! I'm planning that for the (hopefully) 100k sub video, but it won't be "live" but rather full-length following the design steps (~3hrs).
It'll be using KiCad 7 but I haven't quite locked in what I want to make yet..
@@PhilsLab Hi Phil's, it seems good anyway. I have gain a lot of expertise and tips just by watching your videos, specially on those videos where you make a PCB Layout review. 🧐
Phil, how many of these PCBs that you've posted on RUclips are for some company? Or are they all your hobby projects ?
These are all personal projects.
Posting company projects would get me sued I'm afraid :D
@@PhilsLab That's what I thought. For hobby, the quality of your boards are amazing but what amazes me more is the frequency with which you make these PCBs.
This is a pretty nice design, I'm curious as to how the analog side stacks up with something like the Neve headphone amp. Great video as always, keep it up Phil!
Thanks, Zach! I'd be interested in that as well. I just find the price-point to be pretty ridiculous on some of those headphone amplifiers..
@@PhilsLab Honestly it probably performs very similarly. The THX AAA-888 (which is beloved by many in the audiophile community, at least from my experience) is just a power opamp with some standard auxiliary circuitry.
@@PhilsLab The hi-res/lossless audio scams are certainly a contributing factor.
The Neve is based around a TPA6120 amplifier, which was originally an ADSL line driver. It's a great amplifier though, if somewhat cranky stability-wise. There are a number of DIY TPA6120 designs out there that would allow you to achieve Neve levels of performance without the silly price.
Hi Phil, can you do a series on digital RF circuit designs... something like 2.4ghz qpsk transceiver
Good afternoon Sir. Could you explain in a video how to put components on both sides of the board in Kicad? I would appreciate it to much.
Are the ground planes different between each layer at all or do all the vias simply cut through all the layers? I thought there's some benefit to having separate analog and digital ground, but it looks like you're only separating power signals via a series resistor.
As stated in the video, one GND for all - for 99% of mixed-signal designs this is the way to go. The ground planes are stitched together with vias.
Awesome video as always one of the best in the field keep up the great work
Thank you very much!
I'm aware that this is now a past project and your store is a low priority at the moment, but it would be amazing to see another round of assembled NE-XT USB boards. I've been running mine for months and it's been flawless, superior to everything else I've got both in sound quality and reliability. Heck I'd even get a few manufactured myself but I'm not sure if you intended to release the repository for it.
Thank you very much for your kind comment - I'm very glad to hear that you're liking the amp! I actually have a couple assembled boards left. Feel free to contact me via phil@phils-lab.net, should you be interested in them - thanks!
Next time try the capacitance multiplier after a boost converter. With a single transistor you will get a real bang for the buck. Capacitance multipliers are very good at clearing noise in the audio spectrum.
Also you can use any generic buck converter in an inverting buck-boost configuration to get a negative rail. With +/- rails life is much easier in an audio realm, especially if you plan to use op-amp based analog filters. Xonar Essence sound card used 90 cents buck converter designated in a datasheet for a LED flashlight to get a negative rail. 😂 And they achive -114 db of SNR.
Also much MORE ceramic capacitors in parallel after a boost converter are very welcome in audio.
Very cool, thank you, Phil!
About boosting 5V - not the first project, when I'm a bit sad, that PC (not all?) can't give you more than 5V via USB PD protocol.
With all this pi-filters and very nice design, it's enough to be called "audiophile" :)
BB's are very cool - I've worked with PCM1792a.
Usb-pd can deliver up to 20 volt for 100 watt total iirc no? This is of course optional.
Of course this is for usb 3.2 or 4.0. (Maybe 3.1 too but that has up to 20 watts total I think, would have to look that up.)
would like to see something more similar to how Schiit or fiio amps/dacs, more inputs/outputs types
odered one :) I should start digging in boxes for convertor from 3.5 to 6.3mm
Thank you very much, Tomas! I'll ship it out to you tomorrow :)
Hi Phil,
Big fan of your channel...
Just wondering... since it's a 4 layer board and you have dedicated 2 of the plane layers to GND.... wouldn't it be been better to have a split on one of the plane layers, and have power delivered via plane layer or better dedicate a complete layer for power delivery and split for different voltage rails ?
Check out the video on my channel on stackups and buildups for a detailed explanation why I choose this stackup (e.g. return paths, field control, etc.).
Power planes are not needed for this type of design and only become relevant for high-speed digital.
Kickass! If my time weren't taken up with full-time engineering (mechanical) it'd be really interesting to dive into a serious project like this.
Any restrictions on international shipping (Australia here)? It's not like a US-only thing?
Thanks! I ship pretty much anywhere in the world (incl. Australia). Just have it set-up as a flat-rate 15 USD (as this covers pretty much all cases, although sometimes it does cost more).
@@PhilsLab Awesome, it'll be interesting to see how it compares to the Ifi Zen DAC I use currently for audio output.
Awesome looking board !...cheers.
Thank you, Andy!
what program did you use for simulating those paralleled op amps ? Asking since i am not sure how to pick that balanceing resistor that is 1 ohm in your case i want to make a charge pump that gives a few mA of negative voltage and i am thinking of useing 2-3 op amps for buffering my clock signal .
LTSpice usually. However, I took some values from one of the Douglas Self pieces.
Thank you for your awesome detailed videos, I have learned a lot.
Please make a playlist of videos for firmware development for different microcnotroller. Specially I want on chips by WCH, they are very cheap.
hello, great video!
i just want to know why did you wrote "DO NOT PLACE" in description of TC2030 pad in your github schematic/pcb library
Thanks! Believe it or not, some assembly houses will actually blindly order Tag-Connect cables, thinking they are to be assembled.
@@PhilsLab thanks for reply! So what should i do if i want just pad on my pcb? Do i send them a note along side gerber files? Can i still use your pad pcb file?
Brilliant Phil, I just ordered my first 4-layer board today and I needed an ADC for the next project so --- that's a bit handy! :)
Wish I could justify Altium. I'm stuck with KiCad (which is great for what it costs) or TINA which I haven't gotten to grips with for PCB design - and I'd better be quick as demo will expire!
Thank you, Marc! Good luck with your first 4-layer design :)
KiCad is great for sure! I haven't used it in quite some time however..
@@PhilsLab Thanks Phil, I can't tell you how much I've learned from you, Eric Bogatan and Rick Hartley over the last year or two. I've been out of the game for decades and getting back in there is like visiting a whole new planet with SMT being all the rage... it wasn't even "out' when I last designed a board.
Thank you very much, Marc - I'm glad to hear that the content has been helpful.
Yes, certainly a lot of things have changed throughout the last few decades. For me the coolest thing is how much the price of manufacturer and assembly has come down in the last few years!
@@PhilsLab It would be easier if KiCAD (in my case) had better footprint libraries and could send the BOM direct to PCBWay etc. but it's free so I shouldn't really complain.
It's a shame Altium don't do a version for us retired hicks. :)
Would you like to make a video on software as you did for the filters and other argument?! I would like to propuse you the OTA, Over The Air firmware update. Would be very interesting, both for the ESP32, or some wireless capable STM32 MCUs.
I am curious what's your take on using capacitance multiplier circuits (eevblog #1116) instead of using pi filters and ldo. I also did a karaoke board and I ended up using them for filtering the noise from the switching regulators.
Loss of headroom is the issue with cap multipliers (on a 5V rail that's about 15%). Purely passive filters can nail HF noise coming off the USB rails (which are notoriously noisy) whereas "multipliers" seem more attuned to L/F ripple.
Hi phil! Could you make a video using the ATSAMD21 microcontroller family
Hi Jorge, I haven't used those before, but did want to make some more videos featuring other MCUs!
Did you delete yesterday video?
CD quality is 44.1kHz 16 bit not 48kHz, 48kHz 16 bit is DAT quality (or audio for video)
Have you looked into using opa1688 as current drivers?
Not yet - but they look rather neat. Thanks for the suggestion!
@@PhilsLab Another suggestion that I had in mind was for PCM2704 DAC chip. Looks like it has better THD+N performance. Too bad it is not drop-in replacement as pinout is different since it's only a DAC.🤷♂
@@pavlobilous4030 PCM2706 + e.g. PCM5102A is a proven combo.
But, I'd rather see Phil design around CS4398 with differential outputs. Cirrus Logic made a great DAC right there.
Hi Phil, have you ever looked into how soundcards work when it comes to sending audio data from analogue inputs towards the computer and at the same time being able to receive them?
Are there specific ICs that are able to do this?
edit: btw, I really enjoy these audio related videos and thank you for mentioning Dougles' book, I was looking for such a book!
Are you sure your not talking about Line IN? Often the Cyan 3.5mm connector?
@@austayo okay, so I am not so sure you were talking about the Vin pins of the PCM2900 but it seems like it can do exactly what I was asking about :) Either way, thanks for the comment! :D
Thanks for the informative video❤
Thanks for watching, Milan!
25:52 "in any case.." 👌
Thank you
Thanks for watching, Tim!
This is pretty awesome, I'd love to see a design walk through for a microphone amp + USB converter, basically going in the opposite direction for those small few of us looking at the custom microphone space (inspired by this video from DIY perks: ruclips.net/video/LoQu3XXIayc/видео.html&ab_channel=DIYPerks) where the circuitry left a lot to be desired.
I wish Santa to bring you a box with power supply negative rail regulators )
yay
69, noice
I couldn't find Previous Video. 🥲