My Sure IRS2092 Amplifier Build

[donradick]

Re: My Sure IRS2092 Amplifier Build

Bob,

1. How is the Corian sold? How large a piece does it come in?
2. Can it be cut with a regular table saw?
3. How do you polish it?

Thanks.

I just researched this in depth - you can find smallish corian pieces on Ebay pretty cheap that are good for faceplates or backpanels.
You can even find cheap corian cutting boards for larger pieces.

There are a number of threads here on PETT about working with corian. Consensus I got is that carbide router bits work well,
but be sure to wear breathing protection.

HTH,

[donradick]

Re: My Sure IRS2092 Amplifier Build

A good compromise is wood with protruding heatsinks to draw the heat outside the enclosure. Apexjr has some nice heatsinks for $4 that have a lip that makes them easy to flush mount into a wood enclosure. One is probably sufficient for a pair of LM3886 chips, or slice them down to use two smaller heatsinks with one chip apiece.

[ATTACH=CONFIG]45374[/ATTACH]

Thanks Neil - now I see what those heatsinks are good for!
For a Class D amp I'd take a chance with the heatsinks inside the box, but for a Gainclone I want the heatsinks on the outside.

[hongrn]

Re: My Sure IRS2092 Amplifier Build

Thanks Don. I didn't know Corian generated that much dust.

I just researched this in depth - you can find smallish corian pieces on Ebay pretty cheap that are good for faceplates or backpanels.
You can even find cheap corian cutting boards for larger pieces.

There are a number of threads here on PETT about working with corian. Consensus I got is that carbide router bits work well,
but be sure to wear breathing protection.

HTH,

[robertcottiers]

Re: My Sure IRS2092 Amplifier Build

Great deal there Neil.. $4 is dirt cheap.

Bob

[Ron_E]

Re: My Sure IRS2092 Amplifier Build

Bob,

1. How is the Corian sold? How large a piece does it come in?
2. Can it be cut with a regular table saw?
3. How do you polish it?

Thanks.

This site has small Corian pieces. http://solidsurface.com/sheet-materi...rder/sqft_sort

Ron

[hongrn]

Re: My Sure IRS2092 Amplifier Build

Great find Ron. See what you started Mr. Bob?

[robertcottiers]

Re: My Sure IRS2092 Amplifier Build

I may have started it but you guys are going to finish it...Gentleman start your routers :D

Bob..Watching from the bleachers

[jsr]

Re: My Sure IRS2092 Amplifier Build

Some stupid questions from a stupid person (um...me)...so be gentle...

I can't find any power info on the IRS2092 datasheet. From what I've seen with chip amps, the power is directly related to the supply voltage. It appears the IR reference design to get 250W at 8ohm requires +/-70V. The Sure uses 55 to 65V rails, but claims the same power (250W into 8ohms). Is the Sure making that or is it lower power (such as at the +/-56V used here)?

Stupid #2: Does the power supply boost the transformer's voltage? I thought it just provides filtering, but the output of Hong's transformer is +/-40V while the PS output is +/-56V.

Thanks.

[JimHRB]

Re: My Sure IRS2092 Amplifier Build

Regarding #2, the transformer voltage is VAC which is RMS. The DC output is the peak voltage (1.41 times VAC RMS) after rectification and the large caps.

[hongrn]

Re: My Sure IRS2092 Amplifier Build

I hope Neil will chime in to answer your question #1 about the voltage requirement versus output, but yes, the higher the voltage, the higher the output. The Sure amp is built for rail voltage between 55VDC and 65VDC, and exceeding 65VDC will shut it down.

On question #2, transformers are rated in AC voltage, which is rectified to DC voltage through a bridge rectifier. To figure out the DC voltage, you take the AC voltage and multiply it by 1.41. Since the maximum DC voltage for the Sure amp is 65VDC, 65/1.41=46, you can theoretically pick a transformer with 45VAC secondaries for the amp. However, keep in mind that house voltage can vary by as much as 10%, so there's a chance the rectified voltage may go over 65VDC, triggering the amp to shut down. That's why I picked a transformer with 40VAC secondaries, rectified to 56.4VDC (40x1.41=56.4) to power the amp and stay within its voltage requirement. I hope this helps.

Some stupid questions from a stupid person (um...me)...so be gentle...

I can't find any power info on the IRS2092 datasheet. From what I've seen with chip amps, the power is directly related to the supply voltage. It appears the IR reference design to get 250W at 8ohm requires +/-70V. The Sure uses 55 to 65V rails, but claims the same power (250W into 8ohms). Is the Sure making that or is it lower power (such as at the +/-56V used here)?

Stupid #2: Does the power supply boost the transformer's voltage? I thought it just provides filtering, but the output of Hong's transformer is +/-40V while the PS output is +/-56V.

Thanks.

[jsr]

Re: My Sure IRS2092 Amplifier Build

Ha forgot about the rms thing. Told ya it was stupid. Thanks guys.

[[email protected]]

Re: My Sure IRS2092 Amplifier Build

I built one with a 40-40 and ended up with 60.3vdc. Its really nice.

[neildavis]

Re: My Sure IRS2092 Amplifier Build

I hope Neil will chime in to answer your question #1 about the voltage requirement versus output, but yes, the higher the voltage, the higher the output. The Sure amp is built for rail voltage between 55VDC and 65VDC, and exceeding 65VDC will shut it down.

I was just getting ready to answer it...

First off, 250W into 8ohms requires about 45V RMS (Power = E^2/R, so the voltage is Sqrt(250*8), or 44.7V). In order to get a nice sine wave at 45V you need a peak voltage of plus and minus 45V * 1.41, or 63V. There are some complicating issues like maximum modulation index and some voltage drop across the output transistors that may reduce the total output from the 250W theoretical maximum at ±63V, so the amp may not actually deliver a clean 250W unless you have higher rail voltages (like ±70V).

But there are two other things going on here. Look at the first attachment, which shows the measured power for the reference amp at ±70V. In the past, amp output was specified for a low-distortion output, using a sine wave. For a sine wave, you need to divide the rail output by the square root of two (1.41), because that gives you the area under the sine curve (RMS power). At .1% distortion, the reference amp puts out a solid 250W. However, a lot of class D amps are specified for clipped sine waves, so they pick some point on the distortion vs power curve, like 10% distortion. At that level the reference amp puts out close to 350W with 70V rails, which would be 300W with 65V rails. For an 8ohm load, that's pretty impressive.

According to what I've read, Sure set the overvoltage protection for their version of the amp so that it cuts out around 66-67V. So the amp may not output a full 250W with 65V rails with low distortion, but it will be close to 300W at the 10% standard. So, yeah, it's a 250Wx2 amp.


The other thing to consider is the behavior with lower impedance loads. The most important specification to consider for an amplifier is the safe operating area (SOA) of the output devices. The SOA curve for the output devices used in the reference amp is shown in the second attachment. What's most important about this curve is that it shows how you can have high voltage across the device at reduced currents, or high current at low voltages, but you can't have both high voltage and high current (which would be "outside" the dashed lines). The SOA curves are for single pulses, and it's not clear how you would select the curve for a class D amp, where the pulses are repetitive, but the key take-away here is that the amp will be able to deliver more current with low voltage rails. For example, if you run the amp with 50V rails you would only get (50/1.41)^2/8, or 157W for an 8 ohm load. However, at that lower voltage you could use a 4 ohm load and still stay safely within the SOA curve. At 4 ohms, the amps will put out a bit more than 250W, according to the measurements in the IR reference design document. So that's another reason for accepting the Sure product as a 250Wx2 amp.


Kind of a wordy answer, but hopefully the technical issues are clear. The simple answer is that the board is in the 250Wx2 class, and how much power you *actually* get depends on the supply voltage and the load.

[hongrn]

Re: My Sure IRS2092 Amplifier Build

Thanks Neil, you ARE the amp sifu...

[Wolf]

Re: My Sure IRS2092 Amplifier Build

do is use a cove bit on either side of the power switch and a simple chamfer around the perimeter....

Bob

Actually- that would be a 'core-box' router bit, not a cove. Coves imply using an edge with the bearing guide, whereas, the Core-Box bits require you either use a router table or an added/clamped straight edge to guide your router/wood.

Agreed on the Forstner bits....

Later,
Wolf