The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
SP12 Filter control mod!!
Old 10th July 2019
  #31
Gear Head
 
B.Minor's Avatar
 

SP-12 FILTER CONTROL MOD - PART II:

FILTER CONTROLS EXPLAINED

Now, before we start with the technical part, I’d like to add some explanations on the general SSM2044 filter behavior. This might be useful in order to know the background for some of my technical decisions related to the final design mentioned in the last part following in the next days.

A short description on the filter controls:

When it comes to filter adjustment and the interworking between the planned controls in conjunction with the described SSM2044 frequency modulation, the amount of audible sound will mainly depend on the current shape of the frequency "ENVELOPE" curve and the amount of any applied "CUTOFF" frequency. For example, when turning down the "ENVELOPE" knob fully counterclockwise (CCW), its shape will disappear (become totally flat) so that any applied filter frequency (adjusted via the "CUTOFF" knob) will not change anymore over time during sample playback; the time for filter closure will be that fast already at sample start so that you cannot hear anymore how quick filters are closing. Let's figure out some more extremes first.

Having both, the "ENVELOPE" and the "CUTOFF" controls fully turned down simultaneously (hard left), you might not even be able to hear the original sample sound anymore, as filters are now fully blocking any audio, preventing it from appearing at the output. If you now apply some "RESONANCE" amount (sometimes also called "Q") – let’s say halfway turned up – and slowly start raising the "CUTOFF" amount by turning its related knob more and more to the right (clockwise - CW), in fact you are changing the “fixed” filter frequency. As an audible result, your sample sounds will receive a kind of noise decay like a "steel effect" whereas the effect doesn’t change its pitch over time (remember, “ENVELOPE” is still fully turned down). While the pitch for that effect is now solely depending on the current frequency "CUTOFF" setting, the “steel effect” intensity only depends on the current "RESONANCE" setting.

Now, if also slowly raising the "ENVELOPE" setting from left position more and more towards the right side, you can suddenly hear how frequency will start to change over time. In case "RESONANCE" is still turned up halfway as adjusted before, you will now hear a kind of “twitter effect” whereas its pitch changes over time as well. While still being close to the “ENVELOPE” knob’s CCW end, any audible frequency changes over time will be executed very fast (slim shape) compared to results close to the knob’s CW end where the frequency curve will open up more and more (wide envelope) until the filter stays permanently fully open (hard right). In fact, when both, "ENVELOPE" and "CUTOFF" are fully turned to the right, you can be sure that no filtering is taking place anymore; you will get a totally clean and unfiltered sample output signal again, even if you might still have some "RESONANCE" setting applied. In fact, this will have the same effect as if you would have “bypassed” the whole SSM filters (e.g. via previously explained “bypass filter switches”).

However, please note that the amount of any applied "RESPONANCE" setting is crucial and has radical impact on the original sample's output level. Whenever "RESONANCE" is fully turned down (hard left = off), the original sample's output level is retained to its original loudness level; its audible portion is only depending on the current "ENVELOPE" and "CUTOFF" settings, defining the intensity and length of any filtering effect. In case the "RESONANCE" feature is slowly brought into game by turning the related knob slowly to the right, the original sample output level will always start to drop significantly while the "RESONANCE" effect grows heavily instead. Pushing this effect to its limit can get very annoying compared to the original sample's loudness level. Unfortunately there's no way to overcome that behavior unless the upper “RESONANCE” range is somehow limited on the potentiometer at a certain level (which I will in fact do in my personal design). Unfortunately the general filter characteristics just offer a kind of "trade-off" between original sample output level and the achievable maximum "RESONANCE" effect; that drawback is even manifested in the SSM2044 data sheet which still exists online. That said, the "RESONANCE" control should always be treated with most care. If you don't use it, turn it fully down again in order to retrieve the maximum sample output level.

Summary:

To recap, the use of the three controls (per filter) need to be carefully aligned to each other in order to get the desired effect; natively a change on one single knob immediately may influence the possible headroom, intensity or even the remaining range subsequently available to other knobs – you must consider them all as a set of interactive controls which always depend on each other. But don’t worry, after playing around some time with the three filter settings you will soon figure out how the SSM filter design is supposed to work in general.

NoizeGearIch already covered the general control interworking very good in his introduction video, so please refer to it here again.

In the attachment of this part you find some more photos, this time related to the “pre-requisites” mentioned in part I of this tutorial already. In particular, the rectifier on the backpanel was moved to another spot (between power regulator and fuse holder in order to provide the necessary space for a “2x8 control matrix” for which then related holes can be drilled accordingly (just on top of outputs #1 to #8).

To be continued…
(I know – you are already waiting for the final filter control design…
It will follow in the next and final chapter – so stay tuned one more time…)
Attached Thumbnails
SP12 Filter control mod!!-04-rectifier-shift-.jpg   SP12 Filter control mod!!-05-rectifier-shift-b.jpg   SP12 Filter control mod!!-06-rectifier-shift-c.jpg   SP12 Filter control mod!!-07-backpanel-mounted.jpg  
Old 10th July 2019
  #32
Lives for gear
 
Sam an huel's Avatar
Waah many many thanks for sharing your work.
Old 11th July 2019
  #33
Gear Head
 
B.Minor's Avatar
 

FILTER CONTROL MOD - PART III:

FILTER CONTROLS DESIGNED

After introducing the basic SSM2044 filter information and the way the controls work in general, let’s finally proceed to the technical details. I guess that’s what you have been waiting for already. Here's the same obligatory disclaimer as applied to all of my previously published SP-12 mods:

Disclaimer:

Please note that any modifications on your SP-12 should only be performed by people who have technical skills and exactly know what they are doing. I'll accept no liability for any errors in my related descriptions, schemes, drawings or tables, and I'll accept no liability for and damage, consequential damage and accidents caused by the usage of this provided information. Therefore, please apply and use it at your own risk; please don't blame me if you destroy your gear by not having taken enough care. The mod has worked out very well for me; however, that doesn't mean that you will get happy with it as well. You have been warned! ...

How to interpret the instructions:

Please note that for simplicity reasons all following described modifications generally apply to both SSM2044 filter chips (applicable to output channels #1 and #2. Of course you may want to implement the mod to one of those two available dynamic filters only, but I wouldn't see any reason why to do so, especially if you already decided to implement the mod at all. The circuitry schematics (provided in an attached PDF file below) along with a special switch connectivity table will include useful comprised information applicable to both filters along with circuitry details differing from each other, e.g. resistor numbering, position on the board (left/right resistor end to be de-soldered), and so on. For example, the information “R125/R97” refers to “R125” being applicable to SSM #1 circuitry while “R97” is valid for the second SSM #2 circuitry. Both circuits are fully identical from a technical point of view. Take care which circuit you are currently working on and if you have picked the correct components and soldering spots in such a case (left/right ends).

Please also note that the two introduced “triple on/on switches”, one for each SSM filter, will have to switch all three modification paths (for "ENVELOPE", "CUTOFF" and "RESONANCE" controls) simultaneously in order to work properly in conjunction with reverting back to the original design and to obtain a clean separation between SP-12's "original" and the new "customized" filter design options. Anything “in-between” might not work as expected and may be realized on your own risk only.

Let's start with the first (and easiest) of the three filter modifications, the implementation of the "ENVELOPE" control:

As mentioned before, the original SP-12 design just provides a “fixed” filter setting, applicable to both dynamic SSM2044 filter chipsets. As a result of such a "default design", the frequency "ENVELOPE" is always providing the same "standardized" frequency shape/curve as designed by E-mu.

However, by de-soldering the proper end of resistor "R125/R97" from the SP-12 mainboard and putting the first third of our "triple on/on toggle switch" (SW 1) between those exposed points, we are now able to replace the existing 15k resistor by a linear 25k potentiometer in order to achieve adjustment across the full "ENVELOPE" range available, reaching from a slim curve (left/counterclockwise - CCW) to a wide curve (right/clockwise - CCW) as described before. While 0 Ohms (pot is shorted) represents the described hard-left setting, the pot's 25k adjustment means fully cranked up to the right. Therefore, please refer to the related circuit scheme (as mentioned before, this is provided below) when identifying the proper CCW/CW connectivity positions in conjunction with the middle terminal in order to get the correct rotation direction. In addition – and like explained in all following steps from now on as well – switch position "a" will refer to the "original" filter design while position "b" represents the "customized" solution. It can be switched between both options at any time without losing any customized settings.

Next, let's move on to the second filter modification, the implementation of the "CUTOFF" control:

Just by looking at the original SP-12 filter design, you can see that per SSM2044 circuitry there is an internal trimmer "RT6/RT5" available on the mainboard, pre-defining a certain amount of frequency "CUTOFF" to be offered as per "default design". As it was never intended by E-mu to make any frequency control accessible by SP-12 owners, such an internal trimmers was exclusively used to calibrate the filter in order to reflect a pre-defined factory frequency (1.3kHz) while being in self-oscillation (for more details please refer to the "RESONANCE" explanations in the next chapter below). That original factory "OFFSET" adjustment ensured that all SSM filters finally showed the same standardized frequency characteristics which were not supposed to be altered by anyone.

However, there we come into play with our related "CUTOFF" mod. After de-soldering the proper ends of the two resistors "R112/R102" and "R104/R89" from the SP-12 mainboard (please refer again to the schematics provided in the attachment), we are now able to use those exposed points in order to disconnect the existing circuitry to that built-in trimmer path and replace it with our own new components instead. Of course, this is done again by additionally putting the second third (SW 2) of our newly introduced "triple on/on toggle switch" in between. In our case the new circuitry is made of a 22k serial resistor (please don’t forget that one, as it is crucial) followed by a linear 50k potentiometer, leading to -15V. There are several good spots on the mainboard from where such main voltages can be obtained. For example, there are two long traces available just below the SSM chip circuitry, providing plated-through soldering points; one rail for +15V, the other one for -15V. Just check them both out, but don’t mix these two up!

Now, by adjusting the related potentiometer knob, the full frequency "CUTOFF" range, reaching from low frequencies (left/CCW) to high frequencies (right/CW) can be achieved. Again, please watch the circuit scheme with the potentiometer's CCW/CW and its middle connection in order to get the described direction and behavior. This also applies to the introduction of the second switch part where position "a" again refers to the "original" design and "b" represents the "customized" design option.

Finally, let's approach the last of the three filter modifications, the implementation of the "RESONANCE" control:

As you may have noticed as well, E-mu did not include any "RESONANCE" (or sometimes also called “Q”) functionality in their original SP-12 design. On one side of the existing circuitry there is just a 4k7 resistor connected to SSM2044's related chip input pin #2 while on the other side we can just spot a test point (TP) jumper which was formerly bridged directly to +15V. However, this temporary configuration was previously used for the 1.3kHz oscillation frequency calibration during the manufacturing process only; the related jumper was intentionally cut before shipping the SP-12 units out to customers. In other words, when receiving a "virgin" SP-12, no control voltage (CV) is usually applied to the dynamic filter's Q input and therefore no "RESONANCE" feature was ever supposed to be available by default.

We can easily change that situation by using exactly these two (now separated) ends of the cut TP bridge to insert our additional "RESONANCE" circuitry. In fact, just bridging the two mentioned TP points again – but this time making that connectivity "switchable" via the last third of our "triple on/on toggle switch" while using a potentiometer providing a very high value (like 500k or even 1M) – would already do the trick. However, it's not optimal in case you are expecting to get the right "feel" when operating the "RESONANCE" knob. You will also run into an annoying overload of that “RESONANCE” as explained earlier, if not some precautions are taken. In order to apply a more sophisticated solution, we just have to spot another necessary connection point on the SP-12 mainboard we'll subsequently need, namely the analog ground (AGND) test point we'll be connecting to later on. There happens to be a soldered bridge available for AGND on the mainboard which is perfectly suitable for our purpose.

But before continuing, let’s look at some possible approaches first on how to finally realize the desired "RESONANCE" or "Q" circuitry to be inserted. I took a closer look at the original SSM2044 data sheet showing the recommended way any Q CV may be applied to the SSM filter chip. Aside from specific resistor values proposed there, e.g. in order to discard some portions of the lower Q curve, the chip manufacturer also recommends to use a 5k potentiometer as a voltage divider for Q control. In fact, very little current but well-defined voltage levels are essential to make the Q control work properly and to offer the proper Q range to be adjusted. In particular, it might be useful to make sure that a fully cranked-up Q effect doesn't get too annoying compared to the original input's signal level. As mentioned before, the SSM2044 filter design unfortunately works in such a way that the original signal level drops drastically the more the Q intensity is increased. As you can see from the SSM2044 datasheet as well, the Q feedback slowly kicks in at the CCW pot side, increases moderately within the center pot rotation range to be finally climbing up very fast to its maximum oscillation point as the pot almost reaches its CW end position. In order to compensate for that special filter behavior, there is a recommendation in the datasheet to use a "reverse-audio" pot for any Q control.

Now, considering all those recommendations, I started creating my personal design by mainly keeping the "voltage divider" principle described in the SSM datasheet but making some personal changes in order to limit the range. Of course, I applied at least a serial resistor (in my case 8k2) between the +15V "TP13/TP11" and the followed potentiometer’s CW side in order to limit the maximum current entering the pot; however, I refrained from using a serial resistor on the other pot end towards AGND, making sure the original signal level would be fully returning in case "RESONANCE" is completely turned down (= off) again. We also know that most of the available audio or reverse-audio pots on the market can be very expensive and also don't reproduce exact exponential curves due to manufacturing reasons; in fact, they don't even provide very good logarithmic or reverse-logarithmic approximations. Therefore, I decided to use a simple linear potentiometer instead, with a high 250k value as the base, in conjunction with a so-called "taper resistor" providing a 2k2 value in my case, applied between the CW end and the middle terminal of the pot. This simple but very effective trick could be varied by using different taper resistor and/or pot values, allowing one to shape any desired resulting control curve in a much more precise way than it could ever be achieved by just using any available reverse-audio pot. Finally, the middle terminal of the potentiometer had to get an additional serial resistor for limiting the range – in my case the choice was 15k – in order to get connected to the existing 4k7 resistor "R111/R116", finally leading to the SSM chip's Q input (SSM pin #2). By applying all mentioned values above, I could achieve a very useful emulation of an extremely reverse-logarithmic voltage curve (factor B=113) which perfectly rendered the oppositely directed original Q filter behavior into a very usable and smooth Q control curve. To my personal taste, it just offered the proper range limits and "the right feel" I was looking for. In my case, any cancellation of the lower Q curve portion was unnecessary. Please refer to the circuitry scheme for more details on how any components are connected to each other.

In general, no "RESONANCE" effect will be audible at the pot's CCW start position (hard left/CCW), then the effect will drop in, growing slowly the more you will turn the knob to the right. Shortly before the full end position (hard right/CW) is reached, the "RESONANCE" effect also reaches maximum intensity before finally ending up in a permanent tone, representing the beginning of self-oscillation.

Please experiment for yourself using different resistor/pot values if you think you need to achieve alternate results which may be tailored to your personal needs and/or expectations in a better way, e.g. pushing loudness or range limits to another direction. However, whenever using the explained "taper resistor" approach, please note that it's very essential to choose a linear pot type in such a case. The proposed components and their values are just recommendations which worked out perfectly for me. Of course, you could push the possible Q design to its technical limits, e.g. by introducing temperature-compensating resistors (so-called Tempcos) which would allow you to trigger the oscillating resonance feedback in a correct tonal way. However, that would mean to introduce additional components, control voltages or additional internal reference calibration points which would only make the whole thing more complicated – and in my opinion wouldn't really pay off in conjunction with the possibilities the SP-12 offers from scratch.

Final words:

That concludes the information on my personal filter control approach. In fact, it’s not very complicated, but you should take care anyway.

In case you have any further questions, just ask. Otherwise please refer to the attached PDF file where related schematics, tables or other useful information is summarized. I think all explained points are covered there.

Please note that I ain’t gonna upload any related demonstration videos or whatsoever; I’m just adding some more photos of the cables connected to the motherboard and the separate “homebrew” PCB where those cables are leading to. The big advantage of such a design is that you can mount/dismount the control switch/trimmer PCB from the SP-12 backpanel in one piece if required.

I’d appreciate your feedback in case you have re-created this mod as well – or if you have developed any other variant.

One more time here are here the links to my other SP-12-related modifications:

Display Mod
Filter Switch Mod
Filter Control Mod (this thread)

Have fun!
Cheers!
Attached Thumbnails
SP12 Filter control mod!!-08-mainboard-modified.jpg   SP12 Filter control mod!!-09-filter-switch-cables-other-pcb-connectors-top.jpg   SP12 Filter control mod!!-10-filter-switch-cables-other-pcb-connectors-side.jpg   SP12 Filter control mod!!-11-pcb-connectors-cables-top.jpg   SP12 Filter control mod!!-12-pcb-connectors-cables-rear.jpg  

SP12 Filter control mod!!-13-pcb-mounted.jpg  
Attached Files
File Type: pdf SP-12 Filter Control Mod.pdf (163.6 KB, 9 views)
Old 12th July 2019
  #34
Gear Addict
 
tiger001's Avatar
or you could buy a / 2 doepfer A105 (SSM2044) filter modules which give you exactly the same controls, they even can be (real time) controlled with CV
http://www.doepfer.de/a105.htm
the impact on the sound is large and remiscent of that of SP12/00, EII etc
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump