I have embarked on the challenge of reviving a 20-plus-year-old DIY device of mine. This blog post is the first of a series of posts describing the process of bringing back to life my prized Electronic Projects for Musicians phase shifter.
But first, some background:
During my first heavy binge on DIY musical electronics, I built around 25 devices or so. Very few of them survive today.
The ones that did survive were among my favorites, of course. They included a two-channel tube pre-amp, a noise-reduction device, a two-channel limiter, the EFPM dual tone filter, and my favorite – the phase shifter from EPFM.
My copy of EPFM contained sound samples from the devices in the book. Back in the dark ages, the sounds were provided via a piece of flexible plastic that was bound into the book. You tore it out along the perforated line, then fired up your turntable. You dug out one of your vinyl LPs, put the piece of plastic on it, and gingerly placed the needle on it.
Somehow, this set-up allowed the book purchaser to listen to the sound of the effects.
I fell in love with the sound of the phase shifter. It was amazing to me. I made it my goal to build one.
I wisely started out with simple effects. I did make plans for the phase shifter, though. I still have my copy of EPFM. The page containing the phase shifter parts list shows where I jotted down notes about price and quantity of parts.
Wisely, I bought a circuit board from Paia Electronics. I managed to score four of the hard-to-find Clairex CLM-6000 opto-isolators. (Nowadays, the readily-available NSL-32 would work in place of the CLM-6000.)
When Anderson needed a dual op-amp for one of his EFPM designs, he chose the Raytheon RC4739. It was a fine op amp, but had an inconvenience: it was contained in a 14-pin DIP case. Most dual op-amps are built in eight-pin cases. Back then, you still could buy 4739s, but they are extinct today. I used two 4739s in the phase shifter.
I built mine with the specified parts. An aluminum Radio Shack (RIP) enclosure served to contain the device. The power would come from an external 9-volt AC-powered power supply, another project from EFPM. I still have the power supply. The phase shifter and power supply were connected by a plug, which could be separated. The power setup is cumbersome. I hope to change that some day, possibly as part of this project.
Anyway, I decided in earnest the other day to try to get the thing to work. When I put it in the true bypass position, it howled through the amp. In the un-bypassed position, there was no sound.
Initially, I tried to fix a few obvious things in the hope the problem was something simple. But I didn’t get anywhere. So, I decided to rebuild the device. The plan is to keep the enclosure, circuit board, the potentiometers and most of the switches. It will have a 3PDT true bypass stomp switch with an LED indicator light.
Anyway, I took a good look at it. I felt strange looking at work that I had done more than 20 years ago. The solder joints looked good. Hey, the thing had worked at some point!
Nonetheless, I would’ve liked to have had some conversations with my 20-year-younger self about this project
Current me: Was green the only color of wire you had?
Younger me: Do you think I was made of money? It worked!
Current me: Why did you leave that rat’s nest of wires? Have you ever heard of organization?
Younger me: I don’t care how it looked. It worked!
Current me: Why didn’t you use mylar or film capacitors? Those ceramic caps look downright nasty.
Younger me: Do you think I was made of money? It worked, didn’t it?
Current me: Thank you for using the Vector T-42 clips to make connections to the circuit board much easier!
Younger me: You’re welcome.
I proceeded to disassemble it after I failed in my feeble attempts to get it to work. The circuit board came off of the standoffs. All of the connections to the board were de-soldered and removed. The T-42 clips made this work easy. The clips fit into a standard circuit board hole. They should be soldered, but they’ll hold pretty well on a friction fit. Because of the clips, I didn’t have to de-solder one side of the board and then try to quickly pull the wire out of the other. The clips also will make it easier to solder the new wires. A pack of T-42 clips is on my want list. They can make connections much easier.
After de-soldering the wires from the pots and switches, I went to work on removing certain items from the circuit board. I theorized that the integrated circuits (4739 and 4136) had been fried or otherwise had gone bad. So, I planned to replace them.
It was a battle to remove the sockets of the 4739. They had to come out because I planned to replace them with eight-pin op amps such as the NE5532. However, the footprint is a 14-pin design. I have a plan to solve that problem, which I will discuss later. There also is a 4136 quad op amp on the board. It’s not a common chip, but is still available.
I de-soldered the solder-side connections of the sockets as best as I could. I managed to pull the socket out, but many of the metal legs didn’t come out of the circuit board. What was left were small pieces of metal, sticking out of the board.
Further removal of solder on the solder side of the circuit board was fruitless. Then, the light bulb came on! I applied the solder iron tip to the bottom of the socket legs. That sent heat to the solder side. The pieces of metal then came out easily.
While doing battle with the sockets, I accidentally tore the top of a ceramic capacitor. I looked at the schematic and discovered that mylar or film capacitors would be better than the ceramic ones. So, out came all of the ceramic capacitors.
The resistors and electrolytic capacitors remained on the board, as did the Clairex CLM-6000 opto-isolators and the 4136.
Replacing the capacitors won’t be too risky. I don’t want to change too many parts, though, because of the risk of introducing new problems into the circuit. I’ve sourced some NSL-32 opto-isolators in case the venerable CLM-6000s have stopped working.
I bought some parts from Jameco to get the project rolling. Among the parts are upgraded capacitors, a few NE5532 op amps and 14-pin wire wrap sockets. The wire wrap sockets have long legs.
The plan to replace the 4739s with 5532s is based on a product sold by Paia Electronics some time ago. I will insert the 5532 into the bottom of the socket, or on the side without the notch.
I will cut off a small piece of stripboard. I’ll mount the 14-pin wire-wrap socket in the 14 holes in the circuit board. I will then use jumper wires to connect the circuit board signals to the appropriate NE5532 pins.
Anderton nicely included in EPFM a guide for substituting 8-pin dual op amps for the 4739. It should prove very helpful. The NE5532, in addition to being available, is a good value for a low-noise dual op amp. There are better (and more expensive) dual op amps, but the NE5532 is a good compromise of price and performance.
I also plan to order some NSL-32 opto-isolators in case the CLM-6000s don’t work. Digi-Key, sells both the NSL-32 and the 4136 op amp.
However, I’m not going to replace everything at once. The first step will be to insert the 5532 setup. Re-wiring of the leads to power and potentiometers will follow. This time, I’ll use different colors of wire to avoid confusion! I also hope to keep the leads shorter and better organized. I also will replace the ceramic capacitors with an upgraded capacitor.
At that point, I’ll test it. If it doesn’t work, I’ll double-check all of the connections, be sure that power is getting ot the chips, and give it a good “eyeballing.” If it doesn’t work, I’ll assume the 4136 and/or the CLM-6000s are not working. I’ll order the parts and then install them. If that doesn’t work, I’ll start pulling my hair out!