Monthly Archives: July 2014

Sony TC-WR590

A while back I was given a bunch of hi-fi audio equipment by an ex-colleague. He was having a clear-out and offered the stuff to me. A big stack of Marantz gear was cleaned up and is now doing service at a friend’s restaurant, the other thing was this Sony tape deck. Don’t know the year, but it’s fairly high spec. I had powered it up, it made a rumble and a few clicks but that was about it other than the screen coming alive. I put it aside for further inspection.

Welp, pulled it to bits and found the same problem that plagues most old tape decks… bad belts. I ordered a replacement kit for it and a new belt for my record turntable at the same time. Pulled the tape deck to bits to change the belts out.

Tape deck front down

Having found the manual online at HiFi Engine, I started stripping the mechanism down.

Undo motor cage

Nicely designed, with the service technician in mind. Three screws and the motor cage lifts off.

Motor cage off

The main drive belts had turned to the consistency of warm liquorice. I peeled the remnants off the pulleys..

Old belt
More old belt

Cleaned it all up with isopropyl alcohol and Q-tips. Fitted the new belts. Helpfully there’s little claws to put the belt around, which keep the belt outside the edge of the pulley it has to sit on.

Belts fitted

With the motor cage screwed back down, knock the belt off the claw and neatly into the pulley groove. Ditto the main drive belt:

square belt on
Flat belt on

To show it’s all back together we have a wonderfully blurry photograph sorta showing the innards all put in.

inside tape deck

Cleaned up the heads and capstan pinch-rollers. The head mechanism on each deck is fantastic. The clockwork amuses me- it’s an auto-reverse deck so when the direction is changed, the head assembly drops down a half inch and the entire thing flips round 180 degrees then pops back up to meet the tape again.

Tape heads

Finally, a Rolling Stones tape playing, after a little adjustment of the heads. I really could do with a proper set of alignment tapes.

Playing a tape

It’s now in the living room, making the stack of audio equipment I have look even more bizarre. Black face, silver face and grey plastic turntable. Tech through the ages.

Current Loop Interface

This is a little out of sequence (rather, it’s a lot out of sequence), but it’s today’s project.

My Renault has a serial data output stream that provides telemetry from the engine’s state of operation. It operates at 62500 baud, and that rate across a piece of wire a few feet long is soon interrupted and corrupted by the RF generated by the ignition, alternator, various motors and relays and the like. I tried using twisted pair and the results were better but I still had corrupted data.
Somebody suggested to me to make a current-loop interface rather than reading voltage levels.

Why? It’s fairly easy to induce a voltage in a piece of wire- if I take my voltmeter and don’t connect the leads together or to ground, it’ll read about 0.3V AC. That’s just from the radio energy in the air around this house.
What’s more difficult is to induce a current, especially through a resistance. Therefore, if we use the flow of electricity to indicate a signal rather than the presence of electricity, we should be doing better.

So, enter the schematic.

Schematic, current loop interface TX, RX

This was pilfered from the Internet, from various places. I went ahead and took a trip to Radio Shack and picked up an LM324 quad op-amp and an LM724 op-amp to use for either end of the circuit. The transmitter takes two op-amps, the receiver one.


Threw it all onto breadboard. The transmitter on the left, the receiver on the right, connected by the coiled-up wire in the middle.
With a voltage applied, nothing happens. A small current is passing through the loop.

Lights off

Cut the voltage to the input and a larger current flows, switching on the comparator at the far end. Seeing as I had two spare op-amps inside the chip I decided to put an LED on the near end to show what’s what.

Lights on

It’s a fairly compact circuit but I’ll have to build it into a small box, I think.


I just need to get a P-channel FET now. They are fairly expensive and hard to come by locally. I have a PNP transistor in it’s place right now which means the circuit doesn’t shut off completely. Ultimately the transistor will work just fine but it’ll be better with the FET.

I need to put the circuit through its paces next and see what the highest frequency is that I can pass through.

Pioneer SX-650 receiver

I had decided I needed a “proper” hi-fi system, so began looking on the local Craigslist for anything in a decent price range and of a style I liked.

Enter a modestly-priced ($65) Pioneer SX-650, found in uptown New Orleans from a guy in the hi-fi club who bought it as “working” but soon realized it was indeed not, and sold it as “parts”.

It was in good cosmetic condition, despite being grimy and some of the photo-finish on the corners of the wood coming away. I got it home and gave it a good old clean up.


Once acceptable to place in the living room, the top came off, and a rudimentary inspection given. Everything inside was very original looking, nothing hacked up or nasty looking. Brought it up gently on my Variac, connected to a couple old computer speakers.

First power up

At about 60 Volts the beast came alive! Pops and crackles out of the speakers, then I managed to dial in a local radio station. The sound was good, considering. Watching the current, I brought the voltage up slowly. At about 95 Volts the current began to run away, so I killed the power. Not soon enough, the main filter capacitors both boiled and vented. Needless to say, that’s an unpleasant smell and I was not the most popular person in the house.

Main filter capacitors

The caps are fairly large, considering the device. There’s the originals, side by side to a regular “D” cell battery for comparison. At that point I decided that if those ones had gone bad, the rest of the electrolytic capacitors on the board probably weren’t that far behind. Shotgun time! Got onto Mouser, and from a useful parts-list from AudioKarma purchased a big box of capacitors.

New caps

Started the long task of matching up all the capacitors in their bags and writing on each bag the location on the board that it was due to replace.

Organizing capacitors

All set up on the table, there was much of this, removing old, adding in new:

Kitchen table workshop
In bits again

Soon enough I had a box full of old capacitors.

Old caps

Comparing the sound of the device at each step along the way was surprising. I’d replaced the capacitors as per suggestion that some of them by the original design were running very close to their maximum tolerance.

Power amp board

Higher voltage rated ones with better impedance and crosstalk characteristics were fitted. Higher temperature ones were used where I could get them.
Soon, I had a set that was able to be switched on and left on.

FM on

With a rudimentary antenna, 100.3 KLRZ-FM coming in at 5 on the scale. Replacing the capacitors in the FM stage helped a lot in removing the FM stereo “hiss”. In that image, the scale illumination bulbs had all burned. Only the power and stereo lights were operational. I decided to try and see if I could use a fluorescent tube from an old computer screen to light the scale:

Fluorescent tube

It was too blue for my liking, as interesting as it was. That came out, in favor of some flashlight bulbs of the correct voltage and current rating from Radio Shack.
I had then started to try and track down the source of some pops and crackles. There’s a speaker/FET protection circuit built in. It monitors the resistance across the outputs and also times a period after initial turn-on to prevent damage to loudspeakers from the THUMP you get as everything charges up. The audio output goes through a relay, which I removed from the circuit.

Protection relay 1

Isporopyl alcohol to clean up the gold-plated contacts. They were quite dirty:

Protection relay 2

That helped a little bit but there were still crackles when I moved the speaker selection switch, so off came the front and out came the switch, a multi-wafer affair with some very dirty contacts.

Selector switch
Selector switch dirty

That got a clean up, same as the mode switch, which was the same but more complex:

Mode switch 1
Dirty mode switch

Careful use of wet-n-dry paper and DeOxIt D100 spray, along with cramped fingers:

Clean mode switch
All cleaned up and low resistance. Reassembled it in the correct order, then soldered it back in.

Mode switch fixed

Pops and crackles begone! I had some parts on back-order for the record player preamplifier boards (below):
The silicon was also going noisy with age, so in the interim, I built a stereo preamplifier board out of spare (and non matching parts) to a design I found online:
Preamp on breadboard

That allowed me to at least hook up and listen (with much background hum) to my records until the parts arrived:

Records playing

Everything set and tested:

Stereo scope

All set up and well within limits. It should deserve the sticker it received from a previous home, one of the government offices in Los Angeles:

Safety tag

Nice to have a little bit of history with it. Got it all tidied up next to my record player:


Looking and sounding good!

Compaq – the Forgotten images

I found a use for the Compaq. Well, another use, at least.

Ncurses analogue clock, via telnet. Found here.

Compaq showing Telnet clock

I had, however found a couple extra photos of part of the Compaq. They are from my work with the PWM dimmer board for the front lights. There will be those of you who have noticed the move towards the “bright white” LED recently, for useful items such as flashlights and Audi’s daylight running light system. Those little things are BRIGHT. So much so that I didn’t fancy them running at full brightness in a dim room and making the computer’s screen appear dim in comparison.

Therefore, enter a PWM (Pulse Width Modulation) circuit!

PWM circuit

Mmm, NE555 goodness right there. Doubly good because there’s two of them. It’s like being back in high school electronics all over again.
Why not just use a light sensitive resistor to drop the current through the LEDs and dim them that way? Well, whilst that is entirely possible, LED (Light emitting diode) chips have a tendency to dim in non-linear ways and some dim more than others due to the way they are mass produced. They do generally tend to be fairly conformal when driven at their nominal current so instead we shall use the eye’s persistence of vision to “dim” them for us instead.

This is achieved by pulsing the LEDs on and off very quickly. So quickly that their on-off is no longer discernible by the naked eye, but the average appears as differing intensity.
This is where the NE555 timer chips come into play. They are used in one of two design modes. Monostable (think a timer that you trigger and it turns off after a period of time) and astable (blinky blinky like turn signals on a car).
If we connect one into the other, we can make pulse-width modulation. The left 555 turns on and off. The turning on triggers the right hand one to count down a period of time. That period of time is adjusted by the amount of light falling on the light-dependent resistor (LDR).

So, if we turn the left one on off on off


The right one will be triggered for a short period in the dark


and a long period in the light


Which makes it dim in the dark and bright in the light.

PWM on breadboard

(The 556 is two 555’s in one chip. Bonus!)

In the end, it works well.