Korg SP250 Mainboard no audio output

[2phar]

Hi all..

I have a question about probabilities in troubleshooting a surface mount
mainboard in a Korg SP250 Digital Keyboard.

The mainboard uses a SAM9753 ('dream') DSP/controller, with digital
audio & clock signals going straight into a PCM1716E DAC, the stereo
output of which goes thru left/right low pass filters built on a
NJM4580M dual op-amp.

All power supply rails look good.

The mainboard passes its internal memory checks and MIDI loopback
checks, etc., which leads me to think the SAM9753 itself is ok.

I have seen reports of two other Kong instruments with similar
condition.. main DSP / power ok but no audio out.

I intend to report back here if I get to the bottom of it, but I was
wondering given the above what people would be inclined to suspect /
most likely failure of the following... solder, caps, the DAC chip, the
OpAmp, partial failure of the DSP chip? Unit is about 5-7 years old.

[Gareth Magennis]

"2phar" <no...@example.com> wrote in message
news:ljthrt$phg$1...@dont-email.me...

Do you mean no audio from the inbuilt speakers, or line outs, or headphones,
or all?

Is the keyframe working? Check there is MIDI coming out of the MIDI out
socket using an LED inserted into the two pins either side of centre. (Try
both orientations, it will only work one way round)

The internal speakers are muted by a headphone jack into a headphone socket.
Could one of these sockets be faulty thus muting the audio?

If the OP-amp has failed it will have DC on one or both of its output pins.

The Master Volume control is a convenient place to scope for a signal, and
could be the problem.


In other words, look for the simpler things first, instead of thinking of
DSP's and DAC's and stuff.



Good luck.


Gareth.

[N_Cook]

Is there an audio out/in 1/4 inch socket to 1/4 inch external loop send/
return ? which could have a flakey bypass switch

[2phar]

Thanks for the responses Gareth and N Cook!

Sorry.. I should have been a bit clearer to start with.

The output from the OpAmp on the main board goes to the volume pot, and
then on to the separate audio amp board which does all the jacks, line
outs, main amp, muting. The audio amp stage is working.. I can feed
audio into it ok.

I tested the output off the mainboard that goes to the volume pot, and
there is no audio at that point.

There is an on-board self test sequence which I can activate and the
memory/logic test passes, and the MIDI test passes when the midi in/out
ports are looped back, it fails when the MIDI ports are not looped (as
expected).. so I believe the MIDI and basic processor/memory etc part is
probably ok.

Here is the schematic of the main board:
"http://i.imgur.com/hVf4N6i.gif"

IC3=PCM1716E

So far all I know is there is no audio on CON8-1 and CON8-3 (right of
diagram).

Is it typical for both op-amps in the dual op-amp to die together?

I have to figure out how to get the boards running on the bench next so
i can get at the mainboard to test it in situ. I will check for DC
output on the amp.

Thanks again for the suggestions so far.

[dave]

On 05/01/2014 09:27 AM, 2phar wrote: there a dark recess with a hidden
"reset" button somewhere?

Is ther

[2phar]

No reset button that I can find anywhere. There is a startup reset
circuit on the mainboard that does drive the reset line on the DSP and
the DAC (and the CPLD for IO), but I can't see how it could be working
ok for the DSP (which seems to run ok), but fail for the DAC?

On 5/1/2014 1:56 PM, dave wrote:
> Is there a dark recess with a hidden "reset" button somewhere?

[2phar]

On 5/1/2014 10:05 AM, Gareth Magennis wrote:
> Is the keyframe working? Check there is MIDI coming out of the MIDI out
> socket using an LED inserted into the two pins either side of centre. (Try
> both orientations, it will only work one way round)

Nice tip thanks.. Yes.. I am seeing LED activity on pressing and
releasing keys on the keyboard, so that part looks good.

[Gareth Magennis]

Here is the schematic of the main board:
"http://i.imgur.com/hVf4N6i.gif"

IC3=PCM1716E

So far all I know is there is no audio on CON8-1 and CON8-3 (right of
diagram).

Is it typical for both op-amps in the dual op-amp to die together?

I have to figure out how to get the boards running on the bench next so
i can get at the mainboard to test it in situ. I will check for DC
output on the amp.

Thanks again for the suggestions so far.

A quick look at that schematic shows the Op-amp to have a single rail supply
(+9v / 0v) and the output pins will probably normally be at a DC of 5v,
before the DC blocking capacitors.

What I meant to say was a blown op-amp will usually have an output of either
the full +ve supply or the full -ve. In this case +9v or Zero.



Cheers,


Gareth.

[2phar]

On 5/1/2014 3:50 PM, Gareth Magennis wrote:
>
>
> Here is the schematic of the main board:
> "http://i.imgur.com/hVf4N6i.gif"
>
> IC3=PCM1716E
>

> A quick look at that schematic shows the Op-amp to have a single rail
> supply (+9v / 0v) and the output pins will probably normally be at a DC
> of 5v, before the DC blocking capacitors.
>
> What I meant to say was a blown op-amp will usually have an output of
> either the full +ve supply or the full -ve. In this case +9v or Zero.

With the keyboard idle, using a DVM, the op amp power rails at pins 4
and 8 are at 0 and 9 volts, the outputs pins 1 and 7 are both 2.47 volts
and the + input pins 3 and 5 are both 2.44 volts.

Also touching the 3 and 5 pins is producing hum on the amp output.

[Gareth Magennis]

"2phar" wrote in message news:lk0jk1$rgf$1...@dont-email.me...

Sounds to me like the op-amps and downstream are working fine.

You will need to scope the D/A converter then.
If the Data pin 2 shows a signal stream that varies as you press keys, then
it is being given data.
If it doesn't vary or is non existent, then either there is no data from the
CPU, or the faulty D/A converter is shorting it out. You would need to cut
the track to establish which.



Pins 5,3,1 I think should all be carrying clocks.
There should be 5v on the pins 8,9,15,20.
Rst pin 22 should be 5v, as should the group 23,25,28. (faulty reset/mute
circuitry is not that uncommon)

If all this is happening and VoutL and VoutR are not varying with key
presses, then it is highly likely the D/A is toast.




Gareth.

[2phar]

Yes.. so I have been exploring the DAC pins with the DVM, not having
anything else:

pin 5 (ext clock) = 0.78 v
pin 2 (data) = 0.33 v idle, varies from 0.8 v back to 0.33 when note
played [is a assume there is a data stream here]
pin 3 (bit clock) = 1.66 v
pin 1 (left/right clock) = 1.66v

reset is 3.3 volts - DAC spec says logic high is > 2.0 v
all the supply lines are good
zero level output stays at 0 volts
mute/mode lines are hard wired to the correct voltages per diagram

the analog left/right outputs pins 13/16 show 2.44 volts
analog extl/extr on pins 1118 measure 2.42 volts

So yup.. ordering a new DAC.

Any thoughts on desoldering/replacing 28 pin SSOP SMT package?
And why would a DAC within a working board just die? External power surge?

[Gareth Magennis]

Oops,

the digital circuitry is a mixture of 5v stuff and 3.3v stuff so some "5v
readings" I quoted will actually be 3.3v.

In particular the Reset line, which is actually a "Not Reset" line, meaning
a Reset happens when 0v is on this pin. Anything higher is ignored.



Gareth.

[Gareth Magennis]

"2phar" wrote in message news:lk0pal$8m5$1...@dont-email.me...

We must have both been typing at the same time there.


There are many ways to desolder/solder SMD stuff, go on youtube and have a
look, there are some great tutorials there.

My preferred method is to put a new blade into a craft knife and cut through
all the pins first, then you can get the main chip body off the board
leaving the pins behind - occasionally they are glued to the PCB really
strongly.




Gareth.

[2phar]

To follow up this issue..
I replaced the DAC and system is now producing audio again.

Only remaining problem is a few keys on the keyboard not working fully.
It appears the switches are those rubber membrane / carbon pad things..
so lots of potential there for flake outs.

[Cydrome Leader]

yay.

how did you remove the original chip?

[2phar]

I used ChipQuik solder paste along one side and a pretty hot 40 watt
iron to get one side molten enough that i could lift it slightly with a
pick. then it came off easy enough doing same on the other side. There
appeared to be no issue with adhesive holding down the package itself.

Being my first go at SMT, I initially tried to cut the package away from
the pins with a knife but that didnt go very well.. some of the pins
just bent and broke one track and pulled a few pads off the board. That
led to some tedious unbending and repositioning the pads with a 10x
magnifier and bridging one track with some 36 swg copper wire, but got
there in the end. Live and learn.

[Gareth Magennis]

"2phar" wrote in message news:lkbaqc$f0o$1...@dont-email.me...

You need to put just vertical pressure on the pins as close to the package
as you can, so you are not sawing or moving them side to side.
The blade will just click straight through them one by one causing no
damage.



Gareth.

[Rick]

"Cydrome Leader" <pres...@MUNGEpanix.com> wrote in message
news:lkb339$igv$1...@reader1.panix.com...

This works to fix the bad keys.
http://www.ebay.com/itm/KEYPAD-FIX-Permanently-Repairs-All-Rubber-Keypads-/290658760252?pt=US_Remote_Controls&hash=item43ac9cb63c

[2phar]

Thanks.. I have a few cordless phones in bad need of that stuff.

The issue on the SP 250 keyboard turned out to be the PCB the membranes
sit on had gotten pushed down a bit a one end and slipped below the
plastic retaining tabs on the edge, so the tops of the rubber mouldings
were a little too low below the keys.