GC amp problem

[Jamh]

Hi everyone,

I got the mica installed today, and I discovered why the caps were blowing, there was a short from two of the pins of one of the LM chips to the ground.

Now I'm listening to it and have a question: When I first power up the amp, it sounds great, but after a few seconds (maybe 10), a distortion sets it, sounds like blown speakers. Is this related to the fact that it needs breaking in? I think it's getting better, but I'm not sure.

BTW, the mica didn't make that much of a difference. ie. it pretty much sounds the same with both chips on the heatsink that is not grounded, and with the mica and the grounding of the heatsink.

Thanks,
Jam

[Jamh]

A picture of the inside

[Nuuk]

This is a general alert!

Hardwiring items together that are left 'floating' in the air will put massive strain on the joints and result in eventual failure. Alternatively leadouts will bend or sag causing a potential short circuit.

The GC amps lend themselves to hardwiring due to the small component count. This does not mean that they can be abused!

Hardwiring does not need to be a 'jungle'

[runebivrin]

If the amp starts distorting after a while, it's usually because there's no connection between the center-tap of the transformer and power ground.

Rune

[Jamh]

What do you mean by center tap? I added a ground from the bolt that goes right through the toroidal transformer, but that didn't do anything (there's no metal in the middle area). The transformer only has 2 AC leads and 4 DC outputs.

Please bear in mind that I'm just learning. This is a prototype and will not be used for anything. Hopefully if everything goes well I can use the experience and maybe the case.

I'm thinking I fried something when I had the short, and I should start with fresh parts again.

Thanks
Jam

[runebivrin]

Well, I don't know if you've wired with one or two rectifier bridges.

If you've used two - i.e. one for positive rail and one for negative - make sure you've connected the - of the positive rectifier, the + of the negative rectifier, corresponding capacitors, ground and loudspeaker minus in one point: power star ground.

If you've used one rectifier, you will have joined two of the wires from the transformer. This is the center tap. Make sure this is connected with the middle point of the filter caps, ground and loudspeaker minus.

Rune

[Jamh]

I've used two bridges. I'd like a bit more information before proceeding.

I've used one bridge per side: one for the right channel, one for the left channel. Are you saying to connect the - of both bridges to the ground? My ground is a star, and so far it has the following items attached (per channel):

+ All the metal parts, including the heatsink.
+ the - of the speakers
+ the - of the input
+ the - of the 1000 cap (its + goes to pin 1)
+ the + of the 1000 cap (its - goes to pin 4)
+ the - of the 22 uF cap

[Nuuk]

When we talk about one bridge or two, we mean per channel . So you have used one bridge in your circuit (just think of each channel separately to keep things simpler).

You should also have two ground stars, one for signal and the other for power. Take a look at my DD Gainclone pages as it is all explained there.

[runebivrin]

Near as I can tell from your description, you have one transformer winding feeding one rectifier bridge, and the start ground is not connected to the transformer at all. That's just not going to work very well, since there's nothing fixing the ground to half the supply voltage. For that you need a center tap on the transformer winding.
What you have now is not +-16.5 volts, it's just 33 volts.

You may be able to get by with something like two 470 ohm resistors across the 1000uF caps as a voltage divider, but it's hardly recommended.

Judging from the voltages you show in the first post, it might be best to connect - from one rectifier to + of the other, and use that point as ground. That'd give you +-33 volts, which would give you more power as well (provided your cooling is sufficient.)

Rune

[Jamh]

I see what I did wrong. Thanks!