Introducing DSPi | A powerful, user friendly and open source DSP for less than a cup of coffee

[somebodyelse]

Yes, I told him that months ago - about that (good or bad) habit of maintaining parity between versions for different operating systems...

It's probably not reached peak entitlement yet, but complaining about something someone is offering you freely has to be close. Config support has been variable between Windows and MacOS - sometimes one is ahead of the other - but not consistently in either direction. The linux version hasn't kept pace with either if you want anything to grumble about - which you shouldn't unless you've been contributing code - and that was meant as a cross-platform version. iOS and Android users are out of luck entirely, but nobody else has stepped up to implement it for them. Are you volunteering to help keep parity between the config versions for different operating systems?

 

[john61ct]

it’s perfectly fair to accept that, until a certain milestone is reached, Troy can do whatever he likes!

That remains true no matter what milestone is reached, less so (only a little) if you are their employer or paying client

 

[Red_Red]

Perhaps I didn't express myself clearly, but that is obviously my opinion as well.

 

[TOMSON]

I managed to run PCM1808 in master mode with DSPi via I2S bus, and now if I wanted to connect 3 amplifiers on TAS5828M chip that work in slave mode via I2S bus, they will use the clock on the PCM1808 chip board ?

 

[Weeb Labs]

I managed to run PCM1808 in master mode with DSPi via I2S bus, and now if I wanted to connect 3 amplifiers on TAS5828M chip that work in slave mode via I2S bus, they will use the clock on the PCM1808 chip board ?

Yes, that's right.

 

[Enri]

I must confess I had a few whiskeys, but I'm really enjoying DSPi right now with Roon. Please contribute to this project! It's amazing the features they've managed to add to a limited Pico! I'm not asking for anything, I'm just enjoying it immensely. A contribution equivalent to a dinner anywhere in the world would help to recognize this impressive project!!!!

 

[Kessito]

I have been testing dspi on a rpi2350 and i find it gets rather hot to the touch, when reading out CoreTemp, its running around 57 degrees.
This is in open air, no housing, no extra cooling fin. Tambient was 33degrees (yup,here also a hot summer)

I know 57C core temp is fine according to datasheet, but at 75 degrees it will do thermal throthling, and this gets me a bit anxious to put it in a housing.

Do any of you have the same experiences, or got problems with a unit in a housing? or did get any results with cooling fin e.d.?

Thanks,
Kees

 

[Holy Spirit]

I have not found any better thread, so I am posting here. I have this DSP schematics from VituixCAD. There is 2nd order inverting all pass filter. Question: Is using non inverting all pass filter and then inverting the whole channel going to work the same? Google AI says Yes, but I do not trust it. Apparently some DSPs (Hypex filter designer) do not offer inverting all pass.

 

[TOMSON]

After long struggles with the program controlling the TAS5828M amplifiers via I2C and communicating the amplifiers with the DSPi via I2S, a test with an analog input on the PCM1808 in master mode communicating with the DSPi via I2S

 

[Red_Red]

I have been testing dspi on a rpi2350 and i find it gets rather hot to the touch, when reading out CoreTemp, its running around 57 degrees.
This is in open air, no housing, no extra cooling fin. Tambient was 33degrees (yup,here also a hot summer)

I know 57C core temp is fine according to datasheet, but at 75 degrees it will do thermal throthling, and this gets me a bit anxious to put it in a housing.

Do any of you have the same experiences, or got problems with a unit in a housing? or did get any results with cooling fin e.d.?

Thanks,
Kees

Yes, that’s a problem we’ll likely have to face more and more often in the future... where I live, the temperature has dropped below 25°C on only two nights in nearly three months (and an average of 36°C during the day and 27,5°C at night) ; and that’s outdoors ! Indoors, without air conditioning, it’s 6 to 8 degrees hotter. We can choose not to run our electronics, of course, but for living things like us, it’s a different story... And even though air conditioning makes the situation worse, it’s disheartening to have no choice but to rely on it... There’s no controversy here, just plain, hard reality. After more than 50 years of warnings and no realistic global policy to tackle the issue - plus the willful blindness of the greedy at everyone else's expense - this is what we’re left dealing with...

Sorry... sometimes you just have to let it out...

 

[Weeb Labs]

I have been testing dspi on a rpi2350 and i find it gets rather hot to the touch, when reading out CoreTemp, its running around 57 degrees.
This is in open air, no housing, no extra cooling fin. Tambient was 33degrees (yup,here also a hot summer)

I know 57C core temp is fine according to datasheet, but at 75 degrees it will do thermal throthling, and this gets me a bit anxious to put it in a housing.

Do any of you have the same experiences, or got problems with a unit in a housing? or did get any results with cooling fin e.d.?

Thanks,
Kees

Those temperatures are perfectly reasonable but if you are concerned, I would recommend picking up a small heatsink like this one. With those ambient temperatures, you might find it very helpful when we introduce selectable system clocks/voltages in future betas.

 

[Red_Red]

Those temperatures are perfectly reasonable but if you are concerned, I would recommend picking up a small heatsink like this one. With those ambient temperatures, you might find it very helpful when we introduce selectable system clocks/voltages in future betas.

That's also the solution I found to cool off this summer... I applied over a thousand of them to pretty much my entire body... haha, just kidding, of course.

 

[john61ct]

I have not found any better thread, so I am posting here. I have this DSP schematics from VituixCAD. There is 2nd order inverting all pass filter. Question: Is using non inverting all pass filter and then inverting the whole channel going to work the same? Google AI says Yes, but I do not trust it. Apparently some DSPs (Hypex filter designer) do not offer inverting all pass.

Start your own thread – they're free!

This one is specifically about DSPi project

suggested title:
"Question on DSP filters – inverting Allpass "

but ofc whatever you like.

 

[Salt]

After long struggles with the program controlling the TAS5828M amplifiers via I2C and communicating the amplifiers with the DSPi via I2S, a test with an analog input on the PCM1808 in master mode communicating with the DSPi via I2S

 

[Red_Red]

I have not found any better thread, so I am posting here. I have this DSP schematics from VituixCAD. There is 2nd order inverting all pass filter. Question: Is using non inverting all pass filter and then inverting the whole channel going to work the same? Google AI says Yes, but I do not trust it. Apparently some DSPs (Hypex filter designer) do not offer inverting all pass.

Here is the response from another AI (Mistral AI) :
"

Non-Inverting All-Pass Filter + Channel Inversion​

• Non-inverting all-pass filter: This filter passes all frequencies equally in magnitude, but shifts the phase of the signal. The output is not inverted in polarity.
• Inverting the whole channel: This flips the polarity of the entire signal (multiplies by -1).

Will it work the same as an inverting all-pass filter?​

No, it will not be exactly the same.

Why?​

• Phase Response: The non-inverting all-pass filter introduces a specific phase shift. If you then invert the whole channel, you are adding a 180° phase shift at all frequencies. This is different from the phase response of an inverting all-pass filter, which has its own unique phase characteristics.
• Polarity: The inversion affects the polarity of the signal, but the phase shift introduced by the all-pass filter is not simply a polarity flip—it’s a frequency-dependent phase rotation.

Practical Implications​

• If you cascade a non-inverting all-pass filter with a channel inversion, the magnitude response remains flat (as expected for an all-pass), but the phase response will be the sum of the all-pass phase and the 180° inversion.
• An inverting all-pass filter, by design, has a different phase response (often with a 180° shift at certain frequencies, but not uniformly).

Example​

Suppose you have a non-inverting all-pass filter with a phase shift of +90° at a certain frequency. If you then invert the channel, the phase at that frequency becomes +90° + 180° = +270° (or -90°). An inverting all-pass filter might have a phase shift of -90° at that frequency, which is not the same as +270°."

 

[Salt]

Here is the response from another AI (Mistral AI) :
"

Non-Inverting All-Pass Filter + Channel Inversion​

• Non-inverting all-pass filter: This filter passes all frequencies equally in magnitude, but shifts the phase of the signal. The output is not inverted in polarity.
• Inverting the whole channel: This flips the polarity of the entire signal (multiplies by -1).

Will it work the same as an inverting all-pass filter?​

No, it will not be exactly the same.

Why?​

• Phase Response: The non-inverting all-pass filter introduces a specific phase shift. If you then invert the whole channel, you are adding a 180° phase shift at all frequencies. This is different from the phase response of an inverting all-pass filter, which has its own unique phase characteristics.
• Polarity: The inversion affects the polarity of the signal, but the phase shift introduced by the all-pass filter is not simply a polarity flip—it’s a frequency-dependent phase rotation.

Practical Implications​

• If you cascade a non-inverting all-pass filter with a channel inversion, the magnitude response remains flat (as expected for an all-pass), but the phase response will be the sum of the all-pass phase and the 180° inversion.
• An inverting all-pass filter, by design, has a different phase response (often with a 180° shift at certain frequencies, but not uniformly).

Example​

Suppose you have a non-inverting all-pass filter with a phase shift of +90° at a certain frequency. If you then invert the channel, the phase at that frequency becomes +90° + 180° = +270° (or -90°). An inverting all-pass filter might have a phase shift of -90° at that frequency, which is not the same as +270°."

The last two sentences contradict them self, not (+ 270 means the same as -90)?

 

[TOMSON]

What is the procedure for switching from an input implemented using a PCM1808 operating in master mode to a USB signal input? Is it possible to save the signal input configuration? For example, if I have a PCM1808 connected in master mode and switch to USB, the I2S signal settings automatically switch from Slave to Master mode. Because when I switched to USB, the amplifiers stopped working and couldn't detect the I2S signal. This lasted for quite a while. Even after returning to the input with only a USB signal, it still didn't work. I only had some oscillations on the amplifier inputs. One time it showed that the signal was present, the next moment it showed that there was no signal, and I only heard a hum from the speakers, luckily with a small amplitude.

 

[Red_Red]

The last two sentences contradict them self, not (+ 270 means the same as -90)?

Yes, that is - among other things - what I meant! If we aren't vigilant and let our guard down, these "consumer-grade" AIs can make us believe just about anything... yet, despite that glaring issue, I’m inclined to agree with the rest. I’m no expert, I don’t have all the answers, but the whole "phase" concept makes sense to me. So, is it AI manipulation? Was the question not phrased optimally? Is it just wishful thinking?
Basically, my point is that it’s actually necessary to ask for help, just as "Holy Spirit" did. No one is perfect, and - more importantly - everyone is susceptible to influence. Sure, this might not be the best thread to find a solid answer, but it wasn't a bad idea on his part either. A dedicated thread on DSP creation isn't necessarily a bad place for that kind of answer. The real question is: who here is qualified enough to answer it? Not me, that’s for sure. I’ve been using Vituix since almost the day it came out, but I’ve never used that specific type of filtering.

 

[TurtlePaul]

Can a mod clean all of this up into a new thread? I don't want to respond because this is so far off topic.

 

[Red_Red]

Not so much, and "(...) if you're right it's because I prefer to remain silent." ("Faust", Johann Wolfgang von Goethe)

And don't worry, your question won't go unanswered!