The AC RMS voltage is called the effective voltage. Because the voltage is cyclic and swings positive and negative, the RMS AC voltage is calculated to be equivalent to the DC voltage that dissipates the same power over a resistor as the AC voltage does.
But an AC sine wave has a peak voltage that is 1.414 times the RMS value in both positive and negative directions, or 2.83 times peak to peak.
So when an AC is rectified by a voltage bridge and used to charge capacitors, the negative part is folded over (with a full wave bridge rectifier) and the capacitor charges to the peak voltage of the rectified sine wave.
There is some voltage loss in the diodes in the bridge (~1.0Vdc). The Transformer voltage will droop under load (called regulation). The Airlink 80VA toroidal is rated at 12% regulation, meaning that with no load the output voltage with be 1.12 times the rated voltage.
So the transformer voltage (with the amp idling) will be 25Vrms x 1.12 = 28Vrms, which when rectified becomes 28Vrms x 1.414 = 39.6Vdc, but then deduct the 1.0V drop across the bridge diodes = 38.6Vdc.
At full load the transformer will output 25Vrms so x 1.414 = 35.5Vdc, then - 1V drop in the bridge = 34.5Vdc.
All of the above calculations assume the rated input voltage of 230Vac; if it's 240Vac all the voltages will be multiplied by 240/230 or ~4% higher.
But an AC sine wave has a peak voltage that is 1.414 times the RMS value in both positive and negative directions, or 2.83 times peak to peak.
So when an AC is rectified by a voltage bridge and used to charge capacitors, the negative part is folded over (with a full wave bridge rectifier) and the capacitor charges to the peak voltage of the rectified sine wave.
There is some voltage loss in the diodes in the bridge (~1.0Vdc). The Transformer voltage will droop under load (called regulation). The Airlink 80VA toroidal is rated at 12% regulation, meaning that with no load the output voltage with be 1.12 times the rated voltage.
So the transformer voltage (with the amp idling) will be 25Vrms x 1.12 = 28Vrms, which when rectified becomes 28Vrms x 1.414 = 39.6Vdc, but then deduct the 1.0V drop across the bridge diodes = 38.6Vdc.
At full load the transformer will output 25Vrms so x 1.414 = 35.5Vdc, then - 1V drop in the bridge = 34.5Vdc.
All of the above calculations assume the rated input voltage of 230Vac; if it's 240Vac all the voltages will be multiplied by 240/230 or ~4% higher.
Thank you very much for that clear explanation. All I could dredge up from a long ago physics class was the square root of 2 being in there somewhere.
I follow your thinking jaycee, but the readings were taken with the amp behaving itself. I'll try to catch it when it's showing the fault but am spurred on by the findings of the technician in the earlier link as the symptoms are so similar.
I follow your thinking jaycee, but the readings were taken with the amp behaving itself. I'll try to catch it when it's showing the fault but am spurred on by the findings of the technician in the earlier link as the symptoms are so similar.
Last edited:
New caps are in but no change so I'll try swapping out the transformer. Voltage across the big caps read 40v DC BTW, didn't have time the check the output transistors.
I've been offered a toroidal from a friend with these specs:
primary 230v
secondary 24-0-24 V
power 200 VA
Dimensions: Ø104x44mm
Close enough? It's got five wires not six (two in, then two reds and a black out). I'm thinking that's not an issue as the blue/yellow 5th and 6th are common and read as ground on the board, which in my mind means one black wire will do the same job. It did service moving a garage door so I'm guessing it should cope with shaking a couple of bookshelf speakers.
I've been offered a toroidal from a friend with these specs:
primary 230v
secondary 24-0-24 V
power 200 VA
Dimensions: Ø104x44mm
Close enough? It's got five wires not six (two in, then two reds and a black out). I'm thinking that's not an issue as the blue/yellow 5th and 6th are common and read as ground on the board, which in my mind means one black wire will do the same job. It did service moving a garage door so I'm guessing it should cope with shaking a couple of bookshelf speakers.
40V on the rails is well within normal tolerances. At idle, I probably should have disregarded the forward voltage drops across the rectifier diodes, which would have made my calculation 39.6Vdc.
The substitute transformer with be fine as long as it fits. Being bigger (200VA) means it will be 'stiffer' (better regulation) than the original. In effect the slightly lower voltage will be offset by the better performance of the transformer meaning it will be an upgrade over the orginal.
There is a slight risk of poorer noise performance (mechanical and/or electrical), but you won't know until you try it.
You are correct; the convention is that the secondary winding is continuous with the black wire being a centre tap, and the two red wires the ends. The black replaces the blue and yellow wires of the original transformer.
If it works and solves the intermittent fault, stick with it unless you need to give it back, but at least you will know for sure whether the original is faulty.
The substitute transformer with be fine as long as it fits. Being bigger (200VA) means it will be 'stiffer' (better regulation) than the original. In effect the slightly lower voltage will be offset by the better performance of the transformer meaning it will be an upgrade over the orginal.
There is a slight risk of poorer noise performance (mechanical and/or electrical), but you won't know until you try it.
You are correct; the convention is that the secondary winding is continuous with the black wire being a centre tap, and the two red wires the ends. The black replaces the blue and yellow wires of the original transformer.
If it works and solves the intermittent fault, stick with it unless you need to give it back, but at least you will know for sure whether the original is faulty.
The P40 has a single rail supply of -55V for 20W per channel > 8R speakers (20+20W= 40W total rating. It also means it has capacitor coupled outputs and you don't need dual windings on the transformer unless you think a "dual mono" style of construction is worth the effort. For some useful pics, illustrations and legible if unusually arranged schematics, see Paul Kemble's pages: A Paul Kemble web page - the Cambridge Audio 'P' series amplifiers.
I'd spend 5 minutes with a chopstick prodding around the power supply components on the circuitboard before you give up. It's a intermittent fault so something is making and breaking. You don't need speakers connected, just a meter on the output terminals so you can see the DC offset.
Be quite forceful with your taps, even bang the whole chassis on the bench. It you can make the fault appear and disappear, then try to hone in to the spot on the board or the actual component that is sensitive to prodding. Your fault is almost certainly a dry solder point, cracked printed circuit track, lifted solder pad, or similar.
Be quite forceful with your taps, even bang the whole chassis on the bench. It you can make the fault appear and disappear, then try to hone in to the spot on the board or the actual component that is sensitive to prodding. Your fault is almost certainly a dry solder point, cracked printed circuit track, lifted solder pad, or similar.