So last year I bought a pair of L20.5 amplifiers. On first power up one was fine and the other burned up the outputs without a signal.
I finally got around to fixing it today and while testing it with a dummy load (no speaker) I noticed it was whistling. I have never heard an amplifier whistle or hum before. (I have heard transformers or SMPS make noise but not a solid state amplifier board.) With the exact same setup the following do not hum or whistle: L20 V7, uPC1342V, MX50X2, MX50SE and YJ L28. Note that the whistle noise is not coming from the dummy load. It is definitely coming from the amplifier board.
I am not yet sure which component is whistling but I think it might be coming from near the 10 Ohm resistor and the yellow film/safety capacitor. (See the red arrows in the attached picture.) I can't image why any of the components on this board should hum or whistle.
I have also attached the result from the THD test which indicates that the amplifier is basically working properly. Offset is 5 mV. Bias is a little low but I can adjust that later.
Does anyone have any idea why the L20.5 might whistle during a 1kHz THD test? (At 64W.)
Could it be in any way related to why several L20.5 owners have reported sudden failure (blown outputs and more)?
I finally got around to fixing it today and while testing it with a dummy load (no speaker) I noticed it was whistling. I have never heard an amplifier whistle or hum before. (I have heard transformers or SMPS make noise but not a solid state amplifier board.) With the exact same setup the following do not hum or whistle: L20 V7, uPC1342V, MX50X2, MX50SE and YJ L28. Note that the whistle noise is not coming from the dummy load. It is definitely coming from the amplifier board.
I am not yet sure which component is whistling but I think it might be coming from near the 10 Ohm resistor and the yellow film/safety capacitor. (See the red arrows in the attached picture.) I can't image why any of the components on this board should hum or whistle.
I have also attached the result from the THD test which indicates that the amplifier is basically working properly. Offset is 5 mV. Bias is a little low but I can adjust that later.
Does anyone have any idea why the L20.5 might whistle during a 1kHz THD test? (At 64W.)
Could it be in any way related to why several L20.5 owners have reported sudden failure (blown outputs and more)?
Attachments
Last edited:
I’m unfamiliar with the amp in question but it looks like the output zobel network
Check the resistance of it is intact and the network (the cap and resistance) is definitely connected to ground and output
Check the resistance of it is intact and the network (the cap and resistance) is definitely connected to ground and output
This is speculation but I suspect the whistle is likely related to the fault that caused the original failure. The whistle is a from electromechanical stress of the component(s) making the noise, possibly from high current somewhere unexpected. How does the supply current of the two channels compare?
Also due to the construction of capacitors as multilayered, and because dielectric materials are not infinitely stiff, capacitors can make significant noise under some conditions. In this case it may be that the capacitor (or another component nearby) is of faulty manufacture.
If you have an open stethoscope and a conduction stethoscope (AKA a mechanic's stethoscope) you should be able to identify the individual components generating the noise.
Also due to the construction of capacitors as multilayered, and because dielectric materials are not infinitely stiff, capacitors can make significant noise under some conditions. In this case it may be that the capacitor (or another component nearby) is of faulty manufacture.
If you have an open stethoscope and a conduction stethoscope (AKA a mechanic's stethoscope) you should be able to identify the individual components generating the noise.
Well I very carefully tried to damp the sound by touching components. (Very carefully, as there is nearly 120V=2x60V.)
It does not appear to be the Zobel network. (The 10 Ohm resistor or yellow safety cap.) But I can modulate the sound by touching the 0.1 Ohm BPR 5W resistors. Never heard sound from those on an amplifier before. Maybe someone else has?
It does not appear to be the Zobel network. (The 10 Ohm resistor or yellow safety cap.) But I can modulate the sound by touching the 0.1 Ohm BPR 5W resistors. Never heard sound from those on an amplifier before. Maybe someone else has?
Get a wooden spoon handle if you're bothered by the rails 😀, and you can prod about (and really put some pressure on), to your heart's content.
Try resoldering those plate resistors and putting them snug against the board.
You could also try some silicone to damp the resonant component.
Surely the right approach is to fix the faulty component or design fault, not suppress the symptoms?
Did you connect a scope to the output? It's normal to be able to hear music coming from an amplifier at high output levels. You won't hear it with a speaker connected (for obvious reasons) but you will hear it with a dummy load connected. Maybe the amp is oscillating and that what you are hearing? That would explain the failure.
I'm not suggesting leaving it like that once rework shuts it up , but it might pinpoint the dodgy component
Yeah I agree it's agood idea to "see" if there is oscillation , which is why I was thinking he zobel wasn't fully in circuit
I’ve had this before. In my case, I used a 1 nF ceramic cap from the base to collector on the driver transistors on an EF3.
When I output > 40 V pk-pk into an 8 ohm dummy load, the caps sang (whistled) at the test frequency.
Cause: the ceramic in the cap dielectric has a piezo electric effect, so when you apply a voltage across it, the material flexes, and this deformation generates an audio OP. I replaced my ceramic caps with COG types and the problem was solved.
Try replacing the Zobel caps with a film cap or, if you think tne cap is film, try a better quality cap. Zobel caps are typically 0.1 uF, so COG is probably not an option - use film types.
(Do not operate your amp without the Zobel!).
When I output > 40 V pk-pk into an 8 ohm dummy load, the caps sang (whistled) at the test frequency.
Cause: the ceramic in the cap dielectric has a piezo electric effect, so when you apply a voltage across it, the material flexes, and this deformation generates an audio OP. I replaced my ceramic caps with COG types and the problem was solved.
Try replacing the Zobel caps with a film cap or, if you think tne cap is film, try a better quality cap. Zobel caps are typically 0.1 uF, so COG is probably not an option - use film types.
(Do not operate your amp without the Zobel!).
OP obviously did not explain the "whistle" is with or without input signal.
May be a good idea to connect the scope instead of forcing members to
speculate.
If the amp is making these noises all the time, even without signal, it is
oscillation.
On the other hand, driving it with some elevted level and dummy load
you will certainly hear the mechanical stress associated with this.
By the way - everybody knows what L20.5 is, isn't it ?
May be a good idea to connect the scope instead of forcing members to
speculate.
If the amp is making these noises all the time, even without signal, it is
oscillation.
On the other hand, driving it with some elevted level and dummy load
you will certainly hear the mechanical stress associated with this.
By the way - everybody knows what L20.5 is, isn't it ?
The audible noise is only made when the input signal is applied. I only noticed it when I stepped from 16W to 64W output. There is no signal on the output with the generator off. (And no audible noise.)
I had not heard such audible noise from an audio amplifier component before (apart from transformers and SMPS supplies).
At this point I am sure that it is the 0.1 Ohm BPR 5W resistors. I did not notice this sort of noise on the L20 but that has eight 5W BPR resistors. The L20.5 only has two. I wonder if they are a little undersized considering that there are eight output transistors and only two 5W BPR resistors?
The L20.5 is rated by the seller as a 250W amplifier into 4 Ohms with a ±50V power supply. It has eight output devices in a CFP configuration. It is largely based on the Blameless amplifier.
I had not heard such audible noise from an audio amplifier component before (apart from transformers and SMPS supplies).
At this point I am sure that it is the 0.1 Ohm BPR 5W resistors. I did not notice this sort of noise on the L20 but that has eight 5W BPR resistors. The L20.5 only has two. I wonder if they are a little undersized considering that there are eight output transistors and only two 5W BPR resistors?
The L20.5 is rated by the seller as a 250W amplifier into 4 Ohms with a ±50V power supply. It has eight output devices in a CFP configuration. It is largely based on the Blameless amplifier.
Last edited:
Yes, I see that you made it clear somewhere else :
https://www.diyaudio.com/forums/solid-state/323965-l20-5-ap-sys-test-data-design-18.html#post6800262
Some components will sing when stressed under load.
Change the test frequncy and "see" yourself.
Every young guy knows it from early radio experiments.
https://www.diyaudio.com/forums/solid-state/323965-l20-5-ap-sys-test-data-design-18.html#post6800262
Some components will sing when stressed under load.
Change the test frequncy and "see" yourself.
Every young guy knows it from early radio experiments.
That's cos it's normally obscured by the speaker level 🙂
Never heard of it 'til this thread, but I see other audio stuff mostly in my day job. (I don't do diy audio - busman's holiday 😛)
If a component makes noise, then it is physically changing dimension in at least one plane. It's very likely that the dimensional change is also introducing nonlinear distortion into any any signal passing through that component, though whether the level of intermodulation caused is audible or not depends on how bad it is and whether it's buried in all the other contributions of distortion.
Last edited:
If the amp was oscillating badly enough to whistle, the FFT plot would likely look a lot different.
My money is on a ceramic component as the culprit. A power amp and a good PSU should be absolutely silent at any output power level and frequency
🙂
My money is on a ceramic component as the culprit. A power amp and a good PSU should be absolutely silent at any output power level and frequency
🙂
A current coil with loose turns may sound. A capacitor under voltage of an audio frequency - film, electrolytic or ceramic - can sing. A wire wound resistor is also a coil.
Dial resistors from several SMD 2106 connected in parallel.
Dial resistors from several SMD 2106 connected in parallel.
Last edited:
Ceramic capacitors other than C0G are piezo electric and are both microphonic and sing quite strongly at high AC voltages. This is one of the reasons why they are best as decoupling caps and not directly in the signal path (such as the Zobel network).
However all capacitors and inductors will exhibit the effect to some degree as mechanical forces are produced by electric and magnetic fields, and forces deform the component.
However all capacitors and inductors will exhibit the effect to some degree as mechanical forces are produced by electric and magnetic fields, and forces deform the component.
