Hifonics Brutus sounds like ol' shortwave radio

[valterdaw]

Bought dead Brutus from ebay just for fun (BX1500D). It had a driver IC burnt from overheating or reverse polarity, who knows. All FET's seems to be fine... Now it works, puts a full power, a lots of power The only problem that it has a high frequency sound noise (actually very entertaining signal), variable, just like very old shortwave radio when you are tuning it, not loud, and not very noticable with sub, but still annoying. What it can be, and where to look at

[Perry Babin]

If the noise is barely audible and not related to engine speed, it's likely that it's due to the interaction of the two oscillators (one for the PS and one for the class D audio section). Some amps synchronize the two oscillators to prevent this. I don't think it's possible with the TLx94 IC (which is what I believe that amp uses). Amps that use the SG352x IC often have the oscillators sync'd.

If all of the filter caps at the output are OK and there are no shorted inductors, it could be normal for that amp.

[djQUAN]

Originally posted by Perry Babin
If the noise is barely audible and not related to engine speed, it's likely that it's due to the interaction of the two oscillators (one for the PS and one for the class D audio section). Some amps synchronize the two oscillators to prevent this. I don't think it's possible with the TLx94 IC (which is what I believe that amp uses). Amps that use the SG352x IC often have the oscillators sync'd.

If all of the filter caps at the output are OK and there are no shorted inductors, it could be normal for that amp.

I have seen from the TL494 datasheet that it can be synchronized from a master TL494 oscillator but you may have a different idea of synchronizing the amp and SMPS that doesn't allow this possible?

[Perry Babin]

If I'm not mistaken, when the TLx94 IC is slaved, it simply uses the master IC's oscillator instead of its own on-board oscillator.

The on-board oscillator is a sawtooth waveform, it would be somewhat difficult to derive that from the class D section's signals. In a typical amp, you'd have to divide the frequency, produce the sawtooth waveform (could be done from the integrator output on some amps) and then change the level to suit the TLx94 IC's logic section.

[valterdaw]

Originally posted by Perry Babin
If the noise is barely audible and not related to engine speed, it's likely that it's due to the interaction of the two oscillators (one for the PS and one for the class D audio section). Some amps synchronize the two oscillators to prevent this. I don't think it's possible with the TLx94 IC (which is what I believe that amp uses). Amps that use the SG352x IC often have the oscillators sync'd.

If all of the filter caps at the output are OK and there are no shorted inductors, it could be normal for that amp.

Thank you! Noise is a mixed product of several harmonics, each around 10-15 kHz and, unfortunately, it is noticable (may be because, I know it's there). Time to walk through everything one more time I guess

[djQUAN]

Originally posted by Perry Babin
If I'm not mistaken, when the TLx94 IC is slaved, it simply uses the master IC's oscillator instead of its own on-board oscillator.

The on-board oscillator is a sawtooth waveform, it would be somewhat difficult to derive that from the class D section's signals. In a typical amp, you'd have to divide the frequency, produce the sawtooth waveform (could be done from the integrator output on some amps) and then change the level to suit the TLx94 IC's logic section.

so it is still possible to sync when using the TL series oscillators. not what you said in the previous post. (well, not directly but possible.)

[Perry Babin]

OK, it's technically possible but not practical and as far as I know it's not done in any commercially available amplifier that uses the TLx94 IC.

If someone knows of an amplifier using the TLx94 IC that syncs the class D oscillator to the power supply oscillator, please post the make/model. I'd like to see how they do it. I'm always willing to learn something new.

[djQUAN]

Originally posted by Perry Babin
OK, it's technically possible but not practical and as far as I know it's not done in any commercially available amplifier that uses the TLx94 IC.

If someone knows of an amplifier using the TLx94 IC that syncs the class D oscillator to the power supply oscillator, please post the make/model. I'd like to see how they do it. I'm always willing to learn something new.

that I cannot contribute since the only class D car amp that I have owned is a DIY amp2 from 41hz and a coming soon amp8 at 600WrmsPC

[1moreamp]

I looked on the schematic for your brutus and sync'ing is not the problem, as I see it.
Pull your Rca's , does the noise go away ?
Start checking all the caps in the D section around the LM833's and the HIP4080AIP, all the way back to the cmos oscillator.

Also check for a burnt ground resistor back near the power supply you can't miss it as its always overheated and near the toroid.

These normally fail because of lead frame failure. The leads on the output fets literally break off inside the TO-220 case where you can't see them. ( way too much clamping pressure on the fets )
If you attempt to remove the fets from the grip of the sil-pad you will usually see a lead just simple snap off at the case edge.
The failure causes the fets on the output stage to ramp to full turn on, and causes a shorted condition ( the gate lead usaully causes this condition).
I have one I just repaired for this condition, works great now with a new set of outputs and some silicone grease to the pads to give some slip on the cases. Plus I have seen a bunch like this, and it does appear to be the main failure I have seen so far over the last 25 or so I have repaired.

Your problem appears as a noise related issue in the front end of the amp. I have seen similar problems on Tripath designs and other brands,
Funny though the amp does not have a bandwidth to reproduce 10 to 15 kilo-hertz so its got to be passing right thru the output filter network.

check some of your caps up front. I would be very interested to see your findings.

as always I hope this is found helpful

[Perry Babin]

I've seen the legs break off of the transistors inside the transistor body on MANY planet audio amps. The problem with the pa amps is that they have insufficient support for the board where the inductors are mounted. When the amp is subjected to vibration, the mass of the inductors causes the board to flex which causes the legs of the transistors to be stressed and eventually break. When the legs break a couple of things can happen, If the gate leg breaks first, it will cause the amp to shut down (if the protection circuit does it's job). If the drain or source legs break first, the parallel transistors will fail when the amp is driven hard.

I've seen a broken gate leg cause techs problems because the problem can cause the amp to intermittantly shut down.

With the clamping systems used in most amps, I don't think it's possible to have enough clamping pressure to damage a transistor (as long as the pressure is evenly applied to the face of the transistor). I've used heatsink clamping systems that used much more pressure and never damaged a transistor. Most amps use a one 3mm screw per two transistors into dead-soft aluminum. The sink will strip well before the transistors are damaged.

The transistors MUST be tightly clamped. I don't want him to think he needs to leave the clamps loose to allow it to move.