tabs touching, no output

[Matttcattt]

has anyone else found that when the tabs of two gainclone chips are conected electrically, one channel has no output? both my gainclones are mouned directly on the heat sink, but they are on seperate heatsinks. today, i tried connecting both heatsinks together with a pice of metal, and one channel stoped working! i immediatly disconnected power, and felt the chips, they were not even warm, so i tried again, with the same result. has anyone else noticed this? i do not know why this would happen, because both tabs are at -v.

 

[Cradle22]

Hi!

One thing that comes to my mind is that even if you connect both to the same supply, I bet you still use seperated filter caps for supply voltage.

So both ships are not "really" sharing the same supply, because of their own filters...

I will not check this out, but on my P3A I have one toroid, but apart from that totally seperated power supplies... and I bet that my P3A woult not like it very much if I connected V- from one supply with V- from the other...

They still share mass, though...


But maybe someone else can shed more light on this...


Bye,

Arndt

 

[Matttcattt]

no, one chips -v pin is connected directly to the other's -v pin.

 

[raindog]

to what pin is the tab conected (if so?)😀

 

[Matttcattt]

to what pin is the tab conected (if so?)😀

pins 3 and 4.

 

[Jennice]

Hi folks,

I haven't built a gainclone, so I'm taking a wild guess about the typical IC amp circuits here...

Am I right in assuming that the inputs are (somewhere along the electrical path) connected at the audio source or pre-amp?

In that case the circuits have at least one common reference level.
Connecting the heat sinks (and thereby taps) you create a second common level (more or less). Since this is not at precicely the same potential, it causes a fault current, which is sensed by one of the sensing circuits on the chip. This triggers it's alarm state. This, in turn, shuts off the output to avoid self-destruction or un-intentended operation.

Try checking for multiple electrical paths for the same signals/potentials...

Jens

 

[zobsky]

on the side, ..... does anyone have 2 chips mounted on the same heatsink? ..i thoght i saw some pictures on the www with this arrangement

because this is what i intend to do, ..scrounged a working transformer from an old reciever and got my 3845 chips yesterday.

 

[tiroth]

Generally the chips will be isolated from the heatsinks at any rate. These amps usually stay cool enough that you can afford the insulator.

 

[carlosfm]

Matttcattt,
You forgot to say what power op-amps are you talking about.
I think you use OPAs, don't you?

I remember reading on one datasheet (don't remember what) that the tabs shouldn't be used do conduct power.
That is to say they should be isolated.
But if you used a single supply (+ and ground) the tabs would be at ground potential and no problems.

 

[Matttcattt]

Matttcattt,
You forgot to say what power op-amps are you talking about.
I think you use OPAs, don't you?

I remember reading on one datasheet (don't remember what) that the tabs shouldn't be used do conduct power.
That is to say they should be isolated.
But if you used a single supply (+ and ground) the tabs would be at ground potential and no problems.

oops, forgot 🙄

yes, OPA541, free samples 😎

 

[Matttcattt]

is anyone else going to comment? or should i just forget this and just insulate the chips from the heatsink?

 

[zobsky]

hi,

i belive that insulation using a thin mica insulator was the recommended method on some of the diy-gainclone guides

cheers

 

[JOE DIRT®]

are using thermal grease on the tabs???....also have you disconnected the -v hook up wires???.....and tell me the heatsinks are isolated from the chassis


DIRT®

 

[Matttcattt]

are using thermal grease on the tabs???

yes

have you disconnected the -v hook up wires???

no

tell me the heatsinks are isolated from the chassis

i havent got a chassis yet, i just tried connecting the heatsinks together

 

[JOE DIRT®]

maybe the thermal compound is causing enough of a resistance.....also another Q.....is each channel on the same power supply???

 

[Matttcattt]

maybe the thermal compound is causing enough of a resistance.....also another Q.....is each channel on the same power supply???

yes

but i want to know why they wont work with the tabs connected.

 

[JOE DIRT®]

my last kick at the cat Matt...and I may be wrong here...but some manufacturers do connect a power IC substrate to the mating surface and although it is a common connection it wont conduct enough...they refrain you from using a heatsink for a common connection because they know of possible isolation problems
until you tell us some voltage readings and what testing you have have done I`m kinda stuck about your problem as I have not read the data sheet or am familiar with the chip you are using.....but I`m sure one of the audio pro`s here will have a better clue😉


DIRT®

 

[Matttcattt]

my last kick at the cat Matt...and I may be wrong here...but some manufacturers do connect a power IC substrate to the mating surface and although it is a common connection it wont conduct enough...they refrain you from using a heatsink for a common connection because they know of possible isolation problems
until you tell us some voltage readings and what testing you have have done I`m kinda stuck about your problem as I have not read the data sheet or am familiar with the chip you are using.....but I`m sure one of the audio pro`s here will have a better clue😉


DIRT®

ok, ill read the voltages when i get a chance.

 

[Matttcattt]

i have mesured the tabs, from ground. they are at 0.3v from ground. not -15v as they should be...

 

[tiroth]

Is the tab voltage 0.3V for both units when disconnected?

You should list the separate disconnected voltages as well as the common connected voltage.