Mark Bass hates bass

[diydidi]

Monitor the supply rails while slappin da bass. These amps dont have a regulated supply and there is no feedback to the SMPS control IC from the secondary side.
Another thing to check is the smps osc frequency. If its lower than the design intends you can have dropout issues.

[dmeek]

Just fixed a Little Mark 250 today with this problem, C15 1uF/50V was leaky (C29 on other schematics)

This schematic shows C29 connected to +12V as C15 was in the 250.



edit: Maybe there is a mix-up on those schematics. Here is an in-focus-as-I-could-get-it photo of the pcb (Little Mark 250). C15 is the problem cap, C29 is maybe a .1 uF to ground.

[nickb]

There is a second protection mechanism on these. The voltage across the 0.33 ohm source resistors is sensed by the optocoupler IC10. How much voltage is required is not clear as the exact value of the current transfer ratio is not known or well controlled, but if we say we need 10mA in it's diode for it to trip the over current protection then that means 3.2V across the source resistor and a load current of 3.2/0.33 x 3 = 29A. That's rather a max, in reality it will be lower than that for most optocouplers

Now 400W average into 4 ohms is a peak load current of 14.1A so the numbers seem reasonable.

Now I've seen this problem and the reason was one FET (see later) was low gain so making the others take the load current and tripping the current protection. I could not make it fail on the bench so I had the owner come in with his bass and play while I scoped. He really wasn't playing loudly at all yet easily produced peaks that tripped the protection. BTW the amp had been sent to Mark Bass and they couldn't figure it out. The LED D15 is the indication that it's the current limit and not the voltage sensing.

Ideally the limiter should kick in but it's just not fast enough to stop these peaks. In my opinion the current limit needs to be a bit smarter than just peak detecting and should require the over current to be a bit more persistent. Maybe a clipper on the input to the power amp is a better way.

The owner lost confidence in the amp and sold it so I never really found out if putting in a matched set of FETS on the one bad side fixed it.

Other factors are that at bass frequencies and away from speaker resonance, the speaker impedance is lower and so currents are higher.







LM3_IProt.pdf

[drewl]

Hmm...
I had already replaced C29 so that wasn't it.

Grounding the + end of C29 it would not turn the relay on.
Lifting one end of C29 it would engage relay but trip with a small signal and no load

I think one or more of the FET's in the detect circuit may be leaky/bad?

However like Nick said, one of the FET's might not be as matched as the others.

I got them at work where they were used in power supplies and we're probably not matched from the manufacturer.

[nickb]

Just in case I didn't make the point very well above, if a fault LED is on then the problem is not DC voltage on the output. There are two fault indication LEDS that I can see: D16 for over current and D15 for temperature.

[Helmholtz]

@nickb : Just curious, in your post above (#34) I see a large empty frame below the text. Should there be some content?

[drewl]

He pulled the relay to get at the parts as there is very little room to get in there.

Found a way to make it trip with a signal generator, pump in a square wave and it trips at the same level as a bass.

Then I noticed one of the FET's I replaced is a 9140 when it should be 9240.

Found a 9240 from the same batch and it's working with a square wave
Cautious optimism!

We'll see if it works with a bass.

[nickb]

@nickb : Just curious, in your post above (#34) I see a large empty frame below the text. Should there be some content?

It's a part of the schematic showing the over temperature and over current protection. It's showing correctly for me in the post above. Try a different browser or clear the cache.

[Helmholtz]

It's a part of the schematic showing the over temperature and over current protection. It's showing correctly for me in the post above. Try a different browser or clear the cache.

Thanks, but clearing the cache didn't help.
Tried my old laptop (also Firefox) - no success.
Switched to Explorer - same.

Am I the only one?

[eschertron]

I am also seeing the 'not seeing it' thing. Strange.

[The Dude]

FWIW: I don't see it either. Maybe it's a link to a local attachment on Nick's computer and only he sees it?

[drewl]

Can't see it either, thought you meant dmeek's post.

Anyway, cranked a bass through the amp and it's working great now.

[nickb]

I've edited #34 to add a pdf of the relevant bit.