Search

Matt's Projects

Sometimes I make things

Category

Tape Delay Pedal

Project: Analogue Tape Delay – Final Demonstration

OK, so the pedal is finally built. Here is a video! I’m pretty pleased to be finished with this one. The final few weeks have been a bit frustrating and I’m looking forward to moving onto something new.

Enjoy.

Project: Analogue Tape Delay – Assembley

It’s been about a month since my last post. The first two weeks of which were very productive. The last two have been more frustrating.

In the week before I went back to uni I had a big push to try and get the project finished before I came back. This resulted in me staying up until 7am the day I went back to uni. …And I still didn’t get it finished. However, the box had been constructed, the circuit board soldered and installed, the metal work drilled and the inputs mounted.

At 2am the day before I came back to uni, all I had to do was solder on the switches etc and mount them on the board. However, due to a lack of forethought, I hadn’t actually put proper thought into how it would be assembled. As it turned out, it was impossible to mount the switches in the case because you simply couldn’t reach them when you put the lid on. Multiple attempts nonetheless resulted in broken wires and a lot of lost sleep, hair and sanity.

I ended up taking the project back to uni with me to finish off, and I think it’s pretty much there. Unfortunately, exam term comes with other priorities and I don’t have half as much time as I would like. Anyway, now for some pictures.

The front panel I made for the project
The front panel I made for the project
All of the electronics can be seen if you remove the metal panel.
All of the electronics can be seen if you remove the metal panel.
I think the jack connectors look pretty smart.
I think the jack connectors look pretty smart.
The metal panel will be mounted over the bottom hole.
The metal panel will be mounted over the bottom hole.

Project: Analogue Tape Delay – Etching PCB

Another opportunity to test out my CNC!

Sadly, through no fault of the CNC, what should have taken about a 45 minutes ended up taking about 2.5 hours!

The PC I use to drive the CNC is a 15+ year old laptop running Windows XP. This makes it very prone to blue screening and it crashed TWICE during the whole process. Given that it takes about 20 minutes to boot and become stable this results in a massive waste of time. It also meant that each time I restarted the software I had to realign the board which resulted in a loss of accuracy. Nonetheless the board is now fully etched.

I took a slightly different approach to what I have done before. Usually I do all of the drilling by hand after I etch, however I thought it would be neater if the CNC drilled the holes before it etched them.

Not wanting to risk drilling through the PCB and into the CNC, I compromised and decided to have the CNC drill 0.7mm into the PCB, making it really easy for me to drill the rest of the way later on.

The final board! Pretty pleased.
The final board! Pretty pleased.
The board with only the holes drilled. Looks pretty cool.
The board with only the holes drilled. Looks pretty cool.

Obviously in future I want to avoid situations like this. Therefore I have decided to try and get Grbl working.

Grbl is a system that allows any computer with a USB port to control a CNC parallel port interface using an Arudino Uno in the middle. This will save me lugging around a massive old laptop and will hopefully be more reliable ūüôā

Project: Analogue Tape Delay – Modifying the Motor

I’m starting to see a trend of post frequency being highest at the start of a project and decreasing as each project matures. I suppose this makes sense as there is usually a large learning curve at the start, while the rest of the project takes more experimentation.

Until yesterday I had assumed that controlling the speed of the motor would be easy. I assumed it was a normal DC motor and by putting it in series with a variable resistor, I could vary the speed. This is necessary so that I can change the time it takes for a note to get from one tape head to the other.

I thought I ought to test this assumption, so I tried putting different resistances in series with the motor. But what I found was that for a resistance of less than 40 ohms, there was very little change in speed, and for a resistance of greater than 40 ohms, the motor did not turn at all.

Combine this with the fact that there is a hole in the motor allowing you to turn a potentiometer inside it, it became pretty obvious that this was a so-called “capstan” motor – ie it has electronics inside it to regulate the speed.

The motor in question.
The motor in question.

Initial attempts to open the motor up to have a look were unsuccessful and I soon gave up, resolving to buy a replacement DC motor online last night. However, this morning, following the advice of the guys on EEVBlog, I gave it a second shot and, using a drill, managed to open the motor.

The electronics inside!
The electronics inside!

I was right! Now all I had to do was modify the board to bypass this control circuitry and turn it into a normal DC motor. By removing the circuit board, I could find where the motor brushes were soldered on.

Using a Stanley knife, I scratched away the traces to isolate the input connectors as well as the brushes and then proceeded to short the input connectors to the brushes.

Not very pretty! But it works
Not very pretty! But it works

Plugging it back into the tape deck, it works and operates at a much higher speed. However, now I can slow it down with a resistor.

Project: Analogue Tape Delay – Slowly Building the Circuit

So now that I am back from uni for Easter, I have a fair bit more time on my hands. I have begun the process of building the entire circuit on a breadboard. This process has been rather slow because at each point I have experimented with different designs to (hopefully) ensure that I am choosing the best component and structure for my purpose.

mixer_circuit_breadboard
This is the current state of the breadboard layout. I have attempted to make it as neat as possible. (However, I refuse to trim resistor leads).

While I am using dual op-amp chips, in order to make my life easier I am spreading the circuit out by only using one op-amp from each chip for now. The overall circuit will be clearer if I draw it out:

Mixing_circuit_cad
I have attempted to break the circuit up into sections so it is easier to see how each part functions.

The circuit so far takes an input from the tape head, amplifies it and mixes it with the guitar signal. It also drives the cassette player motor.

I did have a problem with noise from the motor being transmitted along the power lines. This was sorted out by adding 10uF decoupling capacitors across the 10k resistors.

In other news I have modified the tape so that I can fit two tape heads in – one to write and one to read.

Modified_tape
I just cut out a little plastic.

Project: Analogue Tape Delay – Writing to Tape

Another short one. There’s not too much more I want to do before I get home from uni next week. Everything will get much easier to do once I can leave everything out on my desk and do some hardware modifications. The last thing I wanted to do to be sure that everything fitted the theory was to try writing to tape.

As I said in a previous post, it is necessary to bias the signal that is written on to the tape. The tape player from which I took the recording head used to use DC bias. It would add 50mV to the signal. Supposedly, so I read, AC biasing is much more effective. Adding a ~2V AC sine wave to the signal should result in higher fidelity sound. So I tried both, and what I found disagrees with this theory.

The method which achieves the best sound quality – according to my tests – is to DC bias the signal by 50mV DC… exactly what the original tape player did. AC biasing resulted in crackly sound that faded in and out. Perhaps it is no surprise then that this is what the original designers used, but I can’t find a way to explain it. Maybe this particular head is not suited to AC biasing? So for now at least, I will use DC biasing.

Project: Analogue Tape Delay – Amplifying Tape Signal

This is just a quick one. I said in my previous post that the signal produced by the tape head was too small to read using my oscilloscope. I put together a simple op-amp circuit with about 100x gain and measured the output to see what it would be like. It worked pretty nicely:

Amplifying the signal from the tape head by 100x resulted in a pretty nice waveform.
Amplifying the signal from the tape head by 100x resulted in a pretty nice waveform.

The signal looked a little noisy so I hooked the output up to my guitar amplifier and it sounded pretty good! At least as good as the sound coming out of the speaker in the tape player I took the tape head from. I’m feeling pretty optimistic about this bit.

You can see the tape head is plugged directly into my breadboard which is then amplifying the signal and feeding it out to my guitar amplifier off screen.
You can see the tape head is plugged directly into my breadboard which is then amplifying the signal and feeding it out to my guitar amplifier off screen.

Project: Analogue Tape Delay – Measuring signals for recording

Today I bought a cassette tape from Oxfam on Cambridge highstreet: “Mozart Symphony No. 41 in C Major”. I reckoned that I could probably put up with listening to it a lot while debugging the recorder. The slight trouble was¬†that the cassette was write-protected so I couldn’t simply record over the old music. As it turns out, it’s not particularly hard to get around this. The only protection is a small hole in the top of the cassette which tells the cassette recorder that this is read only. Cover the hole with a bit of tape and the cassette player now has no idea.

The tabs at the top on either side indicate whether or not this is a write-protected device.

It’s one thing for the tape player to work, it’s another thing to be able to replicate the signals myself and record what I want. So I got out my oscilloscope and took some measurements.

This is the signal in the tape head when writing my whistling to the tape.
This is the signal in the tape head when writing my whistling to the tape.

As you can see, during recording, the signal across the tape head is approximately 100mV in amplitude, and the maximum I saw it reach was about 200mV. Because I had my oscilloscope set to AC mode, what you don’t see here is that the signal is actually DC biased by about 50mV. The reason this is done is the ensure that the electromagnet in the tape head is in it’s linear region resulting in a less distorted wave form. While DC biasing improves the fidelity of the wave form, it can be improved further by AC biasing as I intend to do in my design. You can read more about biasing here.

I also measured the signal across the tape head during playback, however the signal was so small it was barely visible and isn’t worth me uploading a picture. Hopefully given some amplification it will become more visible.

Anyway to finish off the post here is a video of the mechanism in action. In this clip you can see the tape head (centre) being engaged during playback and also the erase head (to the left of the tape head) being engaged during recording.

Project: Analogue Tape Delay – First Parts arrive

So after a lot of searching I’ve managed to get two old tape players to use for parts. One, which can only read, I bought off of ebay for a few quid (and is yet to arrive) while the other, which can read and write, I got from a lovely couple on free-cycle.

I managed to find the service manual for the tape recorder which was very useful as it allowed me to tear it down without risk of damaging it. As it turns out it is almost perfect for my uses.

This was such a delight
This was such a delight

The design of the recorder is such that I can remove the cassette assembly as one piece and simply transplant it into my own box. In fact it is so simple that the only IO I will need is the motor controls, the tape head signal and the switch indicating that the play button is pressed. What makes it even simpler is that all of the controls are dealt with mechanically. The fast-forward/rewind buttons change the gearing of the mechanics to increase the speed of the tape Рno electronics required (although I will want to control it electronically, it is still one less thing to reverse engineer). The pause button simply disengages the tape-head Рno electronics required (this will work nicely as a mute button). I could go on about how perfect this is, but hopefully this will be self evident when I write about how I actually use it.

Create a free website or blog at WordPress.com.

Up ↑