Class D Questions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello
Been thinking about this: 75 volts into 8 Ohms equals 9.375 amps. Amps times volts equals watts, so this comes to 703 Watts in round figures. Multiply that by 2 (to include the other rail, and you get 1406 watts total, agreed?
Mackie state that the Input Power for the DLM12S out of the Wall Socket is 250 Watts;
So do You think that it is safe to try this?
My resistor stack is 4 x 8 Ohms at 200 Watts each -- so 703 Watts into a 200 W resistor? -- I think there will be some heat generated!
Awaiting your answer -- Thanks for all --
 
If I was not sure about output of a power stage, I used some halogen burner for outside lamps in parallel or series, depending on the DUT. With a volt and ampere-meter this gives an impression of what to expect. Not the best, as they increase in resistance while heating up, but cheaper than a huge resitor.
Your resistor will perform much better I guess.

I your case it is important to measure both sides of the PS at the same time, but with ground in the middle.

Maybe first use 8+8+ground+8+8 ohms. Set the scope to AC and look how clean the + and - voltages are under load. If both voltages stay at 75 volt you have 2x350 W.

If this proofes OK, I would measure the other auxiliary voltages, under no load and after the point where the output signal gets fuzzy.

All OK? Fine, now it is the amp section the suspect.

First I would compare the signals at the switching power MOS-fet, there must be something that can be seen. I hope 3 test OK and 1 will look distorted.
Then work back, comparing the same points in both channels, as you know how the "good one" looks, you should be able to find the point where the difference appears first.
So far not very much difference to a A/B amp.
I´m really no D-amp specialist, but it seems to me some timing is wrong.
To be honest, often I don´t understand how something works in detail, I just repair it...
 
Hi Telnet100,

Looked to me like Turbowatch2's suggestion in post 19 was subbing for the main rails -- but of course, I'm not a Class D expert, either.

Love those big 225 Watt Ohmites tapped down to 8 Ohms -- got a couple myself, though not air-cooled. But remember, they're VERY inductive! Fine for loading a well-filtered power supply, but can cause some pretty wild waveforms when loading an amp output -- analog OR Class D.

I would think that the best indicator of Power-supply/vs/Amp-section is that you have 2 amp channels that can be diagnosed independently. Unless you have a +/-75V supply that you can sub, maybe just assume the supply is OK for the moment. Then you can keep the scope probes on ground and use ch2 to allow you to watch both outputs at the same time.

Also, if you have one available, replacing the TL081 would be an inexpensive first cut. That wild oscillation on only the negative excursions kinda hints that way.

Sure am familiar with the pain of a dead piece of gear that your income depends on.

Best Regards,
Rick
 
Hello
Thanks for that --
I can't connect the scope probe's ground to the Zero Volts rail of the PS Section, because that is isolated from the Mains Ground. I also at the moment don't have an Isolating Transformer to supply the scope from - (but must get one) - so I am going to ground the probes to the Zero Volts rail via a capacitor. So far, it looks like the PSU is capable of supplying Plus and Minus 71 Volts into those 16 Ohm Resistors; I have only run it up for a few seconds because of the Power Levels involved. Next thing is to look at those rails with the scope. I don't know whether I like the Idea of dropping the load resistor values to 8 Ohms as suggested, as Power Levels then would then be double. But if the waveform at 350 Watts looks to be quiet then I guess that the PSU would be okay?
The Fan that cools those resistors is at the bottom of the stack, and blows upwards to help dissipate the heat. I will be posting the scope views to help others --
Thanks for all.
 
Back Again --
It seems like that Both Plus and Minus Rails of the Power Supply are OK. Both are delivering 72 Volts under load into 16 Ohms, so, 350 watts apiece. There is a 1.5v P-P Mains Frequency AC Ripple Voltage on the DC output, both rails, so it looks like Mackie might have been tight when they used only 2200uF Mains Filter Capacitors in here. There is No real RF Switching Waveform visible on the output rails.
So, it now looks as though I shall have to look further into the Power Amp Stages. I will once again Separate them out of bridge configuration into singe ended and compare one against the other to see what's what.
Thanks for all --
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.