IRS 2092 Class D Amp Kit - Oscillation or Normal?

[TomCarlos]

MEF Members -

To learn a bit more about Class D amplification, I bought a kit from ebay - LJM L15D KIT IRS2092 Power amplifier kit (with IRFB4019 MosFets). For a tad over $20, it was an interesting build. But without a schematic, it kinda defeated the point. Eventually, I did find a schematic for this board - directly from International Rectifier. The schematic is a match. The only thing that is tricky is understanding the Overload Protect and circuitry around the relay. I do not see the relay per se - I do see the "open" note next to R29A and C9A. I also see a couple Diodes on the schematic D5A and D6A near the output. I believe those are created by combining two 2N5551 transistors connected to one another in an odd way. Anyway, I am working through all that. The amp is working - yeah!!

So, my question centers on the ripple (or oscillation) that I am seeing on the Power Supply. I am using my Peavey 400BH Module Power Supply Project for the plus and minus 50 volts. The ripple (unloaded) looks like one would expect - the typical sawtooth shape. But when I connect the power to the amp board, I am seeing something totally different. It is hard to measure the amplitude and try to discern the frequency. Is this normal? I am wondering if the 2092 chip is causing this.

The amp sounds clean - I am just curious about the waveform on B= and B-.

Thanks, Tom

[Enzo]

It is a digital amp, it works at very high frequency, so I don't doubt it leaves high freq artifacts on your power rails.

The only thing that is tricky is understanding the Overload Protect and circuitry around the relay. I do not see the relay per se - I do see the "open" note next to R29A and C9A. I also see a couple Diodes on the schematic D5A and D6A near the output.

What overload protection around what relay? I see none of that.

[TomCarlos]

Hi Enzo...

Thanks for the reply. I was thinking the ripple wave may have something to do with the IC high frequency.

As for the overload protection, this is where the schematic is not complete. In the photo below, you see the blue relay. One end goes to L1A (22uh inductor) and the other end goes directly to the output.

Just below the relay, there are two 2N5551 NPN transistors mounted where the round portion faces the other. I need to trace out the connections very carefully, but it appears they are connected together at the base. On one end you have a collector of one transistor tied to an emitter of the other transistor. That connection goes to a 100K resistor and then to a 1N4004 diode. The diode then goes to B-. The other two transistors have similar connections but they appear to go to B+. That pair of @n5551 are located to the right of the L1A inductor and below C12A, the ,47uf cap.

I will order another kit (to have a pair of amps in a small box) and when that board arrives, I will take my time to trace out these four transistors and see how they are connected into the circuit. It is alot easier to do as you install the parts, checking components and connection points (at least for the components after L1A) and not having to worry about measurements of other components that are installed.

Unless I find another schematic that is more accurate, I will revisit this post at a later date. It takes awhile to get shipments from the far east!!

Tom

[g1]

In case you don't have this Tom, there is a bit about external protection circuits, but no relays that I can see.

[TomCarlos]

Thanks G1 -

Yes, I have that white paper. Figure 25 is the same schematic I posted.

I wrote to the manufacturer of the PC board and much to my astonishment, he claims that he does not have the schematic for his own board!!! So I will draw this one up when I get the new kit.

[nickb]

Scope shots tell us almost nothing if we don't know the vertical and horizontal scales. So what are they?

[TomCarlos]

Hi Nick...

Getting out my digital scope -

Power Supply, no load, 37mv rms, 120 hz.

Power Supply with amp connected, 55mv rms. But you can see the readout showing the frequency @ 1.258 mhz.

[nickb]

Hi Nick...

Getting out my digital scope -

Power Supply, no load, 37mv rms, 120 hz.

Power Supply with amp connected, 55mv rms. But you can see the readout showing the frequency @ 1.258 mhz.

That's better! Thx.

Is the 55mVpp ripple at full power or zero power? If at idle I wonder what it is a full power. The thing that is crucially different about class D amps is that the power supply ripple directly passes to the the output, mitigated only by the negative feedback. In a class AB amp the power supply ripple (usually) has a much smaller effect. Therefore you do need to have a decent low ripple power supply.

The HF hash you see is just the amps switching noise. It's probably of no consequence but might be an indication that the board's power supply impedance is higher than it could be due to poor power rail design.

[TomCarlos]

Yes, ripple voltages are at idle.

I happened to look at equipment (Harbinger powered monitors mostly) that use the IRS2092. That one too uses a similar power supply as to what I have. So just for the hell-of-it, I might look into a different PS design. If nothing else, good to learn.

[TomCarlos]

G1 - I stand corrected....

Figure 25 of the doc you posted is similar to the schematic I previously posted, but there are some differences. The photo below comes close to the PC board but there are components that are missing. In time, I will add those components, document any differances, and repost a schematic.

Thanks again.

[TomCarlos]

Team MEF:

So I finally received my second LJM L15D KIT IRS2092 Power Amplifier Kit and had the time to slowly trace out the circuit. The totally funky part is the output (after the 22uH inductor). Yes, there is a Songle SRD-48VDC-SL-A relay protecting the output that is triggered by some type of protection circuit. I numbered the four 2N5551 transistors so that I could trace out the connections. For the newbies like me, the internal relay switch closes when power is applied. Otherwise, it looks like an open circuit in the datasheet. A (moderate) Starbucks Coffee card will be sent to the first person that can figure this out the protection circuit and explain it layman’s terms to the rest of us!!

I needed a big sheet of paper to draw this out and since I could not scan and 11x17, I will give you the circuit in three pieces. You can compare what I have vs the circuit in post #10. That schematic comes from International Rectifier and is a typical application circuit.

You’ll notice components that are different, components that are on the board but not on the schematic (which is a typical hookup for the IRS 2092 chip), and some components on the schematic that are not on the board.

Why do this? The kit that I bought from eBay did not come with a schematic. And when I contacted the company that sold the kit, they did not have a schematic. Go figure. For those who want to build the kit but have the schematic too, well now you have it. And it is a good way to get an introduction to Class D amps.
Of course, I could have made an error in my circuit drawing. Back in the old days working at National Semiconductor, we had Beer Bust sessions. This is where the Techs and Engineers would review circuits and the Designer would buy you a beer if you found an error. With today’s Social Distancing, I cannot offer that. But please feel free to challenge my drawing and I will take another look. But once the board is assembled, it gets tricky to trace out.

Any hows, here it is.

[g1]

I'd suggest you follow through the ends of the relay coil again (output.jpg). Current must flow through the coil. Transistor bases can not flow any appreciable current.

[TomCarlos]

Thank you G1 -

Crapola... I forgot to draw a point to B+ where the right side of the relay meets the 390K. So we have 50v on one side of the relay coil and 2v on the other side that goes to the bases of T1 and T2.

(Blame it on the double sided PC board !!!)

Edit: There is .52v on the bases of T3 and T4. The Emitter of T4 is at 0v. There is 1.96v at the bases of T1 and T2. The Emitter of T1 is at .4 volts and the Emitter of T2 is 0v. So it looks like those two transistors are not turned on.

[g1]

The coil still needs a current path to energize. The bases of T1 and T2 can not provide that. Is it possible the pinout is mixed up?

[TomCarlos]

I will look at it again.

The good news is that the board is working!! So at this point, I will be taking voltage measurements and seeing if I can figure out if and how the transistors are turned on or off.