Opamps fried in a Supergreen pre, (SOLVED!)

[lagoausente]

Hello everybody.  Hope someone can give me a light,

Some time ago I built the Supergreen pre. Some months ago I changed the case to put it in a rack unit with headphones amp etc. Have replaced the output opamps NE5534 by the LME49710, what are the single version of the LM4562na.
  I tried it yesterday, and have found that some of the fron opamps NE5532 I had around was wrong.  One not working, other has a lot of noise. Replaced with a new one and worked ok in both channels (have two).
  Well, I wanted to try the LM4562na I ordered sometime ago. It didn´t work, no sound, and the opamp got very hot, and finally fried. I blown four of this opamps on my trying, no one work.  Finally, I tried the last on my Cmoy phones amp, and worked ok, then tried on the Supergreen and fried again.
  I remember that some months ago I tried and comprared the opamp I still have alive in my Cmoy and worked perfectly.  Now, they are fried. But if I put again the NE5532 it works and this is not fried. ???

  Have checked the supply requirements for the LM4562 and it´s 2.5V to 17V, with absolute max rating of 36 volt total.  I power the preamp with four 8,4volt batteries, so have +/- 16.8 volt, max 33.7 or so.

  Well,  I would like to know If it´s posible to protect the opamp to avoid get fried while checking if it works or no, or if I must blown a lot of opamps while I find the problem..
  And the other point is, what I should need to check?  What can be the problem?  What should I check to find out what occurs.
    I remember that this type of opamp worked ok some months ago in the same preamp. The only changes I made was the output opamps, and maybe a diferent battery voltage, but still seems in the correct range..
  Any ideas?

Attached below the supergreen schem.
 

 

[okgb]

not that it is related but there was a mismarked zenar diode on the board
[ the input protection ones i think ] you've got those in correct ?

 

[lagoausente]

I remember that,  D1,D2 D3 and D4, isn´t?  I put them ok in that moment. Although they were to protect from over phamtom voltage peaks, but I tried the LM4562 without phamtom.

  Before the LM4562 gets blown it gets hot a lot, but strange is that is currently stable with the NE5532ap. So it´s dificult to understand.
    Could be any incompatibility using the LME49710 at the output with the LM4562 at the front?  It seems the only change I made since the last time I tried the 4662.
  Umm, wait a moment, I have just realize that have read about the need to remove the cap between pin 5 and 8 in the single NE5534 when use other opamp..., haven´t removed it, yet, but could this cause the fault?  ???
  I´ll try that, but have to order another 4562 to can test it.

 

[Samuel Groner]

Were the DC conditions correct and nothing oscillating? What about current draw? These are *always* the first things to check before you start to even think about listening to it. Without this information it is difficult to give any advise. In any case I didn't see any inherent incompatibility for the NSC opamps.

Samuel

 

[lagoausente]

Were the DC conditions correct and nothing oscillating? What about current draw? These are *always* the first things to check before you start to even think about listening to it. Without this information it is difficult to give any advise. In any case I didn't see any inherent incompatibility for the NSC opamps.

Samuel

  Thanks lot for your reply.
  Well, today a I have bought four NE5532 to try. One of them dead. I mean dead without any heat. I tried aswell a LT1358 what dead aswell (or was already dead).  This is one thing that have happened to me other times.  Are opamps so much delicated or is my preamp?  Do they suffer from electrostatic from hands?  Or maybe because of electric charges in the circuit that remain some minutes?  Or it´s my paranoia?

Your points:
  1-  DC condictions,  could you specify what points must I check?
  2- Oscilating?  Don´t know.  I just can say have heard nothing from the preamp.
  3-  Current draw, what current do you mean, measure the current draw without the opamp?  Or through the opamp?  Must I lift up one of the opamp legs and insert a amperimeter?  Limit the current with a resistor there?
    Extrange here is that from four NE5532 I have bought, one died, but the other three didn´t, and now have two installed.  The question is, when I´ll receive again the LM4562, how can I test and check the fault without burn them?  A current limit resistor in the supply input of the opamp maybe?

 

[lagoausente]

Well, today have switched both channels and seems that one of them is the one that blown the opamp. Today as a 5532.
  Have removed the dual front opamp in both channels and measure voltages. Both channels seem to have the same.
  Edit:
  Have measured the current supply of the front opamp. Current is ok, max 12ma, then stabilize at around 7ma.
  Just switched down the preamp. Removed the amperimeter, then put again the opamp with all legs in, and blown again. Tried that opamp in the Cmoy headphones preamp, and result is that one channel is dead, the other is ok.  Dead is pin 2-3.  So it´s IC1a in supergreen schem.
  went to T2, measured values there between base-colector, base-emiter, and values seem to be ok, similar to the other transistors.  Although I measured the onboard. 
I´m blocked at this point.  Haven´t find no shorcut to ground. Not sure what more to check.
  Well, it´s Ic1a what gets damaged. Current through the opamp is ok, even when it works, even when it has one channel dead.
  what more can I do?

 

[12volts]

As suggested, check for oscillation using an oscilloscope. Any destructive signal will be in the Mhz region not something audible
like you said earlier. Check, then post again with the result. An easy fault to fix.

 

[Samuel Groner]

Any destructive signal will be in the MHz region.

Oscillation alone is unlikely to cause damage to the opamps (power amps are something different); its presence might indicate however other faults.

Are opamps so much delicated?

Not really if you observe a basic rule--never exceed the maximum ratings given in the datasheet. I think I've killed only one opamp within the last couple of years, and I didn't follow the basic rule.

DC condictions, could you specify what points must I check?

* Pin 1 and 7 of IC1--should have the same value, and not more than a couple of volts from ground (also compare with working channel).
* Pin 2/3/5/6 of IC1--should have the same value, and at about +14 V (also compare with working channel).

Current draw, what current do you mean, measure the current draw without the opamp?

Current draw of IC1, e.g. measure voltage drop accross R100/R101.

How can I test and check the fault without burn them?

You should find the fault first (see below). A pair of 1k resistors in series with the inputs will probably help most in this case to protect the opamp. And add cross-coupled 1N914B diodes across the inputs.

Haven't find no shorcut to ground.

What about shorts to the supply rail?

Not sure what more to check.

All resistors close to IC1 inputs. Looks like there's something wrong.

Samuel

 

[lagoausente]

Any destructive signal will be in the MHz region.

Oscillation alone is unlikely to cause damage to the opamps (power amps are something different); its presence might indicate however other faults.

Are opamps so much delicated?

Not really if you observe a basic rule--never exceed the maximum ratings given in the datasheet. I think I've killed only one opamp within the last couple of years, and I didn't follow the basic rule.

DC condictions, could you specify what points must I check?

* Pin 1 and 7 of IC1--should have the same value, and not more than a couple of volts from ground (also compare with working channel).
* Pin 2/3/5/6 of IC1--should have the same value, and at about +14 V (also compare with working channel).

Current draw, what current do you mean, measure the current draw without the opamp?

Current draw of IC1, e.g. measure voltage drop accross R100/R101.

How can I test and check the fault without burn them?

You should find the fault first (see below). A pair of 1k resistors in series with the inputs will probably help most in this case to protect the opamp. And add cross-coupled 1N914B diodes across the inputs.

Haven't find no shorcut to ground.

What about shorts to the supply rail?

Not sure what more to check.

All resistors close to IC1 inputs. Looks like there's something wrong.

Samuel

Best regards for your support.

I suspect that the problem is because of the wrong way I´m suppling the preamp. I´m using four 8,4V recheargeable batteries, but have taken the ground directly to the batteries, so due to diferent discharge rates on each battery, I had finally +15V but -13V.  Is that critical? isn´t?  I have realized that measuring pin 1 and 7 had around 2.8V, but sudenly had 5 volt when some batteries on -V went down. Well, have just checked the opamp in the "supposed ok channel" and it has, aswell, the Ic1B pin 2 and 3 dead.
  Is here the fault?
If so,  I should implement a basic power supply to get the Power ground centered from the total 30-34Volt. I´m in the right way? 

  Thanks a lot.
 

 

[Samuel Groner]

I had finally +15 V but -13 V. Is that critical?

This should not cause problems.

It has the IC1B pin 2 and 3 dead.

What do you mean by "pin 2 and 3 dead"? What are you measuring? I don't think there's an easy way to tell what has been damaged...

Samuel

 

[lagoausente]

Quote
It has the IC1B pin 2 and 3 dead.

What do you mean by "pin 2 and 3 dead"? What are you measuring? I don't think there's an easy way to tell what has been damaged...

Right, I have found another fault. What I was doing was testing the opamp in the CMOY headphones preamp, to see is both channels of the opamp, left and right sound ok, so when I refer to pin 2 and 3 I refer to the amplifier that uses pin 2 and 3 as input.  But I was wrong.  Have realized that the cable I was using have a desoldered conection, so when I check again today on CMOY with a good cable both opamps I have now sound ok for left and right on headphones, but here I have another question.
  That cable was the one I used to connect the preamp output to CMOY. It is a jack to jack stereo.
  The cable that should go to out3 (pin 3 on a XLR) was desoldered, so could be doing:
  1)  shortcout to pin2, so output in shortcut
  2) contact ok, pin3 of preamp would be conected to right channel input of CMoy, while pin2 was to left channel of CMOY
  3) On air, so output would be taken from ground to pin2 ( pin3 disconected)
  Any of these situacions could cause damage?

....................................................................

What I have now.  Have two Ne5532 (other went to trash, maybe some of them was ok).
      1)  One of them work ok on both preamps.
      2)  The other doesn´t work ok on both preamps.
    The one who doesn´t work ok, have this  behaviour. If connect with gain switch in a position of high gain (16 ohm or less) the preamp doesn´t start. If set the gain switch down to low gain (820 ohm or so), the preamp start, and can hear the mic, if wait some seconds, then go again to higher gain, remains working.  I have had this situation before, what solved changed the opamp for a new one.
  Note that the opamp that is ok, works ok on both channels, and the one who haves this behaviour, has it on both channels aswell.

 
  Today have conected ok the output of the preamp. Pin3 soldered to gnd, pin 2 to left channel on CMOY..
  Have to buy new opamps to can continue testing.
Well,  this situation is quite confusing.  I believe on the posibility that I damaged the opamp due to a output shorcut?

 

[Samuel Groner]

Any of these situacions could cause damage?

No, any recent opamp is short-circuit proof. And in any case those opamps which actually connect to the output (IC2, IC3) would have suffered damage.

The preamp doesn't start.

Again: what are the DC conditions in this not-started-up condition? Only this will tell you where's something wrong...

Samuel

 

[lagoausente]

Go;

1-Working:
 
    IC1 (Voltage refered to ground)
Pin1  -2.78
Pin2  13.00
Pin3  13.01
Pin4  -15.7
Pin5  13.01
Pin6  13.01
Pin7  -2.86
Pin8  14.65



Voltage drop across R 101 0.73
                          R100 0.45

Voltage on T1 and T2 legs, refered to ground:
t1
b -0.14
c 13
e -0.76
t2
b 0.08
c  12.97
e  -0.68

 
  2-Not working:
IC1 (V refered to ground)
Pin1  -13.11
Pin2    12.81
Pin3    12.85
Pin4  -15.16
Pin5    12.80
Pin6    12.82
Pin7    13.02
Pin8    14.05

Voltage drop across  R101  1.22 
                              R100  1.06

Voltage on T1 and T2 legs refered to ground:
t1
b 0.27
c 13.49
e 2.93
t2
b  -0.13
c  12.53
e  -0.74

 

[Samuel Groner]

You interchanged the voltages for IC1, right? With pin 1/7 sitting at -13/+13 V the preamp is not going to work...

Samuel

 

[lagoausente]

You interchanged the voltages for IC1, right? With pin 1/7 sitting at -13/+13 V the preamp is not going to work...

Samuel

Yes, sorry, wrong copy paste from the notepad. I´ll edit now.
Just fall ill this morning, so I´m not very alert.

  Could hear the radio for moments, don´t know if that has someting to do with the MHZ oscilation you mentioned.
  Other curious situation,  measuring R13 onboard with supply off , measured near 0 ohm with voltimeter in on position, and gived 4kohm reverting the voltimeter tips. I then measured if there was any voltage there but seemed to be 0 volt.
  In the other channel measured 1.9 kohm, whatever position of the voltimeter tips.  I found that ilogical, and replaced R13, but behaviur didn´t change at all.
  Other thing is, bought new Ne5532, first time on working channel, no problems, then on the failing channel, it worked. the first times but after some switching down and up, started to fail. But then tried on the other preamp channel, and fail there aswell.
  But when take a opamp, and use only in the right channel, never fails, so seems that some damage the opamps get in the wrong channel.

 

[Samuel Groner]

If connect with gain switch in a position of high gain (16 ohm or less) the preamp doesn't start.

From your description and available DC conditions it looks to me as if the gain switch would produce a short in higher gain ranges sometimes; have you verified its behaviour, e.g. by measuring its resistance in all gain settings, interchanging it with the one from the working channel and/or temporarily replacing it with single fixed resistors?

Samuel

 

[lagoausente]

If connect with gain switch in a position of high gain (16 ohm or less) the preamp doesn't start.

From your description and available DC conditions it looks to me as if the gain switch would produce a short in higher gain ranges sometimes; have you verified its behaviour, e.g. by measuring its resistance in all gain settings, interchanging it with the one from the working channel and/or temporarily replacing it with single fixed resistors?

Samuel

Have replaced the gain switch a fixed resitor.  Tried one of the new opamps, worked ok first times, on fourth time or so I switched on the preamp, it start failing. 
  Have damaged two Ne5532 today. I say damaged because as I told before, once it starts failing, then they continue failing in the good channel.  One of them makes some noise, like white noise, then starts,, hear the mic for half second, then white noise again.
  The other opamp just do nothing. No noise, no start, nothing.
  The good channel continue working fine with the opamp that never was tried on the bad.

  Have checked resistor values onboard, and compared values with the good channel, also checked for shorcuts from diferent points on the preamp to +V and -V rails as you mentioned. Haven´t found nothing strange.
  What other things can occur?  could I isolate sections from the opamp to test them alone?  DC values mentioned could have relation with oscilation?  I really don´t know what more to do. It´s really strange. 
  bad soldering in C11-C12 so infinite resistence could get damage?
  Maybe oscilation have something to do with the thing that only starts to fail after some tries, instead of instant fail? 
Other things to consider is that have changed some time ago the position of the bit wima caps c7-c8, what are now sideway, but in both channels. Always C11-c12 are sideway on both channels.
  Other thing to mention is that Cmoy and the preamp shares the chassis as ground, but no supply as they use diferent batteries.

 

[Samuel Groner]

Maybe oscillation have something to do with the thing.

So *is* there oscillation or isn't there? You really need a scope--there's not much point typing here for days if you don't have the tools to check. With the money it would have cost you to pay me just writing the posts in this thread you could have gotten at least one very decent scope from e-bay...

So just to make sure: if you start up the defective channel at low gain it is fully working and the DC conditions are the same as in the working channel?

Samuel

 

[lagoausente]

 

So *is* there oscillation or isn't there? You really need a scope--there's not much point typing here for days if you don't have the tools to check. With the money it would have cost you to pay me just writing the posts in this thread you could have gotten at least one very decent scope from e-bay...

Sorry man, I misunderstood you.  Whe you said:

Oscillation alone is unlikely to cause damage to the opamps (power amps are something different); its presence might indicate however other faults.

I thought that had to check for other things. I can borroy a digital scope. Hope will allow the range required. Sorry for your loosing time. I´ll check and see I can find something with that.
 
 

So just to make sure: if you start up the defective channel at low gain it is fully working and the DC conditions are the same as in the working channel?

  Yes, when it starts, is has the same  values as the good channel, just when it gets damaged when the values change.
  Have received today four new LM4562.  Tried one. R100 fumes. Voltage of 15Volts, so 200ma across it. It occurs in both preamp channels. No sound. 
  I´ll check for scope a see what can find.
  thanks and my apologies.

 

[Samuel Groner]

Just had an idea: what happens if you remove C13/C102/C103?

Samuel