PDA

View Full Version : Hardware reset for Slimp3: Can anyone help with resetting Xilinx chip?



roboele
2009-01-08, 12:38
I have problems with my slimp3 - takes a long time before sound starts, and the sound comes out garbled. If I use the web interface I can get it to play ok, but as soon as I use the remote control, the sound gets halted for every command, and when it resumes, the sound is garbled.

Also have done some research on the forum, wiki etc.
Came across a comment regarding squeezebox on the wiki
(http://wiki.slimdevices.com/index.php/Hardware_Reset_commands):

> Reset Audio Hardware (aka Xilinx Reset)
>
> These steps will often fix problems with garbled sound, or no sound at all during playback.
>

The wiki page does not say if it is possible to do such a reset on a Slimp3, so I sent a question to Logitech support, and they confirmed they have no way to do this, but suggested maybe the wider community may be able to help.

My question for help comes in 3 parts:
1) anyone know a way to diagnostics if this is really the root cause of my problems? Or suggestions on other issues that may cause similiar problems (I already tried to reload the firmware with bootloader, tried different server versions, different PCs as servers, wireless/wired network - made no difference)

2) anyone ever done a reset of a Slimp3 Xilnx chip?

3) on the circuit board, there are soldering points for what looks like a 6 pin connector, and from the Xilinx documentation I can see that the cable used for in-place configuration of the chip use 6 signalling cables - anyone know any more about the "pin-out" for the 6 soldering points, and/or any info on how to get hands on the JEDEC file for the Slimp3?

Any help is greatly appreciated, I love my Slimp3 and would be very sorry to let it go

Regards
Jan Robøle

Mark Lanctot
2009-01-08, 12:50
I don't believe the SliMP3 or the SB1 even have a Xilinx chip - I believe it was introduced with the SB2 as part of the "SlimDSP". The codecs aren't hard-coded, they're coded on the Xilinx which can be rewritten. The SliMP3 and SB1 have hard-coded codecs and can't be changed or reset.

I've read about the socketed CPU in the SliMP3 getting oxidized contacts. It can be extracted with a chip extractor or (carefully) with a jewller's screwdriver IIRC and the contacts cleaned with a pencil eraser or a fibreglass contact cleaner pen. You might want to search for those posts.

SuperQ
2009-01-08, 13:19
The slimp3 has a completely different firmware updater. I've noticed that the PIC chip tends to corrupt the rom over time. Sometimes simply flashing the PIC with the slimp3 firmware updater can help.

http://wiki.slimdevices.com/index.php/SLIMP3

Grab version 2.3 from here:
http://downloads.slimdevices.com/firmware/2.3/

seanadams
2009-01-08, 13:54
The best you can do with SLIMP3 is re-flash the firmware - there is no field programming mechanism for the CPLD.

roboele
2009-01-08, 14:24
thanks for your comments guys:
Mark: I opened the case, and there is a xilinx XC95144XL on the inside ;-)
Sean/SuperQ: Thanks for the suggestion. I have already reloaded the 2.3 firmware. I will try cleaning the socketed chip!

Mark Lanctot
2009-01-09, 07:04
Mark: I opened the case, and there is a xilinx XC95144XL on the inside ;-)

Huh! Out of interest since Sean's responded to this thread, what did the CPLD in SLIMP3 and presumably the SB1 do? Just run firmware?

The hardware comparison wiki page:

http://wiki.slimdevices.com/index.php/HardwareComparison

shows that the decoder in SLIMP3 was a Micronas MAS3507D, in SB1 a Micronas MAS3539, but that in SB2 and later was software-based. Presumably this is SlimDSP, and I understood this was done in the CPLD so that new codecs could be added.

seanadams
2009-01-09, 11:11
Huh! Out of interest since Sean's responded to this thread, what did the CPLD in SLIMP3 and presumably the SB1 do? Just run firmware?


These xc95... CPLDs are very small logic devices relative to most of the parts that Xilinx makes. They are not processors running software, they are actually just an array of simple logic gates. I use them mostly for glue logic, for example inverting or delaying a signal, selecting between different clock sources, converting i2s to left justified, IO multiplexing, etc.

SLIMP3 has the largest CPLD of any of the products - it is has a DMA controller implemented in the CPLD, in order to offload from its tiny CPU the work of pumping audio data from the ethernet chip to the buffer RAM and then serializing it out to the decoder.



The hardware comparison wiki page:

http://wiki.slimdevices.com/index.php/HardwareComparison

shows that the decoder in SLIMP3 was a Micronas MAS3507D, in SB1 a Micronas MAS3539, but that in SB2 and later was software-based. Presumably this is SlimDSP, and I understood this was done in the CPLD so that new codecs could be added.

the CPLDs in each o the products are:

SLIMP3: xc95144 (DMA controller, SRAM interface, bitstream output to decoder)
Squeezebox1: none (relatively fast CPU connected directly to decoder - did not need DMA or glue)
Squeezebox2, 3, Boom, Receiver: xc9536xl (clock selection and s/pdif encoding - more features for handling raw PCM properly)
Transporter: xc9572xl, similar functions as SB2/3, but a lot more of them eg s/pdif input

Mark Lanctot
2009-01-09, 11:19
Ah, so the CPLD is not actually the decoder, the decoder is separate.

Understood, thanks Sean!

roboele
2009-01-09, 15:50
These xc95... CPLDs are very small logic devices relative to most of the parts that Xilinx makes. They are not processors running software, they are actually just an array of simple logic gates. I use them mostly for glue logic, for example inverting or delaying a signal, selecting between different clock sources, converting i2s to left justified, IO multiplexing, etc.



Excuse my ignorance, but does the 'I use them mostly ...' indicate you have been involved in the development of Slim devices?
If so, do you know if it is possible to get hand on the Slimp3 xc95 config (JEDEC files or in other formats)?

Support was not able to help me out, but given the config is available I could check with the Xilinx agent here in Norway and see if they have a rig and is willing to help me out uploading it.

toby10
2009-01-10, 04:37
Excuse my ignorance, but does the 'I use them mostly ...' indicate you have been involved in the development of Slim devices?........

There is a very big clue to your question following Sean Adams' name in his post. :)

roboele
2009-01-10, 10:04
There is a very big clue to your question following Sean Adams' name in his post. :)

ohoh ... <blush>
Sean: apologize for the inconsiderate question, and I really want to congratulate you with a series of great devices!

toby10
2009-01-10, 10:15
ohoh ... <blush>
Sean: apologize for the inconsiderate question, and I really want to congratulate you with a series of great devices!

I hardly think an apology is needed. He seems like a nice guy with a good sense of humour.
I'm sure, worse case scenario, he would have *chuckled*. :)

seanadams
2009-01-10, 14:42
ohoh ... <blush>
Sean: apologize for the inconsiderate question, and I really want to congratulate you with a series of great devices!

I don't know off the top of my head where it might be.. the thing is, reprogramming it is not a simple affair and it's very unlikely that it would help anyway. The JTAG header is a not a standard footprint so it's not as simple as just plugging in a cable, and what you're describing sounds like an IC failure, not a corrupted flash of some kind. I'm afraid there's not much hope for it at this point, unless you can see a physical failure like a loose or shorted pin.

roboele
2009-01-11, 12:20
I don't know off the top of my head where it might be.. the thing is, reprogramming it is not a simple affair and it's very unlikely that it would help anyway. The JTAG header is a not a standard footprint so it's not as simple as just plugging in a cable, and what you're describing sounds like an IC failure, not a corrupted flash of some kind. I'm afraid there's not much hope for it at this point, unless you can see a physical failure like a loose or shorted pin.

I have done some more testing incl multiple reload of the 2.3 firmware.
Seems like i have a consistent pattern:
The unit is working ok after power up as long as I don't use the remote control but use the web interface to turn on the unit and select music to play.
As soon as I press any button on the remote, the slimp3 halts for maybe 15 seconds, and then continues to play, but with grabled sound, sound a bit like it is skipping packets but it does so for hours. In addition the volume seems to be maxed.
When the device flips to this 'garbled' mode it stays like that until the unit is powered down (ie stand-by, not physically disconnected from mains).
If I then turn the device on again with the web-interface, it works ok again - until the next keypress on the remote control ...

Unless anyone have some bright ideas I guess I have to chuck the remote away ;-)

wandering_andrew
2009-01-14, 15:34
Hey folks,

I'm a squeezebox newbie, just got a slimp3, sans remote, for free yesterday and have been playing with it ever since.

I had severe issues with audio distortion at all times. It was more prevalent with music and was barely detectable with some all-talk Internet Radio stations.

I was able to resolve this by:
-Removing the Microchip processor, cleaning all contacts and reinstalling
-Going into SqueezeCenter settings, Advanced, Network, and set max UDP size to 1000 (default is 1400).

Just thought this may help someone down the road.

roboele
2009-03-24, 14:51
Final note on this issue: After cleaning the chip etc etc eventually my slimp3 died alltogether. After a periode of denial, I ordered a squeezebox classic as replacement. When I received the unit, I got the old and the new power adapters mixed up. To my big surprise - with the new power adapter my slimp3 is working perfectly fine! So it turned out it was my power adapter that was broken, not the slimp3-unit ...

Thanks to everyone trying to help me out here!