Amplifier get into protection with no load

SOLVED Amplifier gets into protection with no load

Is that a normal behavior?

I'm testing a pioneer sx-850 with no load, feeding a 135mV RMS 1000Hz sinewave to both channels (input sensitivity is rated 150mV), and the protection activates when volume knob reachs 2 o'clock. Turning down to 1 o'clock, it works again.

Tested with the scope, never goes into clipping, the protection activates before any signal clipping. Both channels driven, the signal looks perfect.

No DC at the outputs when clipping, I've just adjusted it to 0.2mV.
Protection board is receiving power. It's just doing it's job.

Should I proceed with a small load, let's say 15 ohm or more?

Edited: It's not a consistent fail: I've driven one of the channels to MAX volume without protecting itself. The other channel protects earlier, but once this happens, next time the protection activates earlier... I think I'll recap the protection board to trust it. I've replaced the differential pair this morning, DC offset looks perfect.

Thanks!

You sound like you know what you are doing but what happens with a load(speakers or a high wattage resistors)?

I don't want to plug a load until I trust the unit...

I'm monitoring the output voltage, the "good" channel (L) can do 20 or 22V AC RMS at the output without protecting.

The "bad" channel (R), protects when reachs 10 or 11V RMS. If I leave it at 9.6V at the output, it works fine. When I push it a little more, "CLICK", and protection engages...

Sounds to me that you're asking quite a bit of the receiver. I assume "rated sensitivity" means a signal of 150 mV can drive the amplifier section to full output power (whatever that is for an SX-850... which is quite a bit, if memory serves). 135/150 = 90% of rated sensitivity. My simple mind assumes this would correspond to the power amplifier endeavoring to output 90% of full (continuous) power at zero attenuation. Now, I agree that open-circuited, it should be happy enough... but clearly it isn't.

Couple of things come to mind (and, bear in mind, I am not very solid-state savvy at all!)

1) Are you sure there is no DC offset in the sine wave signal you're feeding the preamp?
What input are you using? Is your input signal going through active preamp stages or mainlined to the power amp?
I am wondering if you're passing a big slug of DC into the active stages of the receiver, and the protection is aware of the transgression and trying to protect itself against it.

2) The amplifier expects a load -- my guess is it would be happier (certainly not any unhappier) with one. Is there any good reason NOT to put a noninductive (i.e., pure resistive) load (8 ohm is fine) capable of dissipating adequate power (100W) on the amp for testing purposes?

I don't know what an SX-850's protection circuit looks like, but a bad smoothing capacitor on a DC sense line could act like this. If that were the cause, it would most likely be sensitive to frequency: at lower frequencies, it would probably go into protection at lower output levels.

Good luck,

chazix

I've tried connecting to the AUX in, and to the Power amp IN (removing the pre-main bridges), with the same result. Reaching 10.5 Volt AC in one channel engages the protection, but just in one channel. The other can be driven to 22 V AC, no problem. This discards a lot of variables (all the circuits common to both channels)

The problem happens even with the speaker selector disconnected (both A and B "off"), that discards a problem at the speakers connectors, but not at the speaker selector switch.

I don't measure any DC at the output even when the amp is protected (measuring before the relay, of course), I can double check for that.

I'll check how the signal looks at the relay terminals. I guess it will be OK. I suspect some component out of specs in the protection board. THERE ARE two 0.22 caps in the sensing circuit, I'll replace them.

There are coupling capacitors at the power amp input, so any DC should be blocked there.

To discard a "source " unbalance, I just plug the same RCA cable to one channel, and then to the other. My source is my Phone playing "Test Tone " app. Never had a problem with that.

I'll measure some voltages at the protection board comparing both channels and see what happens.

You are right, a load shouldn't hurt if the protection is working, but there is something wrong in one channel, I'd like to find that first. Some service manuals advice to connect a load only when all the tests without a load were OK. And here something is not OK yet.

I'll do some more testing tomorrow, my main question was if the protection should engage without a load, but it seems normal operation.

I don't know what an SX-850's protection circuit looks like, but a bad smoothing capacitor on a DC sense line could act like this. If that were the cause, it would most likely be sensitive to frequency: at lower frequencies, it would probably go into protection at lower output levels.

Click to expand...

Perhaps at 4500Hz the protection was not so sensitive, but I didn't pay that much attention to frequency. I think there is something in this direction...

take a look around q1 q2 protect board .. looks like they are over current sensing circuits .
also check emitter resistors

Thanks. I'll check those transistors, voltage at the base when protected, and replace the 0.22uF caps just in case.

I think emitter resistors are fine. I'll measure them again just in case. It seems they can be measured in circuit.

test/replace q2 .. or maybe pull r2 r4 and test like that .

Replaced Q1 and Q2 at the protection board, fitted some BC550 I had at hand, now I can push it to clipping (clips at 32.8V RMS without load), no protection !!!!

I'll check with a load now, but I think it's fixed !!!!

THANKS!! I've learned the protection circuits can fail too

UPDATE: with a 10 ohm load, 1KHz signal clips at 25.6V, that makes 2.56Ampere, 65.5 Watt before clipping

One more question... How can I check the protection is working, without risking the amp?

Good question. One thing is I doubt transistors would cause this problem, I owuld think caps. Reason being is that it filters the output to DC, otherwise the AC would trip it. I would think C1 & 2 open or high ESR.

The only real method I see is to take out all the outputs and apply a nine volt battery (or similar) to the speaker output. Try both polatiries of course. It should kick out.

That will only test the offset protection though, not the current protection. Looking at the circuit, IF the offset protection works, the current protection should as well. The input at 6, 5, 7 & 8 that tell it about the offset also tell it about the overcurrent status.

IF the overcurrent does not work and you test it, you might fry it. Howevr I believe it quite unlikely. It looks like if those diodes or resistors are open the thing should tend to false trigger rather than fail to trigger, but I can't be sure with just a quick look. It does look that way though.

Maybe instead of a short circuit you could test the overcurrent with an extremely haevy load. this way it should not fry the outputs, just make them hot. Like four eight ohm speakers on a channel. While that is hard on the amp, a dead short is worse.

I dunno, the offset thing is not that hard, maybe someone else around here has some ideas for the current protection. I would be very trepidatious to just short it out because in case it doesn't work, there goes a set of outputs.

I could try to move the DC offset with the DC offset adj trimpot, and see what happens if I go out of specs...

Transistors at the differential pair were "tired", lot of noises, and very different hFE between them... I think the protection circuit transistors could be in the same condition.

The fact: with the new transistors it works OK, and with the old ones, very unstable (actually, the owner had 2 problems: intermittent Pops, and intermittent protection)

I didn't change the caps, to isolate the problem. After the new transistors, my next move would be new caps . But the problem was fixed.

i noticed earlier the sx-750 manual explains how the protect circuits work .. i didn't compare circuits but going from memory they look very similar

But, can it be tested? or the fact that the relay is open at power on, and engages after some seconds, is enought to confirm it's working?

I guess a good protection circuit should fail open, and close the circuit only when everything is fine. ..

What about parallel some large resistor with the one at Q1, to "unbalance " things?

I've had a similar problem, and that turned out to be low frequency oscillation which only happened with no load and loud volume. It could be detected by hearing the transformer hum more loudly and the power meters went through the roof.

i would think over-current should kick in if you turn the volume up too high .
did you read the 750 manual ?

That it does what it's supposed to do on power up, for me, tends to lead me to think it's working as it should.

However, the true test is to crank it up and short the outputs. That'll tell you if there is actual protection or just the illusion of it.

Do you feel lucky? Well... do ya?

I've closed the cover, I'll leave it alone by now...

The transistors at the protection board are new, the unit works, 65WPC before clipping, I don't want to risk it testing further. Almost sure, at a home environment, it will never be so stressed...