[sdiy] Untangling the Oberheim OB-8 filter

Thu Jun 15 01:08:17 CEST 2017

"But in this case I can't see where the "inverted phase" is coming from. What am I missing?"

The VCAs in the chip are all inverting, I think! 

all that gain applied to the input mixer with frequency (already pointed out), anyone know why?

Chris 

Sent from my iPhone

Sent from my iPhone

Sent from my iPhone
> On Jun 14, 2017, at 7:05 AM, Tom Wiltshire <tom at electricdruid.net> wrote:
> 
> Hi Magnus,
> 
> It doesn't really "use" 4 OTAs though, does it? My reading of it was that the first OTA+buffer is being repurposed as basically an op-amp mixer - they're using its virtual ground node, but not much else of it. The fourth OTA+buffer doesn't really feature in the circuit at all, except as a left-over from the four-pole version (so it's hanging off stage 3)
> 
> Please correct me if that's not right though - I'd like to have this diagram as accurate as I can.
> 
> I've got another question about the SVF circuit too; How does that BP feedback work? There are two routes from the BP stage (stage 2) back to the input, one via the 100K+1M resistors, and the other via the Resonance VCA. I've seen similar arrangements before, where one feedback path acts to increase the resonance, and the other (with inverted phase) acts to damp it. But in this case I can't see where the "inverted phase" is coming from. What am I missing?
> 
> Thanks,
> Tom
> 
> ==================
>     Electric Druid
> Synth & Stompbox DIY
> ==================
> 
> 
> 
> 
>> On 14 Jun 2017, at 09:31, Magnus Danielson <magnus at rubidium.dyndns.org> wrote:
>> 
>> Hi,
>> 
>> The 2-pole SVF might look a bit strange.
>> 
>> Notice how the first OTA, stage 1, is used to scale the input and feedback with frequency and that it the feedback point is from Stage 2, which is really the first resonator stage. It looks like an off by one, but it isn't. It uses 4 OTAs in total, so two 13700s would pull it off.
>> 
>> Cheers,
>> Magnus
>> 
>>> On 06/12/2017 11:16 AM, Tom Wiltshire wrote:
>>> Ok, update time.
>>> 
>>> I've redrawn the schematic in way I find easier to understand:
>>> 
>>>  http://electricdruid.net/wp-content/uploads/2017/06/OB-8FilterSchematic.png
>>> 
>>> Once I'd got it like this, it's easy enough to take away the irrelevant parts for each of the two versions:
>>> 
>>>  http://electricdruid.net/wp-content/uploads/2017/06/OB-8Filter4PoleLowpass.png
>>>  http://electricdruid.net/wp-content/uploads/2017/06/OB-8Filter2PoleSVF.png
>>> 
>>> I think these make it pretty clear what's going on.
>>> The original voice schematic is here for comparison:
>>> 
>>>  http://electricdruid.net/wp-content/uploads/2017/06/ob8-sm-24.jpg
>>> 
>>> I left a certain amount of components on the 2-pole and 4-pole diagrams because I wasn't completely convinced that they could be removed as irrelevant. If you see anything on those diagrams that clearly doesn't do anything, please say so. And of course, any other errors you spot. I've been staring at this thing for too long to be able to see it clearly now!!
>>> 
>>> Thanks,
>>> Tom
>>> 
>>> ==================
>>>    Electric Druid
>>> Synth & Stompbox DIY
>>> ==================
>>> 
>>> 
>>> 
>>> 
>>>>> On 11 Jun 2017, at 20:23, Magnus Danielson <magnus at rubidium.dyndns.org> wrote:
>>>> 
>>>> Hi Tom,
>>>> 
>>>> It was Juergen that got me interested in the OB-8, and when I found one, I found it in Australia. With it came a complete service manual with a lot of Enginering Change Orders (ECOs). It didn't take long before I scanned the full set and put up on my webpage. That is when the manual hit the web.
>>>> 
>>>> I have receiver one or two ECOs from others that provided them, I think one came from Kevin "Synthfool" Lightner.
>>>> 
>>>> It bothers me that both of these friendly guys left us.
>>>> 
>>>> Anyway, hopefully all the schematics acts like inspiration to you all, and also help to maintain these machines operating.
>>>> 
>>>> Back in the days, there was essentially nothing online, and then through luck and sometimes secret handshakes you could get some more schematics. People like me scanning things and sharing changed that situation and provided a richness. The other week I got the Electronotes everything package. What a lovely set of material.
>>>> So much to read, so much to think about.
>>>> 
>>>> Cheers,
>>>> Magnus
>>>> 
>>>> 
>>>>> On 06/11/2017 01:15 PM, Tom Wiltshire wrote:
>>>>> Thanks Magnus.
>>>>> 
>>>>> Redrawing the diagram is exactly what I'm currently engaged in, so as soon as I've got something worth showing, I'll post it. I totally agree that it seems like the best way to make it clearer what's going on.
>>>>> 
>>>>> Is any of Juergen's OB-8 stuff online still? That sounds interesting. Do you have a link?
>>>>> 
>>>>> Regards,
>>>>> Tom
>>>>> 
>>>>> ==================
>>>>>   Electric Druid
>>>>> Synth & Stompbox DIY
>>>>> ==================
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>>> On 11 Jun 2017, at 10:57, Magnus Danielson <magnus at rubidium.dyndns.org> wrote:
>>>>>> 
>>>>>> Hi Tom,
>>>>>> 
>>>>>> The resonator control cell either do the feed forward (2-pole) or feedback (4-pole). This is the less obvious one for the 2-pole scenario, while it is very simple and classic for the 4-pole scenario.
>>>>>> The output switches also helps.
>>>>>> 
>>>>>> If you redraw the CEM3320 as separate stages, it becomes much easier to unndestand what each contributes to.
>>>>>> 
>>>>>> Juergen Haible did his own modifications to his OB-8, and this also included a high-pass capability.
>>>>>> 
>>>>>> Cheers,
>>>>>> Magnus
>>>>>> 
>>>>>>> On 06/09/2017 04:55 PM, Tom Wiltshire wrote:
>>>>>>> Hi All,
>>>>>>> 
>>>>>>> I've been doing some research on filter designs that use analog switches to reconfigure the filter (as part of a series of blog posts on multimode filters that I'm working on):
>>>>>>> 
>>>>>>> http://electricdruid.net/multimode-filters-part-1-reconfigurable-filters/
>>>>>>> 
>>>>>>> Among these is the Oberheim OB-8, which uses a 4016 and a 4053 to mimic the effect of the two independent filter circuits in the OB-Xa (one four-pole lowpass, one 2-pole state-variable, used as lowpass only). The circuit looks like this:
>>>>>>> 
>>>>>>> http://electricdruid.net/wp-content/uploads/2017/06/ob8-sm-24.jpg
>>>>>>> 
>>>>>>> I'm struggling to untangle this. It's safe to assume that the two separated circuits are close to the OB-Xa designs (they are - that much is clear). I've already documented those here:
>>>>>>> 
>>>>>>> http://electricdruid.net/cem3320-filter-designs/
>>>>>>> 
>>>>>>> Despite having a good idea what I'm looking for, I haven't yet been able to get the two different OB-8 designs untangled. I've got the schematic open in a image editor with multiple layers, so I can delete irrelevant parts of the schematic from the "4 pole" and "2-pole" layers and so on. This is getting me closer all the time, but I'm not there yet. I think it's also the case that in 2-pole mode there are a number of components hanging off things that are connected but basically irrelevant, since they're there for the other circuit, and similarly for 4-pole mode. That makes life more of a mess.
>>>>>>> 
>>>>>>> Any advice on how to untangle this? I'm feeling a bit like I might have to redraw the schematic into a clearer layout first, and then deal with separating the two versions, but that seems like adding another step.
>>>>>>> 
>>>>>>> Thanks,
>>>>>>> Tom
>>>>>>> 
>>>>>>> ==================
>>>>>>>  Electric Druid
>>>>>>> Synth & Stompbox DIY
>>>>>>> ==================
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> _______________________________________________
>>>>>>> Synth-diy mailing list
>>>>>>> Synth-diy at synth-diy.org
>>>>>>> http://synth-diy.org/mailman/listinfo/synth-diy
>>>>>> _______________________________________________
>>>>>> Synth-diy mailing list
>>>>>> Synth-diy at synth-diy.org
>>>>>> http://synth-diy.org/mailman/listinfo/synth-diy
> 
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy