Modifications to the TTSH
There are numerous modifications you can do to the TTSH. Some will require modifications to the PCB, some will require modifications to the front panel. If you are considering any modification, you should thoroughly understand what is required and what you may gain. The table below only covers the 6 mods that I implemented and may provide additional insight before committing to them. Some, have additional costs for materials, etc.
|Modification||PCB and/or Panel mods||Additional materials||Complexity Rating|
|AC coupling||Cutting 4 traces on the PCB||4 film type capacitors||Pads exists on pcb for capacitors. Cutting traces between pads is fairly easy and can be done with an Exacto hobby knife.|
|External clock for electronic switch||Running a separate wire between resistors and Jack||6” wire||Really easy to implement. Cutting trace near jack is fairly easy and can be done with an Exacto hobby knife.|
|Gate Booster||Cut one trace on PCB and requires additional wire connections to jacks||This is a separate PCB with components and can be purchased as a separate kit||Fairly easy, and does not really risk damage to other components.|
|Oscillator Sync||Requires additional wire connections to Oscillator cores and modification of the main panel||This is a separate PCB with components and can be purchased as a separate kit||Highly complex and risky modification. It requires running shielded cable between the additional PCB and the oscillator cores. Cores can be damaged if wiring is incorrect. This modification also requires drilling holes into the front panel to fit switches.|
|Wave shaper||Requires additional wire connections to Oscillator cores and modification of the main panel||This is a separate PCB with components and can be purchased as a separate kit||Highly complex and risky modification. It requires running wires between the additional PCB and the oscillator cores. Cores can be damaged if wiring is incorrect. This modification also requires drilling holes into the front panel to fit
5 new jacks.
|Op Amp types||May require component changes on main board or module boards depending on Op amp selected.||Op Amp, capacitors, resistors as required||May degrade, modify or improve sound. Op amps should be tested individually, with sound examples recorded and compared.|
Documentation is readily available for all these mods and others through the online forum. I have shown the five I implemented here only to provide additional information. I highly recommend the first two on the list. They provide additional benefit, without additional cost or risk (if you are careful). The AC coupling capacitors also come included with the Synthcube kit. These two mods are also completed in the field for many Arp 2600s!!
The purpose of the gate booster is so you can use the TTSH with lower voltage Eurorack or modern digitally based sequencers, controllers, etc. The gate booster basically boosts lower (5v) gate signals to the required 10V levels. This is worth putting in, and can save headache down the road. I hate having to set my controllers to use a 10 or 12V gate, as this can damage digital synths. Having them all set to 5V means I don’t need to worry about where they are plugged in. Simple to build and really easy to install, I would highly recommend this mod.
Harder mods that require panel cutting or drilling
Of all the mods, the oscillator sync and wave shaper require the most rework. The image below shows the area where panel modification is required. For the wave shaper, you are adding 3 jacks to VCO1 (sub oscillator, Triangle and Sine waves) and 2 jacks for VCO 3 (Triangle and Sine waves). They are unlabelled in the photo. Also, is a hole or cutout for the sync switches (labelled as ‘sync’ in the image). One for VCO 2 and one for VCO3.
The oscillator synch requires some planning. This board has multiple configuration possibilities, but incorrect connection could damage components on the VCO cores. You also have to fit the switches into the front panel, and you have less than 12mm between the panel and the main pcb, so you need to ensure any switches A) fit in the space and B) Do not short out connections on the main pcb.
I decided, I wanted VCO 2 and/or VCO 3 to be synched (slaves) with VCO1 (master) when switches are placed in the ‘On’ position. The mod kit also came with toggle switches, but they were 13mm under the panel, so I ordered some sliders (eSwitch micro) from Digikey and used those instead. They also match the existing sliders on the panel.
Osc Sync Installed
You need to find a place to mount the PCB. I ended up mounting it to a plastic M3 threaded post replacing one of the speaker M3 nuts. The TTSH has multiple places all over the main board where power can be taken from. You will also be working with coax cable to eliminate the possibility of signal bleed. My personal opinion about this modification is “Meh”. It certainly synchronizes the oscillators perfectly (maybe too perfectly), but I prefer a subtle detuning between VCOs. This mod makes the oscillators sound a little wider, but not as phat as slightly detuned ones do. I am neutral on this mod, and would not implement it in an existing TTSH. But adding into a new build workflow is the only reason I did it.
The wave shaper is a great modification and basically makes VCO 1 and VCO 3 the same as VCO 2. In fact, the sine wave that this mod provides is a near perfect sine wave form. The original Arp design found on VCO2 only, is a very rough looking sine wave. Additionally, it gives a -1 octave sub oscillator to VCO 1, which is totally cool and really expands the capability of the synthesizer. This board also mounts to the back of the VCF and fits easily into an existing or finished Arp.
Using different Op Amps.
One of the things I planned out was to use various different Op Amps in the audio path. There is one that should be done, and that is to use a TL071 or TL081 in place of the LM301 for the Lag Processor. This is marked in the main board.
A couple of backgrounders here that need to be considered:
It was mentioned in the preceding build section that the original Blue Marvin and Grey Meanie (Steel cased Arps) utilized the MC1439G. These were high quality Op amps but had a output voltage that could not exceed 20Vpp. When the Arp exceeded that max voltage point, the Op amp could latch up. In many cases they would get destroyed and reliability became a factor for Arp. So, they modified existing Arps (when they broke) and started using the LM301 throughout the entire synth. But for specific LM301s that lay in the audio path, the schematics call the SL19986 spec to “select an op amp with lowest possible input offset voltage”
Back in the day, the internet was not a source for information. Most engineers and designers I know, used the Source Books or the Op Amp Cookbook for finding alternatives. According to the 1972 National Semiconductor Sourcebook, the functional alternative to the MC1439G was the LM301. The functional alternative for the MC1539G was the LM201 or LM101. The MC1539G is a higher temperature rated version of the MC1439G,
Price-wise, the MC1439G sold for about $6 each back in the 1970s, the LM301 was about $1.50. When you need 30 Op amps for each 2600, you can certainly see how the costs of Op Amps could effect selection of which one you could get away with. The fact is, overall the Arp 2600 was not a big selling synth. At over $2500 per unit, it was hard for an average musician to afford. However, they sold more 2600 units with the LM301 in the audio path than they did with the ones having a higher quality MC1439 based audio path. Some claim the early ones sounded better than the later units, but the Arp 2600 also seemed to build a mythical reputation over time. Has this really changed from today? I think the filter change did more to effect the sound of this synth than anything else.
Now, according to the 1973 Motorola Source book however, the functional alternative to the MC1439G was not the LM301, but the LM218/318 (with the corresponding alternate for the MC1539G being the LM118).
So, what is the difference between an LM101/201/301 series and an LM118/218/318 series Op Amp? Well, the price for one. The price for the LM318 was substantially higher that the LM301 and maybe that was enough for the Arp engineers to choose it instead.
With the exception of the AD711, all of the Op Amps below have a gain of about 40db @ 20KHz. I am also considering others like TLO71 or LF411, but not really sure if either would benefit the audio quality. Back in the day, the LM301 was the least expensive Op Amp of the ones listed (the current AD711 has been updated, and still expensive).
|Op Amp||Typ Input Offset Voltage||Max Input Offset Voltage||Slew Rate||Voltage Output @20,000 Hz|
|LM301||2.0 mV||7.5mV||10 V/uS||+/-12V @ 10pF comp capacitor|
|LM101/201||0.7||2.0||10 V/uS||+/-12V @ 10pF comp capacitor|
|AD711||0.2||0.5||18-20 V/uS||25 Vpp (+/-12.5V)|
|MC1539G||1.0||3.0||34V/uS||20Vpp (+/- 10V)|
Using this device would mean using the original OpAmp found in the first 2600 models. To use them, you would have to modify the compensation capacitor used in the LM301 and replace it with the capacitor/resistor pair (top drawing). Basically, reversing the suggested fix in the service manual. I actually have a bag of MC1539G Op Amps. These were all purchased for the LS Arp project, and come in the 8 pin metal can package. These are the only Op Amps that are no longer made. And they are becoming rare so you can expect counterfeits on eBay. There are some large US based electronic suppliers, that can provide these and they certify them as original. They will cost you though, anywhere between $10 and $25 each and will require you to still build a test rig and test each one to ensure it will work. One caveat: These devices are well over 30 years old and may not have the longevity they once had, under use. They also do not solve the peak to peak voltage problem. They clip or max out at 20Vp-p and many cases take the output over that.
The LM118/218 are almost impossible to find in a DIP8. All three (Lm118/218/318) do come in the SOIC 8 package, but only the LM318 is easy to find. These would also not use the 10 or 30 pF compensation capacitor, and would require a different resistor/capacitor (see spec sheet) to provide compensation.
The AD711 is expensive (about $5 per), if you purchase it in the SOIC 8 package. it is stupid price if you need the DIP8 or metal can. This does not need any compensation capacitors at all.
The LM101 only comes in the metal can and is now over $25 per unit. The LM201 in the metal can is also really stupid high. However, it also comes in a SOIC8 package for under $3 each! Furthermore, it is a direct replacement for the LM301 so it will use the same compensation capacitor. Better yet, is that the LM201 has a lower offset input voltage than the LM301 and with proper selection, these can be remarkably close to less than 0.1mV
Testing Various Op Amps
The problem with testing out various Op amps, is that I needed an easy way to swap them out, without having to unsolder capacitors and add resistors, etc. I wanted an easy way pull one out, and plug another in. There are also some tight places between some cards and the main board, so the height can’t exceed 12mm. Good news is that other than the MC1539G, all of the Op amps come in a SOIC8 package and by using an adapter board, I could build small assemblies, complete with the required compensation circuit. Example: for LM301s in the audio path, I just neglected to install the 10 or 30pF capacitors on the main board and installed them between the appropriate pins of the LM201s on the SOIC8 adapter cards instead. Alternatively, for the LM318s I used SMT capacitors and resistors soldered to the pins, then added a wire to connect them.
For the MC1539G can packages, I was still able to mount them to this adapter as they had holes located in the traces. You just have to be careful and ensure they match pin for pin, so pin one of the can matches pin one of the IC socket this will plug into.
Don’t go crazy here!!
Remember, sound improvements or detriments can be obvious or very subtle. And my preferences for how a synth sounds and my hearing levels are very different to yours! You need to test Op Amp replacements in one area at a time, and then test it thoroughly and through the entire range of functionality. Example, start with the 4012 filter first. This is the easiest and most noticeable. With the LM301, my 4012 would have distortion when I panned the frequency with all 3 VCOs at full tilt. In my build, I tested LM301. Selected LM201s, selected LM318s, AD711 and an MC1539G. Then move on to the VCA, Ring Mod and preamps.
I am still testing other Op Amps, but have so far found that in the 4012 filter, using a selected LM318 (IOV at 00.01mV) provided a very noticeable improvement to the sound of the filter. It sounds cleaner, wider and far less distortion when all 3 VCOs are fully up. Also, Op amps MAY change some of the calibration settings, so you should develop a workflow to install, test, check calibration, test, record sounds, repeat. Recording the output on an oscilloscope will also ensure your own biases are not interfering. it allows you to compare between Op Amps and physically see the waveform. Finally, compare Op Amps. against the LM301 only. The LM301 is the reference. In fact, try selecting Op Amps for lower IOV and see if there is a difference as well.
I really think everyone should experiment with using different Op amps. Some changes do nothing!! Some, even though it is a better Op Amp actually do narrow or flatten the sound, while some actually improve the sound without killing the grit.
If you wanted to build a TTSH, this may be your last chance. The market right now is being flooded with cheap, consumer versions of vintage analog mono synths and the TTSH is nearly 3 times the price in its unbuilt form. The consumer music industry, like other consumer industries, has to innovate to stay competitive. And sometimes innovation entails the reconstruction or reimagined versions of things that existed a generation ago. Eventually, all of these clones will end, as people move to the “Next big thing”. Having said that, don’t expect an original Arp 2600 to fall in price. It won’t happen as it will always be a collectors item. Fact: A 2020 Ford Mustang GT will blow the doors off anything made in the 1960s. But a 1969 Mustang Shelby will still fetch a million dollars regardless of that fact. Consumerism is, what it is. (as of this writing I just saw a 1976 Chevy Chevette sell for $10K, and they were $2600 brand new).
The Behringer 2600 and the Korg micro 2600 will no doubt be big selling synthesizers, but I guarantee you a TTSH will sound differently and has the ability to sound much better. You can modify the TTSH at will, no so easily with the Behringer or Korg.
So, 40 years after this adventure began……