Copyright 2009-2020 by djg. All Rights Reserved.

Wonky Gibbon Ramblings

The Katana 50 MkII – Making Great, Greater – Adding an FX Loop

Posted on October 07, 2023 by danny

This was a project inspired by user ColorMaker on the website. In this post ColorMaker posted a video showing a modification he’d done to his Katana 50 MkII which added an FX loop.

In every review of the Katana 50 MkII – the one criticism everyone has of this otherwise excellent amp, is the lack of an FX loop.

So Mr ColorMaker had done it and in so doing, proved it was possible – and he’d helpfully posted a block diagram showing roughly what was involved but, unfortunately without the detail.

Now despite a request for the circuit diagram posted on the website nothing was forthcoming. And I can’t say I blame him, having to support these things can be a right pain in the butt – I’d have to work it out myself.

Over the last few weeks I have thus been thinking about it, slowly gathering information, did a circuit diagram, translated it to a stripboard layout, bought the bits, soldered it up – and blow me – it works!

The one thing I will say at this point, is that this is not a project for a novice or suitable as a first project in hobby electronics. That’s not to say it electronically hard, it really isn’t – but – if you don’t know what stripboard is nor which end of a soldering iron to hold – this is not the project for you. It’s an awful lot of amplifier to be messing up.

You also need good eyesight. I used to find working with stripboard easy but I’m in my 50s now and my near vision isn’t as good as it was.

To be clear then, if you choose to do this and knacker your amp – or the circuit doesn’t behave as you expect and you don’t know how to debug it – don’t be sending me your complaints or requests for help. If you have a go at this – you do it at your own risk in the knowledge that if you mess up, you yourself will have to sort it out. You’re on your own, I’m not helping you! And I say that with love, but you know, life is too short.

Are we all good?


So here’s how the project went then.

I found a circuit diagram for the Katana 50 MkII here. The important information is on p40 in the top left hand side of the diagram, it shows a connector coming from the main digital board onto the amplifier board. As the signals on it have already been through the D/A process, they’re analogue. This is your point of interception and is why we’re in business with this project. There’s no track cutting involved or anything on the PCBs, just a cable. Opening up the Katana I found the connector – it’s the white and black bunch of wires, cable tied, in the image below, going from the top board (which is all the digital stuff) to the small bottom board (which is the 50W power amp module that drives the speaker).

So my FX send / return loop would effectively be inserted into the mono analog signal going across that connector. Pass me the wire cutters! Ok – maybe not – we need to do some planning first…

A further innovation ColorMaker had quite rightly done, was to insert two buffer amps into the circuit. As you don’t know how much current the upstream circuit can provide nor how much the downstream one (ie: your pedals) require, it’s polite to buffer these with an OpAmp at each end.

Also he’d discovered that the output voltage level of the amp on the connector can be too high for your pedals (ie more than 9V peak to peak), meaning the signal would clip and get distorted, so smartly put a signal cut switch in, to give the option of attenuating the signal as it goes out on the FX send and then boost it back to its original scale, on its return. The switch is there to turn this feature on and off depending on whether it’s required or not.

As an aside, you could do the attenuation part using resistors arranged as a voltage divider, but as we’re buffering with an OpAmp it makes sense to use that for the purpose since, it’s there and available!
On the return side, there is no option but to use an OpAmp to boost the signal anyway.

So on that basis both send and return sides each buffer using a standard inverting OpAmp configuration, for which the standard gain formula is: -Rf/Rin

As I wanted to achieve a cut factor of 4, resistor values of 1k2 and 4k7 respectively are appropriate.
Similarly on the return side we want a boost of 4x so the same resistor values can be used again, just swap them around ie: 4k7 and 1k2 respectively.

Now as we are using Op Amps – they need some power, ie: +Vcc, -Vcc and Gnd. Going back to that connector in the Katana, I could see from the diagram that it has supply rails on it. Unfortunately it didn’t indicate the voltage on them, so I had to open it up again and measure with a digital multi-meter.

The connector pins on the connector from the Katana Preamp to the Katana Power Amp are as follows
Pin 7 is the audio signal
Pin 5 is +Vcc (28V)
Pin 2 is -Vcc (-28V)
Pins 3&4 are Gnd

So we’re going to splice a 5 wire bus out of the main bus going between the two Katana boards that will branch off to our FX loop, comprising +Vcc, -Vcc, Gnd, Signal in and Signal out (with Signal In and Out of course being the two cut ends created when you cut, the wire from Pin 7) .

But wait a minute, +/- 28V you say? That is a lot, though not altogether surprising, as it is driving the power amp module through to the speaker – so of course it needs some beef. But yeah, supply rails of +/-28V is a lot. This leads to a problem as I’m not aware of any OpAmps that can run off supply rails that high. A bit of a google search did reveal one part that might have managed it but the information was scarce and in any case I couldn’t find a supplier.

SO – I decided to use my usual “goto” OpAmp for audio, the TL07x series, which is a very low noise part suitable for this sort of thing. It’s been around for decades and very easy to source (Amazon, Farnell etc)…

In this case as we need two Op Amps, I would be using the TL072 which contains 2 OpAmps in one 8 pin DIP package. This can run off supply rails up to about +/- 18V – so we would also need to build a voltage regulator circuit to bring the +/-28V down to a rather more manageable +/-12V which the TL072 can handle. The classic parts for this are the 7812 and 7912 voltage regulators, also decades old and easy to source.

In most of the circuit diagrams online for these regulator parts (which I’ll let you google), they have modulation capacitors both before the input and after the output. There are big ones for smoothing the AC mains power sinusoidal shape out (which you would need after a transformer / bridge rectifier), and smaller ones after the regulator for handling output ripple and then tiny ones for handling transient spikes. However – as we’re tapping into nice clean DC rails that Roland Boss have already cleaned up for us rather than the mains, you only really need the ripple ones which rather simplifies your parts list.

So – stuff all of that in a plastic box (thus avoiding any accidental short circuits from a metal box), add a switch and some suitable mono jack connectors – and you have an FX loop.

Here is my circuit diagram for the FX loop stuff.

And another for the 7812 and 7912 power regulation. Note: The pins for Vin and Gnd are swapped between the 7812 and 7912 so be careful when laying out the stripboard and wiring them up. Also you will need small heat sinks on those two parts or the purple smoke inside them is likely to come out. Electronics: it’s all about keeping the purple smoke inside the components…

I did think about providing sketches for my strip board layouts but I actually think it’s important you create your own, otherwise you get let off having to understand the circuits – and that path leads to amplifier death or something you can’t debug.

Here’s a picture of my little power regulator board. The big connector plug has +/-28V and Gnd going in, and +/-12V coming out, all on the one connector. I wish I could have found some nicer more compact connectors instead of that great big thing – but I didn’t have the time.

Here’s a pic of the main board before it got stuffed in the black plastic box (which also ended up containing) the regulator board

And what the back of my Katana looks like with the loop in place. NB: Set the box back a bit, so that the switch doesn’t catch on things and get broken when you’re carrying the amp around – gigging etc…

And here is the parts list:

Some Veroboard (aka Stripboard)
1x TL072 – Dual Op Amp
1x 8 Pin Dil package holder
2x Mono 1/4″ Jack sockets
3x 4k7 resistor
3x 1k2 resistor
2x 10k resistor (to “pull” the inputs down to 0V when disconnected)
1x DPDT Toggle Switch
Plastic Case

For the voltage regulator circuit:

Regulator 7812 (Vin max 35V)
Regulator 7912 (Vin max -35V)
2x 100uF on output to smooth ripple
2x heat sinks

Over to you – enjoy.

Things I would do differently if I was doing this project again? I’m glad you asked! … :

  • I used 7812 and 7912 voltage regulators because I happened to have some lying around. This means that the TL072 that uses them would have max peak to peak input and output signals of +/- 10.5V (since the TL072 can only get to within 1.5V of its supply rails). Now – I think that when you have the gain and volume knobs cranked right up on the Katana, plus a few effects – that could cause some clipping distortion. In fairness I haven’t heard any as yet, but given you could just as easily use 7815 and 7915 regulators (+/-15V) or even 7818 and 7918 regulators (+/-18V) as the TL072 is good for supply rails up to +/- 18V – this would give you a bit more headroom.
  • I’m not convinced the reduction switch is worthwhile. I can’t hear any difference in the amount of noise between the two positions (either clean or with gain), which is the only reason you’d elect not to cut the send and boost the return voltage. So I would be inclined to remove the switch, which removes two resistors from the circuit and reduces the wiring complexity – and just have it always wired to a ratio of 1:4
    • ie: Cut circuit : Rin=4k7, Rf=1k2
      Boost circuit: Rin=1k2, Rf=4k7
  • Talking of that cut / boost ratio of 1:4, we’re trying to get the signal down to something that will go though a pedal which runs off a 9V battery. Worse, if we assume your pedals also can’t get within 1.5V of the battery supply voltage (and I’ll bet they can’t) it probably means a peak to peak max signal of 6V (ie: +/-3V). Given we’re starting from a signal that could potentially be +/-16.5V you might want to increase the resistor ratio to around 1:6 rather than 1:4, both cut and boost side. Having said all that – as I say, I haven’t heard any distortion as yet so I could be being overly cautious.

Now, I’m going to say this one more time – if you haven’t done this sort of thing before – this probably isn’t the best project to start your journey in electronics. It’s a rather nice amp to be destroying if you get it wrong – and something of a pricey mistake.

This is why I have been deliberately a little vague in my descriptions here, to deter newbies. Not to mention that providing literal step by step instructions would be excruciating. If you can cut through the fog and work out what needs to be done and design your own stripboard layout from the circuit diagrams, then you probably have the necessary skills to do the project without damaging your Katana – and in so doing make both it and yourself feel fabulous. However, if you can’t – don’t.

Have fun!

Oh – and one final word of thanks and acknowledgement to ColorMaker. It’s one thing to build the circuits for this when someone has told you it’s possible – but ColorMaker did it from scratch with no help or clues from anyone else. That’s worth a round of applause – good work!

Leave a Reply

↑ Top