Servicing Classic QUAD 50 AmplifiersFollow article
How do you feel about this article? Help us to provide better content for you.
Thank you! Your feedback has been received.
There was a problem submitting your feedback, please try again later.
What do you think of this article?
Venerable 1960s transistor amplifier favoured by the BBC and British Rail gets an overhaul.
The Quad 50 is a single channel transistor audio amplifier of somewhat unusual design, in that it uses an output transformer — which putting the increased weight and cost aside, provided it with certain benefits in terms of flexibility and robustness. The amplifier was designed with professional applications in mind and became part of a standard studio monitoring solution at the BBC, where it was designated AM8/11 and paired with the now legendary in HiFi circles, LS3/5A loudspeaker.
Other Quad 50 users included British Rail and London Underground, where thanks to its rock-solid reliability it was put to use in public announcement systems. Nowadays the amplifiers are appreciated by HiFi enthusiasts for their “valve-like” qualities, which is no doubt due to the use of a sizeable audio output transformer much like might be found in a valve amp.
In this article, we take a look at replacing aged electrolytic capacitors which can be a source of noise and lead to other problems, before proceeding to set up biasing. There are various scans of the Quad 50E Operating Instructions available online, with these including a schematic diagram and basic servicing instructions. Some copies have also been updated to contain details for the slightly different Quad 50D variant, while there are also numerous websites with servicing info.
Next, let’s take a closer look at the amplifier.
The Quad amplifier was available in two variants, 50D and 50E, with both being identical in every respect, other than the 50E variant being a little more configurable. The flexibility of the amplifiers came about through the use of an output transformer with a total of eight independent secondary windings, which could be connected in series and/or parallel in order to drive different loads.
Above we can see a fragment of the schematic diagram for the 50D, where pairs of secondary windings are already connected in series, thereby reducing the number of permutations that are possible through linking/cabling at the output connector.
Whereas in the 50E we can see that all eight windings are individually brought out to the connector.
In total the supported outputs/loads are:
- 17V / 5.8R (50E only)
- 25.5V / 12.5R
- 34V / 23R (50E only)
- 51V / 50R
- 68V / 85R (50E only)
- 102V / 200R
Quad 50D output transformer with black wires pairing secondaries.
Of course, it would also be possible to take a 50D and internally reconfigure this for one of the above intermediate configurations, since the windings are brought out to tabs and these are simply strapped at the transformer, rather than being brought out to the speaker connector.
In addition to more output configurations, the 50E features a socket for an optional 600R balanced input transformer.
Plus also a potentiometer that can be adjusted to reduce higher input signal levels.
The mechanical constructions are also robust, to say the least, with a substantial frame and cast aluminium front panel, plus a rather large heatsink at the rear.
The end result is an amplifier that is rated at 50W RMS output and which is unconditionally stable into any load. Indeed, Quad stated that “The amplifier is virtually proof against misuse and no harm can be caused by factors as gross overload, continuous operation on short circuits, heavy reactive loading, inadequate ventilation, etc.”
Some Quad 50 amps are fitted with a small round Bulgin mains connector, while others utilise an XLR style mains connector made by Cannon — neither of which will be up to modern-day safety standards. The audio input and output connections are via multipole connectors variously referred to as Plessey, Painton, Jones and, in the USA, Cinch. In the UK these were originally made by Painton, who were bought out by Plessey and marketed as “Multicon”. Regardless of the name used, these are becoming increasingly difficult to find.
The Quad 50E was in production from 1966 to 1983 and so even with a later model, electrolytic capacitors are likely to be well past their best. Particularly considering that the PCB has a power resistor which tends to get very hot indeed and heat the board up, with elevated temperature shortening the life of such capacitors since the electrolyte is lost at a faster rate.
The replacement capacitors were selected as follows:
- C1. 12uF / 63V
- C6 + C9. 47uF / 50V
- C7 + C8. 4.7uF / 50V
- C13. 100uF / 16V
The physically large 5000uF PSU smoothing capacitor equivalent series resistance (ESR) was checked, rather than just immediately replacing this, since these are much less common to fail and a replacement would be far smaller and look somewhat lost in the chassis.
With the capacitor taken out of the circuit and using a Peak ESR70 meter, this measured at 0.03R, which is pretty good for a capacitor of its spec and so it was not replaced.
The three trimmer potentiometers however were replaced, with new 100R and 2K2 ones that we had in stock, since those fitted were the original and now quite dusty, open frame carbon track sort.
Another component that is sometimes swapped out is the large 180R wire-wound resistor (R32) located in the PCB cut-out, which is prone to baking the board. Typically replacing this with a modern metal-clad type, mounted to the chassis and cabled to the PCB. However, such changes are a tough call as it generally means drilling the chassis, which is an irreversible modification. This may be revisited in future if the heating proves to be problematic over time.
Since components were replaced and in particular, trimmer pots, some setting up was required next.
RV1 must be adjusted such that 5.5v is measured between the emitter of Tr3 and ground. A convenient point to connect a meter is the anode of C8 — which is connected to the emitter of Tr3 — and then one of the many ground points, such as the terminal of the bridge rectifier with green wiring, else simply the amplifier chassis.
Next, the output stage quiescent current or bias must be set. This should be done when the amplifier is cold and with no input. There are two approaches that may be taken. The first — the Quad specified procedure — involves removing links on the primary of the output transformer and inserting an ammeter in its place. This can be seen above, is a little fiddly and can easily result in accidentally melting insulation. The second method instead involves measuring the voltage across R30 and R31, for Tr8 and Tr9 current respectively.
Using the first of the aforementioned methods, RV2 and RV3 are each adjusted for 30-40mA, with the ammeter inserted in place of the links for Tr8 and Tr9 respectively. With the second method, the links are left in place and RV2 and RV3 adjusted for 15-20mV across R30 and R31. The latter approach does result in a small error, but it should be negligible.
Searching online it seems that some will set a bias current multiple times the maximum specified value, in pursuit of a lower THD figure. However, while these are remarkably robust amplifiers and may survive such treatment, it was decided to go with the original Quad specified parameters. Furthermore, an amplifier of this design — featuring an enormous transformer in its output stage — is never going to win any competitions when it comes to THD and, actually, this is probably part of its charm.
Quad 50 amplifiers are really a thing of beauty and hold particular appeal due to their unusual design, flexibility and rock-solid reliability. Following a simple service, with a handful of electrolytic capacitors and a few preset potentiometers replaced, then bias set, they should hopefully run trouble-free for a few decades more.