2x250 IRS2092 amp build

[neildavis]

Re: 2x250 IRS2092 amp build

I'm more confused then ever!
http://sound.westhost.com/project04.htm

Project04 is a solution for a different problem. It uses separate filters/reservoirs for each channel to keep interaction between left and right channels to a minimum. That only works when you have separate power connections to each amp. However, the 2x250 board uses a shared power supply connection, so project04 is not relevant.

[[email protected]]

Re: 2x250 IRS2092 amp build

I'm more confused then ever!

http://sound.westhost.com/project04.htm

[Pete Schumacher]

Re: 2x250 IRS2092 amp build

Pete,

With independent secondaries, wouldn't you want to do this instead of tying them in series (please see enclosed drawing)? I think you would preserve the current of each secondary pair for each channel. Thanks.

That will work because both supplies are floating. The previously linked PDF had a ground running from the caps of the individual sections which was quite odd.

[hongrn]

Re: 2x250 IRS2092 amp build

Pete, Neil, thanks so much guys. This has been an exciting learning session. Just when you think you got it finally, you ain't got it. Ben, Wolfie, thanks for your timely input.

Hong

[neildavis]

Re: 2x250 IRS2092 amp build

Pete,

With independent secondaries, wouldn't you want to do this instead of tying them in series (please see enclosed drawing)? I think you would preserve the current of each secondary pair for each channel. Thanks.

Pete's approach is the most common configuration, used in many, many amplifiers. However, there are some claims that the dual rectifier approach will reduce noise due to saturation currents in the transformer (see this link). Either solution will work, but there may be some cases where the dual rectifier approach is advantageous. Pass later explains that:

As I said, this is only an issue when there is an imbalance of windings and net DC current, which is not a typical situation.

This situation should not be an issue for this amplifier and these transformers, so I'd second Pete's recommendation to use the simpler approach (unless, of course, you already have a working dual rectifier solution .)

[hongrn]

Re: 2x250 IRS2092 amp build

With dual secondaries of 45V each, you wire the secondaries in series and choose the center tap as ground. Then you run the dual power supplies off that with a rectifier bridge for each channel. Each end will be your + and - 65.

[ATTACH=CONFIG]46972[/ATTACH]

Pete,

With independent secondaries, wouldn't you want to do this instead of tying them in series (please see enclosed drawing)? I think you would preserve the current of each secondary pair for each channel. Thanks.

[donradick]

Re: 2x250 IRS2092 amp build

Guys, I'm really enjoying this discussion!

For my first build, I bought a power supply board, but sooner or later I'm
sure that I'll try building a PS from scratch. Good tips and analysis here.

[Pete Schumacher]

Re: 2x250 IRS2092 amp build

Thanks for chiming in Pete. Please see the enclosed schematic. This is how I wired my IRS2092 amp, and I get no humming, buzzing, and can play as loud as I can into a 4 ohm load. Since the Sure board requires a separate PSU for each channel, I used a toroid with dual secondaries, feeding each into a bridge rectifier. Are you saying that this wiring scheme is incorrect, and I won't get the rail capacitance I expect to get? Would you correct this by just having one bank of reservoir capacitors per PSU, then the center tap would go the board, to the chassis ground as well? Thanks.

Hong

That schematic has me scratching my head. When you ground the two PSUs, you've basically connected them together and they're referenced from the same common. That means the negative rail of one will cause a current flow into the positive rail of the other.

With dual secondaries of 45V each, you wire the secondaries in series and choose the center tap as ground. Then you run the dual power supplies off that with a rectifier bridge for each channel. Each end will be your + and - 65.

[hongrn]

Re: 2x250 IRS2092 amp build

Thanks for chiming in Pete. Please see the enclosed schematic. This is how I wired my IRS2092 amp, and I get no humming, buzzing, and can play as loud as I can into a 4 ohm load. Since the Sure board requires a separate PSU for each channel, I used a toroid with dual secondaries, feeding each into a bridge rectifier. Are you saying that this wiring scheme is incorrect, and I won't get the rail capacitance I expect to get? Would you correct this by just having one bank of reservoir capacitors per PSU, then the center tap would go the board, to the chassis ground as well? Thanks.

Hong

[Pete Schumacher]

Re: 2x250 IRS2092 amp build

You're looking at 2 separate PSUs, rectified by 2 separate bridge rectifiers that Lowell put on the same board. Each PSU has +/- rails, and its own ground. Both grounds are connected together, then to the chassis. The + rail of one PSU is connected to the - rail of the second PSU to form the center tap, so the end result is + rail, center tap, - rail. With the ground probe of the DMM on the center tap, Lowell should be able to measure -63V/+63V.

Hong

If each of those capacitor banks is for one power rail, one for (+) and one for (-), and if each of those bus wires represents the full voltage of the rail, connecting two of them together to establish GND, then the short wires between the caps are floating, and the caps are in series.

And if what you said about each PSU having it's own + and - rail, and it's own ground, where is the ground in each PSU connected to the caps? And if you did connect each ground together, you can't then connect the positive of one to the negative of the other. The grounds would have the be floating or you would have large current flow from the Pos to Neg since they share a ground.

Do you have the schematic of the dual PSU including the transformer?

[hongrn]

Re: 2x250 IRS2092 amp build

You're looking at 2 separate PSUs, rectified by 2 separate bridge rectifiers that Lowell put on the same board. Each PSU has +/- rails, and its own ground. Both grounds are connected together, then to the chassis. The + rail of one PSU is connected to the - rail of the second PSU to form the center tap, so the end result is + rail, center tap, - rail. With the ground probe of the DMM on the center tap, Lowell should be able to measure -63V/+63V.

Hong

[[email protected]]

Re: 2x250 IRS2092 amp build

The pic I posted does not have the ground as it was added later. My apologies. In this pic you can see the wires coming out from under to a star ground behind the yellow shrink wrap. It is like Hongs schematic.
There are two identical rails fed by two Bridge Rectifiers. The minus from one goes to b- the positive from the other goes to b+. the other minus and plus are combined and go to the ground tap on the amp.

Sorry I caused so much confusion.

[Pete Schumacher]

Re: 2x250 IRS2092 amp build

My apologies. This pic does not show the ground wire to case, but it is present just like Hongs and the other schematic.

There is no way that this connection is giving you a set of 8 caps in parallel on each rail. Unless you're connecting all those short segments together with a bus wire, you've got caps in series with one another.

Here's how each section should be wired (assuming each bank is for one power rail).

[[email protected]]

Re: 2x250 IRS2092 amp build

My apologies. This pic does not show the ground wire to case, but it is present just like Hongs and the other schematic.

[scottsehlin]

Re: 2x250 IRS2092 amp build

Here is another example. These are polarized caps. I used two of them with the + of one and - of the other joined to form the 'ground' for the amp.

I have attached a drawing of what I think you have based on your pictures and description. If that is correct, you have two 4700 uF capacitors connected in series on each parallel leg, which will give you 2350 uF, but double your voltage rating. 4 of those legs connected in parallel gives you 9400 uF. 9400 uF may be enough filter capacitance for that amp, in which case you have a lot of voltage safety margin, which isn't necessarily a bad thing.

Notice the difference between this schematic and Hong's. In his case, the ground wire is routed between the two capacitors in a 4x2 array. That gives him 4 times the capacitance of each capacitor value between the positive voltage and ground, and the same between ground and the negative voltage.