Oh dear! Gainclone and bust drivers

[ericallan]

I've built 2 gainclones before with no problems and thought I'd have a go at a BrianGT kit I bought about a year ago. I made it dual mono with 2 seperate rectifier bridges and 2 transformers. Tested the voltage and got +/- 35 on each side, so went on to connect up the amp boards (basic design with just the resistors and 1500uF caps. Tested the offset and got a few millamps on each side, but when I connect a source and speakers up, I got a loud buzzing from both speakers and then a high pitched whine from one channel. This stopped when I turned the source off.
However, I have 2 burnt out mid/bass Kef B110 drivers which aren't cheap to replace.
I assume the amp was oscillating but I daren't use it again and don't have access to an oscilloscope. Shall I throw it in the bin?

[Sander]

Please don't throw it in the bin. Instead of that mail it to me.

Do you have relatives with a scope? Maybe you can ask them.

[Vikash]

measured offset in milliamps or mv?

I think throwing it away is a bit drastic. Some careful debugging should get it sorted. Persevere you must young DIYer!

[PauSim]

Originally posted by ericallan
Tested the offset and got a few millamps on each side, but when I connect a source and speakers up, I got a loud buzzing from both speakers and then a high pitched whine from one channel. This stopped when I turned the source off.
However, I have 2 burnt out mid/bass Kef B110 drivers which aren't cheap to replace.

Milliamps or millivolts?

When I kept burning my Cresta 2 drivers I sent them to a local winder to recoil. It cost 25 € against the 180 € the KEF distributor asked for a pair of new drivers.

[ShinOBIWAN]

Originally posted by ericallan
I've built 2 gainclones before with no problems and thought I'd have a go at a BrianGT kit I bought about a year ago. I made it dual mono with 2 seperate rectifier bridges and 2 transformers. Tested the voltage and got +/- 35 on each side, so went on to connect up the amp boards (basic design with just the resistors and 1500uF caps. Tested the offset and got a few millamps on each side, but when I connect a source and speakers up, I got a loud buzzing from both speakers and then a high pitched whine from one channel. This stopped when I turned the source off.
However, I have 2 burnt out mid/bass Kef B110 drivers which aren't cheap to replace.
I assume the amp was oscillating but I daren't use it again and don't have access to an oscilloscope. Shall I throw it in the bin?

Sorry to hear about that.

The KEF B110 drivers are not only expensive but they are increasingly rare these days.

I hate to say I told you so but you really should have tested the amps into a dummy load first and then into some cheap drivers that you don't care about.

Only when you've tortured the amp for a good while should you connect the expensive stuff.

Thats amp building 101.

[richie00boy]

Not only are they rare, but there are a few different varients of the B110 and you must get the correct one.

As shinobiwan has said, not a nice lesson to learn.

Did you use a zobel network? Did you have a DC blocking capacitor at the amp input?

[ericallan]

Originally posted by richie00boy
Not only are they rare, but there are a few different varients of the B110 and you must get the correct one.

As shinobiwan has said, not a nice lesson to learn.

Did you use a zobel network? Did you have a DC blocking capacitor at the amp input?

I've had about a dozen pairs of B110s go through my hands - but this were the last. KEF were repairing them not too long ago, so its just the repair bill that stings.
I didnt use a zobel or DC cap - thats a lesson learnt. It didnt sound like DC though - more like a very bad earth.
One question - I assumed that the earths being tied at the rectifiers was enough, but do they need to meet later in the circuit as a star ground?
Cheers for the info so far - I'll try to get pics here tomorrow.

[richie00boy]

A pop of DC as you plug in a source (and a lack of protection from this at the input) can be all that is needed to knock the chip amp into oscillation, especially as it will be over sensitive to this with the lack of zobel network.

If you have used a twin bridge PSU the centre of the caps is your star ground and should be the only ground return point in the system. Were your input sockets isolated from the case? Was your case connected to mains earth? Was mains earth connected to the star ground?

[ericallan]

Originally posted by richie00boy
A pop of DC as you plug in a source (and a lack of protection from this at the input) can be all that is needed to knock the chip amp into oscillation, especially as it will be over sensitive to this with the lack of zobel network.

If you have used a twin bridge PSU the centre of the caps is your star ground and should be the only ground return point in the system. Were your input sockets isolated from the case? Was your case connected to mains earth? Was mains earth connected to the star ground?

The pop of DC would have come from the cheap portable cd player as it was turned on - it was all quiet before that. The input sockets are insulated but no earth connection to the case yet and the mains earth was not connected to the star ground.

And to the earlier question - it should have read millivolts.

I dont like to moan about such nice pcbs, but the build instructions still need some way to go to make a foolproof amp.

Anyway - managed to get some piccys;

[richie00boy]

If your CD player spits out DC it doesn't matter so much if you an a capacitor at the input of your amp. Did you have such a capacitor?

You should stick with insulated input sockets, do not earth/ground them individually.