What voltage rating do you think these caps are in the Sequential Circuits Pro-1 VCF?

[Script]

Schematics are here
http://www.synthfool.com/docs/SequentialCircuits/Pro_One/
But really need to look at the datasheet of CEM3320.

Pro 1 equivalent circuit with short description is a bit down this page:
https://electricdruid.net/cem3320-filter-designs/


I don't really know how to read that, but I'd say the Pro 1 is not an ARP with their sealed mystery components (yes, the ARP SMD reissue is pretty close but does indeed sound different).

I'd try the cheapest 150p caps I can get and then audition against higher quality 'poly capacitor' -- and then decide and call it a day.


[added] Also no guarantee that the CEM3320 and V3320 do behave exactly the same.[/added]

 

[ccaudle]

Possibly more work than you are prepared to undertake, but you can get a decent idea of the dielectric if you remove one of the capacitors and measure the capacitance at various temperatures (e.g. room temperature, after being in a refrigerator, with a hair dryer or other hot air source warming it well above room temperature). When you get some data on the magnitude and direction of capacitance change with temperature you can compare that to datasheets for C0G, Y5V, X5R ceramic caps to see what was likely used.

 

[MaxDM]

I'm doing some tests right now with a 1 meg resistor going into four 150 pF caps shunted to ground, passing music through it into a High-z input.

I have three types of disk caps.

One is an Erie 150 pf, probably 250 or 500 V, diameter is about 14mm. These sound the 'deepest' and has a smooth and dark top end.

Another is a generic 5mm 50v cap. It has less deep bottom, does not cut off the top end as much as the Erie, even though by a small amount.


I have a third 5mm type that must be the wrong value.. it was given to me as 150 pF but it's not attenuating high freq. nearly as much as the other two. Maybe it's actually 15 pF, with the modern 150 code.

 

[soapfoot]

Lots of instances when people have changed caps in VCF filters and resonant quality was affected.

You can't connect 'one of the caps' because it is a 4-pole filter, so that all the caps are equally involved and work together as a unit.

How much of a change is another thing, of course, and it will be interesting to find out, because I plan on doing it to other SMT synths as well if it is a significant difference.

I think it sounds patronizing to say that 'I can't hear a difference' if I'm paying attention.. if the difference is subtle enough that I have a hard time noticing it, that is an indication that I should not invest any more time into doing the same with other synths, and probably will put the synth back to original state.

Will need my watchmaker's glasses, cause those caps are very small...

Re-reading your post, I see now that you’re not trying to repair a fault, but are rather trying to make the Behringer copy sound more like the original.

Have we verified that the SMT versus through-hole caps are the literal only difference in the circuit implementation?

identical IC part numbers, identical component values, etc? I just want to make sure we’re targeting our attention in the most productive direction.

The thing with (presumably) class 2 ceramic capacitors is that they have a high voltage coefficient and a high temperature coefficient.

“better” class 1 ceramics (even SMT devices) are, on the other hand, pretty optimal in these characteristics. This can lead to performance that’s quite different in some applications!

I’m not trying to be patronizing, I’m just trying to gently suggest that, by focusing on voltage rating of these particular capacitors, we may not be focusing on the thing most likely to get you the results you’re after.

 

[mjrippe]

the 150 pf caps that you claim are pointless details are part of a four stage/pole OTA-based Voltage Controlled Low-pass filter with a resonant feedback loop.

It is the heart of the sound of the synth, because those caps are what shape the sound by ‘subtracting’ harmonics from the waveforms.

Fair enough, before seeing the schematic I was inclined to think they were stabilization or de-coupling caps.

I'm doing some tests right now with a 1 meg resistor going into four 150 pF caps shunted to ground, passing music through it into a High-z input.

I have three types of disk caps.

One is an Erie 150 pf, probably 250 or 500 V, diameter is about 14mm. These sound the 'deepest' and has a smooth and dark top end.

Another is a generic 5mm 50v cap. It has less deep bottom, does not cut off the top end as much as the Erie, even though by a small amount.


I have a third 5mm type that must be the wrong value.. it was given to me as 150 pF but it's not attenuating high freq. nearly as much as the other two. Maybe it's actually 15 pF, with the modern 150 code.

Like tube rolling, cap swapping can be very subjective.

 

[MaxDM]

I've replaced the caps and now the synth sounds more correct, to my ears, and has a broader sound. Might be due to lack of selection of components as well, which spreads out the knee of the filter at lower resonances, slightly.

The high voltage 1960's ceramic disk caps sounded the best, but there was no way to fit them in.

 

[JohnRoberts]

To approach this objectively measure the THD of the SKU you want to mimic, then measure the THD of your unit. This will give a crude idea of distortion magnitude. Better yet a null test will tell what the difference sounds like...(not which one is different, but if there is a difference to pursue).

Have fun... I can't recall ever hearing somebody who longed for cheap ceramic dielectric sound.

JR

 

[soapfoot]

Have fun... I can't recall ever hearing somebody who longed for cheap ceramic dielectric sound.

I've found applications--certain Valco (Kay and Gretsch) guitar amps for one--where I much prefer it.

 

[JohnRoberts]

OK now I heard someone say they prefer it....

enjoy...

JR

 

[ccaudle]

The high voltage 1960's ceramic disk caps sounded the best, but there was no way to fit them in

The term "ceramic" covers several different materials with quite different properties. You could possibly find smaller modern equivalents which were closer in performance if you were able to narrow down "ceramic" to a more specific formulation, or measure the behavior as I suggested and look for devices with close behavior.
Ceramic devices are rated by how their capacitance changes with temperature, but the capacitance also changes with applied voltage, DC and AC. Unfortunately there is no standardized rating system for voltage dependency of capacitance, but in general higher voltage devices are less sensitive, and devices which are larger than the smallest available for a particular capacitance are less sensitive. It could be that the devices you prefer have a lower, or at least different, distortion profile in your application, or that the frequency response varies more or less under voltage changes than the devices which you did not prefer. Hard to say without measuring both, so hard to suggest alternatives for any future mods you might want to try.

 

[MaxDM]

To approach this objectively measure the THD of the SKU you want to mimic, then measure the THD of your unit. This will give a crude idea of distortion magnitude. Better yet a null test will tell what the difference sounds like...(not which one is different, but if there is a difference to pursue).

Have fun... I can't recall ever hearing somebody who longed for cheap ceramic dielectric sound.

JR

I tried measuring frequency response, which was identical, and THD, which was higher by a fraction of a percentage in the low frequencies for the older and bigger caps.

In the low pass filter, the bigger caps seemed to be darker and with more bottom by a hair. More of a textural difference than anything. I wonder if there is some hysterisis going on there.

next up will be another VCF that originally had polystyrene selected caps. Need to order.

the SMT caps made the filter brighter overall.

 

[soapfoot]

It'd be interesting to know the class of the SMT caps.

If they were C0G/NP0, they would have been much more temperature-stable, and their value would've stayed close to "what's written on the tin" regardless of operating temperature.

Class 2 ceramics can radically change value above or below the expected operating temperature, as much as +20% or -80% for typical parts. This is a pretty serious swing, which is part of why you don't see these parts used in audio circuits as signal coupling caps too terribly often (they're widely used for power supply decoupling, though).

Class 1 ceramics also might have exhibited less harmonic distortion than Class 2 ceramics, which (depending on their use in circuit) could have manifested audibly in some way.

Without some controlled tests, it'd be hard to say exactly what's going on, but just for the sake of putting it out there... it's almost definitely not a case of "surface mount = bad."

 

[Script]

And it's most certainly also not about that one magic component that makes all the difference. Way too many variables in entire circuit.

Yeah, obsession with sound, that's what synths.and tone modulation is about And the filter, well, tube rolling guitarists or sax players obsessing over their mouthpieces and filing their reeds ? Sure why not.

Makes me wonder how two 150p mixed with two other values (160p or 75p) would sound -- welcome to circuit bending.