Home Cinema active Subwoofer malfunctioning

[KeepItSimpleStupid]

That's the problem. I thought that couldn't happen because Q8 was pulled. LOW POWER STUFF FIRST! If anything, you could have started to measure some stuff in the power on MUTE section. I gave you some ideas and you gave me one or two measuremenst, but not them all where Q2 or something driving Q2 was a problem. I want you to cut the trace and add a 1K resistor between ON of the switch and R44/R40. This will add a slight delay to the ON position. For now, also put two diodes in series for D5.

HFe can be measured with the Hfe DMM function at least for the small signal transistors.

For your pennance buy a cheap 2n3055 transistor. If you have a vise, you can squeeze it at the top very hard or you can cut the top off with a Dremel tool. This will show you how fragile the transistors are. Once the temp reaches about 150 C, the transistor is gonner.

Magic Smoke is precious. As you found out, once you let it out, the part ceases to function. The corrolary, of course, it Magic Smoke makes electronics work.

 

[lemofus]

I have a few spare transistors here, including some bust ones, I will dremel one of those...

I found 7 C2240 transistors, 3 new, 4 recycled, and found the following hFE readings with the DVM:
New 1 = 550
New 2 = 427
New 3 = 405
Old 1 = 465
Old 2 = 533
Old 3 = 542
Old 4 = 580

Measuring the old transistor hFE was a little harder, since the pins weren't long enough, I had to stick in some pin extensions and then try to get a good contact as best as possible...

Any preference which pair I choose for Q1 and Q2? Alternatively I could also go instore with my DVM and ask to test 10 or so transistors, and buy the closest matching ones...? Need to get a replacement Q11 and Q8 anyway, although those will stay out for now !

 

[KeepItSimpleStupid]

550, 542; diff = 8 or about 1.4%; You can't get any better than that.
533, 542; diff = 11

Which two were in there?

Remember the A,B and C designations and the Hfe range for each in the datasheet? Does this range fit? It's easier for you to find that info than me.

 

[lemofus]

Which two were in there?

No idea which of the old ones, unfortunately.

Remember the A,B and C designations and the Hfe range for each in the datasheet? Does this range fit? It's easier for you to find that info than me.

No idea what the A, B, C designations are. hFE range is between 200-700, broad scale...

I assume the value does not matter, and that what really matters, if the smallest delta between Q1 and Q2?

 

[KeepItSimpleStupid]

GR is 200 to 400 and BL is 350 to 700. I thought somewhere there was an A,B and C suffix for the2SC2240. e.g. 2SC2240A, 2SC2240B or 2SC2240C or a 2SC2240BL and 2SC2240GR

It could make some amount of difference and could impact the sizes if R4 and R6 causing more or less conduction of Q5. In any event, adding the bias pot will fix that.

Add the second diode in series with D5 and add a 1K resistor between ON and R40/R44.

I need you to look at R35 and Pin #1 of the OP amp (pre-amp ckt) with no audio in signal. Actually look at both side of the diode (D1?) with respect to ground. Also look at it with an audio signal. Look at the AC signals as well under both conditions. That OP amp could be oscillating.

What's that OP amp type. I need to check if it's unity gain stable.

 

[lemofus]

Ah! OK! The C2240 go as follow:

New 1 = 550 BL6C
New 2 = 427 BL6C
New 3 = 405 BL6C
Old 1 = 465 BL1L
Old 2 = 533 BL6C
Old 3 = 542 BL1L
Old 4 = 580 BL1L

Add the second diode in series with D5 and add a 1K resistor between ON and R40/R44.

Will do. Need to get the additional new board section back on, and on the long run, will glue gun it to the normal PCB1...

I need you to look at R35 and Pin #1 of the OP amp (pre-amp ckt) with no audio in signal. Actually look at both side of the diode (D1?) with respect to ground. Also look at it with an audio signal. Look at the AC signals as well under both conditions. That OP amp could be oscillating.

What's that OP amp type. I need to check if it's unity gain stable.

This is once the previous step has been achieved, I assume?

 

[KeepItSimpleStupid]

The IC part # would be useful now. With all the transistors out, like Q1, Q2 and Q8 and Q9 nothing nasty can happen so it doesn't matter.

PS: managed to fix the wireless issue. Moved LOCAL SERVICE to the "...DCOM... " group. This I'm happy with.

 

[lemofus]

What's that OP amp type. I need to check if it's unity gain stable.

BA4558N
147 L06

This is what you meant by IC part #, right? This is "U4" on the diagramme (although I probably have to edit that)

I need you to look at R35 and Pin #1 of the OP amp (pre-amp ckt) with no audio in signal. Actually look at both side of the diode (D1?) with respect to ground. Also look at it with an audio signal. Look at the AC signals as well under both conditions. That OP amp could be oscillating.

I have Q11 and Q8 pulled out; I assume I can do those readings just by leaving the amp on "stand-by" RED light; with audio signal, not sure what would happen with those pull out? Or do you mean for me to plug a source in, get some music going, and do some testing on the pre-amp bit without the amp bit being switched on? Possibly the signal from the input would be used here?

PS: managed to fix the wireless issue. Moved LOCAL SERVICE to the "...DCOM... " group. This I'm happy with.

Wow. Good news. It always seems so simple once it's fixed, but it's often a nightmare to figure out the fix... Had no idea this could have fixed it!

 

[KeepItSimpleStupid]

I have Q11 and Q8 pulled out; I assume I can do those readings just by leaving the amp on "stand-by" RED light; with audio signal, not sure what would happen with those pull out? Or do you mean for me to plug a source in, get some music going, and do some testing on the pre-amp bit without the amp bit being switched on? Possibly the signal from the input would be used here?

Q8/Q11 OR Q8/Q9 should be enough to disable the amp. Additionally pull Q9 just to be safe. No speaker. Amp should be able to turn on without damage.

 

[lemofus]

OK, Q11, Q8 and Q9 pulled out. The chosen Q1 and Q2 with close hFE values are back in. Indeed, the amp switches on fine, no magic smoke thanfully.

Now, before I do something silly, just want to double check:

Add the second diode in series with D5 and add a 1K resistor between ON and R40/R44.

Literaly, I have to cut the conducting part of the PCB between the ON switch and R44, and stick in a 1k resistor? I guess this is pretty irreversible... I just go at it with a knife, or dremel it? Secondly, 1k resistor, 1/8, 1/4, 1/2 or 1W? any preference?

Extra diode in series seems less stressful, more reversible

 

[KeepItSimpleStupid]

OK. The diode is easy. It should add a little more of a delay. It's reverseable and just adds component height.

Adding a resistor looks tough without cutting the trace, but if you don't want to. Don't. You can do it by lifting Q2, R40 and R44. Add a 1K resistor on end at the pad for R44 and then connect the dots with pieces of resistor leads. Reverseable. You can make it permanent later.

or

Make a cut away from the switch pad. Dremel or hobby knife and you have two options:
1. Tack solder a resistor on the foil side. You can insulate part of the lead if you have to to existing points or
2. Get a 1K surface mount component, remove some resist and solder.

It's reverseable just by bridgeing the trace. If the gap isn't too big, it can be bridged just using solder.

The non-destructive method might work.

1/8 or 1/4 W is fine.

As to a cutting method. All you have to do is score a line through the trace, but the switch will have a large pad though.

 

[lemofus]

OK, I cut the trace, just wanted to make sure there was no misunderstanding before doing so; and added a 1k resistor. Also added a second D5 in series. Now, measurements:

I need you to look at R35 and Pin #1 of the OP amp (pre-amp ckt) with no audio in signal. Actually look at both side of the diode (D1?) with respect to ground. Also look at it with an audio signal. Look at the AC signals as well under both conditions. That OP amp could be oscillating.

No Audio signal, but amp ON:
R35 - Pin#1 = 0.4V
D1 (OP side) - GND = 0.5V
D1 (R36 side) - GND = 0

Audio signal going through + amp ON:
R35 - Pin#1 = Anything from 0.02V up to 1.2V. Even in manual range, jumps a lot.
D1 (OP side) - GND = between 0.08V and roughly 1V, quite a bit of change there.
D1 (R36 side) - GND = approx between 4-5V. Variance here seems quite plausible, probably changes with music signal...

For the D1 (OP side) - GND side, I assume I got the measurements right. I suspect D1 on my schematic is actually D5 on the board? Some things are not coherent between the board, plan (and possibly my plan). On the preamp section, for example, there is only D5 and D6. can't seem to find anything else in the way of diodes !

 

[KeepItSimpleStupid]

Perfect.

Check that there is a noticeable delay between when you turn the switch to on and when the light changes color. The same delay should exist when you apply music. The 1k should add a delay when you move the switch to on. That probably wasn't there before.

This still leads me to suspect Q2 based on the other measurements you did.

That said remove Q1 and Q2 and check the gain and compare against the datasheet, otherwise just replace Q2 and possibly Q3 since you don't have a way to measure leakage.

Actually have an idea: With the switch in the OFF position and no music, measure across R41. Probably millivolt scale.
Also do the same for R38. This would be an attempt to determine which transistor is leaky. The leakage current would be I=V/R providing the transistor is not turned on. Gains being different than the suffix value 2SA1015 pdf, 2SA1015 description, 2SA1015 datasheets, 2SA1015 view ::: ALLDATASHEET ::: would also be a clue.

So what I'm saying is that you may be able to determine which transistor is leaky by an in-circuit voltage measurement. Then by comparing the expected gains (use suffix and datasheet) with the measured gains to also tell. Finally, you can "give up" and just replace them both because you don't have a leakage tester or you can try to "create a leakage tester" with a batttery and a large resistor 10K-1M ohm through E and C, base open. Measure the voltage across the resistor to get a current. Should be probably 0.1 uA or lower.

Although the extra diode increases the delay, it does not change the fact that if it was left in auto mode and Q2 was leaky, there might not be a delay. Thus the 1K resistor makes the ON mode behave like the Auto mode in terms of delay. So, if Q2 were bad, there may be nearly no delay in auto mode but with the added resistor, it will exist in the ON mode which it didn't before.

 

[lemofus]

Hmmm, no noticeable delay when I switch it on... Can't see any change when I switch between the 1K resistor and the 2 ends shorted (ignoring the resistor). Strange?

 

[lemofus]

Well, on the plus side, I do have both a spare C1815 and a spare A1015Y since I assume you asked me to get those at some point! So if Q1 or Q2 are problematic, easy fix. I'll test them for hFE.

 

[KeepItSimpleStupid]

Only the OFF position will discharge the cap. The only real tests are:
1. OFF(wait) to ON(watch for light)
2. OFF(wait), switch to AUTO and ADD music immediately.

OFF discharges the timing circuit, Q2, I believe, is charging it slowly, hence no difference between ON and AUTO, but there should be a delay between OFF/ON.

I'll lave to see if I can calculate the delay.

 

[KeepItSimpleStupid]

Mmmm. I calculated about 50 ms for 1 diode and 100 mS for 2 diodes. If you put in one of your 5.6 V zenier diodes (reverse the polarity) you should get about 0.6 sec.
Values are aprox: I used 0.5, 1V and 5.6 V. At 0.1 seconds you can't tell. Hopefully my calcs are right.

Replace it with a Zenier diode installed backwards compared to the regular diodes.

 

[lemofus]

OK, so with the AC power on, but the amp on stand-by, there is still no noticeable delay when I shift the switch from standy-by to ON. If, however, I have the AC OFF, switch ON, and turn the AC on, then there is a slight delay during which the green LED gradually gets brighter.

 

[KeepItSimpleStupid]

Doubt you can see a tenth of a second. Put the 5.6 V Zener in the two diode's place, but reverse the orientation. You should have 5.6 V to ground at the input of the OP amp. If you put it in like th regular diodes, only about 0.6 V.

 

[lemofus]

sadly that is with the zener in there, in reversed position! I put it in the opposite way of how the original diode was... now I'm baffled!