Meridian M2 powered speakers

[Luke]

Hello,

I have a friends Meridian M2 speakers. They have an electronic crossover and biamped I believe.
The fault is that the speaker tweeters have a very loud hiss. If I take the amplifier module out and put it in the good speaker the fault moves.
This indicates its the onboard amps or crossover that cause the fault. It gets quite warm so I thought it would be good to replace all the small electrolytics on board (6 in total) and tidied up a few dry joints.
Theres alot of stuff in this small amp so was wondering if anyone has some pointers/experience with this product.
Havent had a good listen yet but they have excellent build quality and I think warrant a bit of time spent on them.
I would appreciate any help.

[Luke]

BTW the dc offset for both amplifiers is 100mv (quescent current 100ma) and 120mv (quiescent current 27ma) on the faulty unit. There is one pot for each of the amps which seems to adjust the quiescent current. The offset and crossover frequencies are obviously not adjusted by pot.
It also has a daughter board which I may swap. This has 3 x LF353 opamps and a 5532, I believe these are regulated.
Is 120mv too much offset for a tweeter? It normally takes a few minutes for the hissing to start.
All the power resistors seem to measure fine and appear to not drift.
Is it possible that the transistors ahve gone noisy? I recal having this problem with a preamp once and the fix was to replace all the transistors.
Any ideas people?

[Shaun]

It's difficult to say what's wrong, but my first guess is that it is some type of oscillation. I think that 100mV is high, given that offset can be trimmed (aim for around 10mV when you readjust). When the tweeter hisses, does the treble increase as well? If so, it may be a faulty resistor. Try to isolate it down to a lower level, such as the amplifier or crossover. Sorry, you have not given sufficient detail; perhaps some pics will help.

[Luke]

Originally Posted by Shaun

It's difficult to say what's wrong, but my first guess is that it is some type of oscillation. I think that 100mV is high, given that offset can be trimmed (aim for around 10mV when you readjust). When the tweeter hisses, does the treble increase as well? If so, it may be a faulty resistor. Try to isolate it down to a lower level, such as the amplifier or crossover. Sorry, you have not given sufficient detail; perhaps some pics will help.

No it doesnt appear to have a pot to trim it, as i said it seems to be for quiescent current.
Havent looked yet for oscillations, Ive no been able to look at this for weeks, I wont bore you. It seems I have narrowed it down to what appears to be the crossover board. It has 4 opamps, alot of r and c'c. I tested the capacitors I could and they dont appear to be leaky, most sit on 19.x Meg which I think is ok. Some I need to unsolder because they are across a resistor. Itested all resistors and they seem fine. Could it be the opamps? They are 3 X LF353 and 1 X 5532, I wonder if they are causing this problem? I will check for dry joints, but solder looks really good.
How do you test a crossover?

[jez]

Probably a faulty LF353

[Luke]

Thanks Jez. It was looking to be an opamp I guess. Do you know if I can replace the lf353 with 5532 opamps? Thats all I have. Maybe i look at the spec sheets, not sure this part is still available.

[VivaVee]

Originally Posted by Luke

Thanks Jez. It was looking to be an opamp I guess. Do you know if I can replace the lf353 with 5532 opamps? Thats all I have. Maybe i look at the spec sheets, not sure this part is still available.

Simple answer, NO. But read on...

The LF353 is a JFET input dual opamp which has much lower input bias current specs than the NE5532. So as a general rule, you could replace the NE5532 with the LF353 but not the other way around.

The LF353 is available ex stock from Farnell or you could look at the AD712 which has improved specs, lower noise in a pin-for-pin compatible package.

From your question, I'm guessing that you are in unfamiliar territory so I wouldn't recommend going any further than the reliable and unity gain stable AD712. More recent opamps will need much more care and attention when dropping them into an old design.

[Luke]

Thanks Alan, good plan, and they are only a couple of bucks. But Im curious, what difference would a couple of mA of bias current make if I chose to use the 5532 opamps?

[VivaVee]

Originally Posted by Luke

Thanks Alan, good plan, and they are only a couple of bucks. But Im curious, what difference would a couple of mA of bias current make if I chose to use the 5532 opamps?

Crossover circuits will typically use high value resistors so that small value (and therefore cheaper and closer tolerance) capacitors can be used. So the low leakage JFET opamps are a good choice there. Depending on the actual topolgy used, the active filters may also have voltage gain. The combination of high value resistors plus gain will lead to greater voltage offsets.

My recommendation is based on a 'better safe than sorry" philosophy. If you really know what you are doing then you can make a lot of things work...

[Luke]

ok so in summary, if I use the same part, Im less likely to have high dc offset.
im convinced, will order some and hope this fixes the hum.