Soldano Preamp MK2
I’ve been a fan of Soldano amps for many years. A decade ago I built a 5 watt version of the legendary SLO 100 and loved it but eventually sold it because of lack of space in my new apartment. Then I built a single channel version of the preamp, loved it but sold it after few years because it lacked the versatility of the 3 channel version. So now, many years later, I’m itching for SLO sounds again and I’ve set the goal to build the 3 channel version of the preamp in a compact format with simplified switching that doesn’t involve expensive opto-isolators. As a starting point, I used the official SLOCLONE schematic, took the preamp circuit from “Input” jack to “Send” jack and added two Volume controls at the end.
Layout
With this build I wanted to try a different point-to-point layout technique that is challenging and requires a lot of planning, but should yield very compact layout with short leads between the components and following the signal path closely to avoid any risk of excess noise or feedback. It uses point-to-point layout with components arranged in two planes that are parallel with each other and orthogonal to the tube sockets. One plane carries components that go directly from tube sockets to the ground bus (mostly cathode resistors and bypass capacitors, grid reference resistors) and the other plane contains the rest of components that go to tube sockets. The second plane uses a terminal strip that is used to connect the components to the controls that sit on the other side of the terminal strip. Additional components may run from the terminal strip to the ground bus (ground reference resistors, snubber capacitors, etc). I got inspiration for this layout after seeing a similar technique used in some 50s ad 60s NASA tube equipment in the Smithsonian National Air and Space Museum.
Mods
SLO preamp is almost perfect as it is, so this time I didn’t bother too much to modify it beyond recognition. As mentioned before, I changed the way channel switching is implemented, mainly to simplify the build and reduce the cost. I applied the same relay switching to Mark IIc+ build before and it’s been working for 4 years already with no issues.
I omitted the tone stack because I plan to use this preamp either with a graphic EQ or connected to the laptop where EQ tweaking possibilities are endless, but it’s simple enough to add the stock 3-band EQ before the volume pots. It would be slightly harder to do separate tone stacks for each channel, but it’s perfectly doable with another DPDT relay.
The second change I made is a switchable bright cap on the OD channel gain control, a.k.a. switchable Warren Hayes mod. That way you can go between the stock (bright) mode and the Warren Hayes mode (less harsh highs when the gain is rolled down).
One practical change I made is to wire the tube heaters in series and power them with 12.6V instead of the typical 6.3V parallel mode. The reason behind it is that doubling the voltage effectively halves the current, so there’s less EM radiation around heater wires. Another minor reason is that series wiring uses only terminals 4 and 5 on the sockets and allows for a cleaner layout. The downside is that you cannot use Russian tubes anymore. Also, with 12.6V, the current draw of the pair of 100 ohm resistors in the virtual center tap is too high (63mA with 0.8W dissipation!), so larger resistors should be used. A pair of 220 ohm resistors will dissipate around 0.36W, so I went with a pair of 330 ohm 1/2W resistors drawing around 0.25W, just to be safe. If using 12.6V heaters, another alternative is to elevate the heaters by connecting the reference DC voltage to the unused center tap (pin 9) of on of the tubes instead of referencing both sides through a pair of equal resistors.
Finally, I made it possible to choose between having completely isolated channels or leave the normal channel bleed into the overdrive channel like it does on the original amp. Clean signal is out of phase with the overdrive signal and is usually quieter, so it will effectively cancel-out some of the overdrive signal. The difference is subtle and may or may not be intended by the designer of the circuit. I’m leaving it up to you to decide which version to build. The layout below is drawn with channels isolated, but it’s easy enough to change it to have clean the normal channel bleed into the overdrive channel. Just bridge the 3rd and 4th terminals from the right. That way, normal channel will always go straight to the V3a triode without being affected by the channel switch.
It is worth mentioning that I changed the power supply slightly compared to the schematic. I reduced resistor values from 15K to 4.7K to reduce voltage drop and have voltages similar to the original with the 275VAC secondary and added an extra RC node to improve filtering. Duncan’s PSU Designer app confirmed that these capacitor and resistor values should work well, giving the correct voltages without any noticeable ripple. 4.7K resistor and 22uF capacitor in each node form a low pass filter with cutoff frequency around 1.5Hz which is more than satisfactory for our application.
Switching Options
If footswitch operation is not needed, the simplest way to implement channel switching is using a DPDT switch that switches between different signal paths. The layout is optimized so there are no long leads carrying signal to and from the switch that would pose a risk for noise or feedback. In this case, just use the layout as it’s drawn, ignoring the “Footswitch Circuit” section.
If footswitch operation is needed, the DPDT switch used to switch between the two signal paths can be easily replaced with a miniature DPDT latching relay that is powered from the heaters secondary (or a separate secondary). The relay can be switched either using a footswitch pedal or with a panel-mounted SPDT switch when a footswitch is not plugged in. Note that miniature relays need between 150mW and 500mW of power (DC) for the coil to switch. Heaters secondary must be able to provide extra 0.4-1VA AC (depending on the relay type) or a separate power transformer should be added to power the relay. The AC heater voltage needs to be rectified and then filtered and regulated before we can use it to power the relay. Omron G5V-2, Finder 30.22, Takimisawa RY12W-K or similar relays will work fine here. Choose relay voltage and regulator voltage according to the heater voltage leaving at least 2-3V extra for the regulator to work properly and taking bridge rectifier diode drops into account. 12.6V AC heater winding should be enough to drive a 12V DC relay. 6.3V AC heater winding should be enough to drive a 5V DC relay (probably not enough for 6V, though!). Use a switching jack to disable the panel-mounted channel switch when a footswitch is plugged in.
Footswitch circuit uses a bi-color LED to toggle between the two colors designating the selected channel. A P-channel MOSFET is used to switch between the two LED colors. Alternatively, two separate LEDs can be used the same way with cathodes tied together. For my build I wanted to toggle between red and blue color and have a matching knobs on pots that control each channel. So I got a common-cathode RGB LED and just used the two diodes ignoring the green one.
Parts
- The enclosure measures 250 x 180 x 80mm and is made from sheet aluminum, powder coated in black
- Power transformer is a custom wound 20VA toroidal with 270VAC @ 40mA and 12.6VAC @ 0.6A secondaries
- Filter capacitors for the power supply are Panasonic NHG 450V electrolytics
- 1uF capacitors for cathode bypass and output capacitor are Panasonic ECWFE metallized polypropylene film capacitors rated 450V (total overkill for cathode, but it was easier to get all the 1uF capacitors of the same kind)
- 22nF coupling capacitors are Russian Paper-in-Oil caps rated 400V
- 1nF and 2.2nF caps are some 630V poly film caps I already had
- Small capacitors in the picofarad range are NP0 ceramic disk capacitors rated 500V
- Signal resistors are 1/2W metal-film Xicon, plate resistors are 2W metal-film KOA and power supply uses 2W metal-oxide KOA resistors
- Pots are 24mm Alpha
- Jacks are closed-style with isolated threads, so I can control grounding to the chassis
- Switches are Miyama
- Tube sockets are Belton Micalex VT9-ST-C with an integrated shield
Construction
I started by mounting the three sockets on the tube bracket.
Power supply is built on a separate terminal strip.
All parallel resistors and capacitors are wired together to simplify wiring to the sockets later.
Heaters are wired together with the 220K resistor on the third tube before the power supply is mounted on the bracket.
Then the second layer is populated with cathode resistors/capacitors.
Finally the top layer is populated together with the ground reference resistors that go between the top layer and the ground bus. I also wired the leads for input and output jacks as well as the lead that comes out of the normal channel’s Gain pot. All shielded leads’ shields are directly soldered to the bus. Teflon coating prevents the heat from melting the insulation and shorting the inner conductor with the shield.
From the side you can clearly see the three separate layers of components. At this stage, the main module is ready to be tested, first without and then with the tubes attached. We can test if heaters light up, check the heater voltage and drop it with a pair of resistors if it’s higher than 12.6V. Then we can test the power supply and make sure that it outputs the expected voltages and we can do the basic testing of the circuit by measuring plate and cathode voltages of all triodes. SLOCLONE schematic lists all important voltages, so we can use those as a reference.
The switching module I put together on a piece of terminal strip with narrower terminal spacing to save some space, as there are no big components to mount here. The three leads on the left that go to the LED(s) are color coded to make the wiring easier.
Front panel components are installed with corresponding resistors and components mounted directly on the pots and switches.
Testing Procedure
Since we are building the preamp in modules, one of the main advantages is being able to test modules individually before putting it all together. It’s easier to test a smaller module and it’s definitely easier to fix any issues before the whole circuit is put together in a tight box. I organized testing in three separate stages:
Switching circuit: quite simple to test once the LED is soldered to the three leads coming out of the board. I connected a 12VAC supply to the rectifier, verified that the LED glows blue. Then connected the two leads that will later go to the switch/jack to verify that the LED color switches to red.
Front panel: one lead comes from the main module into the Clean/Crunch switch and there are two leads coming out – one out of the Normal gain pot and one out of the Overdrive gain pot. We can test the panel wiring by connecting guitar signal (or signal generator or any audio signal) to the input and monitoring each of the two outputs separately. Respective pots and bright switches should affect the sound and there should be a difference in levels between Clean and Crunch on the Normal output. Overdrive output should be louder than both Clean and Crunch.
Main module: this is the heart of the beast and it’s more complicated to test, but still it’s better to take few basic tests while it’s still out of the chassis.
- With the tubes unplugged, check the heater wiring for shorts. Resistance between the two heater leads should measure only the total series resistance of the two virtual center tap resistors (2×100 ohm on the layout). If it shows resistance close to zero, there is a short somewhere. If it shows an open circuit, something is not connected properly.
- Check the main power supply for shorts by measuring DC resistance between the B+ line and ground.
- With tubes plugged in, connect the heater voltage to heater leads and make sure that all three tubes are lighting up and verify that heater voltage is close to 6.3/12.6V. In my case it was closer to 13.6V, so I used a pair of 1.2ohm resistors (value calculated using Ohm’s law) to drop the extra volt and bring heater voltage down to desired voltage.
- Connect the high voltage secondary through a fuse and/or light-bulb current limiter and take the voltage readings: all B+ voltages, heater reference voltages and plate/cathode voltages on all 6 triodes.
Final Assembly
After all the individual modules are tested, it’s time to put it all together. First I installed the transformer noise bracket, followed by the transformer, and the power line (AC cable, fuse holder, power switch). Then I fixed a 5-terminal strip next to the power transformer that carries the pair of 1.2 ohm resistors that drop the heater voltage, followed by the switching board. The next step is to install the main module and install the jacks on the back plate.
The final assembly step is the trickiest and it involves connecting the front panel components to the rest of the circuit. The space is tight, so I had to solder everything with the panel detached from the rest of the chassis and then screw it in place after everything is soldered.
To finish if off I designed the faceplate in Corel and had it laser engraved at a local shop.
And finally the complete build with the faceplate installed. The name “ZLO” is clearly a word play on “SLO” and in Serbian it means “Evil”, thus the devil’s horns on the logo.
How Did it Turn Out?
In one word – awesome. Firing this puppy up for the first time felt like meeting an old friend. There were no issues whatsoever and it worked perfectly from the first try. There’s no excessive noise – on par or better than VST amp simulations I run on the laptop. There is no switching noise at all, which was my primary concern with the relay switching circuit. And all the great sounds that SLO is known for are there. The brutal lead channel that can do anything from heavy blues to metal. The crunch which reminds me of plexi tones that can cover my need for Blackmore or Kotzen tones and a nice clean channel. All channels respond very well to overdrive at the front, so together with my Ultimate Overdrive I effectively get 6 different channels.
Another thing I love about it is that it brings out all the nuances of the instrument. Digital amp simulators are fun and I’ve used them most of the time in the last couple of years, but they do tend to have sound on their own that sometimes masks the character of the instrument. With SLO preamp it feels like guitar plays a much bigger role in the overall sound and I can get even wider range of sounds by switching guitars. I will definitely integrate SLO into my everyday rig and use it in conjunction with VST effects and all the goodies that are available through the DAW, but still have it be the foundation of my sound.
What Would I Do Differently?
It took a lot of planning to do this build, so I didn’t leave too much rooms for errors. Still, looking back, there are few minor things I would change to make it go even smoother. First, I would move the main module back for 1-2cm to leave slightly more space between the front panel and the main module. It would make the final assembly easier. Another potential change would be the orientation of the front panel switches. The two bright switches go up-down and the two channel switches to left-right. It may be counter-intuitive to someone that does not know that. It made sense to me at the time, but I’m not sure anymore.
Further Mod Ideas
- True Bypass: in its current form, the preamp is not true bypass. Should you need it to be true bypass for whatever reason, a switch or relay can be added close to the input jack to bypass the circuit.
- Shared EQ: I intentionally omitted any EQ because I plan to shape the output later, but you can easily add a 3-band EQ before master volume pots. There might be enough room to do it with the same sized enclosure, especially if using 16mm pots for the tone stacks.
- Dedicated EQ: a bit more complicated and would probably require a bit more space but still perfectly doable. Using another relay in parallel with the main channel switching relay, we can toggle between the two tone stacks.
- Switchable Boost Stage: SLO reacts to booster at the front very nicely, so it’s not a terrible idea to consider integrating a booster stage at the front to push the tubes a bit harder. This is a good application for LND150 solid state gain stage that can be powered from the latest B+ node. It would probably be a good idea to add a dedicated RC power supply node for the LND150 to avoid oscillation. This would push the SLO closer to Peavey 5150 territory with the extra gain stage. The coupling cap after the boost stage should be tailored to tame the low end and prevent the bass from becoming “boomy” with higher amounts of gain.
Audio Clips
All clips recorded through audio interface with VST effects and cabinet impulses instead of a mic’d cabinet.
Ibanez RG 2550
LePou HyBrit (Marshall) simulator set to clean and Rosen Digital 5150 and Mesa Oversized Recto cabinet impulses. One of the impulses is delayed by 3ms to make the sound more “complex” and creamy, somewhat simulating double-tracking. Added a touch of reverb and delay VST effects.
LePou Marshall simulator set to clean and Rosen Digital cabinet impulses. Double tracked, panned hard left/right.
Ultimate Overdrive (gain fairly low, bass cut maxed) in front of Soldano MK2 Preamp (lead channel, gain at about 1/3rd) into Sansamp section of Fly Rig 5. Added RedWirez G12M impulses and V-shaped EQ in VST.
Similar settings as above with a bit more gain and delay from the Fly Rig.
Headless Mini Telecaster
LePou Marshall simulator set to clean and Rosen Digital cabinet impulses.
LePou Marshall simulator set to clean and Rosen Digital cabinet impulses.
Bancika MK1 Guitar
LePou Marshall simulator set to clean and Rosen Digital cabinet impulses.
Hi. It looks like the layiout does not show the output as connected to ground? Also, with the input and output it looks like maybe these are shielded wire, so there would be ground wire connected at both ends, but shield only connected to ground on one end? Awesome project by the way!
Layout drawing shows three blue shielded wires. The first two are grounded on both ends with small green wires, must have missed to draw one side on the output wire. It should be grounded to the ground bus. All jacks are isolated, so no ground loop is formed.
Hi mate, I’m not sure if anyone else noticed, but it looks like your relay is wired incorrectly. As the normal in is the output pin for v3 grid and the same on the other side of the relay.
Alex
Hm, not sure I understand. I built it as drawn and it works.
Thanks for sharing, what an awesome project!