Mackie 408m half loud

[Sowhat]

I have a Mackie 408m powered mixer with the complaint that's it just dropped in volume suddenly. I decided to look at the mixer section later and decided to focus on the amplifiers first. I pulled the mixer section and injected 440hz sine signal into the class H power amplifiers one amp at a time. I could see that amplifier #1 was performing good and I could run the output into clipping at 36VAC into 8 ohms clean and easy… I think that's about right? Amplifier #2 however never got above 18VAC into 8 ohms and was somewhat flakey looking on the scope… sometimes a good sine and then a lumpy one and then good and so on.

Although I still need to trace and check this board from supply on up I have my theory on where this diagnosis will end up. I think it's in the variable power rail and/or power level detection circuits for amplifier #2. I don't think the rails are tracking the signal level or are indeed tracking but unable to cause an effect on the rail voltage. With no high main rail available the amplifier tries to run without adequate rail voltage resulting in a weak anemic output… quite a theory, eh? Anyway, amplifier #2 is not running right and this theory could be complete nonsense.

Any class H experience, ideas or suggestions?

[nickb]

Seems like a good place to look first. It's a little odd the both HV rails are not switching. Mackie ship these with independent switching for the + and - rails so you might like for a common cause. This might help 408 SSWI.pdf

[Enzo]

If rail switching isn't working, as soon as signal gets large enough to need the extra rail volts, the distortion will be tremendous. Or is that what you were describing.

I could be completely wrong, but assuming you are going in the front panel PA IN jacks, I might check to see that the signal at the actual input to the power amp is full strength. Is there any low level circuitry between the jack and the PA board? Like an input buffer fed by the jack?

[Sowhat]

If rail switching isn't working, as soon as signal gets large enough to need the extra rail volts, the distortion will be tremendous. Or is that what you were describing.

I could be completely wrong, but assuming you are going in the front panel PA IN jacks, I might check to see that the signal at the actual input to the power amp is full strength. Is there any low level circuitry between the jack and the PA board? Like an input buffer fed by the jack?

I think that's what I'm seeing, possibly some of the worst crossover distortion I have seen and it's dependent on the signal voltage... low pretty clean high no so much.

I actually removed and disconnected the entire mixer for now. I'm going straight into the amplifier board pins to make my signal connections. I needed to remove the mixer so I could get good access to the amplifier(s) board inside that cabinet without disassembling everything.

[Sowhat]

Mackie ship these with independent switching for the + and - rails so you might like for a common cause.

First thing Tuesday morning I will look at the detector portion first, maybe that U6 op amp is dead.

[Enzo]

Or just look under the commutators to see if the voltage is switching in and out.

[Sowhat]

This might help [ATTACH]46051[/ATTACH]

Thanks, but I have a 20 page version of this manual, wish I had the full manual which is hundreds of pages apparently, but 20 pages it seems will have to do for now, I think it's enough to get the job done... afterall, this ain't rocket science, or is it?

[Sowhat]

Been having problems with my camera but I got a shot of it face on so we all know what unit is being talked about. I'll get some good internal shots up on Tuesday.

[Jazz P Bass]

Thanks, but I have a 20 page version of this manual, wish I had the full manual which is hundreds of pages apparently, but 20 pages it seems will have to do for now, I think it's enough to get the job done... afterall, this ain't rocket science, or is it?

Did you open the pdf file?

It covers the commutation theory.

[Sowhat]

Did you open the pdf file?

It covers the commutation theory.

Yes I did... like I said, I already have that document but the one I found has more information... something like 20 pages. Commutation is a new phrase for me in relation to these amplifiers. I understand commutation like in regards to thyristor reset triggering but I never heard the term applied to a class G or H amplifier. I have read the 808m/808s Power Amplifier Theory of Operation a bit and get the gist of things, they never use the term commutation but it sure sounds good.

Anyway, this is all academic... a 193 amplifier board used in the 808 series has class H topology and supply while the 204 amplifier board used in the 408 series appears to be plain old class AB push pull despite Mackies class H claim. Guess my little theory is indeed nonsense, it's some consolation that I have a working identical power amplifier on the left side of this board for comparison testing.

[nickb]

Can you provide a scope shot showing the problem?
Are you using a purely resistive dummy load?
Is the problem affected by the load impedance?

[Sowhat]

Can you provide a scope shot showing the problem?
Are you using a purely resistive dummy load?
Is the problem affected by the load impedance?

Scope shots? I'll take some tomorrow when I get back in the shop.
Resistive load? Yeah, 400 watts at 4 ohms, 200 watts at 8 ohms and 100 watts at 16 ohms.
The signal voltage tracks the loads pretty much as you would expect... I didn't do the 16 ohm one however. The distortion shape remained the same, just bigger or smaller... or then again not, sometimes it just looks like a little sine wave, maybe depending on when I energize the amplifier... I dunno, something's awfully screwy!

Tomorrow I have some tests to do and I'm sure the problem will reveal itself in short order. So far I have only done the most preliminary of checks, first of which was to check for DC on the outputs, unfortunately I didn't do that, I went straight to the signal and power measurements. I bet if I check for DC on the output I just might find something interesting with amplifier #2.

[Sowhat]

I went ahead and yanked the amp out of the case and started checking some stuff.



All supplies are good, the quiescent currents are good, both amps match, nothing looks to be a problem at first glance but this is the scoped output:



The scope is reading across the load with a 10X probe and the scale is 2V/div, so... on the right we have a little less than 70VPP, (I can't attest if my scope is calibrated well or not but it's in the ballpark). That's right at the point that clipping starts... the top of the rock, so to speak., that's a good working amplifier and just about spot on to the specifications. The left however is much weaker and it's getting the same amplitude input signal as the right... looks like about 18VPP or so, just a whisper.

That radical waveform behavior I started off this thread with is in the stereo channels 7 and 8 of the mixer and that seems to have damaged this amp, I have senior moments every now and again, getting things mixed up a bit, sorry. The mixer is another problem to be attended to later but this amp may very well have suffered input stage damage. Unfortunately, I don't have a schematic but I think I can trace it through. With the amp up on the amp bench now, it's easy to get to everything. Problem now is I gotta go and get my flu and pneumonia shots at the doctor so I will have to wrap thing up for today and start back in tomorrow.

[The Dude]

If memory serves, and sometimes it doesn't so take this with a grain of salt, that series has triacs on the output jack board. It's a crowbar circuit like the Peaveys. It's possible you have a leaky triac and the amp is trying to drive that load. Check the resistance right at the speaker output jack with no speakers or load attached and compare the two sides. See if maybe the bad side measures some low resistance. If so, try temporarily disconnecting the triac and see if your output returns to normal.

[Jazz P Bass]

Schematic: 408-808-power-amp-ps.pdf