Archive for the ‘SDIY’ Category

SEM Modifications – LFO Section

September 30, 2011 Leave a comment

Location of LFO Signals

As part of some design decision ages ago, Oberheim opted to only provide a sine wave output for the LFO section even though the triangle and square waves are available in the circuitry. Bringing both of these out is easy and well worth doing. The square wave is a little more work as it is present as a timing element in the circuit and its voltage is well beyond usable ranges as it swings pretty much completely between the +/18.5v power. For this, we need to construct a simple divider. Take a look at the picture of the component side of the SEM where I’ve marked the locations you need to bring out. In both cases, they are on pin 6 of the 741 ICs.

Attached Triangle and Square Wave Output Mods

The triangle is very straight forward, all you need to do is attach a 1k resistor to pin 6. The square requires you to make a small voltage divider between pin 6 and ground to bring it down to standard modular voltage levels. If you are unfamiliar with a voltage divider, you should make sure conversant now. You see these all over schematics and you also will regularly need these in your mods. Wikipedia has a good entry on the topic and I use this handy calculator for determining the correct resistor values. There are many such pages out there. In this case, you need a 1k8 resistor attached to pin 6 of the IC and a 1k resistor attached to ground. You then need to attach a wire to both of these resistors. I started by twisting the resistor leads together with the wire and then soldered and applied heat shrink to the join. I then used wire to connect the two resistors and used heat shrink to make sure they would be no shorts. This is the same method I’ve used throughout on all the breakout mods. It should be clear from the image.

Categories: SDIY, Synths Tags:

SEM – The Machined Panel Arrives!

September 29, 2011 Leave a comment

SEM Panel Ready for Wiring

If my work space is any indication, the hardest part of any project is the panel and enclosure. I have all kinds of fully stuffed PCBs sitting around that just need a proper panel or housing and they would be finished. I’m not even talking about one or two projects, a quick look around reveals 4 such PCBs and I know for a fact there are at least as many in various storage boxes hiding from my direct glare. I used Frontpanel Express in the past but while the product was good, I found using their software left something to be desired. in particular, if you wanted a dial gauge of some sort, getting it to line up properly was a frustrating exercise in 1/128″ increments. Even then it seldom was perfect. Equally frustrating was that the cost was just high enough to be on the edge of tolerable. Fortunately, I have found a better way and projects are now slowly ending up behind panels again much more quickly.

The key is to use HPGL graphic files for you text, gauges, etc. and import these into the Frontpanel Express software. The cost is considerably less and you have much greater control over graphic elements. This is probably no great revelation to any of you who have used Frontpanel Express. I had heard about this from the first time a made a panel. Unfortunately, HPGL is not the friendliest or widely supported format out there. I experimented with multiple methods on the web: using old HP plotter drivers, printing to file, and reimporting; saving to SVG or other vector files and uses translation programs, etc. None of them worked even close to satisfactorily. Typically one of two things occurred.  Either the output vectors were choppy because there were not enough points stored for the curve or the scale was off.  The later was particularly frustrating because a couple of the methods to get HPGL eventually yielded smooth outputs, but I could never get the scale right.  It always seemed off by just a slight amount so the gauges were always ever so slightly off.

The absolute best solution I’ve found is Corel Draw.  It supports HPGL natively, outputs perfectly, and is really easy to use and in many ways optimized for this kind of technical layout.  I also have Adobe Illustrator I still always use Corel for my panels (but Illustrator for pretty much anything else).  I’ve also heard that Inkscape supports HPGL but I’ve not tried it so I do not know.  In Corel, the trick is to scale everything up so that you get sufficient output points for all the curves to be smooth.  Just to keep the math easy, I scale everything up 10x larger than my desired final output.  Then I import the HPGL into Frontpanel Express at 10%.  Unlike my previous attempts, the scale is dead on perfect. I still use Frontpanel Express’ software to specify the milling of holes etc. but Corel does all the labeling, graphic, and gauge type stuff.  One really great thing about Frontpanel Express is that it is trivial to have them mill “d-holes” on your panel.  If you are using banana jacks, spend the extra dime for these.  You jacks will never turn on the panel again.  I even added Chinese labels to my panel as graphical elements and they came out great!

Categories: SDIY, Synths Tags:

SEM Modifications – VCF/Filter Section

September 29, 2011 3 comments

The filter section needs less modification for my project than both the oscillators (because there is only one VCF) and the LFO section.  I love the SEM filter as it is and as such, I really have no desire to upgrade any of the components or make any modifications that might alter the sound in any way.  Like the oscillator section, I do want more modulation sources than the SEM provides with the stock config on the rear Molex connectors.  The two CV inputs to the VCF are found on pins H1 and H2 while the EXT modulation source is on H3.  Just like the oscillators, I want the inputs on one of the pots of the front panel to be switchable from going straight to the VCF modulation or routed to the EXT modulation input.  See the post on the oscillators for more about how to make the leads, soldering, et.c

The only other wrinkle is that I have also added a waveshaping circuit that takes the outputs from the oscillators and outputs both triangle and sine waves.  I’ll have more information on this circuit in a future post.  I want these to have their own dedicated volume control and feed directly into the filter just like the pulse and saw waveforms.  I love the alpha pots with the push pull switches like the ones I used for the oscillator section to route the cv to either the modulation or EXT modulation inputs.  I use these all the time and they will be used on the panel to control the volume of the triangle/sine waveforms and also to select which is being feed to the filter for each oscillator.  Unfortunately, these pots only come in a linear configuration so I’ll have a later post showing how to make them behave with an audio taper by adding a resistor to change the curve of th pot response.  Looking over the schematics and probing around a bit on the SEM shows that there is a group of 100k resistors on the upper left part of the board where we will need to add a 100k resistor in parallel with the ones already on the board to add the additional CV input to the VCF.  Towards the middle-left part of the board, there is a 220k resistor where we will need to add 2 additional resistors for the sub-oscillator inputs.

Categories: SDIY, Synths Tags: ,

Korg SDD-3300 Triple Delay – Display Upgrade

August 12, 2011 4 comments

I love my Korg Triple Delay.  I love it so much I actually have three of them.  You might not need three, but you really should own at least one.  I’ve had one since the early 90’s, I bought a busted one on eBay for parts that turned out to have a trivial issue, and the third on eBay because it was just too good of a deal to let someone else get it.  The last one was crazy impulsive but there you go.  There are only three things about the 3300 I don’t like.  Everything else about it is magic. 

If you are unfamiliar with the unit, it is a three input, three output, three delay unit monster.  Each delay unit’s time can be up to 500ms.  You can route any of the inputs 1,2,3 into any of the three delay A,B,C units in any amount you want.  Additionally, you can route the ouput from any delay to any of the three outputs or back to any of the three delay units.  It also sports two LFOs which can be set to a locked phase relationship which makes for great chorus effects.  If you like your feedback like me, this unit is like nothing else.  It is an early digital delay with a ton of character and a great sound to back up the impressive feature set.  Unfortunately, like I said there are three things I don’t like about it.  First, the output is rather low.  Perhaps some reader who is smarter than me will look over the Service Manual and tell me how to boost the output.  Secondly, the units do have a good bit of transformer hum/whine.  Lastly, the displays aren’t great and pretty much every unit I’ve seen the backlights are completely dead.  I only have a solution to the last of these three complaints – the display.

First off, there are tons of drop in replacements for old LCDs with EL backlights on the market.  I’ve been slowly upgrading my displays as I work on gear.  This is not without pitfalls as the pile of VFD and LED displays on my desk left over from this project will attest.  There are two main problems; first, just because a display says it is compatible with the chipset on your old display, doesn’t mean that it is.  In the case of the 3300, I had several LED backlit displays from two different manufacturers that did not work.  I would get the top line of text but not the bottom.  It turns out that the newer chips had tighter and faster timing requirements than the display in the 3300 and thus the problem.  I also had a couple of VFD displays (my favorites) that simply would not power up or display anything at all.  I suspect there was not enough available current in the 5v line on the SDD to drive them properly.  Again, it also could have been a chipset compatibility problem.  I did finally find a display that works.  It has two things going for it (besides the fact that it works).  First, it is an LED backlit model so it should have an extremely long life and secondly it was really cheap.  VFD displays can run up to $60 for the nice Noritake ones and having 3 units to repair I wasn’t looking forward to the bill, especially if they didn’t work.  The displays I’m using are from 411 Technology Systems part number SSC2F20DLNB-D.  They are a steal at $11.49 each.  There are a couple of problems though.  Being LED backlit, they are thicker than the original display and the front panel of the SDD won’t attach with them mounted the way the old display did.  They also have separate 0v and 5v power inputs for the LED backlight which are not jumpered to the 0v and 5v inputs on pin 1&2 of the standard display.  Neither of these is a deal breakers but it means you should expect a frustrating hour cursing at how your fingers are too large (that part only lasts about 10 minutes but it will piss you off).  I recommend you pour yourself a beer.

Replacement LCD with jumpers and current limiting resistor

The first step is easy.  Open the case of the SDD and remove the existing display.  Carefully remove the wire bundle for the display from the main board but gently but firmly pulling on thewire housing (not the wires themselves) until it comes out.  Pin 1 is the one farthest from the front the SDD and in all three of my units it was a brown wire.  The connector is keyed so you can’t put it in backwards but you need to keep track of pin1 so you can wire your new display.  There is also a pair of larger wires that provide voltage to the EL inverter.  These end in a connector on the mainboard.  Unplug this with the display as you won’t need it anymore.  You can cut the wire bundle off the old display and then restrip and tin the wires or unsolder the existing ones from the display.  It’s up to you.  If you flip the display over, you will see pin 1 and 2 labeled on the back.  Resolder the wires from the old display to the new making sure you keep the pins the same.  The odd numbered pins are closest to the display as you look at it from the front and the even pins are on the outside.  It really would be best to install a pin header here and crimp terminals onto the wires and put them in a standard housing.  It really would be.  You can see from my pics though that I didn’t do it.  That’s because these displays arrived yesterday and I was so surprised that they actually worked, I could not wait until next week to place an order and wait on UPS to get them finished with so I just plowed ahead and soldered them.  Once you have resoldered all the wires (or populated a crimp housing which would be better) you will see that there are more pin positions on the new display (16) than the old display (14).  This is because the new display has two more pins for a 0v and 5v supply for the LED backlight.  The easiest thing to do is just jumper the pins by connecting pin 15 to pin 0 and pin 16 to pin 2.  This will provide the voltage you need to drive the backlight.  If you do this you will likely significantly shorten the life of the backlight.  You should install a current limiting resistor between pin 15 and 0 to lower the overall brightness of the display.  Use a resistor between 10 ohm or 20 ohm.  I found the 20 ohm was a good pick and the display was still really bright.  You could go even higher if you want your display darker.  Without a resistor, the display is extremely bright.  You can see how I attached the resistor in the photo.  I always cover resistors and components midwire like this with heat shrink so it doesn’t short out to anything.

Now is where the cursing comes in.  The new display has the same hole configuration as the old display.  However, if you mount it in the same way the old display was, you won’t be able to put the front back on the unit.  You need to mount the display *behind* the brackets and use nuts to hold it in place.  To make matters worse, you can’t actually get the current display behind the brackets without a modification of some sort.  I opted to bend the bottom left post inward (see picture) which gave me enough room to slide the display behind the remaining three brackets.  Alternately, you could completely disassemble the unit and then you could have the room to mount the display to all four brackets and also bypass the cursing that is about to happen.  However, three brackets hold the display fine and probably save you an hour plus of disassembly and reassembly.  I opted for the three bracket route.  There are also 3 capacitors behind the display that might give you problems with having the display sit flat behind the brackets.  I was able to gentle press them flat to the PCB with no problems on the three units I upgraded.  If it is not possible to get them flat enough, you will need to unsolder them and remount them on the other side of the board.  Just make sure you get the polarity right if you do this.  Also, If you have to go this far, you probably best ought just go ahead and disassemble the unit.

Things in the way of your new display

Now, to mount the display is frustrating.  The screws that came with the unit are long enough but you will need to get three nuts to hold the display against the brackets.  Take one of the screws with you to your local hardware store.  Not Home Depot as they are evil and will make you purchase a box of these and you only need three.  Go to your good old local hardware store that has a decent selection of hardware in bins and buy four of the correct size nuts.  Unfortunately, I’m old and can’t remember the exact size I bought so you are on your own.  I said to buy four because you are likely to lose one during this process and you really don’t want to go back to the hardware store a second time.  You only need three if you are foolish, brave, or talented.  I am none of these things – I lost two to the darkest, most distant space under my workbench.  Getting these nuts on tight on the back of the display in a tight space that you can’t reach well is what causes the cursing.  You need these to not only attach, but to attach tightly in order to hold the display in place.  The problem is the brackets themselves are threaded and that makes it difficult to get the nut taken up tightly unless you start it as soon as the screw pokes through the back and don’t let it turn while tightening.  I first screwed the screws all the way down from the front so that the display was lined up with the screws through the mounting holes.  Next I took a small allen wrench and applied a small piece of double sided carpet tape to the smaller bent part.  This holds the nut in place and gives you something to lift up on to make sure the display is tight against the brackets.  Back one screw out enough that you can use it to get the nut in position.  While applying light pressure on the nut to keep it on the screw shaft, back off the screw so that the nut is all the way tight against the bracket.  Now, while applying enough pressure to the nut so that it doesn’t turn, tighten the screw all the way down again pulling the nut tight to the bracket.  Repeat this two other times.  I also tried putting a piece of the carpet tape on my finger to hold the nut and that worked somewhat but a couple of the brackets are just a little too out of the way.  I needed my allen wrench tool on one or two for each unit I did.

Too Much Awesomeness: SDD-3300 x 3!

Now all that you have to do is get the whole thing put back together and back in the rack.  You might want to check the screws you took out of the unit to make sure none of them are stripped or wrong before you go to the hardware store so you can get replacements if you need to.  One of my units was owned by a guitarist and he apparently had lost the screws that went to the rack ears and decided that big #8 woodscrews were close enough to the metric machine screws that are supposed to be used.  I’m not sure how he got this to work but it certainly took a lot of effort to force the woodscrews into the case.  I had to get out my tap and die set to rethread the hole and then replaced them with the correct screws.  What is it with guitarists?  My buddy Brooks and I have long said that we need to release a high-end effect rack that comes in a limited guitarist’s edition that has the rack ears hack sawn off and comes with an empty paint can to put the unit on.  If you are a guitarist and think I’m out of line as you know all your appropriate fasteners and their correct uses, I would suggest that you are playing the wrong instrument.  As to the end result, as you can see they look absolutely great!

SEM Modifications – VCO Section

August 3, 2011 Leave a comment
SEM VCO Section – Top of PCB

A quick look at my proposed panel should make you realize that there are more connections than the SEM accomidates on the molex connectors and features available than are not  on the SEM.  There are a total of 3 frequency CV inputs on the panel and only two on the SEM (not counting the EXT input) and there is no pulse width modulation input on the SEM.  The addition of sine and triangle waveform outputs will be handled in a later post, this post concentrates on how and where to add the additional cv inputs – the frequency and pw mod.

The SEM has provision for two frequency mod inputs and an input that routes to the EXT of both VCO1 and VCO2 on its molex connectors.  The frequency mod inputs for VCO1 are B1 and B2 with the VCO1 EXT source on A1.  For VCO2, the frequency inputs are D1 and D3 with EXT on E1.  A quick look at the SEM VCO schematic(taken form Kevin Lightner’s Synthfool site) clearly shows where the external modulation points enter the VCO.  If you look at the inputs B1,B2 on VCO1 and D1,D3 on VCO2, you see that they just run through a 100k resistor summing node before getting mixed together the 741 opamps A2 and A7.  All we need to do to add additional frequency inputs is attach additional 100k resistors in parallel to those already there.  You can add as many as you need – I needed only one additional.  I could have used the EXT input without having to add any more inputs, but I find that I generally want to have the mods active without having to engage the EXT.  In order to maximize the usability of the SEM though, the second CV input on the panel is routed through a pot which has an integrated push pull switch.  When the pot is pushed in the mod runs through my additional input.  When the pot is pulled out, it routes to the EXT mod location on the SEM front panel.  This gives me the best of both worlds.  Make sure you are attaching your additional resistors on the correct part of the PCB so that they are before the 741 input and in parallel with the other 100k input resistors.  Oberheim hand matched the 100k resistors on the SEM but I am not going to bother as I do not need my extra modulation input to track precisely as I intend to use the existing 100k matched set for v/oct inputs and my additional input with an attenuator for LFO modulation.  the original SEM needed to be able to precisely track incoming pitch CV as in a 2 voice there was a sequencer for one of the inputs and in the 4 and 8 voice there was a programmer that needed to have predictable inputs.  If you need this kind of precision on your additional inputs, you will need to hand match your 100k resistors to the ones on the board or just replace all of them with hand matched ones.

Mods all attached and completed for the VCOs

Resistors mid-wire with heat shrink

The pulse width modulation is a similar add-on.  If you follow the schematic to where the PW mod internally to the SEM comes in, you will see another summing opamp (A5 on VCO1,  A10 for VCO2).  This time, use a 47k resistor attached as part of the summing resistor nodes.  You can where all the mods need to attach in the picture.  I’ve shown the top of the board to give you some landmarks.  You can attach to the components on top of the board but I tend to instead make all my modifications to the bottom surface of the PCB for several reasons.  First, it is really easy to melt a nearby wire or worse a near-by component.  The underside of the board simply gives you more room to work.  Secondly, these types of mods are easily reversible, putting your SEM back to its stock factory state.  Having these on the bottom not only makes reversal easy, but also repairs and undoing them in the future if you should choose to do so for some reason.

There are a couple of ways to go about attaching the resistors.  First, locate where you are going to attach the resistors on the bottom of the board.  This is best done with a multimeter using the continuity function (where it beeps when you have a connection).  Put one probe on the top of the board in the locations shown in the picture and find a corresponding solder pad on the bottom.  Once you do this you will need to solder the new resistors to these locations.  I used to make a small loop in each end of the resistor and trim them close.  Then I would solder to the appropraite pad, then attach the wire.  I have found over the years that any movement in the wires tends to fatigue the resistor lead rapidly and you will end up with it breaking off.  Now, i still form the loops but instead I attach a short length of wire to one and and then solder it to the board.  This simple extra step provides some flexibility and will not break off the way soldering a resistor directly to the board will.  Remember, solder is a physical connection, not a mechanical connection and these sorts of mods are relatively fragile.  After you have made your wires with the appropriate resistors, make sure you use some shrink wrap to cover the resistor.  You don’t want these shorting out on you once you are finished.  I use clear shrink wrap so I can see the values of the components.  I’ve included a close-up picture of a resistor attached to the SEM with the wires attached so you can see.  I’ve never had one of these fail.

Categories: SDIY, Synths Tags:

SEM +/-15v Secondary Power

July 23, 2011 Leave a comment
+/-15v Aux Power Board

+/-15v Aux Power Board

My plans for the modifications to my SEMs require the addition of two pcbs.  One is the dual waveshaper board that will add sine and triangle output waves for the two SEM oscillators and the second is a CV processing voltage board that adds 8 inverting/attenuating controls as well as a pair of inverters for the envelopes to add negative going envelopes to the SEM.  I already have a 5u modular system and so all my DIY stuff is designed to run on+/-15v and tapping off the SEM after the power regulation to use its regulated +/-15v seemed more than a little inelegant even by my standards.  I really didn’t want to redesign the boards (although I had to call in massive help on the waveshapers – more on that later) so I decided to add a +/-15v regulator board to provide me the necessary power for the auxillary boards.

One thing about me that is fairly constant is that I’m willing to spend extra cash on a project if it is going to either save me time or likely be of higher quality.   This is true because if you are getting into DIY, it isn’t going to save you anything over purchasing a unit where economies of scale come into play.  A quick look through the Mouser, Digikey, Newark catalog should make it painfully obvious how these companies do not exist to service the likes of you and me where standard pricebreaks beigin in the 1,000s of peices.  Additionally, I despise  working or any kind of strip/proto board.  For my power needs, I noticed that Ken Stone has a power supply board , the CGS66, designed to use an AC wallwart at 18v AC and deliver regulated +/- 15v DC.  I ordered four of the boards and built them minus the AC filter stage so they really are just 15v regulators with standard amp type connectors for the remaining modules.  The SEM in my casework leaves enough room that these boards are mounted to the back panel.  You can see the parts I left of for the AC filtering in the photo on the right.  Works great and now I have power for the other boards.

Categories: SDIY, Synths Tags:

SEM Power Section

July 22, 2011 Leave a comment
SEM Power Section

Always, always take the time to replace the electrolytic capacitors in your vintage synths if you find yourself opening them up for a repair anyhow.  It is relatively cheap and your synth will generally sound and behave better after you have given it a fresh set.  Electrolytics have a limited lifespan and therefore *all* your vintage babies probably are due.  Go ahead and splurge on some nice high temp and high quality caps to ensure you don’t have to do this for another twenty years.  The SEM is particularly easy as there are only two electrolytics on the whole synth.  To open an SEM, there are three srews on the PCB to remove and then you have to pull the two boards apart.  I hate this part.  They are connected with a series of pins around the edges and it takes a good bit of force to get them seperated.  This is another bad design decision as there is no way to seperate the boards without flexing them at least a bit.  I have had solder connection go bad from this process so try and minimize the number of times you remove the PCB.  If you are following along at home, try and open it, make *all* the modifications I’m documenting and then close it up rather than pulling it apart for each one.  There are not many mods to do (at least here) and chasing down intermittent solder joints is not fun.

There is not much to the power section on the SEM since the main work was handled off the SEM and in my case in the external supply as you can see in the schematic.  I’ve pointed out the interesting bits in the pic.  I’ve had two occasions over the years where a malfunctioning SEM was due to one part in this section.  Note the white Molex power connector – this is one I had to repair as it had broken off.  If your SEM is acting at all flakey (or if it is completely dead), start by checking the voltages coming out the power section first.  If they are not correct, your problem is likely to be here.  Once when I was probing around with a 5v control voltage trying to find where to insert the pulse-width modulation mod into the VCO1 circuit, the probe slipped and shorted across something on the board.  My head was turned at the time looking at the scope so I didn’t see it.  The result was a non-functioning SEM.  The gate LED would flash constantly and there were some random gates going on but none of the voice sections were outputing anything correctly.  The second time was more horrific.  It was late and I was trying to troubleshoot a problem that had derailed my day and I accidentally reversed the power supply voltages.  This of course released the magic smoke in the circuit and the SEM was completely dead.  In both cases, the voltages coming out of the power section, particularly the negative side, were way off.  As all the parts are readily available from Mouser and cost a total of something like $5 I decided to just replace the whole power section’s ICs and caps.  In both cases it turned out to be the 2N3638A transistor and therefore an easy fix.  I am not sure if I just got lucky or if the design of the SEM is such that the transistor sacrifices itself in the case of shorts or other power problems. I would not replace anything besides the electrolytics as a matter of course, but if you are having issues, check the voltages in the supply section.  You might get lucky and have an easy fix!

Categories: SDIY, Synths Tags: