Design of a Korg Nutube Amplifier Part 4: Build and Test
This is the penultimate post in our series about creating a guitar pedal style amplifier with the Korg Nutubes. We previously left off with ordering PCBs and the nerve-racking wait associated with it. Due to the somewhat unreliable nature of the Mark 1 Human, everyone gets a PCB wrong now and again and it’s that certain knowledge that causes those sleepless nights. This tension is also the reason it’s so exciting the day PCBs arrive and we had already ordered a kit of parts to build up our amplifier, so we set straight to work on this.
Building up the circuit was a case of soldering the through-hole parts. This is somewhat of a novelty since as design engineers we spend most of our time soldering metallic dust onto pads we cannot see — otherwise known as surface mount components. Having been a while since we soldered up a circuit that you can see with the naked eye, it was a rather therapeutic hour.
The good news is the CAD cells look great: the Nutube fits perfectly, as do the FETs, pots and connectors. The TO220 LM317 is a little on the large side and ideally could do with smaller pads, however, it’s far from the end of the world!
Build done, it’s time to power up the design.
One thing that is often neglected and is quite possibly the most important part of a design is the test — quality testing is what sets apart a project from a design. We’ve already done quite a bit of testing in Part 2, so let's start by checking these waveforms and that both our build and PCB are performing as expected.
Before we get started we must adjust the bias to around 2.5V, just like we did in Part 2. It’s best to not connect the circuit and use a DVM for this. The 22k pots are set to 50% out of the factory which will give 1.5V Bias which is not enough.
First, we adjust all of the control pots from the 50% level, setting both gain and volume to max, and the tone to min.
Looks like the amplifier is working!
The output is around 6 times the input, which is acceptable for a distortion pedal. We are not running the tubes for maximum gain; the 10k resistors R9/R4 and R10(lowR in the case above)/R6 will be acting like a pot down to the inputs and changing these to a higher value should increase gain if desired, but note there needs to be sufficient plate current for operation.
One thing to note here, there is some slight distortion to the output, we found playing with the bias can help.
By adjusting the bias to around 2V we can remove the distortion, but as previously noted in Part 2, this may be a desired effect and we can play with this once we get to using the device.
Next the tone circuit.
The image above shows min to max adjustment of the tone circuit @ 1kHz. This should have a profound effect on the audio and make the bass far more prominent.
Finally, we need to know the max/min input to output gain.
From this, we can see we don’t have quite as much gain as the first prototype (this had around 10 total), but there is plenty of adjustment. At the minimum gain and volume, we get an attenuation of 0.2 i.e. 1/5’th of the input. At the maximum settings, we get around 8-9 times the input signal, which is plenty for the purposes of a “distortion pedal”.
We also checked the impulse response using a square wave, again the system erred on the over-damped side, which is excellent. However, we neglected the most important thing: does it work for a guitar…
At this point, we have to admit we have no idea how to play a guitar, so it’s time to cheat. We have a phone and YouTube with plenty of excellent guitar solos on there. We need a cable from 3.5mm stereo to 6.25mm mono these are relatively common.
Next, we needed to check the peak-to-peak output of the phone and luckily it had an output of 0.2V pk-pk at max volume, which is roughly that of a guitar pedal!
All that is left is to test.
Throughout this series, we’ve been discussing options for expansion and improvement. The tone circuit, for example, can be adjusted by changing C10 to 47nF. This would increase the cut-off response. By changing R6 and R4 to a higher impedance gain could be improved. Adding the input impedance amplifier Q2, C13 & R19, the input impedance would be less critical and a high impedance output would be attenuated less.
Given two of these circuits, this design could become a nice pre-amplifier for a transistor stereo.
Alternatively, using a single stage could give a similar effect for a stereo application but with a single tube.
Being an engineer from the age of transistors it’s been a blast working with something so different. In some ways the humble valve is so much better than a transistor — two words: it glows! The four-year-old in us all loves things that glow.
Hopefully, this series has given some insight into the inner workings of a valve amplifier and inspired new projects. For those who would like to try building their own, the DesignSpark PCB database, Gerbers and a bill of materials can be found on GitHub.
CommentsAdd a comment
January 31, 2018 15:20
Discrete components! Valves! 1/4" jacks! Absolutely outstanding. It brings back my days of youth. Projects like this enable understanding of first principles of electronics and physics, and much more fun than programming a PIC!
Bring them on! I wonder if the sound gets better as the tube gets run in? Like the old Marshall amps.
January 31, 2018 08:42
A good project with a pretty specialised device. A technical comparison of how different amplifier types behave when turned up to 11 would be really interesting. The youtube reviews of the Ibanez Screamer - a formerly solid state guitar pedal that's now gone NuTube - by guys who can play their guitars at https://youtu.be/QU0rFqPNr0Q and https://youtu.be/aD_0byuj0I8 are really interesting, if lacking electronic detail!
January 9, 2018 08:37
I like this project because it introduces me to a new technology. I agree though, that I would love to hear this thing cranked up to eleven so I can hear its overdrive characteristics; scream, baby, scream!
January 8, 2018 14:04
I don't want to look negative but, to make a relatively linear pré-amp for acoustic guitar is something much easier with an operational amplifirer.
Where I think this device coul be very interesting was not addressed in this case. It would be a design where the non-linear characteristics of the nutube could shine.
For that, it should be designed an overdrive pedal, for electric guitar (not acoustic) perhaps with an op-amp as a first stage, with variable gain, to drive the nutube from linear to highly distorted.
In overdrive pedals filters are usually used to control/shape de amount of high order harmonics created by distortion. That should also be part of such a pedal.
At last, the output should be connected to a guitar amplifier (clean channel).
Only in this kind of scenario the potential of this devices could be analised.
I have designed several guitar pedals and a full DSP amplifier, and I am a guitar player. So, I will be very glad to share my expertise in the matter if someone wants.
January 8, 2018 13:53
Nice demo, but (as I'm sure all the bedroom guitarists were hoping to hear), what nice warm crunchy sounds do you get when it's overdriven? Surely the main point of a tube pre-amp is to create a "Marshall-in-a-matchbox"? Go on, turn it up to 11...