diyAudio (
-   Solid State (
-   -   Cambridge Audio 640A Overheating / Distortion (

nickalmond 17th March 2013 12:34 PM

Cambridge Audio 640A Overheating / Distortion

Hoping someone can help me diagnose the problem with my Cambridge Audio 640A v1 amp - I have had it for about 4 years now and recently (6 months ago) it began severely overheating until the protection would kicked in after about 15minutes of it being powered on. To temporarily 'solve' this, i had a 12v fan running through a temperature sensor to cool it down. This ran fine up until last month when the left channel began intermittently cutting out - to resolve i simply wiggled the volume control a little and it sprang back into life. Not wanting to damage my new speakers (B&W 684's) i decided to buy a new amp which is working perfect.

I am hoping to get the Cambridge unit up and running again as it was a great amp during its first few years of use and would like to do the work myself to hopefully learn a little more about them.

I had opened the unit and found the heatsink for the faulty channel was rising to around 80 Degrees Celsius before cutting out - with or without an audio input. I took the PCB and heatsinks out, checked over all the contacts, removed some thin plastic sheets that sat between the heatsink and the thermal paste (not sure if this was a wise thing to do), cleaned it up and re-applied new thermal paste. All contacts seems in good condition.

All this made little difference to the heat output so after reading for a couple of days on various websites, I decided to open the unit up again and have another 'play'. I adjusted the BIAS pot to a lower value which made a seemingly huge difference to the temperature of the left channel (faulty) heatsink - I left the amp on for several hours and no overheating occurred (still a little hotter than the heatsink for the right channel).

I connected a cheap hifi speaker to the working channel and audio played perfectly as expected, followed then by the faulty channel and got huge distortion throughout the entire range when playing back any audio. Working backwards, I turned the amp off, adjusted the BIAS pot back to what i believe was its original position, reconnected the speakers and still the same heavy distortion along with the overheating.

This is where i am up to and not sure if it is simply a case of replacing the pot or if there is now a bigger problem - also not sure how to correctly test the pot.

The only additional note i can add is that I tested the voltage across two other pins of the chips noted 'SAP15' (sorry, don't know their correct name but please see first image) - the voltage jumps around between 2v and 6v through the working channel but stays fairly idle at around 2v on the faulty channel.

Apologies for the long post but i really appreciate any help anyone can give me. Happy to do further testing - have a decent multimeter to hand and fairly good at soldering. I should also note that I only have a basic knowledge of electronics so apologies if it seems like i am asking daft questions.

Many thanks,

Service Manual;


Plastic Pieces Removed;

davidsrsb 17th March 2013 12:49 PM

The SAP15s are the output transistors with built in temperature measurement diode, hence 5 pins
Was this amplifier built with the (expensive) multi-turn pot?
The "plastic pieces" are insulators for the output devices. You need them!

nickalmond 17th March 2013 01:03 PM


Brilliant. Thank you, will pop the plastic pieces back in.

The pot is a multi-turn one, yes. I counted 10 full turns which reduced the temperature significantly to stop the protection module kicking in. It is currently back in what i believe is its original position (counted 10 turns in the opposite direction).

Thank you :)

jaycee 17th March 2013 04:06 PM

It sounds to me like the faulty channel is oscillating. Does R44 (left) or R50 (right) look burnt?

Also, yes, the "plasitc" bits are actually mica insulators, as the back of the power transistors are connected to the power. If you bolt them straight to the heatsink, you will create a short circuit. Hopefully, you haven't powered the amp up without these in place!

To refit the mica washers you will need to clean the white heatsink paste off of everything (the insulators both sides, the heatsink and the back of the transistors), and apply a thin layer of thermal paste between the heatsink and insulator, then between the washer and transistor. I use a toothpick to spread it out. You need a THIN layer only - no more than a pea sized blob.

nickalmond 17th March 2013 05:16 PM

Hi jaycee. Thank you for your reply. I have checked both R44 and R50 and both look perfect.

I have indeed powered it on without the insulators (oops!) - it was turned on approximately 5 times for a period of about 30seconds. Both speakers connected to the left and right channel. The right channel continued sounding perfect. The light channel had the distortion. I will make sure the insulators are replaced before it is next powered on for testing.

Could the problem with overheating be anything to do with the BIAS pot or am i way off here? As i understand it, the BIAS limits the amount of current sent through on 'idle' - as in the original thread, the left side was overheating and therefore suggests to me that too much 'power' is being sent through. Reducing the pot results in less heat. If this is even a remote possibility, i would be happy to try replacing it and re-post my results.

Thank you again for your reply, its much appreciated.

Mooly 17th March 2013 05:35 PM

If you fit a link (to short out) across C13 that will force the bias current to zero (assuming the outputs are OK).

Have you measured the DC offset yet across the speaker output ? Should be zero volts DC give or take a few 10's of millivolts.

nickalmond 17th March 2013 05:56 PM

Hi Mooly.

Thank you, that will give me something to go off - will short C13 and check the left channel again.

I did measure the DC offset earlier on today after reading another post on this forum - with no speakers connected, i had 4.7mv through the faulty left channel and 11.8mv through the right working channel.

Thank you.

Mooly 17th March 2013 06:00 PM

The offset is surprisingly low for a DC fault... which is good :)

Maybe your problem is the pot (have you measured it when turned to minimum resistance, it should read as a short), or perhaps one of the diodes built into the output devices is faulty. Shorting C13 should give zero quiescent current and the outputs should be cold. The amp will still work and play but with crossover distortion present.

jaycee 17th March 2013 06:11 PM

You got lucky - the anodising on the heatsink insulated them enough. If you remove the transistors, you should always apply fresh paste.

The bias pot adjusts the idle current. Think of it like the idle speed on a car engine. It should be set to allow just enough current through the transistors to get rid of crossover distortion.

Too much bias will result in the transistors getting hot, but doesn't explain the distortion. I would say something else is wrong rather than just simply a fault in the bias circuit. These amps are not really very robust and can go faulty especially when driven hard.

nickalmond 17th March 2013 06:19 PM

Hello again :). With your help, it appears i am making some progress - I have shorted C47 with a piece of wire (C47 being on the left channel) and temperature of the heatsink is right back down to a more than acceptable level which i guess indicates that the BIAS pot has some form of fault?

The heavy distortion however is still present on the same channel (only started after i adjusted the BIAS and removed the heatsink). I will take the PCB back out again and re-seat all the connections and the heatsink just to make sure i haven't missed anything.

Would it at all help if I recorded the distorted sound?

Also, do you know where i can find a replacement pot? It doesn't provide much information on the pot itself - would Cambridge Audio be the best place?

Thank you again :)

All times are GMT. The time now is 01:55 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio