turn-off transients with LM1875

[Dennis Hui]

Hi there,

Has anybody experienced loud turn-off transient
with their GC's?

I put together one GC channel with an LM1875
yesterday. It's in inverted configuration with
the addition of a 1R/0.22uF Zobel and is running off
a -12.6-0-12.6 transformer. I have no turn-on
transient problems and the DC-offset is a reasonable
14mV. The turn-off transient, however, is producing
a pop that is loud enough for me to worry
about my test speaker...

The LM1875 doesn't seem to have the turn-on/off
transient suppression of the LM3875. So I'm wondering
if this is what I should expect, or is it (more likely)
indicative of some flaw in construction on my part?

Any suggestions or comments are welcomed.

Thanks,
Dennis

 

[ashwin]

pop and thump

Hi Dennis,

I've got a 1875 amp(non-inverting) too, and there are two turn-off problems with it: a turn off thump, and a turn off pop. The thump I got rid of by using relays in series with the speakers, but the pop still remains. I suspect the thump is due to the 1875's supply ramping down on turn off, and the pop is probably due to arcing in the power switch which injects noise on the supply rail before the relays can turn off. My GC is very susceptible to line noise, because I hear pops when lights in another room are turned on or off!
Experts, is my guess on the cause of the popping right? If so, how does one get rid of it? There must be some simple method (without using huge power supply filter coils etc.), because even my brother's boombox doesn't notice lights going on and off. Would ferrite beads on the supply cable help? (I'm no expert, so don't laugh at my suggestions )

- Ashwin

 

[ashok]

Noise from the power switch?

Hi Dennis,
Some audio equipment can be sensitive to transients at the power on/off switch. Turning the system off would create transients at the switch contacts. This also generates an RF burst. Some amps can pick this up and produce a sharp crack at the output.

This sounds different from a "pop". The "pop" is more "rounded" and so may be due to one rail dropping voltage faster than the other. You might need a scope to check this out.
Try disconnecting power at the secondary side of the transformer ( which goes to the supply diodes and capacitor). If the noise goes away, the problem may be at the mains supply switch.

Cheers.

 

[Nuuk]

Try disconnecting power at the secondary side of the transformer ( which goes to the supply diodes and capacitor).

Am I right in suggesting that when doing the above, BOTH rails must be 'cut' at the same time?

One rail to my GC was cut when a wire broke and the result was the full voltage appeared on the speaker terminals.

 

[carlosfm]

The LM1875 doesn't seem to have the turn-on/off
transient suppression of the LM3875.

No, it doesn't.
I found out that too.
I'm making an LM1875 GC and yesterday I was testing.
After disconnecting the mains plug from the wall, the amp continues playing for about 3 seconds, and then... PLOC!
But it's not so loud, and it seams independent of the volume setting.
It continues playing for about 3 seconds and I only have 1000uf caps, bypassed with 0.1uf...
And a weak transformer, from an old HP Deskjet PSU.
This will end up on what probably will be the smallest GC in the world.
I'll open a new thread one of these days with details and pics.
Oh, I have 0.0mv DC offset.
With the 50k multi-turn pots.
Nuuk, this time I removed the pot and measured: 36k.
With a direct wire I had 40mv!
The only chip where I couldn't use a pot to control DC offset was the LM3886.

 

[jonclancy]

The Smallest GC..........

Carlos,

how small is small????????

Quite small or really small.



Pics pse!!

Jon

 

[carlosfm]

Re: The Smallest GC..........

Carlos,

how small is small????????

Quite small or really small.



Pics pse!!

Jon

No box yet, but it will be small.
And as for the pics, you'll have to wait that I finish the film, because I don't have a digital camera.

 

[peranders]

Have you tried to have a VDR across the primary windnings. Maybe this transient killer can kill some bumbs or cracks if the problem comes from sparks of the mains switch. I do not rocemmend to break the power at the secondary side. Not a good idea, due to safety reasons.

 

[Dennis Hui]

Thanks for all the comments.

Cheers,
Dennis

 

[George Kubitsch]

Turn-off transient

Hi Dennis!

I made my inverted GC ( with LM 3875) about two months ago.
I have met the same turn-off problem, which is caused by oscillation.
I have got a lot of advises in this forum, but none helped.
I think the problem is with the quality of the bypassing 1000u
capacitors on V+ & V- ( on pin 1 & pin 4 ).
I changed the Low ESR 1000u type capacitors, which was applied by
me at first, to normal BC (Philips) 1000u caps and the problem has
gone at once.

I hope this helps
George

 

[JOE DIRT®]

a cap across the switch contacts also helps...I use .01uf 250vac ceramic

DIRT®

 

[fedde]

Or make mute switches for the speakers. The original GC has them too. I also found that an oscillation causes turn-off transients.

Fedde

 

[ashwin]

a cap across the switch contacts also helps...I use .01uf 250vac ceramic

DIRT®

Yes, that's what finally worked for me. I tried a spice sim, and I found that a resistor in series with the cap (say 470n in series with 330ohm) brings down the spike drastically compared with just the cap. But in my actual amp the cap alone worked just fine.

Now I need to rig up a good EMI filter...