[sdiy] Akai AX80 weird PWM behaviour on Voice 8 when Saw wave enabled

james meagher w.james.meagher at gmail.com
Wed Jul 11 21:21:50 CEST 2012


Thanks Adam ... I just ran downstairs while the spaghetti was boiling,
and I'm pretty sure from some readings I just took with my scope that
you have led me in the right direction.

When I check IC802 (IC102 in the Block Diagram you referred me to), I
get a nice lookin' Saw Wave consistent with the other voices on pins 1
and 6, a level of PM(W) consistent with the other voices coming into
pin 5. What comes out of pin 7 however is not consistent with the
other voices. Here the depth of the PWM is much great than that of the
other voices and you can actually see the Pulse collapsing entirely -
which it does not do with the other voices. If I turn the PWM off, the
PW of the now static Pulse is also not consistent with the other
voices - it is much narrower. With all the other voices a setting of
50 (out of 0 to 99) for the PW will yield a 50% duty cycle square -
with voice 8 there is no pulse whatsoever until you reach a setting of
about 25, and by 99 it is almost approaching the width of a true
square, but still not quite. With the other voices a setting of 99
yields an extremely narrow pulse like a setting of 1 does. However,
the second I turn off the Saw Wave on DCO 1 of Voice 8, all of the PW
and PWM functions start to operate properly like all the other voices.
It's only activating the Saw that causes the problems to the width of
the Pulse.

Of further note . . . while rushing to check all the legs of IC802
while my spaghetti boiled I accidentally shorted pin1 and pin2. There
was a small spark. I thought perhaps I'd fried something up good. But
shockingly from this point on all the previous inconsistencies with
Voice 8 disappeared and its PW and PWM operated like the other voices
. . . until this morning when things went back to the way they were .
. . malfunctioning.

I have noted that meter readings on T804 (after the Saw Wave leaves
IC802) were much higher than on the other voices. What appears to be
between the Saw leaving pin1 and both entering pin6 and going to T804
is the Saw Wave Level adjustment pot. Florian told me to check the
level of this and also to look for bad soldering on its pins. The
levels I checked looked consistent with the other voices and I
resoldered all the pins - the problem still persisted so I figured
that pot wasn't the issue and left it alone until now. I decided to
swap this pot out with that of another voice to see if perhaps there's
something in the pot not functioning right that I didn't catch the
first time around (not unlikely!), but unfortunately the pot itself
lost a pin during removal... In fact it's shell just kind of crumbled
around one of the pins and it fell out.

So that's where I'm at. Now I have to order a replacement for VR802
and hope I'm selecting the right one. If I'm reading the schematic
correctly it says 47k, and in the parts list it says 3P 0.05W 104. I'm
sure that's enough to go on. I won't lie, with my lack of knowledge
and experience I find electronics parts websites a wee bit
overwhelming!

Many thanks again Adam for leading me in this direction with the Voice
Block Diagram and pins around the SLICE. Now I feel like I have a much
better understanding of what's going on and what all these components
are actually doing. I feel like I'm learning!!!

Cheers,
James






Sent from my iPhone

On 2012-07-10, at 7:30 PM, Adam Inglis <21pointy at tpg.com.au> wrote:

> James when you first mentioned this problem, I believe that you didn't have a copy of the service manual.
> Now that you have one, have you been through the calibration procedure as outlined there, with your scope etc?
>
> On the voice block diagram on page 52 of that service manual pdf you supplied, there is a block in each DCO path labelled "Slice" - it appears to take the sawtooth wave and convert it to a pulse wave, receiving a pulse-width control signal from the CPU.
> What does your scope show you at the output of this block in DCO1 of a good voice compared with the faulty voice?
>
> Adam
> On 11/07/2012, at 4:24 AM, james meagher wrote:
>
>> Okay, since my last post I've eliminated that this issue might be
>> caused by IC704 and IC802 and also that there might be something amiss
>> with the Voice 8 saw wave trim pot level setting or soldering.
>>
>> So I did comparative readings between Voice 8 and Voice 4 across the
>> board and noted many anomolies . . . unfortunately I've not been able
>> to translate this information into a solution because I'm just not
>> there yet with my understanding of schematics and electronics in
>> general. I'll think I'm on the right path and then realize I've got it
>> all backwards. So I've decided to share my readings and ask for
>> guidance.
>>
>> Again, for reference here are some links to the AX80 SM, the Voice
>> Board Schematics and Voice Board PCB alone, and my complete
>> comparative readings between Voice 8 and Voice 4 and when necessary
>> all the other voices:
>>
>> AX80 Voice 8 complete readings 57KB PDF:
>> https://www.dropbox.com/s/o5ayzenvavpmzha/AKAI%20AX80%20VOICE%208%20readings.pdf
>> AX80 Voice Schematic 11.3MB JPG:
>> https://www.dropbox.com/s/bhi1314hv6qxi0v/AX80%20Voice%20Schematics.jpg
>> AX80 Voice PCB 16.9MB JPG:
>> https://www.dropbox.com/s/v73wr6ql0sr0506/AX80%20Voice%20PCB.jpg
>> AX80 Service Manual 34MB PDF:
>> https://www.dropbox.com/s/88xnbxv9a44iqx7/Akai%20AX80%20SM%20compressed.pdf
>>
>> I will list here only the readings with anomalies. To see all the
>> Voice 8 and Voice 4 component readings download the PDF I made from
>> the link provided above.
>>
>> Wherever anomalies occurred in the readings all other voices were
>> consulted to see which voice was indeed aberrant. On top of
>> documenting the issue with Voice 8 (with which there are many aberrant
>> readings), it appears I’ve also unwittingly discovered an incongruity
>> between Voice 4 and the other voices . . . though whatever this latter
>> issue is, I’ve not noticed any audible difference between Voice 4 and
>> the other voices.
>>
>> I took all readings from left to right as I looked at the board in the
>> AX80 (component side up). If comparing my readings to the Voice PCB
>> image from page 61 of the Service Manual PDF they will be backward
>> (mirrored) as though I read them from right to left. In all instances
>> vertically mounted components were read from top to bottom.
>>
>> Any two readings with the word “to” in-between means that the reading
>> fluctuated and I did my best to record the upper and lower value of
>> the fluctuation. Readings with N-off and N-on were used when playing a
>> note with the voice in question caused the reading to change. Where
>> there is no N-off and N-on values playing a note with the given voice
>> had no visible effect on the reading for that component.
>>
>> I’m sure that there is a standard way of presenting all of this, and I
>> apologize for not knowing it . . . but I’ve done my best to be clear
>> and thorough and will be more than happy to answer any questions. I
>> apologize for the length of this email, but I imagine if I want to ask
>> for help I at least need to clearly supply as much information as
>> possible.
>>
>>
>> Voice 8 anomalies:
>>
>> Resistors
>>
>> R805  -0.003  /  -1.224
>> R405   0.000  /   0.124
>> Only voices 2 and 4 have a positive value on the 2nd reading for this
>> resistor. The average of the other voices (1, 3, 5, 6, 7, 8) is -0.800
>> . . . so Voice 8 is below average, but is at least a negative figure
>> like 5 of the other voices.
>>
>> R819  2.158  /  -4.71
>> R419  0.043  /  -4.84
>> All other voices resemble the values of voice 4
>>
>> R820  2.109  /  4.93
>> R420  0.043  /  4.93
>> All other voices resemble the values of voice 4
>>
>> R821  2.793  /  2.115
>> R421  0.018  /  0.042
>> All other voices resemble the values of voice 4
>>
>> R824  -10.00 to -12.00  /  -11.5 to -12.5 (fluctuating)
>> R424          O.L           /  -1.8   to   -3.0 (fluctuating)
>> All other voices resemble the values of voice 4
>>
>> R825  0.60 to 0.90  /  0 (fluctuating)
>> R425  2.54 to 2.86  /  0 (fluctuating)
>> All other voices resemble the values of voice 4
>>
>> R831  4.10 to 4.45  /  0.59 to 0.64 (fluctuating)
>> R431  1.70 to 2.00  /  0.24 to 0.29 (fluctuating)
>> All other voices resemble the values of voice 4
>>
>> R832  0.011  /  0.58 to 0.63 (fluctuating)
>> R432  0.005  /  0.24 to 0.29 (fluctuating)
>> All other voices resemble the values of voice 4
>>
>> R833  0.58 to 0.63  /  0
>> R433  0.24 to 0.29  /  0
>> All other voices resemble the values of voice 4
>>
>> R836  0.49  /  1.91 to 1.88 (fluctuating)
>> R436  0.00  /  0.043
>> All other voices resemble the values of voice 4
>>
>> R839  N-off -0.057  N-on 0.050  /  1.79
>> R439  N-off -0.053  N-on 0.058  /  7.32
>> In this case only Voice 6 was like Voice 8 and all other voices were
>> like Voice 4
>>
>>
>> Caps
>>
>> C810  0.5 to 0.9  /  0
>> C410  2.5 to 2.9  /  0
>> All other voices resemble the values of voice 4
>>
>> C812  0.52 to 0.58  /  -0.02 to 0.03
>> C412  0.24 to 0.29  /  -0.02 to 0.01
>> All other voices resemble the values of voice 4
>>
>> C817  N-off 2.4 to 2.5  N-on 2.4 to 2.9  /  -0.02 to 0.02
>> C417  N-off 1.9 to 2.0  N-on 2.0 to 2.1  /  -0.02 to 0.02
>> All other voices resemble the values of voice 4
>>
>> C818  0  /  0
>> C418  N-off 1.2  N-on 2.0  /  -0.82 to -0.85
>> All other voices resemble the values of voice 4
>>
>>
>> Transistors (I might very well have the legs labeled incorrectly. If
>> they are incorrect they will at least be incorrect consistently). As
>> they are mounted in the board I read them from top to bottom.
>>
>> T804  D) 1.92  G) 2.6     S) 1.92
>> T404  D) 0.04  G) 0.04   S) 0.04
>> All other voices resemble the values of voice 4
>>
>>
>> Diodes
>>
>> D805  2.55  /  9.81
>> D405  0.05  /  9.81
>> All other voices resemble the values of voice 4
>>
>> D807  0.4 to 0.9  /  -9.0 to -11.0
>> D407  2.6 to 2.9  /  -1.8 to -3.0
>> All other voices resemble the values of voice 4
>>
>>
>> That's it for Voice 8.
>>
>> As for the rarer aberrations I found between Voice 4 and the other
>> voices - they are as follows:
>>
>> Resistors
>>
>> I've listed this one in both places:
>> R805  -0.003  /  -1.224
>> R405   0.000  /    0.124
>> Only voices 2 and 4 have a positive value on the 2nd reading for this
>> resistor. The average of the other voices (1, 3, 5, 6, 7, 8) is -0.800
>>
>>
>> Caps
>>
>> C809  -0.064  /  -04.82
>> C409  -0.064  /  -13.72
>> All other voices resemble the values of voice 8 in this instance
>>
>>
>> Transistors
>>
>> T801  B) -0.009  C) -04.82  E) -0.003
>> T401  B) -0.002  C) -13.72  E)  0.092
>> All other voices resemble the values of voice 8 in this instance
>>
>>
>> Diodes
>>
>> D806  2.29  /  -4.17
>> D406  4.60  /   5.06
>> All other voices resemble the values of voice 8 in this instance
>>
>>
>> That's all for now folks. Thanks so much for your time and interest!
>> This once dead AX80 is now almost 100% back on its feet and I can’t
>> wait to solve this last issue and get playing it!
>>
>> Cheers,
>> James
>>
>> On Sun, Jul 8, 2012 at 2:48 AM, Bob Weigel <sounddoctorin at imt.net> wrote:
>>>
>>> On 7/8/2012 1:38 AM, Veronica Merryfield wrote:
>>>>
>>>> Welcome to the list James!
>>>>
>>>> On 2012-07-08, at 12:23 AM, james meagher wrote:
>>>>
>>>>
>>>>>
>>>>> looking most suspiciously at IC704 (a 14 pin TC4011BP) and somewhat
>>>>> less suspiciously at IC802 (an 8 pin NJM4558D). I'm suspicious of
>>>>>
>>>>
>>>> 4011 is a quad two input nand chip. TC is the manufacturers code and BP is the type family.
>>>> The NJM4558D is a high current output op-amp.
>>>> Mouser should have both but from memory DigiKey don't carry the NJM4558 part.
>>>>
>>>>
>>>>>
>>>>> Again, if anyone wants a PDF copy of the AX80 Service Manual feel free
>>>>> to give me a shout. It's not the best PDF of an old service manual,
>>>>> but it's quite legible.
>>>>>
>>>>
>>>> Send it over.
>>>>
>>>> Veronica
>>>>
>>>> _______________________________________________
>>>> Synth-diy mailing list
>>>> Synth-diy at dropmix.xs4all.nl
>>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>>>
>>>>
>>>>
>>>
>>> If anything is unclear I have a good hard copy also
>>>
>>> Haven't heard the problem you're talking there and too tired to figure it out til I finish some projects here.
>>>
>>> --
>>> Check out sounddoctorin.com for songs, a studio tour, help
>>> repairing old synthesizers and parts or thought provoking
>>> articles that could help make your eternity!
>>>
>>>
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at dropmix.xs4all.nl
>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy



More information about the Synth-diy mailing list