Event Studio Precision Repair

zumazon

Member
2008-02-01 12:24 am
Hi Guys! First post in this great forum so forgive me if this is at the wrong place (please let me know and i will move it). In particular I'm not sure if i should post this here or in the chipamps section.

I have a pair of Event Studio Precision 6s, active. One of them is broken - it has extremely high sensitivity (much louder than the other one), and about 3 secs after powering on, the woofer starts to move (and sound like a heart beat), even without input.

I exchanged the amps and concluded that the only thing at fault is the amp. On closer inspection, three capacitors next to the LM3886TF chips are fried. Two of them are surface mounted. The other one is a ceramic and reads "108" I think, so it is a 0.1pF capacitor. Problem is, I don't have anyway to find out the capacitance of the two surface mounted ones.

Q: can I use higher capacitance in place of lower ones? If so then I just need to solder one with the highest capacitance I can get my hands on. Or can some one here that uses the LM3886TF to build their own amp shed some light as to what the capacitance of the capacitors near the chips should/could be? I found that you can buy 2000 pcs in total of the capacitors in 32 different values for dirt cheap, $15 shipped. But I'm not sure whether testing wrong values would damage the rest of the system...

On a related note, the fuse looks "oxidized", not sure if it is broken. The circuit board says T4AL/250V which I gather means slow blow, 4Amp Low
 

Sangram

Moderator
Paid Member
2002-09-25 11:01 am
India
The SMT caps are most likely the decoupling caps, 0.1 uF not pF. Be sure that they are caps, because really small SMT caps usually do not have the value written on them IME, only resistors do. Only electros and Tantalums carry markings. You have photos?

The 3886 is a very widely published design here, and the schematics are similar with small tweaks in component values and topology. I am pretty sure that a commercial monitor would implement the manufacturer schematic, which you can get off the National site.

With a magnifying glass you can figure out which traces/pins connect to which components, and therefore which components are fried.

Your problem may extend beyond the faulty caps/resistors though, not sure that replacing them is a solution that will help for long. Unless you can figure out what the exact problem is, troubleshooting from here is going to be a problem. You might want to contact Event and see if they'll send you a service manual. That would help a lot.
 

zumazon

Member
2008-02-01 12:24 am
Thanks, ShinOBIWAN!

I will post pics of the amp tomorrow. I tried getting the capacitance from an authorized repair shop but of course they declined. Tried asking Event, no luck either. Btw, Event's tech support seems to be a one man team, Michael...
 

zumazon

Member
2008-02-01 12:24 am
Hi sangram,

Thanks so much for helping. When I posted the reply I didn't see your post (new member so require mod approval).

More details (to answer your questions and support the picture to come later):

I'm pretty sure the SMT ones are capacitors because the circuit board labels them as C#, where # is some number, and all the other non SMT capacitors (eg the 10000uF for the power) are similarly labeled. I measured the SMT to be of size 0805 (or 2012 metric). Tried googling, apparently the color of these means nothing.

The 108 I was talking about is a disc ceramic. I tried looking thru the schematics from chipamp.com but non of them are of 0.1pF which seems really small. Actually it looks more like 10B but I think capacitors should all have 3 numbers and the lettering is optional (tolerance).

I can see that the 3886s have a few pins close to the caps that look burnt but that could be just the board. I have ordered them anyways in case I need to replace them. Are all the 3886s the same? they have different numbers preceding them.

The fuse looks "oxidized", and is labeled T5/250D (what does D mean here?), whereas the circuit board says T4AL/250V. One reason I think the other parts are fine is that before I could get my hands on a T4AL, I used a fast 3.15A one, and it did not break.

Thanks again and I will post the pic soon.
 

Sangram

Moderator
Paid Member
2002-09-25 11:01 am
India
Hi,

No problem.

If the labelling says C, then they're probably 10uF electro/Tantalum caps, awaiting your pictures as that will tell us what they really are.

3886 comes in T and TF types. T is non-insulated (metal tab) and TF is insulated (plastic tab). Substituting a 'TF' part with a 'T' part has issues with isolation of the chip from the heatsink. Also a TF part will not necessarily drop into a 'T' because the mechanical dimensions vary and may cause poor contact between the chip and sink.

Download the 3886 datasheet from National, it'll give you a decent idea of the considerations you need when replacing the chips.

Let's see some pics, and don't worry, you'll be fine. I'm just a hack, once you put some pics up the real experts can come in and help you out :)
 

zumazon

Member
2008-02-01 12:24 am
Here are the pics! High res, too large to attach here.

This is the circuit board:
http://img162.imageshack.us/img162/8110/p1010161anvh4.jpg

The chips are:
(from top to bottom)
JM41AK LM3886TF
PM38AK LM3886TF
JM33AQ LM3886TF
JM41AK LM3886TF



This shows the disk cap that probably burnt out (i wanted to remove it but i was destroying the code in the process so i left it there for the time being):
http://img116.imageshack.us/img116/861/p1010166anav1.jpg

It reads like 100 to me now...

This shows the backplate:
http://img208.imageshack.us/img208/1396/p1010174anky0.jpg

Thanks! If I had known this forum b4, wouldn't have bought the Events.
 

Sangram

Moderator
Paid Member
2002-09-25 11:01 am
India
Well, it doesn't really look too bad just yet.

The burnt SMT caps were probably 0805 size 0.1 uf ceramic caps, and their absence may be causing oscillation in the output. I still can't see anything printed on the SMTs, but it's a safe assumption seeing where they are located on the board.

The disk cap looks a little worse for the wear, and it looks like 104 from where I see it, so another 0.1 uF cap.

Seeing that the caps that did get burnt are in the area of the woofer amp, I would change out both those chips just to be sure. It may have been heat-related, telling from the little flash on the chip.

As for the chip numbers, don't worry about the first line of numbers, all are 3886TF chips, just from different batches.

Edit: You take good pictures.

Edit2: 4 LM3886 for each channel of the monitor, I might actually want to pick up a pair for my little studio. If they were slightly better built, that is.
 

zumazon

Member
2008-02-01 12:24 am
Thanks! I will replace all the caps with 0.1uF ones then, and replace the two 3886s that look burnt. Unfortunately I would have to wait for about two weeks for the 3886 and the smd caps, so can't do anything in the mean time (other than removing the chips themselves).

Thanks again sangram! You are most helpful.
 

ShinOBIWAN

diyAudio Member
2004-02-25 9:13 pm
UK
zumazon said:
Thanks! I will replace all the caps with 0.1uF ones then, and replace the two 3886s that look burnt. Unfortunately I would have to wait for about two weeks for the 3886 and the smd caps, so can't do anything in the mean time (other than removing the chips themselves).

Thanks again sangram! You are most helpful.

Quick tip for removing the chips without wrecking the PCB. Cut the legs and then remove each pin individually. It sounds like common sense but before someone told me this, I was actually trying to de-solder multiple legs and remove the whole thing which was a real PITA.
 

zumazon

Member
2008-02-01 12:24 am
Thanks ShinOBIWAN for the tip. I'm just about to start removing the chips.

But before i do that, i am looking to order the smd ceramic caps, and there are diff types like NPO, X5R etc. Which one should I order? I can try replacing only the caps first if i can get them before the chips.
 

Sangram

Moderator
Paid Member
2002-09-25 11:01 am
India
AFAIK you're not gonna get np0 caps in that value. Actually nothing except X7R and X5R, either work in that position.

To remove chips non-destructively I use desolder braid, works like a charm except if the holes are through-plated.

For these kind of boards though the destructive process works better. The heat bakes the adhesive between the tracks and the boards, and the tracks can come off. Use extreme care whichever way you go.
 

Sangram

Moderator
Paid Member
2002-09-25 11:01 am
India
50 is fine for the voltages we're dealing with here, but 100 can't hurt. The chips are probably at about 30 volts in a bridge configuration for each amp, or 35 if they're in parallel.

Try and get a high-temperature type if possible, the D-K catalog has a few industrial grade parts that may help. X7R parts will survive 125 degrees - the 0805 Murata 0.1uF/50 X7R caps are digikey part no. 490-1666-1-ND, and they are cheap.
 

zumazon

Member
2008-02-01 12:24 am
Guys, I got the chips and replaced them (in addition to the caps). Tough job since I couldn't remove the backplate (the power socket and the switch was holding the board to the plate).

No popping now! Thanks to sangram and ShinOBIWAN!!

One slight problem tho - the input sensitivity is stuck at quite a high level, and the sensitivity knob does not work at all; this one is at least 10db more than the other, so it hisses quite a lot. Any one has a clue which part(s) I'd need to replace to fix this? The HF trim seems to be bad too, where as the LF trim seems to be partly working...

For your convenience, again here is the circuit board b4 repair:
http://img162.imageshack.us/img162/8110/p1010161anvh4.jpg

Here is the backplate:
http://img208.imageshack.us/img208/1396/p1010174anky0.jpg
 

Sangram

Moderator
Paid Member
2002-09-25 11:01 am
India
Sounds like the feedback loop has gone shot. I would bet on a bust Ri or Ci.

I can't seem to figure out a Ci, but the Ri for the two affected chips seem to be R43/56 (the 1K resistors - 1001).

At this point I would hesitate to go further without a schematic or the actual circuit board in hand.

A simple injection tester will tell me all, I would inject it at various points to find differences between the working speaker and the defunct one.

I am now beginning to think it may be some chips in the low-level signal area (the 5532). Some more probing will get you the answers, but I think this is about when I'd get a tech to really troubleshoot it.