Help with YJ IRS2092 200W Class D Amp (long)


2005-02-17 3:44 pm
I'm completely clueless when it comes to solid state and chip amps, and I sure would appreciate help on a project of mine. I'm a hollow state guy, and as such, I'm an avid DIY tube dude. I have a new project that I got into, a DIY turntable project. This led me to wanting to have a variable frequency 3 phase motor supply for an old Papst 3 phase motor. I was hoping that someone on this forum may be able to help me with trying to implement these class D amps for this project.

First a little background on what I want to do. The old Papst motors are one of the finest motors for turntables ever, they are a 3 phase delta wired hysteresis synchronous motor. They run the best on 3 phase power, and the speed is locked to the mains frequency, not the voltage. They can be run on single phase power, by the use of a "faking capacitor", however there is some cogging, power loss, and excess heating when run this way. So with a variable frequency 3 phase supply, I can dial in nearly exact RPM, change TT speed without additional mechanical linkage, have more power, and a cooler running motor.

The plan is to use a 3 phase signal generator, that I can dial in and change frequency, set the output phase angles, and vary the output voltage to a certain degree. The output from the 3 phase generator goes to 3 of the YJ IRS2092 200W Class D Amps, the output from the 3 amps is fed to three step up transformers. The secondaries of the 3 transformers are wired to a 3 phase delta configuration, and that feeds the 3 phases of the delta wound motor.

I'm such a newbie to the chip amps, that I bought a 50V 8.4 amp 420 watt supply to run the amps. I tried to run them, but they wouldn't fire up. Now I know that I need a 50V dual rail supply +50V and-50V, which I have found at parts express for $19.95 each, I'm planning on buying them, but need to know if these amps will do what I want them to. I don't want to be throwing good money after bad. Unfortunately these amps came with absolutely no documentation, just three amps wrapped in plastic, in a box, no schematic, nada. So this brings up several questions.

Question 1: Have I already damaged these amps by hooking them up to a single rail supply? All the smoke stayed in the components, so hopefully they will work when I get the right supply. I find keeping the smoke inside is key to audio success.

Question 2: What range of input voltage will trigger the amps, and what kind of gain can I expect from these amps? From what I see, a gain of ~ 30 is typical with chip amps. If so, around .38VRMS will drive the amps to the output voltage I need ~ 11.5VRMS.

Question 3: What's up with the input of these amps? Seems as if they have a three pin connector there, and is labeled 5.1V beside the connector, the 2 outer pins are labled In, and the center pin GND. Why 2 Ins? I'm used to seeing one In, and one Gnd. When I test the pins with a meter the 2 pins test in continuity so seems to me I could wire just 1 of the outer pins to the signal generator output, and just leave the other alone. BTW anyone know where I can get the appropriate connector to hook to this amp, as it is a bitch to try and solder a wire to the pin in such a small space.

Question 4:What does the small blue pot with the Philips adjuster, on the input side of the board do?

Question 5: Will my plan for a load work? Plan: Most fail when they try this, as the amps either blow up, or shut themselves down via overload protection. This is because, if you just hook up the transformers as loads, the the amps see that as pretty much a dead short on start up, until the cores magnetize. When the cores magnetize, the amps then see the reflected impedance of the motor through the transformer as the load. To overcome this, I plan to use transformers at a 1:20 step up ratio. This should reflect back about a 4.5 ohm impedance, from the motor through the transformer. If I then put a 50 watt 4.1 ohm resistor in series with the transformer, as a load, I should get a start up impedance of 4.1 ohms from the resistor, with an additional 4.5 ohms impedance when the cores magnetize. With a 11.5VRMS @60hz across the combination load of the transformer (when energized) and resistor, should give me about 6VRMS on the transformer. With 6 volts on the primaries I should see 120VRMS on the secondaries.

The power requirements of the amps should be plenty. The combination of the 1.4 amp dual rail 50V supplies and these amps, even working into an 8.6 ohm load should give me 70 watts continuous per amp. The power consumed by the resistor and the motor is about 15.4 watts per phase, so the amps and the supplies should be way more than adequate for the job (I hope).

Here's a link to the amps: YJ IRS2092 200W Class D Amp Mono Amplifier Board YJ | eBay

Here's a link to the signal generator: Three Phase Sine Signal Generator 0 to 360 ° Frequency Meter Counter 1 00 200kHz | eBay

Here's a link to the supplies:

Any help would be greatly appreciated.

There are quite a number of threads on this site devoted to the IRS2092 amp; while the circuits may not be exactly like yours, they will be very similar and you should be able to trouble shoot some of this using those schematics. The IRS2092 is a class D controller/driver and uses external FETs for the outputs. It is a half bridge design, bipolar DC supply, but single ended output (non-BTL configuration).

The fact that DDS signal generator is .35VRMS leads me to believe the output is taken directly from the DDS chips and is not buffered. These chips are very sensitive to voltage and can be taken out with a single voltage spike even (some times especially) if AC coupled. If the outputs are directly connected to the chips, you should protect them with reverse connected clamping diodes to the 5V (or 3V) rail and ground. You should also buffer them through an opamp.

I would suggest getting the amps to work into an 8 Ohm resistive only load as a first step. Driving inductive loads can only complicate the troubleshooting process. When everything is well behaved and stable, you can work on reactive loads.
At the risk of seriously undermining the original poster's intent … why not use a variable frequency 3 phase PWM commercial motor controller? Nevermind… it is likely that they don't have anywhere near the frequency accuracy control that you're looking for.

Sure, it'd work. Pyramid's advice about using an 8 Ω load is OK. It could be anything from 30 Ω down to 4, but note that the lower the ohmage the more expensive the resistance. Since one expects that the synchronous motor will be working off near-mains voltages, you could just use some 50 watt incandescent bulbs as the test load. Cheap. Uncomplicated. Visual.

The Papst Aussenlaufer motors came in either 110 ⁵⁰/₆₀ Hz or 220 V 50 Hz models. Since driving 220 V is pretty much outside of what a "class D" amplifier is going to muster (without causing the design to use a threesome of step-up transformers!), I'm assuming you'll be using 110 V devices.

Now while as you state, the rotational rate is dependent on frequency, not voltage, (which in turn implies that you could drive one at a substantially lower-than–110 voltage), be aware also that all of the nice qualities of the nominal ratings (acceleration from a stop to full speed, and amperage drop) change too.

Indeed … as I'm sitting here thinking about this, I'd pretty much want to run the whole thing off a triplet of 110 : 24 volt transformers. And, BTW - you do not need 3 amplifiers! using the "delta" wiring for the transformers, you can do it and maintain 120° leg phase with only 2 amplifiers. "Left" and "Right" channels so to say.

In fact, if you could stumble upon a pair of dirt-cheap "mono block" amplifiers of ANY kind (digital or analogue), and set up your 3-phase generator right, with the 24 → 110 transformers in place, you could test out your theory fast and easy. Doesn't matter the efficiency, as long as you're just testing. Then, if you determine that it works well, then and only then commit to finding the fit-for-purpose solution.

I can't imagine that the Papst Aussenlaufer motors nominally take more than a handful of watts, and perhaps 50 to 100 during out-of-phase startup. Think about the low-voltage thru step-up transformer solution, sir.

PS: there is no such thing as a one way transformer. Therefore, a 110 V primary, 24 VAC secondary is just as readily used as a 24 V primary, 110 v secondary. There will be some voltage loss (because the nominal transformer isn't 100% efficient), so you might have to drive the 24 volt side with 26 volts to actually see 110 V on the "output". But that's a small price to pay - since you can use commodity transformers directly. And they're dirt cheap. And you only need transformers large enough to handle the nominal running power. ANY transformer can handle 10× its rated power … for a few seconds at a time. They heat up a bit, but then it all settles down to nominal in the moments that follow.

I'd probably go with DELTA primary and WYE ("Y") secondary winding hookup configurations. By doing this you easily convert from 24 VAC (peak) to 110 VAC (RMS). Actually, its part of the 3-phase spec.