I must admit that after some time mainly soldering SMD's, the TH parts are a pita.
Especially de-soldering a part. I had to do that for the EEPROM as I used SOIC on a DIP8 adapter, the height became a problem as I had fitted a DIP8 socket.
I had to remove the socket and solder the adapter PCB directly to the FreeDSP board.
Especially de-soldering a part. I had to do that for the EEPROM as I used SOIC on a DIP8 adapter, the height became a problem as I had fitted a DIP8 socket.
I had to remove the socket and solder the adapter PCB directly to the FreeDSP board.
Is there any in depth tutorials/guides/manuals for sigmastudio?
I've only "played around" with the software, my boards are not yet fully populated.
I intend to buy/build a measurement mic and use that with REW.
This is not going to happend during the next few weeks though as I don't have the time.
I've only "played around" with the software, my boards are not yet fully populated.
I intend to buy/build a measurement mic and use that with REW.
This is not going to happend during the next few weeks though as I don't have the time.
I'm not a fan of the layout of the SMD A version - and I don't need balanced for the other SMD version. Because I need to make changes to the board anyways, I'm currently working on converting the Classic into a KiCad project, and then changing out most of the TH into SMD. E.g. all of the ceramic caps and resistors will be SMD. I'm going to keep the layout virtually the same, except:
1) Jumpers for setting the PLL Modes. If you want to accept a master clock from the source as I do, this is likely necessary.
2) I'll probably use the standardized connector as it makes sense to do so.
The SQ on the classic board seems pretty good. I didn't measure a sinewave from it, though I could and compare against the minidsp if there's interest in that. I suspect that the analog section of the FreeDSP Classic board is probably better than the MiniDSP. It's certainly more versatile (e.g. jumper selectable analog input voltage, and the ability to configure it from Sigma Studio - which by itself makes it far superior to the MiniDSP).
For my project, USB will be the main audio source. I had intended to use the CM6631A - it allows for two ~49mhz clocks (one for 44.1 and one for 48). It then provides a ~24mhz master clock output (depending on the Fs of the audio source). The problem is that this chips appears to require an NDA to be able to generate firmware - so I'm going to need to look elsewhere. Anyone have suggestions for a USB->I2S IC that:
1) Acts as I2S master.
2) Accepts two ~49mhz clocks
3) Will generate ~24mhz clocks (based on the source Fs)
4) Provides a GPIO for source clock change indications (I need to mute the output when changing between a 44.1khz output and a 48khz output, as reclocking the DSP will generate digital noise).
5) Is USB asynchronous
6) Reasonable licensing for use in an opensource project
Placed an order for a small (3pcs) run of the FreeUSBi boards from oshpark.
That gives me both options for programming as I also have one Arduino Nano and one more on the way.
Still waiting for a few more parts to populate the FreeDSP boards.
I'm considering using a linear PS as I have a few 0-9VAC 1200mA EI transformers at home already.
Maybe just a diode bridge followed by an LT1083 or similar to drop the voltage a bit before the onboard reg?
That gives me both options for programming as I also have one Arduino Nano and one more on the way.
Still waiting for a few more parts to populate the FreeDSP boards.
I'm considering using a linear PS as I have a few 0-9VAC 1200mA EI transformers at home already.
Maybe just a diode bridge followed by an LT1083 or similar to drop the voltage a bit before the onboard reg?
Here's a good place to start. Also search for sigmastudio on youtube.
https://wiki.analog.com/resources/tools-software/sigmastudio
I did a first serious listening of classic freedsp powered with four channel gainclone and did an A/B comparison with similar speakers with passive x/o attached to 2-channel GC.
There is a possibility that I made a mistake somewhere since the whole setup is DIY, but short of that, freedsp performance is a bit disappointing, to say the least. I am using a very simple circuit in sigma studio, with 2 2-way crossover blocks configured in identical way. Most noticeable degradation comes in the "imaging", which is simply non-existant in freedsp. I will keep checking for errors and comparing, but in the meantime, I would like to hear any similar listening impressions between passive xo and freedsp.
There is a possibility that I made a mistake somewhere since the whole setup is DIY, but short of that, freedsp performance is a bit disappointing, to say the least. I am using a very simple circuit in sigma studio, with 2 2-way crossover blocks configured in identical way. Most noticeable degradation comes in the "imaging", which is simply non-existant in freedsp. I will keep checking for errors and comparing, but in the meantime, I would like to hear any similar listening impressions between passive xo and freedsp.
The reason I turned to freedsp in the first place, is that it should be more/less similar to miniDSP. MiniDSP is used as a crossover in highly acclaimed Linkwitz speaker designs:
LXmini Challenge
Even if it's used as a room compensation only, degradation it introduces (allegedly) in my setup is just unbearable.
LXmini Challenge
Even if it's used as a room compensation only, degradation it introduces (allegedly) in my setup is just unbearable.
DSP allows for much more flexible and powerful crossovers compared to traditional passive ones.
That's true, though I still need to learn how to use the programming software, I am less than new to this.
I do feel that in many(most?) cases less is more, by that I refer to the extra signal conversions.
I could go fully active with my dipoles, IF, I get a digital input to work (and preferably, am able to take I²S to D/A's of my choice).
I use a DCX2496 for the woofers now as my diy active x-over w dipole compensation died and I had long before that lost touch with the guy that designed it for me.
I didn't have the schematic or any information left, so I bought a new DCX2496(the same device the active x-over replaced).
That sounds like something is wrong with your setup. I'm not sure which FreeDSP board you're using but the ADA1701 used on the classic is the same chip used on the minidsp. That project you referenced appears to be using a minidsp with the builtin a/d and d/a as opposed to an I2S a/d and d/a. Given this, I'd expect FreeDSP to outperform the minidsp: the analog filters are better on FreeDSP than the minidsp.
The only time I've heard real noticeable degradation is when using an external clock which had path stability issues (e.g. a wired using a long unshielded cable).
In addition, for both the FreeDSP and the minidsp its important that you use a proper power supply. Powering from a cheap USB supply, or a computer over USB will introduce unacceptable noise levels. This is true for both minidsp (which I was disappointed with until I realized the power supply was the problem) and FreeDSP.
Finally, if you really don't like the A/D or D/A sound from the onboard ADAU1701 converters you can always use external ADC and DACs over I2S. Really the possibilities are endless.
I think it is. Single channel sounds pretty good, but stereo doesn't. It might be some error I made on freedsp board. I will do the full check-up and report back.
Sent from my MI 5s using Tapatalk
And for a different opinion from Sigfried Linkwitz (Crossovers):
"Crossovers may be implemented either as passive RLC networks, as active filters with operational amplifier circuits or with DSP engines and software. The only excuse for passive crossovers is their low cost. Their behavior changes with the signal level dependent dynamics of the drivers. They block the power amplifier from taking maximum control over the voice coil motion. They are a waste of time, if accuracy of reproduction is the goal."
I only read the quote, not had my coffee yet lol, but I don't think I said passive x-over is better.
However I do think there's a cost/trade-off for DSP as long as you have digital to analog to digital (DSP) to analog. If filtering is analog then after the DAC is fine, if it's digital, as in DSP, I for one would prefer to keep the signal digital from source to after the digital signal processing.
But again, there's pros and cons with everything, well almost, and that is one of the things that keep diy interesting.
However I do think there's a cost/trade-off for DSP as long as you have digital to analog to digital (DSP) to analog. If filtering is analog then after the DAC is fine, if it's digital, as in DSP, I for one would prefer to keep the signal digital from source to after the digital signal processing.
But again, there's pros and cons with everything, well almost, and that is one of the things that keep diy interesting.
There are just so many speaker manufacturers who still make passive xos. In theory, dsp should be better, but I just don't hear that for myself. I checked my setup and no errors found. I did tweak dsp crossover and it sounds pretty good now, but passive setup has something almost magical in it, and imaging is noticeably better. Now, the speakers are different, but drivers are more expensive ones on dsp side, so the advantage should be for freedsp.
Sent from my MI 5s using Tapatalk