[slowpogo]

I have been using a TREAD with 24v AC-DC wall wart to power my PIMETA. The other day I was sorting through some old junk, and found a nice 30v, 400mA regulated power supply that was for an old computer printer I don't use anymore. It's the kind that sits on the floor, has separate cables for the wall socket and the device.

The plug is actually compatible with my amp. I was initially keen on using this printer PSU so I could try running the amp at 30v.

But what happens when you plug an already regulated power supply into a TREAD? Is it dangerous at all, or will the power just be that much more "regulated"?

 

[error401]

There's nothing to worry about. TREAD has a bridge rectifier, so there's not even a polarity you need to observe.

Also, if your 24V supply is unregulated, it's pretty likely that it's putting out around 30V under the load of a PIMETA, you should measure it and see.

 

[slowpogo]

I could only get about 27.5v out of the wall wart. Now I know it's safe; will the power running the PIMETA actually be *better* with this dual regulation? I mean, will it be closer to STEPS performance than TREAD?

 

[error401]

If you compare the schematics of the TREAD and STEPS, you'll find that the basic regulation circuit is identical (the components surrounding the LM317). The main difference is that the STEPS has much more pre regulator capacitance (there isn't a CRC or CLC or anything), and the transformer is board mounted. Most everything else is identical. The capacitors are a sort of preregulator in that they reduce the AC ripple and are there to supply current during transients - both tasks that a regulator would be even better at.

If your regulated supply is any good, it should eliminate most of the ripple coming from the AC mains, so the LM317 on your TREAD won't have to work as hard to eliminate it. What ends up happening is that the pre-regulator is responsible for regulating the line and removing the majority of the noise induced there, which is a task the TREAD would normally do. This makes it easier for the TREAD to handle load transients (when current demands change due to a bass note or whatnot, the regulator must be able to keep up with them), and improves regulation.

I think that if your regulator is fairly decent and the power cables are sufficiently thick, it should be possible for such a setup to outperform the STEPS, as a regulator is certainly better than a large capacitor. There's enough local capacitance on the TREAD that it should easily make up for the distance it must be run from the real supply, but I'm not an expert on this. Either way, performance should be improved.

Whether it's audible or not with the PSRR of opamps is another story entirely...

One more note: it sounds to me like your regulated supply is of a linear variety - but if it's a switcher it will probably produce a lot of high frequency noise. One of the shortcomings of the ancient LM317 and LM78xx series regulators is that they don't have much bandwidth and will pass most of this noise untouched. It's above audible frequency, but it can cause stability problems and unwanted interference signals at audio frequencies.

 

[slowpogo]

Thank you, I learned a lot there. One question though, what does "PSRR of opamps" mean?

 

[Jambo]

Power Supply Rejection Ratio. It is a measure of how much of the ripple present on a device's supply is presented on its output. If it's all of it (1:1), then the PSRR is 0dB. I think error401 is implying that decent opamps have good PSRR generally.

 

[ruZZ.il]

Power Supply Effects on Noise Performance talks a bit about Power Supply Rejection Ratio. It's also the first thing that came up on google

error401: are regulators really better than capacitors at providing for quick current draws? they're still limited by the supply current... (I haven't done my homework here yet.. )

and ultimately, the better supply is the one that, while functioning properly of course, sounds better

 

[error401]

Quote:

Originally Posted by ruZZ.il /img/forum/go_quote.gif error401: are regulators really better than capacitors at providing for quick current draws? they're still limited by the supply current... (I haven't done my homework here yet.. )

Gonna make me substantiate my claims eh...I'll run some sims at home, but in most cases I think they probably are. A capacitor bank does (often) have lower output resistance, but nonexistent line/load regulation and little noise rejection capability.

A linear series regulator can be understood as a controlled variable resistor in series with the load. What you end up with in this model is a resistor divider across the input voltage, with the load on the bottom. Feedback is applied so that the series resistance is varied to keep the divider ratio appropriate for the desired output. This is of course greatly simplified, but the model holds fairly true to real devices.

Now, if we have a bad source with high output impedance and no input capacitors on the regulator, it will still be able to cope - to an extent. If we have say 5R input impedance, the regulator can still have a very low output impedance because it can accurately adjust its own series resistance to keep the voltage constant and provide the necessary current even during load changes - until it needs the series resistance to be lower than 5R plus whatever its minimum internal resistance is to provide the requested current. At that point, you'll start seeing it fail to regulate.

In theory, capacitance is only necessary to compensate for the limited bandwidth of the regulator, so that the capacitor can hold the output at the proper value until the regulator can catch up. In reality we place it on the input to do two things - reduce extreme ripple on the input to reasonable levels, keeping the regulator from having to swing wildly and improving regulation by keeping the output element 'close' to the necessary value; and to help deal with fast transients that pass right through the regulator because it's too slow.

I might run some sims when I get home that will prove/disprove/show whether my thinking is correct. You might read Walt Jung's articles on regulators, which can be found on his website.

 

[balou]

ruZZ.il: no, and that's why you most often have two cap banks: one before the regulator to prefilter some ripple from the power supply, then the regulator, and after that another cap bank (or just a single cap, depending on implementation) to provide energy storage for short time current draws. as far as I understand it, the power draw is only limited by the equivalent series resistance of the caps, which is quite low and which in turn makes the max current quite high

edit: exactly same-time posting with error401, and different answer

U=R*I, let's say U=15V and R=0.018R (Panasonic FM 1000uF 25V), resulting in I=833 Amps max.
But looking at the datasheet of the LM317, output impedance seems to be in the same region.
Capacitors at the output might help making the amp an easier load, and thus leading to a better regulation

 

[ericj]

The STEPS also has a lot of AC line filtering.

And fwiw i doubt that your 30v 400ma brick is regulated - I'm guessing that it's just like the piles of 30v 400ma bricks for HP inkjet printers that I see at every thrift store. I've got a few of them.

 

[slowpogo]

From what I can find the supply is definitely linear, but nowhere does it say whether it's regulated or not. It's more than twice the size of my 24v unregulated wall wart, if that means anything.

At the very least it's not a wall wart and it's 30v..

 

[ruZZ.il]

thanks for the replies to my q too. I'll go over it again and give it some more thought

 

[mono]

Quote:

Originally Posted by slowpogo /img/forum/go_quote.gif I have been using a TREAD with 24v AC-DC wall wart to power my PIMETA. The other day I was sorting through some old junk, and found a nice 30v, 400mA regulated power supply that was for an old computer printer I don't use anymore. It's the kind that sits on the floor, has separate cables for the wall socket and the device. The plug is actually compatible with my amp. I was initially keen on using this printer PSU so I could try running the amp at 30v.

There is no reason to believe you will gain anything running a Pimeta at 30V instead of 24V. THat is, unless you have some very unusual high impedance cans that need this extra 6V swing on the output.

Quote:

But what happens when you plug an already regulated power supply into a TREAD? Is it dangerous at all, or will the power just be that much more "regulated"?

What are you talking about? The 30V 400mA supply isn't already regulated, it's a bog standard wall wart, a transformer then bridge rectifier and capacitor, not regulated. The question is just non-applicable.

Quote:

I could only get about 27.5v out of the wall wart. Now I know it's safe; will the power running the PIMETA actually be *better* with this dual regulation? I mean, will it be closer to STEPS performance than TREAD?

There is no dual regulation. Hooking up a 30V wall wart after the tread will result in (30V/110V * 24V =) 6.5V. There is no purpose in doing it.

Also, Steps is not that much better than tread, the difference is in steps having more ac filtration "IF" you need it (Not everyone does), and steps being self-contained instead of a separate wall wart. For a Pimeta, frankly a Steps is overkill.

Quote:

From what I can find the supply is definitely linear, but nowhere does it say whether it's regulated or not. It's more than twice the size of my 24v unregulated wall wart, if that means anything.

What supply is definitely linear? A 30V 400mA printer PSU isn't linear. It's just a basic wall wart as described previously, whose input to the printer is then later regulated in the printer.

Presuming you have the Tread and are using it, there is no reasonable further stage tacked onto it to power a Pimeta. The next improvement OR change in sound would come from different Pimeta opamps, different pimeta buffers, or different amp entirely. PImeta just isn't a fragile amp that succumbs much to PSU noise in the first place, with it's opamps having feedback and good PSRR, and having a ground channel, it is more immune to PSU noise than practically any other amp (topology).

 

[slowpogo]

Quote:

Originally Posted by mono /img/forum/go_quote.gif There is no reason to believe you will gain anything running a Pimeta at 30V instead of 24V.

Some people think certain opamps sound better at higher voltages. Just wanted to test this myself.

Quote:

Originally Posted by mono /img/forum/go_quote.gif What are you talking about? The 30V 400mA supply isn't already regulated, it's a bog standard wall wart, a transformer then bridge rectifier and capacitor, not regulated. The question is just non-applicable.

Forgive my idiocy...

Quote:

Originally Posted by mono /img/forum/go_quote.gif What supply is definitely linear? A 30V 400mA printer PSU isn't linear. It's just a basic wall wart as described previously, whose input to the printer is then later regulated in the printer.

Several sites have that model of power supply for sale, described as a "linear power supply." It is made by Lucent. Why four separate web sites are falsely selling it as "linear" if it's so clearly not is beyond me.

 

[mono]

Some opamps may seem better at higher voltages, but that's contrasting some minimal voltage like a 6-10V range with 24V. That is, unless the desired sound is coming from the opamp heating up more but personally I feel that when I thought I heard a difference from opamp heating, it degraded the sound rather than improving it.

You could always pop open the 30V supply, an autopsy should easily determine what type it is. Since I've had a few seemingly similar HP printer supplies that I have opened, and since the need for a linear supply is not present on a printer, and since it wouldn't be the type of linear we're using even if it were (printer going from low power state to high power state with 30V * 1.41 - 1.4 = 40.9V, and this ~ 11V difference at a few hundred mA would be a large thermal problem in a sealed wart.

Besides that, the other crucial issue was this is a step-down wart, whatever you had regulated from the tread, once put through the HP PSU transformer is then down to 110V/30V(+), approx 1/4th of the input voltage is output.

Maybe I am misunderstanding what you are trying to do. IF you took the HP PSU transformer and directly from that output, wired it into the tread input, you would have higher voltage. I was thinking you were also going to continue using the present transformer you are using with the tread in addtion to the HP PSU transformer. IOW, only one of the two transformers can be used.

As for why some 'sites might call it a linear supply, I could be wrong and it could be linear but given the non-linear examples I have, I feel it is just a listing mistake as many just don't know the difference and think a PSU might only be either linear or switching.