[poolorpond12]

I thought that I read a post where someone asked about an input buffer. Has anyone done this? My thought is that an opamp before the volume pot would provide a stable input impedance. It would be easy to correct for offset bias too, without altering the amp.

The only question that I have is if there is enough power in the on board supply for an extra opamp. I was thinking of tapping off the on board supply with a couple jumpers.

 

[amb]

Quote:

Originally Posted by poolorpond12 I thought that I read a post where someone asked about an input buffer. Has anyone done this? My thought is that an opamp before the volume pot would provide a stable input impedance. It would be easy to correct for offset bias too, without altering the amp. The only question that I have is if there is enough power in the on board supply for an extra opamp. I was thinking of tapping off the on board supply with a couple jumpers.

There is enough power onboard for an extra opamp but I am unsure just what "problem" are you trying to solve by doing this? Can you elaborate on what you mean by "provide stable input impedance" (with respect to what?) and "correct for offset bias"?

 

[PinkFloyd]

Quote:

Originally Posted by guzzler Sorry to be contrary here, but I prefer to cut before soldering, as a) you get a much neater finish on the board b) you don't need to heat the component as long (as you have a smaller mass to heat) c) and the solder doesn't wick up the leg. (basically why "a" is the case) You need to apply a little pressure from the other side to hold it flush, but by no means hard. Most components will actually stay in by themselves, as the leads go into the sockets quite tightly

I second that also. I tend to bend the legs of the component so they make a good "mechanical" connection with the pads first and then I snip the legs flush with the edges of the pad. "Then" I solder which leaves a lovely neat finish, this method of soldering is great if you're going to leave the component in situ and provides a very reliable joint. If you're going to be experimenting with lots of different components (say capacitors) then you may want to adopt the pin witness method amb suggests as it's a lot easier to desolder a component that has been soldered "pin witness" standards.

BTW, reading this thread (whoah It's long!) and want to build a M3 as it looks like a super amp where can I get a board?

 

[MisterX]

M3 boards are available from AMB's site:
http://www.amb.org/shop/

 

[GregVDS]

I would like to know if someone incorporated VU meters in an M³ build.

How to do that? take L and R signal, and then after? Assuming the VU meter sensitivity is 0-200µA (130µA = 0VU) and resistance is 250 ohms, what should I do (348-8425 part number RS-components)?

Thanks,

GREGVDS

 

[amb]

Quote:

Originally Posted by GregVDS (VU meters) How to do that? take L and R signal, and then after? Assuming the VU meter sensitivity is 0-200µA (130µA = 0VU) and resistance is 250 ohms, what should I do (348-8425 part number RS-components)?

I guess this depends on what you're trying to accomplish with the VU meters (actually measure something or just for visual amusement).

If you want to be pedantic, the IEC and ANSI standard for VU meters is to display 0VU at +4dBu (which corresponds to a bit over 1.5Vrms into a 600Ω load). If such a meter is to be connected to the output of a headphone amp, you'd probably find that most of the time it shows little motion at all because the output voltage while driving headphones at average listening levels is far lower than that (see below).

Therefore, you might say that you really aren't trying to conform to any standard. Then the question is, what should the "0VU" reading mean? Maximum output of the amp? Some arbitrary level where you get the most meter activity?

Headphone amps can output a very wide range of voltages depending on the situation. The amp is capable of delivering many volts of output prior to clipping, yet depending on the type of headphones you use the "typical" level could be as little as millivolts (for example, the Grado RS-1, a low impedance headphone, is rated to output 98dB SPL with just 1mV, and that's quite loud). Higher impedance headphones tend to require higher voltage to reach the same loudness. Headphone sensitivities also vary from model to model, so at the very least we're talking 2 to 3 orders of magnitude difference in voltage levels between those extremes. That's over 60dB of dynamic range.

The meter you specified has only a usable display range of just a bit over 20dB. If you set the meter up to get a useful range of motion (for whatever it's worth) with a Grado, it might be quite a useless range for a Sennheiser.

As such, I think it should be clear that VU meters on a headphone amp are of dubious value.

As for the particular VU meter you linked to, it doesn't look like a proper VU meter (good VU meters are a lot more expensive than that). I defer you to the following site for more info (courtesy of Elliott Sound Products) about VU meters and their drive circuitry:
http://sound.westhost.com/project55.htm

 

[GregVDS]

Ok,

So no interest in a headphone amp because of the very different output levels!

I was sudden nostalgic of these moving needles, I was fascinated by them when I was kid.

If I follow correctly, one should callibrate the circuit powering the VU for a given headphone, based on its sensitivity and impedance? and if we want to display Volt, the needles will practically never move.

It was just a bad idea

GregVDS

 

[skyskraper]

i would have thought with a vu meter that running it before the headphone amp would be the go. there are a few circuits out there for driving meters.

 

[GregVDS]

This would be nice, but the VU meter would only reflect the dynamic of the signal, not the dB.

This is nevertheless a great idea, I will continue to think about it.

thanks for the suggestion!

GregVDS

ps: I received the Mouser pack today, and have practically everything. I only miss the L1 for the STEPS PSU, and forgot the crimp terminals for the molex connectors, I was thinking the connector was already equiped with. I will begin the solder as soon as possible but I'm stuck at a conference all the end of this week, and the WE seems already full, so, nothing before next monday I fear

 

[skyskraper]

ok it was late and id had a long day. run it before the amp section but run it from the output of the pot. iirc the few vu meter circuits i looked at when i was considering doing this had the ability to calibrate via a trimpot the gain/response/whatever of the meter. again it wouldnt be accurate. i gave up when i realised that accurate meters would cost nearly as much as the rest of the amp

 

[MisterX]

Quote:

i gave up when i realised that accurate meters would cost nearly as much as the rest of the amp

Unless you go for one of the annoying flashy led level meter things..... those are pretty accurate and very easy to implement.

 

[skyskraper]

even they wouldnt be as accurate as a proper BBC specced peak program meter.... thats what i want

 

[MisterX]

Quote:

Originally Posted by skyskraper even they wouldnt be as accurate as a proper BBC specced peak program meter.... thats what i want

It's all for visual entertainment anyhow.....

 

[skyskraper]

thats why i ended up putting one of those nightrider dancing led kits on there with pink blue and purple led's. so pwettty

/sarcasm

 

[Dougigs]

One thing that would be worth your time, and might provide useful information, is a clipping indicator:

http://sound.westhost.com/project23.htm