Amplifier is toast -why?

[madires]

The 1500W are PMPO at full moon For a proper power measurement connect a dummy load with the matching resistance, attach a scope and feed the amp with a 1kHz sine wave, Turn up the volume until you start seeing distortion on the scope. Turn down a little bit, just before the distortion starts. Then read the voltage from the scope: P = U^2/R

[jaycee]

Seen many of these cheap shitty rackmount amps. I bet it's only got a pair of TIP2955/3055's in it per channel.. maybe two pairs if youre lucky.

In short, it blew up because it's a cheap piece of shit. Bin it, or gut it for parts.

[michaelr]

These readings also are anecdotally confirmed by the 2 Ohm heater element not heating up.  1.5 kW, even 150W, should have gotten red hot!

If the amp was grossly overloaded, it likely blew up much faster than a resistor would have managed to heat up significantly. That heater element does not get hot instantaneously, even when powered with 1.5kW!  On the other hand, all a transistor needs to die is exceeding the maximum ratings for some milliseconds, depending on type. So your analogies are waay off here.

I would tend to agree, although the amp did not fail immediately.  It was about a minute or two, until it went south, with full power setting across a 2 Ohm load.  I blew three internal resistors, for sure. 

I've attached a couple pics, but I'll apologize for the quality up front.  As soon as I get a new amp, or repair this one, I'll get a better camera and update the photos attached. Viewing the image with the circuit board on the right, the visible damaged resistors are on the lower right hand section of the board.  What else could have fried?  I don't know yet, but am open to suggestions as to where to look.  Again, I apologize about the quality.  It is the best I could do right now.

[IanB]

I've attached a couple pics, but I'll apologize for the quality up front.  As soon as I get a new amp, or repair this one, I'll get a better camera and update the photos attached.

There's nothing wrong with the camera. You just need to make sure the pictures are in focus.

[Twoflower]

Michael If you really want to repair that 2x1500W Amp you should check all transistors in that region plus replacing the burned resistors. The good thing is that you have blown one channel. The other can be used to read the resistor values.

Yes the cam is a bit aged (I couldn't resist to check the embedded meta data). For a fixfocus cam like you have only light will help a bit.

[jaycee]

Wow, a single pair of TO-264 devices per channel. That's about 80W RMS into 8 ohms (safely) at most. They are probably 2SA1943/2SC5200 - or rather, cheap nasty copies of them.

Pretty much what I expected.

[michaelr]

Using a magnifying glass and good light, I've tallied up some of the major components, with their markings, if anyone may find it useful/interesting.  Might as well make this a good crash course on amplifiers...emphasis on the crash part! 

White blocks [4EA] - BPR56, 0.47 Ohms J, KOA595
    o Non-inductive Ceramic Cement Resistor

Large black Box[1EA] - GHong MB, 5A 250 VAC, 5A 30VDC, OTX14F3 L-DC24V-2A
   o Relay?

Large black chips mounted to heat sink[4EA] - Toshiba TTA1943. 408 [2EA] & TTC5200. 408 [2EA]
   o TTA1943: Transistor Silicon PNP Epitaxial Type - 100-W high-fidelity audio frequency amplifier output
stage
   o TTC5200: Bipolar Transistor - Amp Trans 15A 150W 230V BJT NPN PWR

Small black chips mounted on heat sink [2EA] - KSE340
    o Bipolar Transistors - BJT NPN Epitaxial

Medium sized chips just in front of heat sink [4EA] - A1837 [2EA] & C4973 [2EA]
    o Mosfets?

Large 4 pin black chip near capacitors [1EA] RRD GBJ 2510
    o Glass Passivated Bridge Rectifier

Based on the large transistors, would this suggest each channel could put out 250W ideally?  With losses that might put it near it's 880W PSU (110V x 8 Amp fuse). 

[Richard Crowley]

If the PNP output transistor is only rated for 100W, then that is your theoretical limit.  You do not ADD the two ratings together the PNP and NPN transistors are the top and bottom half of the output pair. The maximum you can expect from it is the LOWER power rating of the two devices.  And that is the theoretical ideal, not what you should expect in the Real World. 

That is a cheap 50-75 W domestic amplifier dressed up to look "professional".  The people who wrote the description of that gadget have a dart-board on their office wall.  When they threw the dart, it hit "1,500 Watts" and they all had a good laugh.   

[michaelr]

If the PNP output transistor is only rated for 100W, then that is your theoretical limit.  You do not ADD the two ratings together the PNP and NPN transistors are the top and bottom half of the output pair. The maximum you can expect from it is the LOWER power rating of the two devices.  And that is the theoretical ideal, not what you should expect in the Real World. 

That is a cheap 50-75 W domestic amplifier dressed up to look "professional".  The people who wrote the description of that gadget have a dart-board on their office wall.  When they threw the dart, it hit "1,500 Watts" and they all had a good laugh.   

Those dirty rascals!  Well, now their board is on EEVBlog. Ha!

Clarification: The TTA1943 is recommended for 100 Watt output stage, but it does have a collector power dissipation of 150W, at room temperature.
http://docs-europe.electrocomponents.com/webdocs/143d/0900766b8143d8f9.pdf

[Paul Moir]

Medium sized chips just in front of heat sink [4EA] - A1837 [2EA] & C4973 [2EA]
    o Mosfets?

A1837 = 2SA1837 (or equivalent).  The "2S" numbers are owned by one company so the prefix is often dropped or substituted when making a compatible part.
C4973 = 2SC4973:  same idea.
Worth knowing! 
https://en.wikipedia.org/wiki/JIS_semiconductor_designation

[michaelr]

Michael If you really want to repair that 2x1500W Amp you should check all transistors in that region plus replacing the burned resistors. The good thing is that you have blown one channel. The other can be used to read the resistor values.

Yes the cam is a bit aged (I couldn't resist to check the embedded meta data). For a fixfocus cam like you have only light will help a bit.

The manual for my Kodak EasyShare 3.0 Mp camera was copyrighted in 2005....  You got me, even though I have no clue how you figured that out.  It is time for an upgrade!

The resistors that match up to the burnt out ones, on the sister channel, are 10 & 12 Ohms, most likely carbon film based on their appearances.  I'll check out the transistors next.  Fixing this heap will boost my confidence a bit; and give me more experience.  It's my beater now.

[bson]

     RMS Power @ 2 Ohms: 1,500 Watts

Audio power amplifiers are rated in peak power, not average.  Audio generally has a peak-to-rms crest factor of 10:1, and it needs the peak power to reproduce audio waveforms, but the continuous average power is much less.  Clip indicators tell you you're exceeding max peak power, not average.  You need to be very careful when reproducing plain sinusoidals due to their 1:1 crest factor, which makes it very easy to overpower any audio amplifier.  It won't clip, although many amplifiers have thermal shutdown protection as well that might kick in.  (And, no, I seriously doubt it specs RMS Power of 1500W...)

[Jwillis]

 Just to compare. A vintage power mixer I'm working on now has four pairs of 2SC3182 and 2SA1265 complementary power transistors rated at  70 watts each.But the total rated output power mono  is 300w rms at 2 ohms ,200w rms at 4 ohms and 100w rms at 8 ohms .That's on a heat sink 16 inches X 4 inches X 2 inches.
Your unit has only 2 pairs rated at 100w and is supposed to output ,I'm guessing, 1500 watts per channel at 2 ohms?
I would hunt down the person that sold that to you and make him eat it.

[michaelr]

Medium sized chips just in front of heat sink [4EA] - A1837 [2EA] & C4973 [2EA]
    o Mosfets?

A1837 = 2SA1837 (or equivalent).  The "2S" numbers are owned by one company so the prefix is often dropped or substituted when making a compatible part.
C4973 = 2SC4973:  same idea.
Worth knowing! 
https://en.wikipedia.org/wiki/JIS_semiconductor_designation

Got it.  So, 2SA1837 is a PNP transistor; and 2SC4793 an NPN transistor.  That will help with testing.  Thanks!

[michaelr]

The 1500W are PMPO at full moon For a proper power measurement connect a dummy load with the matching resistance, attach a scope and feed the amp with a 1kHz sine wave, Turn up the volume until you start seeing distortion on the scope. Turn down a little bit, just before the distortion starts. Then read the voltage from the scope: P = U^2/R

Simple enough.  I will do that as soon as I get my hands on a working amp. 

Would this power level be considered safe for prolonged used; and could this technique be used at other frequencies?

Thanks!

[IanMacdonald]

The usual reason for a burnup like that is that the b-e junction of the output transistor has gone open circuit. The (unheatsinked) driver transistor and its emitter resistor then have to handle the full output current.

When repairing an output stage you need to be thorough about replacing every damaged component. Otherwise you will have a repeat failure. Also, after the repair you need to test with a current limited supply (see dim bulb tester) otherwise if you have made just one mistake you can blow the whole lot again. 

For any kind of professional use, an amp that can't stand full power output for a reasonable time is a liability. If a dropped mic or howl round blows the amp and shuts down the show, is that any use? 

As I've said, the main reservation I would have is about using 2-ohm loads. Unless it's high end kit which specifically states that is permissible, you are risking a blow doing that. The majority of top brands do NOT recommend 2-ohm loads anyway.

In any case, with the possible exception of bass guitar it's rare to need more than 50Wpc RMS into decent PA speakers. These are more efficient than typical domestic speakers, and that sort of input will be more than loud enough for a small indoor venue.

[madires]

Would this power level be considered safe for prolonged used; and could this technique be used at other frequencies?

That depends on the amp's design and thermal management. When you perform the power measurement, also check the temperature of the power supply, and the driver and output transistors. The power is specified for 1kHz usually, but you can also check other frequencies if you like.

[capt bullshot]

Would this power level be considered safe for prolonged used; and could this technique be used at other frequencies?

Most audio amplifiers (except the real professional ones) won't stand that power for longer than a minute or so. The heat sink and the power supply (transformer) usually isn't designed for this. Once I've read a rule of thumb: an audio power amplifiers supply / heatsink is designed for about 70% of its nominal sinewave power. Appears to be true for quality gear, but I've seen worse. It can be overloaded (100% of its nominal power) for a short period time. Prolonged use will overheat various components until shutdown by magic smoke. That's the usual deal with "normal" audio power amplifiers. If you run the amplifier with a light load (higher impedance than 8 Ohm, open load), it'll stand longer. Increased frequency at high output voltage levels may cause the ouput R/C damping to burn (especially the R). Audio amplifiers nearly always have a R/C from output to ground for frequency stability added.

So if you need an rugged amplifier for lab use, either get a professional one or really know what you do when using a consumer audio amp.

[IanMacdonald]

Most audio amplifiers (except the real professional ones) won't stand that power for longer than a minute or so. The heat sink and the power supply (transformer) usually isn't designed for this. Once I've read a rule of thumb: an audio power amplifiers supply / heatsink is designed for about 70% of its nominal sinewave power.

Actually a myth, and easily busted on the testbench. Under sinewave conditions the max heating effect on the output transistors occurs at some point below max output.

They are effectively dropper resistors, and when there is no voltage drop then no heat is produced. Of course that never arises with a sinewave output. In fact, the voltage across each transistor follows a cosine law, being the supply minus the output voltage. Therefore the heating effect follows a squared cosine law across the half cycle. (V²/R) 

Thus for small reductions in output power, the increase in transistor heating follows a square law. Reducing the output moderately will cause the output stage to run considerably hotter than at full sinewave power.

Only when you reduce the output to less than 50% or so is the current sufficiently lower that the total heat produced starts to fall.

It's possible that the mains transformer could overheat under continuous full output, but I've never seen one burn out.

The thing that will kill a domestic grade amp on stage, is hooking too many paralleled speakers to it. A 2-ohm load is a definite no-no. 4 ohms, stretching things. 8 ohms, OK. With a very low impedance load and loud but not max volume the thing will get roasting hot.

[janoc]

If you are planning on repairing that (could be a good educational exercise, even if the amp is crap as such), I would also look very carefully into the protection circuitry on that board (if there is any!).

You may want to make sure that there is at least over-temp and over-current protection for the end stage, also possibly some zeners to clamp the drive voltage, etc. That could save you from having to fix it again later.

[David Hess]

Most audio amplifiers (except the real professional ones) won't stand that power for longer than a minute or so. The heat sink and the power supply (transformer) usually isn't designed for this. Once I've read a rule of thumb: an audio power amplifiers supply / heatsink is designed for about 70% of its nominal sinewave power.

Actually a myth, and easily busted on the testbench. Under sinewave conditions the max heating effect on the output transistors occurs at some point below max output.

The first graph below from Bob Cordell's Designing Audio Power Amplifiers shows the relationship between maximum power output and maximum power dissipation.

The second graph shows the load line for both a resistive and typical reactive speaker load drawn against the safe operating area specification of the transistor.  Keep in mind that the safe operating area including secondary breakdown decreases as the junction temperature rises.  So for reliable operation, the transistors need to be derated based on worst cast operating temperature which will occur at about 1/3rd of maximum output power.

If a power amplifier is designed for lowest distortion, then often multiple output devices are used in parallel to prevent excessive Ft and Gm droop in which case sufficient power derating may already built in.  There is also some advantage to using multiple lower power output devices in parallel to achieve a lower case-to-heat sink thermal resistance.  Consumer level designs are unlikely to do either of these things for reasons of cost so are unlikely to handle a 2 ohm or maybe even 4 ohm load at high continuous output power.  Even an 8 ohm load may be pushing it unless additional cooling is used.

[james_s]

Is there equipment more suitable for amplifying these types of wave forms, that won't break the bank?

I got the impression you were feeding it the sine wave to test if it works - not because that's your primary use case. What are you wanting to do with such an amp then?

What should be and what is are often not the same thing. I've seen more than a few audio amplifiers that were far from bulletproof. Pushing a sine wave at full power is brutal service and while the amp *should* shut down in time, quite a few of them won't. It's not uncommon for the output transistors to sacrifice themselves to protect the fuse.

[xani]

Rule of thumb: Any audio kit that has "PROFESSIONAL" printed on the front is unlikely to be anything of the sort (Much the same goes for test equipment!).t

& for TV monitors, cameras, etc----"Professional" was always regarded as several levels below "Broadcast".
Towards the end of my association with such things,  their performance was often ( just) to Broadcast standards, but their construction was always "El Cheapo".

Those are 2 different things tho. One is producent (or people reviewing hardware) dividing their catalog into "tiers" like

• cameras for filming your vacation and your kid
• cameras for "professionals" (as in "people doing it for money", not neccesairly experts) and more serious amateurs
• specialistic cameras (high speed, sport cameras, etc)
• broadcast/movie/other highest tier devices

The other is "let's put some big numbers and nice sounding words like "PROFESSIONAL" on the case so somebody clueless might think they are getting a bargain buying it".

Other reason is so one amateur can brag to another, same reason why some devices brag about random features on the case of device, like cameras having big embossed 3CCD, or writing zoom numbers and number of megapixels on camera case

[xwarp]

Yes, I was testing to see if it works; but I also intended to use it as a signal generator amplifier.  Although, it would have been a nice perk that it could have been used as an audio amplifier also, if the need ever arose!

A signal generator amplifier? What frequencies are you talking about?