I AM D v200, Fx Audio d802, optimisation and TPA3116

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Q5 - yes i love this little piece. Its the "old" version based on STA350BW. The new (=pro) version seems to have a different chipset.

Btw. - my signal processing board (freedsp/ADAU1701) may be booted and controlled via i2c - but it can boot also directly from the eeprom without need for a µP. If "soft configuration" of the powerstage is really unavoidable - aren't chips with such a "selfbooting feature" out there?
 
You would need to write software to monitor the digital inputs and depending upon the input states send commands via i2c (using WIRE library) .... The specification of the software requirements would actually take more effort than writing it.

Yes i believe you. But legions of crashing telephones, amplifiers, tvs, car systems and even washing machines do not - and still continue crashing or doing something else weird. Until Dr. Servicemen gives them a brain wash - in newspeak called "firmware upgrade" :D

Should i really like this for my homebrewn active speakers?
 
Yes, the software industry has managed to hang the whole world to their update-drip. People who stare out rectangular eyes in rectangular heads through rectangular eyeglasses perpendicular to their rectangular models on rectangular screens and claim to do God's work .... they just want our best, but that they do not get! :D
 
You are already running software?

Yes i believe you. But legions of crashing telephones, amplifiers, tvs, car systems and even washing machines do not - and still continue crashing or doing something else weird. Until Dr. Servicemen gives them a brain wash - in newspeak called "firmware upgrade" :D

The ADAU1710 (in freedsp) appears to be a DSP ASIC running a sophisticated signal processing software framework. You can program it using SigmaStudio. The ability to read from EEPROM or i2C on boot up is just a feature in the bootloader software. A DSP is a variant of a microprocessor optimised for signal processing.

SigmaStudio allows you compile a DSP block diagram to a binary image that can run on the target Analog Devices processor.

:trapper: How would you rate the software in the freedsp product?
 
STA350BW

Q5 - yes i love this little piece. Its the "old" version based on STA350BW. The new (=pro) version seems to have a different chipset.

About the only problem I have with the STA350BW is that the output is open-loop. There will be no attempt made to correct any distortion generated in the output filter. Could be issue if cheap components are used. However, the cost, number of options available and the ability to accept a PCM input via i2s make the chip mighty attractive for a DIY class D amp.
 
:trapper: How would you rate the software in the freedsp product?
Not sure what you exactly mean - there is no "firmware" in the freedsp - you have to program the ADAU1701 by your own. Exactly this gives me the freedom to do what i want to do - while the fixed processing in the STA350 does not as the predefined processing by minidsp&co.

Unfortunately this freedom is also limited. One thing you can not do, for example, is a wavelet decomposition and further processing on different samplerates. This means - i can't - if you know how to do, please tell me.

In opposition to that - the external micro gives me nothing than burden - just to control unused and unwanted features of a certain chip. I accept this features may be useful for somebody else as well as a micro might be. But there is no law nor any reason to say things are good as they are and everybody has to love it. :)
 
Last edited:
My question though was about audibility, not measurement.
The wild oszillations near the peaks in Trevors measurements looks for me like signal related but not harmonic distortions - something the human ear is very sensitive for. So i simply assumed this MUST be very bad.

But you are right. Sorry for answering a question of experience by just a assumption. :)
 
SMSL amp quality

My question though was about audibility, not measurement.

Someone did mention that the SMSL amp started to sound harsh at high audio levels. At normal levels they were impressed with the audio reproduction. I assume they were not excessively over driving the amp.

One has to hope that the SMSL amp has a well designed output filter with quality parts. I guess if there is a problem one could improve the output filter for a start. Looks like Trevor Marshall has been down this track. The STA350 does exactly what I want.
 
Member
Joined 2008
Paid Member
FYI, I finally received the Popu Mercury Pro that I ordered over a month ago.

No mods yet, just getting used to the initial sound.

I actually thought it wasn't going to make it, so earlier this week I got impatient and ordered an SMSL Q5 Pro (from a domestic supplier). I made some comments on it in this thread: SMSL Q5 Pro Mods.

When I first took the Popu out of the box I was impressed: nice weight and heft to it, nice enclosure. Of course you can put junk in a nice box, and clearly this community is the first to know that. But if nothing else, it inspires a little confidence.

But then I pulled out the included 36V laptop-style power supply. This is by far the lightest (in terms of weight) power supply I've ever felt. I honestly think the cables weigh more than the PSU. No confidence in that!

Anyway, enough about the weight of the devices.

I've spent a little time with both the Q5 Pro and the Mercury Pro. I was at first somewhat disappointed by the Mercury. I hate to use the "v" word, but things did sound a bit veiled. Not offensive, but in comparison the Q5 just had more definition, "blacker blacks", space between instruments, etc.

However, the Mecury does have a wide input voltage range, 12-36V. And I have a Meanwell GS90A15 15V PSU handy. Obviously not the last word in power supplies, but it's a branded PSU from a generally reputable manufacturer. Listening to the Popu now, on this power supply: noticeably better. It could very well be expectation bias at play here. But I'm prety sure that stock 36V power supply is a joke.

I haven't yet done any extended critical listening, so these are still initial impressions. But in this brief period, with a better PSU on the Popu, I'd say the two devices are about the same. And I haven't even thought about mods yet!

I'd like to know if the SMSL can tolerate voltages other than 19V, because then I could play the PSU swap game with it as well. I also want to try the Popu with an Astron SL-11 power supply (13.8V linear regulated); I just have to make a cable for it.

I can already tell the Popu ought to be easier to mod, only because it has a bigger enclosure. The SMSL is tiny. If I mod it, it probably won't fit back in the case. The Popu isn't big by any means, but it's roughly twice as thick as the SMSL.

I haven't taken either apart yet. When I do, I'll post some pics.
 
SSM3582 might be an option.

If "soft configuration" of the powerstage is really unavoidable - aren't chips with such a "selfbooting feature" out there?

This IC has been discussed elsewhere in this forum. Analog devices make a class D amplifier that can be controlled by i2c or alternatively using ic pins. From page 30 of the data sheet:
The SSM3582 can be operated in a standalone hardware control mode without any I2C control. The same ADDRx pins used to set the I2C device address are used to set the functionality of the device.
See more details at: SSM3582 Datasheet and Product Info | Analog Devices.

Might be useful? There is an evaluation board as well.:smash:
 
...
Wiring the ESP8266 to the D802:

Near the OLED display there are 5 solder pads (GND, SDA, SCL, 5V, 3.3V), here we have the I2C bus. Also we need one pad near the rotary encoder (on a DMM you should get about 4.3V here). On my photo I soldered some pin headers here.

Wiring ESP -> D802:

GND -> GND
VCC -> 3.3V
GPIO2 -> SDA
GPIO0 -> SDL
CH_PD -> Pad near rotary encoder (need a diode here!)

Put the ESP8266 near the D802 MCU should be a good place.


Optionally:

Break out the red LED from the ESP8266 using a small screwdriver. This LED is to indicate power on, but otherwise useless and only draws some mA.


How it works:

If the CH_PD of the ESP8266 goes to high, the ESP8266 will start running (blue LED of the D802 is on). The pad near the rotary encoder gives about 4.3V when the blue LEDs are on. (The LEDs are switched by a transistor, so we have already about 0.7V drop here, 4.3V = 5.0V - 0.7V). So we use simply a diode to get another 0.7V drop here and the CH_PD will get about 3.6V, which should be ok for the ESP8266. On the 4.3V pad there is also a 1k resistor, so this pad is perfect to indicate the D802 is switched on.

On start up, the program waits 5s, then reads the conf D register (0x03) from the STA326 (0x1a), clears the ZDE bit (auto mute) and write it back to the STA326. After that the program reads conf D again and checks if the ZDE is still disabled. The program gives some output to the RX/TX pins, so you can connect it to a serial monitor (115200bps) to see if everything is ok. At least you can see the RX/TX output by some flashes of the blue LED on the ESP8266, so you can see if the program is working.
After that, the ESP8266 goes to sleep an will only wake up by reset or CH_PD goes to high. The WIFI will be completely disabled.


With that trick we can set some other options of the STA326, but maybe that is not reliable.

That sounds great! I am looking for a way to control the volume over WLAN.
Do you think that would work?
Does the microcontroller in the D802 talk to the STA326 when there is no input on the controls? That might collide with I2C commands from an external microcontroller.

Thanks for your sharing your work!
 
Thanks. globulegl thinks the V200 beats the D802 when using USB input, but of course the D802 has a mediocre USB interface, so this is not a fair comparison.
It appears, then, that STM chips have the edge on TI chips at this stage for entry level FDA, but of course nothing stays the same for long ...
 
Thanks. globulegl thinks the V200 beats the D802 when using USB input, but of course the D802 has a mediocre USB interface, so this is not a fair comparison.
It appears, then, that STM chips have the edge on TI chips at this stage for entry level FDA, but of course nothing stays the same for long ...

That is the conclusion I have reached personally **based on my own listening preferences**. To my ears the STM chips can give a more open, fast and detailed presentation which suits me. Other folks may disagree but I'd characterize the V200/Ti as a little smoother and weightier in the bass and I could understand someone else preferring that alternative presentation.

That all said, the D8 with its 'at slight extra cost' xmos option, has upped the game for USB interfaces.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.