SAA7220 filter replacement project
Posted 20th September 2016 at 12:24 PM by abraxalito
Some years ago I reverse engineered the coefficients of the Philips SAA7220 filter, just for fun. At that time the microcontroller I implemented the filter on (LPC1114) didn't have a fully implemented I2S interface so I needed additional logic to get the data in (an array of 74HC595s) and I used the SSP to output a serial stream, albeit not reliably as I didn't implement a foolproof method of generating WS. Hence the filter wasn't ready for primetime.
Now though STM has introduced a Cortex M0 MCU which does have the dedicated I2S ports to realize a pukka emulation of the SAA7220. At least that's what I'm hoping - I know the code worked on the LPC1114 but I seem to recall I needed a slight overclock - there is a report that the STM32F072 can be overclocked but I shall need to check this out for myself. One major advantage this replacement will have is going to be lower power consumption - something of the order of 100mW or under.
In order to kick this project off, I've ordered off Taobao some of these boards (35rmb a piece). Stay tuned for updates all you TDA1541A lovers
Now though STM has introduced a Cortex M0 MCU which does have the dedicated I2S ports to realize a pukka emulation of the SAA7220. At least that's what I'm hoping - I know the code worked on the LPC1114 but I seem to recall I needed a slight overclock - there is a report that the STM32F072 can be overclocked but I shall need to check this out for myself. One major advantage this replacement will have is going to be lower power consumption - something of the order of 100mW or under.
In order to kick this project off, I've ordered off Taobao some of these boards (35rmb a piece). Stay tuned for updates all you TDA1541A lovers
Save
Total Comments 8
Comments
-
Interesting!
Once you do this bottom up, why not interface directly on the front end to a S/PDIF flow, without a receiver chip??Posted 13th November 2016 at 07:40 PM by triode_al
-
Sorry, you lost me here! S/PDIF out is being generated by the SAA7220 that's the chip we're replacing....Posted 14th November 2016 at 01:11 AM by abraxalito
-
Sorry, I forgot my thinking!
Still I am interested what comes out & how this can be shared.Posted 19th February 2017 at 03:50 PM by triode_al
-
Glad you're interested, I've been working on other things as there seems to be so little interest.Posted 20th February 2017 at 12:04 AM by abraxalito
-
I'm interested and I think a lot of TDA1541A chip lovers will also be interested if it will improve the sound.
No problems if it takes time and you have other things to do.Posted 7th April 2017 at 03:19 AM by ekidnah09
-
I'm fairly sure (although I've not listened to a TDA1541A) that NOS sounds best but probably the reduction in power supply noise will give some SQ advantage to 4X OS with the ARM CPU.Posted 9th April 2017 at 10:03 AM by abraxalito
-
Hi,
Any progress?Posted 19th May 2017 at 03:02 PM by giro1991
-
No - in the absence of some end-users cheering me on I'll be working on another couple of projects (a DAC and an amp). Once I get back into microcontrollers and programming them this project is likely to go ahead even in the absence of strong demand though. Can't say when that'll be though...Posted 19th May 2017 at 11:21 PM by abraxalito
