LED Lighting for a Vintage Amp

Some HiFi amplifiers continue to perform nicely regardless of their age; some others, such as the one described in my earlier instructable (https://www.instructables.com/Noisy-NAD-Amp-Fix/) suffer, like people, from ailments due to ageing.

What I describe here is an upgrade/repair of a vintage Technics SU-7 amp from about 1983, which means that it is more than 40 years old now, and still going strong! It has only one flaw: The beautiful, backlit, analogue power meters are only partially illuminated, because some of the old-fashioned, oblong, incandescent festoon bulbs are burnt out. As is commonly known, the main feature, or rather, the flaw of incandescent lamps, apart from their low efficiency, is their limited lifetime.

However, in the early eighties of the last century, only rather dim LEDs in red and perhaps green colour were available, and with a hefty price tag, at that. Their yellow, amber, blue and white varieties were born at a much later date. The only way for a white (-ish) illumination, then, was using incandescent bulbs.

Your amplifier might have a similar problem. But its interior will, of course, be different - anyway, I am sure you can extrapolate the thoughts given here to your own make and type or, at least, find some inspiration for your own solution.

There are two of these festoon bulbs per meter that are directly soldered on a dedicated PCB per channel. In the 1st picture above you clearly see that the left one is burnt out, and the meter's right half is brighter. The 2nd picture above shows the two left-channel bulbs with the the top cover removed - one of them is dark, the other one still working.

I have, btw, not the foggiest notion why the original bulbs are covered in a transparent, blue-coloured rubber sleeve...

When you want your amplifier to keep its original status, finding replacement bulbs having exactly the same size, the same voltage and current consumption, and the same pinout is the only way to go. Finding direct replacements can be difficult, but for this amp at least, complete sets can be found e.g. in online auctions. They are not exactly cheap, and in addition they have the two inherent drawbacks I already mentioned - low efficiency (resulting in a certain amount of heat), and low life expectancy.

I beg to differ, since I prefer modern replacements of the LED flavour with all their advantages (high efficiency, low heat dissipation, very good longevity, moderate price). And when I realized how the bulbs were mounted here on two dedicated PCBs, short pieces of an auto-adhesive LED strip automatically and immediately came to mind. I even have a leftover piece of an LED strip in my junk box from an earlier project (https://www.instructables.com/The-Metamorphosis-of-an-Old-Flat-Bed-Scanner/) - meaning it's free of charge for me :-).

The warm-white LEDs on the strip are arranged in groups of three, connected in series, and each group has its own, series-connected current limiting resistor of 150 Ω - see the 1st picture above. Supplied with the specified 12 VDC (direct current), each group consumes approx. 20 mA. All the groups on the strip are connected in parallel. It is easily possible to shorten the strip by cutting it, as long as you cut between the groups. For ease of use, next to the marks indicating the location of cuts, there are solder pads for connecting the supply wires, clearly labeled '+' and '–'.

One single group of three LEDs has just the right length to be used as a replacement for two of the festoon bulbs, as shown in the 2nd picture above.

The front view of the amp is shown in the 3rd picture, with only one of two incandescent bulbs working on the left channel (and its non-uniformly lit meter), and the right-channel meter with the new LED strip .

When zooming in on the circuit diagram of the amp's service manual (4th picture) you see that the four bulbs in question (two each on the M and N PCBs) are powered from the 11.5 VAC (11.5 Volts alternating current) winding of the mains transformer. LEDs unfortunately cannot be used with AC - some more modifications are required for using them here.

In order to convert from AC to DC, an arrangement of four diodes in a configuration called 'full wave' or 'bridge' rectifier can be used, resulting in a pulsating DC. When adding a smoothing capacitor, this ripple (that might lead to a flickering light output of the LEDs) is reduced. The RMS value from the 11.5 VAC (Volts AC) transformer winding results in a DC value that is larger by a factor of about 1.4 (the square root of 2), i.e. about 16 VDC. The diodes in the bridge rectifier cause a voltage drop (aka a loss) of about 1.2 to 1.4 V, resulting in about 15 VDC - still too much for our 12 VDC LED strip.

Therefor, each LED strip needs an additional series resistor of at least 150 Ω in order to not overload the LEDs. Experiments with different resistors showed me, however, that the LED strip is too bright by far anyway, which allowed me to add an additional series resistor per LED strip of 2200 Ω (or 2.2 kΩ, often written without the 'Ω' unit in place of the decimal point as '2k2'). This reduces the brightness to about the original brightness of the incandescent bulbs, and the current through the LEDs becomes about 2.5 mA only, a total of 5 mA for both strips. This is only about 1/100 (one hundredth!) of the original incandescent bulbs' consumption (440 mA) - in clear: almost nothing, and still only 1/8 (one eighth) of the nominal consumption of the LEDs used here. The consequence is that these LEDs will live forever, at least :-)

The circuit diagram of my new arrangement is shown in the picture above. You can see that the rectifier (four 1N4001 diodes), a 22 µF capacitor with a voltage rating of at least 25 V, plus the LED strip and its additional series resistor are located on the first ('master') of the replacement PCBs only. The second ('slave') PCB contains nothing but the LED strip and its series resistor because it is supplied from the first PCB.

The two wires from the transformer's 11.5 V winding are connected to the master PCB, and the slave PCB is connected to the first one by two wires as well, these going from + to + and from – to – (no surprises here).

The step above shows the theory. Now we want to implement the ideas. I found that the two original PCBs (1st picture, taken from the amp's service manual) were not really useful for the new circuit, so I decided to make replacements from matrix board by copying the originals' outline to have them fitting into the originals' locations. The 2nd picture shows the original PCB with the festoon bulbs above, and below my replacement 'master'. The 3rd picture shows both new boards looked at from both (component and solder) sides, designed with the 'Lochmaster' software (https://www.electronic-software-shop.com/lng/en/electronic-software/lochmaster-40.html?type=N&opt=%7B11%7D22&language=en). This software is not too comfortable to use but still allows designing and changing your hardware in an easier way than with squared paper, pencil and rubber.

Thanks to the simplicity of the circuit, I used some bare matrix board without any copper layer and did the wiring with some bare, 0.6 mm dia. wire on the solder side (not visible in the photographs). Only for the connection to the LED strips I used some pieces of PTFE-insulated, thin wire. Space restrictions dictated that all the components are located near the bottom end of the boards, and there are two areas on each board where no components must be installed due to mounting hardware.

Some darker spots on the new board show you that I had the components arranged differently in the prototype, and had to move them in the final version due to the space restrictions mentioned above.

The two different, 'master' and 'slave', replacement boards with the two interconnection wires are shown in the 4th picture.

Please note that, before installing them, you need to carefully insulate the bare-metal mounting locations by using, e.g. some Gaffa tape in order to avoid short circuits.

As you can see from the last picture, the new LED lighting illuminates the lovely meters evenly - and that's exactly what we wanted to achieve.

The cost of the material used was approximately three times nothing, since all the components came from my junk box. A set of replacement bulbs would have been much more expensive.

On the other hand, I used quite some time for planning and implementing my solution - but seeing the result I feel that this time was well invested.