burnt voice coil damages amp?

Status
Not open for further replies.
i think my woofer has a voice coil damage. it doesn't scrape when i push it with hand but does make rattling noise when bass is loud. now, the vssa (peeceebee version) channel connected to this speaker keeps damaging the vas transistors- after about 30 min it makes a static noise then the dc offset goes way high.
so is it safe to assume that the defective woofer is damaging the amplifier?
 
I suppose, it's vice versa. At some point, DC offset goes high, possibly degrading the voice coil. Well, as long as it's damaged enough, its impedance goes down, loading the amplifier excessively. And so on...

Looks like you don't use DC offset / overload protection. Helps a lot (to both protect the amp and the speaker).
 
found the problem with the speaker- lose crossover soldering. rattling mustve been caused by the driver connecting/disconnecting with vibration. that would damage the amp, right? it would've been same as varying the load while the amp is on.
 
A speaker disconnecting momentarily while being driven will inject a high back emf into the amplifier's output terminals when the connection remakes.
One usually fits back emf diodes across the output to supply rails to help prevent reverse breakdown of the output stage.

You will often see these fitted on schematics, but many omit them because they don't know what they are supposed to protect against.
 
A speaker disconnecting momentarily while being driven will inject a high back emf into the amplifier's output terminals when the connection remakes.
One usually fits back emf diodes across the output to supply rails to help prevent reverse breakdown of the output stage.

You will often see these fitted on schematics, but many omit them because they don't know what they are supposed to protect against.

Do you think the back EMF is what's causing the VAS transistor failures or is it more likely to blow output devices? It seems odd to me that it could be causing a failure that far back in the amplifier circuit.
 
A HF feedback signal could easily cause the amp to go into oscillation.
That oscillation could overstress the next weakest link in the long chain from -IN to OUT.
The VAS may well have been that weakest link. It has to swing from rail to rail when not oscillating and could have to swing from double rail to double rail when actually oscillating.

That's why I use diodes on the supply rails to output and to power ground and also use the Pi version of the Thiele Network.
I see these additions as cheap protections that many others think are not required, ever.
 
Oscillation was my suspicion as well. I had another look at the PeeCeeBee schematic. VAS transistors are doubling as driver transistors. I agree with Vzaichenko, this amp needs a good protection system. There doesn't seem to be much in the amplifier itself to stop destruction if anything goes wrong.
 
I just read the first page (out of **hundreds**😱 ) and found this:
There is no compensation capacitors and zobel on-board. If found to be necessary then ......
.

After such an unauspicious beginning, why waste time reading more? 🙄

So yes, maybe you are finding some compensation and stability are needed ... the hard way of course.
 
I recommend to pay attention to what JMFahey is saying.

There is a big difference between a cool module and an amplifier, ready for being used with the different real speakers. I suspect you didn't try to connect oscilloscope to the output to see is there are some signs of oscillation or other artefacts there.
I would also check at what quiescent current (and what temperature) the VAS transistors are running. Having only 10 ohm degeneration is rather dangerous - they may start overheating, provoking some oscillation and then just break... do you use some local heatsinks for the VAS transistors?
 
There are also extremely tough VAS (MJE15032/33 for example) as opposed to really fast (VAS 2SA1837/2SC4793 for example) ones. Your vendor may have opted for ultimate fidelity instead of toughness. More speed, more accuracy. More safe operating area, tougher. Besides the flyback diodes recommended above, you might also put in better heat sinks on the VAS if not damaged already. Note TO-126 and TO220 cases have different pin order.
The flyback diodes are 3 amps ones minimum (say MR802),first one line to positive rail, not-line to speaker out inside the zobel coil. Other one line to speaker, not-line to negative rail.
 
Last edited:
PeeCeeBee has been built by a number of people on this forum without issue. If your build is damaging the VAS transistors then there is something else wrong. May be layout, wiring, your personal implementation.

You should not need 'tougher' VAS devices. Please post details about your particular build, especially a current 'as built' schematic.
 
Well I can't help unless you post or link to the schematic diagram. Experienced people around here know all the classic diagrams by name, I don't.
As vzaichenko said, if you made any variations of the schematic, draw them out. Paper & pencil is fine with me if your camera will resolve the words.
 
Hi donovas,

A few pictures could have saved you much headache and speculation. People are falling into problems even when building age-old, proven, robust designs. I urge you to do what indianajo said in the post above. Also, you don't need those reverse-parallel diodes with this amplifier because the MOSFETs already have them inside.

_________________________________

Hi JMFahey.

I agree. Very inauspicious Post#1748.
 
Status
Not open for further replies.