¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io

Muddy audio in AF power amp


wn4isx
 

The new transistors are oscillating at about 1GHz, the frequency is frequency modulated by the incoming audio!
?
The cure was a ferrite bead on the base and collector (now that's a first for me) and a 10pF SMD capacitor between base and emitter.
I grabbed my Tiny SA on a wild hunch, found a spike and borrowed a Signal Hound USB spectrum analyzer to confirm the oscillation.
?
The semiconductor industry constantly "improves" their products.
Original 1N4004 switch slow enough to be a minimal EMI source, modern 1N4004 switch fast enough to be EMI sources from hell.? The silicon is purer, the junction is thinner and the turn off time is a lot faster.
Someday really soon I hope to post a simple test device to see how much commutation noise a diode produces.
?
?
?
?
?


 

On Mon, Dec 9, 2024 at 02:46 PM, wn4isx wrote:
The new transistors are oscillating at about 1GHz, the frequency is frequency modulated by the incoming audio!
I am trying to figure out if this is the start of a new conversation, or if it is meant to be a reply to another one.? Without any context, I assume it is the former.
?
(It is quite challenging for me, trying to keep up with so many conversation in every interesting group.)
?
With regards to RFI, oscillations, and changing,silicon, we do live in interesting times.? :-(
?
Andy
?


 

¿ªÔÆÌåÓý

Exactly as I suspected. That means THERE IS harmonic distortion on a simple assembly.

Thanks for confirming!

Can I ask if you noted any of the harmonics at the audio part of the spectrum?

?

Nuno T.

From: [email protected] [mailto:[email protected]] On Behalf Of wn4isx via groups.io
Sent: 09 December 2024 19:47
To: [email protected]
Subject: [electronics101] Muddy audio in AF power amp

?

The new transistors are oscillating at about 1GHz, the frequency is frequency modulated by the incoming audio!

?

The cure was a ferrite bead on the base and collector (now that's a first for me) and a 10pF SMD capacitor between base and emitter.

I grabbed my Tiny SA on a wild hunch, found a spike and borrowed a Signal Hound USB spectrum analyzer to confirm the oscillation.

?

The semiconductor industry constantly "improves" their products.

Original 1N4004 switch slow enough to be a minimal EMI source, modern 1N4004 switch fast enough to be EMI sources from hell.? The silicon is purer, the junction is thinner and the turn off time is a lot faster.

Someday really soon I hope to post a simple test device to see how much commutation noise a diode produces.

?

?

?

?

?


--
Nuno T.


wn4isx
 

There weren't any harmonics but the audio sounded, and I hate this discriptor, muddy, smeared.
There might have been a very weak comb effect.?
?
I fed band width limited white noise, sharp cut off at 20,000kHz, 9dB/octave, ran the amps at 10W and the mid range, ~1000 to ~4000Hz was "depressed."?
?
In one of his good amps, the peaks of the white (random) noise form a nearly straight line, in the 'bad' amps, the midrange is down about 6dB.?
?
I've seen noise spectra of many amps and never anything like this. I'd have expected the upper frequencies to be messed up.
?
And the need for a ferrite bead in the collector lead as well as the base and a 10pF cap between base adn emitter....
?
This was a dozzy!
?
We modified enough transistors for 5 amps, they all worked fine and then we went out for a nice steak, his treat.
?


 

¿ªÔÆÌåÓý

Thank for the feedback.

?

Hope you enjoyed the meal! J

?

Nuno T.

?

From: [email protected] [mailto:[email protected]] On Behalf Of wn4isx via groups.io
Sent: 09 December 2024 22:43
To: [email protected]
Subject: Re: [electronics101] Muddy audio in AF power amp

?

There weren't any harmonics but the audio sounded, and I hate this discriptor, muddy, smeared.

There might have been a very weak comb effect.?

?

I fed band width limited white noise, sharp cut off at 20,000kHz, 9dB/octave, ran the amps at 10W and the mid range, ~1000 to ~4000Hz was "depressed."?

?

In one of his good amps, the peaks of the white (random) noise form a nearly straight line, in the 'bad' amps, the midrange is down about 6dB.?

?

I've seen noise spectra of many amps and never anything like this. I'd have expected the upper frequencies to be messed up.

?

And the need for a ferrite bead in the collector lead as well as the base and a 10pF cap between base adn emitter....

?

This was a dozzy!

?

We modified enough transistors for 5 amps, they all worked fine and then we went out for a nice steak, his treat.

?


--
Nuno T.


wn4isx
 

While the steak was great, helping a friend solve the "impossible" was a lot more rewarding.
?