DOD FX-25 versions

[Fancy Lime]

Hey everyone,

I've been experimenting with a DOD FX-25 based filter. In doing so I was looking at a few schematics here: http://topopiccione.atspace.com/PJ11DODfx25.html
I noticed that on version posted there, which seems to be the "original" version is very different from the others in that the inputs of both OTA stages are flipped which completely changes the feedback loops, no? I'm talking about this schematic:
http://topopiccione.atspace.com/pjimages/DODfx25old1.sch.gif

Is that correct? The rest of the filter section is more or less the same with the exception of a slightly different arrangement of the resonance control. Seems awfully strange that that should work. Or is that an error? Maybe someone with a better understanding of SVFs can enlighten me?

Thanks,
Andy 

[ElectricDruid]

Sorry to not be that person with the better understanding, but is that schematic even an SVF? I'm hunting for the feedback paths that I expect to see, and I can't find them. Instead, I'm seeing two OTA-C lowpass filter stages. Could that be right?

Tom

[Mark Hammer]

While drawn in a slightly different way, version 3 appears to have an error.  The 10nf cap to ground shown for each OTA stage in the other drawings appears to be missing in filter stage 1.  Version 3 (FX-25B) has a blend control, presumably to add in some bottom, along with the swept bandpass, making it more amenable to use with bass.

[Fancy Lime]

Sorry to not be that person with the better understanding, but is that schematic even an SVF? I'm hunting for the feedback paths that I expect to see, and I can't find them. Instead, I'm seeing two OTA-C lowpass filter stages. Could that be right?

Tom

The second one, yes. That would be a normal low pass. But the first stage is configured as a Schmitt trigger with a hysteresis control inexplicably called "Range" and a slew rate controlled by the control input. Or am I getting this wrong? If one were to flip both OTA stages however, it would be a normal SVF, wouldn't it?

Andy

[ElectricDruid]

Sorry to not be that person with the better understanding, but is that schematic even an SVF? I'm hunting for the feedback paths that I expect to see, and I can't find them. Instead, I'm seeing two OTA-C lowpass filter stages. Could that be right?

The second one, yes. That would be a normal low pass. But the first stage is configured as a Schmitt trigger with a hysteresis control inexplicably called "Range" and a slew rate controlled by the control input. Or am I getting this wrong?

Ok, I dunno if you're getting it wrong. All I can do is tell you what I see and you can see if it agrees with what you're looking at. I'd missed first time that the two stages use different paths, and that clearly matters.

Second one has 22K feedback wrapped from the output to the input end of R4 (on the -VE input), where the signal from the first stage also comes in.
First stage does the same thing, but replaces the 22K with a 100K variable, and takes the feedback to the +VE input instead - so quite different.

Still, I'm not seeing any feedback around *both* stages that I'd expect if this was an SVF.

If one were to flip both OTA stages however, it would be a normal SVF, wouldn't it?

That's just it. Where's the feedback from the output of the second stage to the first stage?

And no, a "normal SVF" would have both stages the same, either inverting or non-inverting. This might be a "not normal SVF" though...couldn't rightly say!

Tom

<edit>Or there could be one or more errors in the schematic and we might be discussing something that isn't even real...it's not that unlikely.</edit>

[Fancy Lime]

Hi Tom,

you're right, there is one more resistor missing that decouples the direct signal path from the feedback loop, causing the two to collapse in to one. Have a look at this schematic:
http://topopiccione.atspace.com/pjimages/DODfx25.sch.gif
This looks correct to me. In fact I just build exactly this filter section and it works as advertised. If we take this and move P2 to the left of its upper node, replace R15 with a jumper and flip the inputs on both OTA stages, we get something that looks like the other schematic, don't we? What used to be the upper node of P2 collapses into what used to be the node between R15 and R14, splitting the global feedback into two local feedback loops. I don't see how that would work at all because the feedback is not separated from the forward signal path anymore.

Or there could be one or more errors in the schematic and we might be discussing something that isn't even real...it's not that unlikely.

That's what I thought at first too until I found several posts that seem to suggest that they built one of these from the schematic in question, although it was not 100% clear from the context if they may not have used one of the other schematics. Also, judging by youtube comparison videos (yeah, I know, but hear me out), the oldest version seems to sound noticeably different from the others. Mostly louder and more resonant. Hence my curiosity. But there are other differences between the versions, which may very well be responsible for that. So I just thought I checked if someone can tell me if I'm getting it wrong or the schematic is wrong. But if your not getting it either, I'd say the current score is 2:1 that it's the schematic.

Cheers,
Andy