[sdiy] Oberheim Xpander Res fail

[Csaba Zvekan]

Hi Tom,

Thank you for your help. I have immediately taken out the LM311 for inspection and saw that there was some solder resin debris . I cleaned it and put the LM311 back in the place. At first the VCO, PW etc wouldn't  tune at all then later when the contact spray has settled and dry a bit some of the parts came back to live . However my problem still remains. I continued and followed the digital path of the OSC signal going into the U916 (74L367) checked inputs and outputs . Cleaned the IC and sockets and continued with the Output that goes into U904 (74LS245) . Same there checked input and output all looked fine during a Res tune some bits were sent . And again on the processor side as well. The whole chain from the LM311 to the U907(68B09) seems fine . 
Now I will have to go backwards . I did replace all the CMOS 4051 earlier U102-U107 . So that can't be it .
Maybe it's Voice 1 that is bad or would it say RES 1 VOICE 1 failed or something. It just says "fail" . This doesn't make any sense. :)
So next is I'll check out th U816 especially the RC circuitry around it . I suspect the problem lays there . But who knows 
Thanks you Tom for pointing me in the right direction.  


Csaba
On 15 Jan 2012, at 14:32, Tom Wiltshire wrote:

> Hi Csaba,
> 
> You said that it doesn't tune the resonance, but that it doesn't report which voice, right?
> 
> If it's failing on the resonance in general (rather than on a single voice) perhaps it's the comparator that turns the oscillation into a signal that the uP can measure. There's a LM311 on page 57 of the service manual (pg 56 of the pdf) that looks like chief suspect for this job. This has inputs from all the voices, and the 'OSC' output from it goes to pin 14 U916 ( a 74LS367 on page 51) on the Xpander voice board processor page. That's a 3-state buffer and the signal comes out at pin 13, which is connected to D1, one of the uP data bus lines.
> 
> Still, if the comparator were totally dead, it shouldn't tune the oscillators either, so maybe you're right looking at the CV generation. But then again, if the voices *work*, the resonance CV generation has to be ok. The place to check Res CVs is U816 (a 4051) on page 53 of the pdf. That chip or the associated R/C networks could be a problem.
> 
> Perhaps you could try looking at the DATAIN* signal on pin 15 of the '367. That controls the buffer state. You could also check the MUTE signal on pin 6 of any of the 4051's that control the panning. Since you're not hearing the tune up routine take place, that should be working and going low.
> 
> That's about all I can find that's related.
> 
> Tom
> 
> 
> 
> 
> On 15 Jan 2012, at 10:54, Csaba Zvekan wrote:
> 
>> Thank you Tony,
>> 
>> I just checked the Xpander again took Patch 07 "Spheric2" went ahead and adjusted VCF settings to self oscillate . Played all the notes from C1 to C8 and it seems just fine. Also played one note in several registers to compare them and they sound all fantastic :). 
>> So I think my problem is not in the Voice / Filter section rather it is around the DAC. Or  it can be what ever is responsible for either generating the necessary voltages for the "Res auto tune procedure DAC (which I think is fine tune as the synth sounds perfectly OK). The other variant I can think of is the microprocessor section where it's sensing the 'oscillation break point' and simply can't . 
>> Like Tom explained in an earlier mail . Thank you for that as well.
>> 
>> What I'm trying to figure out is what parts are involved in this "Res tuning " procedure.  I believe somewhere there in that chain lays my problem.
>> 
>> Csaba
>> 
>> 
>> On 15 Jan 2012, at 10:42, Oakley Sound wrote:
>> 
>>>> Even the percussive patches with high resonance work excellent.
>>> 
>>> Have you checked that all voices go into resonance in the same way? That is, set up a simple self oscillating filter patch and check that all six voices can be heard and each can be made to oscillate over a wide range.
>>> 
>>> I had a T8 once that refused to do the filter tuning part of the calibration because one of the VCFs didn't oscillate over the full audio range. This was due, in the end, to a bad capacitor in the feedback of the filter.
>>> 
>>> Tony
>>> 
>>> www.oakleysound.com                    Oakley Modular Synthesisers
>>> _______________________________________________
>>> 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
>