Amp Related Circuit Basics – Transformers, Rectifiers and Capacitors

Print Friendly, PDF & Email

Amp Related Circuit Basics – Transformers, Rectifiers and Capacitors
I thought I would write a set of Posts on some basic aspects of electronics that are common to most circuits in one way or another, but leaning more toward actual valve amp circuits where possible. This is to help fix things in my mind as much as impart knowledge. Wikipedia is great for ease of reading, to get a general if not in depth understanding of basic principles, without having to take a full course in Electronics. If you can get your head around capacitors and inductors at even a basic level, it will help a lot in understanding the nuts and bolts of most audio circuits, which is why I have included so many links to different aspects of their functionality below. For now, a bit of history of valves is always a good start – they really are fascinating with a certain charm and character that solid state just doesn’t have – for me anyway.
http://en.wikipedia.org/wiki/Vacuum_tube
http://en.wikipedia.org/wiki/Valve_amplifier
Valve manufacture and operation videos, Westinghouse and Mullard:

http://www.youtube.com/watch?v=GDvF89Bh27Y
Mains Transformers

The obvious place to start is with mains transformers, which are generally used for mains isolation and the stepping up or down of mains 240V (here in the UK) AC voltage/current to another value, usually less than 240V but sometimes higher in the case of valve based equipment like guitar amps and old valve radio and TV sets. Here’s a common schematic view of a valve amp transformer feeding the AC rectifiers via the first secondary winding, and the valve heater elements via the next lower voltage, higher current secondary winding:

Sometimes, there is a 6.3V mains bulb across the heater winding, as in the Marshall:

Or an LED with a 6.3V Zener diode across it as in the Fender Champ 12:


They are constructed in different ways, sometimes with many windings that output different voltages, usually around the same core of an iron based chassis as a squarish lump or can be toriodal, circular types. Higher output, more expensive types are made of iron/nickel/cobalt alloys like “Alnico”.


The strength and efficiency of a magnetic coil can be quantified in terms of its flux density – so many thousand lines of flux per square inch etc.
The secondary winding usually steps the voltage down for most household appliances like lamps or computers, but for valve amps is steps up to the higher value that is required for valve operation, and can be as high as 500V or more depending on circuit design below, and power output requirements. Later, the Champ is seen giving out 504V DC.
http://www.chambonino.com/construct/const17.html
Trannys designed for a 1000W Valve amp!:

The reason the DC side of a valve amp can reach such high DC levels is partly due to the step up windings obviously, but the thing that has to be remembered is that AC voltages are usually expressed as an RMS (root mean square) value, which – without getting too maths heavy here – is an approximate expression of its equivalent DC value for doing work in an electrical circuit. For example, a light bulb fed by a constant DC supply of 240V should give out the same average amount of light as one fed by 240V AC.
This means that RMS AC actually has a peak voltage level to its sinusoidal waveform that is higher than the RMS value by about 1.4 times (square root of 2), which means the 240V UK voltage peaks at about 339.5V. For a derivation of the maths proof of average AC power compared to a DC circuit, see here:
http://www.doctronics.co.uk/rms.htm
If you have forgotten what a sine wave looks like:

When a large smoothing capacitor is in the chain as part of the AC to DC rectification process, it will get charged to the higher PEAK value, which is why you have to be extra careful about discharging the DC side capacitors before working on a valve circuit board.
A clear example of where the transformer, rectifier diodes and smoothing capacitors sit in a valve amp is in the Aston amp schematic:

The expected voltage values at the testing stage of that amp are (+/- 10%):

TP2 is at the top of C14 – 313V DC
My actual readings for the last column were:
265V AC
3.6V AC
3.6V AC
Test point 1 – 238V DC
335V DC
323V DC
314V DC
170V DC
155V DC
1.11V DC
0.39V DC
25.7VDC
Now can see how the 335V DC comes from the secondary step up windings of this particular transformer, to charge capacitor C14. You may see variations of 20-30V difference from amp to amp (hence the 10% testing allowance) depending on the transformer, and circuit resistances etc.
The original Fender Champ circuit is similar except for a centre-tapped mains secondary and 2 diode rectification, as all these basic 5W – 12W kit amps are based on older audio circuits, like the Williamson amplifier
http://en.wikipedia.org/wiki/Williamson_amplifier#Design
and Mullard 5-10
http://en.wikipedia.org/wiki/Mullard_5-10
amplifiers, and others, usually without recognition, leading to the original 1957 Fender Champ:

And the more recent Champ 12:

This pic is a little unclear (above) but the 240V export model section here shows wires grey/orange/blue/green/yellow/black with the secondary wires (red/red/ and green/green):

This amp kicks out a big 504V DC between the secondary red wire to chassis earth!

Transformers are inductors, which have a reactance to AC ( a change in a steady DC state) measured in Ohms, as they create a back voltage in opposition to the one passing across them, as well as a magnetic field around themselves, which can induce electricity in another wire.
http://en.wikipedia.org/wiki/Reactance_(electronics)
This reactance is frequency dependent, so they have uses in frequency based functions in circuits, such as radio/TV station tuning and audio power impedance matching. They have a time period characteristic like capacitors, based on resonant frequency aspects,
http://en.wikipedia.org/wiki/Frequency
http://en.wikipedia.org/wiki/Resonant_frequency
http://en.wikipedia.org/wiki/Resonant_circuit
http://en.wikipedia.org/wiki/Tuned_circuit
http://en.wikipedia.org/wiki/Regenerative_circuit
http://en.wikipedia.org/wiki/Autodyne
http://en.wikipedia.org/wiki/Heterodyne
http://en.wikipedia.org/wiki/Superheterodyne_receiver
which depends on their inductance value measured in Henrys, when used in charging/discharging circuits with capacitors and/or resistors.
Mains Rectifiers:

Valves or solid state components can be used for rectification. The GZ34 rectifier valve is a common example of a tube diode.

http://www.ampmaker.com/store/JJ-GZ34-rectifier-valve.html
If 4 diode solid state rectification is used, as in the Aston amp, they usually look like this (if not all in a discrete component):

A way to remember how they are wired in the circuit is the DC load sits in the middle – exactly as in the Aston schematic above also:
RectFilter.jpg
As you can see, as the AC voltage swings +ve and –ve, it can only go one way through the load = DC.

Once you have drawn this out yourself a few times you should remember it. You can see the effect of the smoothing capacitor as it discharges during the AC drop in voltage, keeping the DC level more constant across the load. As a general rule, the bigger capacitance the smoothing or “reservoir” capacitor has, the better the smoothing. You may see values as little as 8uF going up to 32uF, 47uF or 100uF, with 350V-600V working limits.
Another point is that 50Hz AC mains is in the low bass audible frequency band for humans, so any ripple here that makes its way through to the output stage will be heard as a low hum. A larger capacitor here will help pass this AC bass signal to earth, so can be viewed as part of the tone control for an amp.
The Fender Champ uses 2 diode rectification, and the mains transformer is usually centre tapped for this method to work, or set one diode to earth as the mid point, so the circuit can achieve the required valve voltage levels and still benefits from full wave (both half cycles) rectification:


The Marshall Mercury uses a single diode so can only achieve rectification for each positive half cycle of AC – current cant flow through the load during the negative half cycle:


This is the inefficient version as it only converts AC to DC every half cycle, so 50% of the mains is unused. Larger value smoothing capacitors are a must for this method so there is plenty of stored DC available to fill in until the next half cycle charge up.
There are some good visual examples of different ways these rectifier circuits can be drawn and applied, showing the phase and half cycles of the AC, here:
http://www.allaboutcircuits.com/vol_3/chpt_3/4.html
Mains Electrolytic Smoothing Capacitors


I have already mentioned these above, but a look at the circuits will show some common values of electrolytic types used for the AC rectification stage. These are the ones that need to be drained before working on a circuit to avoid lethal shocks. The values on the side state the charge capacity, e.g. 100uF at 450V DC working voltage. This means they should withstand that maximum voltage across them without failure, with the polarity the right way ONLY! They can explode if the polarity is reversed or they get shorted out when charged up. There is usually a stripe of arrows down the side pointing to the negative terminal.
For all capacitors in general, the value of the capacitance (quantity of charge it can store in microfarads) also dictates the time period of charging and discharging when connected in either series or parallel with inductors and/or resistors. This characteristic is used in frequency related aspects of circuits from radio/TV channel tuning to audio tone control circuits, or dynamic sound effects like tremolo (change in amplitude) and vibrato (change in frequency). (See the Tremolo section in the Marshall schematic, bottom left).
This can be created by capacitors and resistors in a resonant circuit, producing Amplitude Modulation of the output signal – there is a great java applet of AM and FM here:
AM/FM Applet

They also block DC so are intrinsic to pre and power amp stage signal routing. They prevent the (usually) higher, unwanted DC voltage found at a transistor emitter or valve anode being connected to the input (transistor base/valve gate) of the next stage, which would happen if connected directly, so that only the AC component of the signal passes to the next stage as C2 and C6 do below:

How Capacitors Work:
http://farside.ph.utexas.edu/teaching/302l/lectures/node46.html
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capac.html

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html
Farads.jpg
The important thing to understand in larger amps is that the voltage rating of a particular single cap – say 450V max which was common in the 60s – is not exceeded by stacking them in series, as the total voltage is now shared between them. These often came as large cans holding 2 or 3 caps in one housing, so it was easier to stack these 450V rated caps to handle larger voltages for higher power amps , as in a JMP Marshall example here:
50mF_yellow.jpg
As the caps are equal 32uF values each, only 250V is across each one – not the full 500V HT.
SeriesCaps.jpg
As the top caps lower plate is connected to the bottom caps upper plate so the same thing electrically, the two caps appear as one cap, still of the same original surface area, or plate size, as a single cap, so can still only store the same maximum amount of charge as one on its own. However, this total charge is now distributed equally between both caps, so their combined charge storage value of both together is now halved, or 16uF.
1/Ctotal = 1/C1 + 1/C2 = 1/32 + 1/32 = 2/32 = 1/16 so Ctotal =
16uF.
This would be quite a low value for first stage smoothing for an amp of this current draw potential, so NOW, another two in series are often seen added in parallel to the first pair, so their combined values in parallel ADD, to give 2 x 16uF in parallel to get back to effectively 1 large rated cap of 32uF in value, that can stand a high voltage of 500V DC.
In this case, 4 x 32uF, 450V rated caps have to be used to get the same required 32uF value of one cap that could stand the higher 500V+ voltage.
Nowadays, we can get much smaller physical size caps, that are larger capacitance values and can stand much higher working voltages.
————————————————————–
DC power and AC RMS values maths, courtesy http://www.doctronics.co.uk/rms.htm:
RMS AC Power and DC.docx