"Superbal" VU buffer
- Thread starter NewYorkDave
- Start date
Help Support GroupDIY Audio Forum:
bcarso
Well-known member
[quote author="Kit"]Very nice, any particular reason for the "superbal" ?
For such a "simple" task, a standard instrumentation amp could do the trick.[/quote]
Well, cheaper for one thing. And the parts are multiple-sourced.
Dave, I guess those compensation caps are from a tried-and-true example? That is the tricky thing about this topology---the phase shift/delay through the second opamp.
For such a "simple" task, a standard instrumentation amp could do the trick.[/quote]
Well, cheaper for one thing. And the parts are multiple-sourced.
Dave, I guess those compensation caps are from a tried-and-true example? That is the tricky thing about this topology---the phase shift/delay through the second opamp.
Brian Roth
Well-known member
How about this variant?
http://www.brianroth.com/library/vu-buffer.pdf
I've used that same input stage design for a variety of purposes. I believe it "looks like" an input transformer with a 20K primary impedance and a grounded CT.
NOTE: I show a 3k9 resistor in series with the VU meter...it might need to be 3k6.
Bri
http://www.brianroth.com/library/vu-buffer.pdf
I've used that same input stage design for a variety of purposes. I believe it "looks like" an input transformer with a 20K primary impedance and a grounded CT.
NOTE: I show a 3k9 resistor in series with the VU meter...it might need to be 3k6.
Bri
Hmm, i dont see it.
Maybe we are talking about different things but my idea was basicly to loose the inverting opamp that sits on top of the other and be done with it. It wont have the exact same Z for + and - inputs but this isnt a mic-pre....
bcarso
Well-known member
[quote author="Kit"]
Hmm, i dont see it.
Maybe we are talking about different things but my idea was basicly to loose the inverting opamp that sits on top of the other and be done with it. It wont have the exact same Z for + and - inputs but this isnt a mic-pre....[/quote]
TL072s are very cheap duals. Resistors are almost free.
What inamp do you know that comes anywhere close? Or are you calling a four-resistor diff amp config an inamp? When you say inamp I think of some Burr-Brown etc. device, like an INAxxx---perhaps that's my misinterpretation.
The four-R diff amp does have equal Z to common-mode signals, which is generally the most important characteristic, so I agree it would be adequate in most cases
Hmm, i dont see it.
Maybe we are talking about different things but my idea was basicly to loose the inverting opamp that sits on top of the other and be done with it. It wont have the exact same Z for + and - inputs but this isnt a mic-pre....[/quote]
TL072s are very cheap duals. Resistors are almost free.
What inamp do you know that comes anywhere close? Or are you calling a four-resistor diff amp config an inamp? When you say inamp I think of some Burr-Brown etc. device, like an INAxxx---perhaps that's my misinterpretation.
The four-R diff amp does have equal Z to common-mode signals, which is generally the most important characteristic, so I agree it would be adequate in most cases
NewYorkDave
Well-known member
This is meant to go across the balanced (well, floating actually) output of a mic preamp, so balanced impedances are very important.
bcarso
Well-known member
[quote author="NewYorkDave"]This is meant to go across the balanced (well, floating actually) output of a mic preamp, so balanced impedances are very important.[/quote]
The only possible enhancement I can think of would drive the VU meter and resistances in push-pull, using the other TL072 output as well. You could dispense with your coupling cap, probably, unless you want that to limit low frequencies and/or are anticipating d.c. offset in the source, and double the ripple frequency induced in the power supply while removing ground currents in the dual rail case. You will also only need half the gain, or double the series meter R's.
The only possible enhancement I can think of would drive the VU meter and resistances in push-pull, using the other TL072 output as well. You could dispense with your coupling cap, probably, unless you want that to limit low frequencies and/or are anticipating d.c. offset in the source, and double the ripple frequency induced in the power supply while removing ground currents in the dual rail case. You will also only need half the gain, or double the series meter R's.
Brian Roth
Well-known member
Well, no one shot any holes in my balanced input gizmo, so I Shall Carry On <g>.
Bri
Bri
Matt Syson
Well-known member
Hi Brian
I am not shooting holes in your circuit but I wonder if it is really necessary to load the source so heavily as 10K per leg.
Same could be said of the superbal design posted earlier. It is a VU you are buffering and you won't be listening to it so take the input resistors to 100K or maybe more. If you suffer a touch of HF roll off then a tad of capacitance across the 3K6 VU resistance should sort it out although with only 100K resistors on the input it should be fine.
Driving the meter differentially from the Superbal sounds a good wheeze to limit ground currents but it may be best to be nice to the meter and limit the available voltage swing as over 20dB more than it's full range is a trifle antisocial!
Proper VU meters can have suprisingly good responses. Some Sifam types I have (the expensive type!) read flat from a few Hz (where the needle flaps about) to about 100KHz on sinewave.
Using a pair of opamps as simple buffers is simpler but consideration should bre made for the possibility of overdriving the inputs so there should at least be some series resistance, again say 100K to limit current if the buffer is unpowered while signal is present.
Matt Syson
I am not shooting holes in your circuit but I wonder if it is really necessary to load the source so heavily as 10K per leg.
Same could be said of the superbal design posted earlier. It is a VU you are buffering and you won't be listening to it so take the input resistors to 100K or maybe more. If you suffer a touch of HF roll off then a tad of capacitance across the 3K6 VU resistance should sort it out although with only 100K resistors on the input it should be fine.
Driving the meter differentially from the Superbal sounds a good wheeze to limit ground currents but it may be best to be nice to the meter and limit the available voltage swing as over 20dB more than it's full range is a trifle antisocial!
Proper VU meters can have suprisingly good responses. Some Sifam types I have (the expensive type!) read flat from a few Hz (where the needle flaps about) to about 100KHz on sinewave.
Using a pair of opamps as simple buffers is simpler but consideration should bre made for the possibility of overdriving the inputs so there should at least be some series resistance, again say 100K to limit current if the buffer is unpowered while signal is present.
Matt Syson
bcarso
Well-known member
[quote author="Brian Roth"]Well, no one shot any holes in my balanced input gizmo, so I Shall Carry On <g>.
Bri[/quote]
I like that config for being bulletproof and having no common-mode distortion. At very high frequencies there's a little degradation in CMR although this can be compensated for and another few octaves pulled out.
Noise is not optimal but who cares most of the time, and especially in this application.
Bri[/quote]
I like that config for being bulletproof and having no common-mode distortion. At very high frequencies there's a little degradation in CMR although this can be compensated for and another few octaves pulled out.
Noise is not optimal but who cares most of the time, and especially in this application.
Brian Roth
Well-known member
Matt, I figured that 10K (unbalanced connection) and 20K (balanced) would be plenty high for pro gear, but your point is well taken. I kept to the lower values to minimize any effects from stray capacitances.
Bri
Bri
Matt Syson
Well-known member
Hi Brian
I am sorry but to load it by this much, nearly 0.25 dB if the source is 600 ohm is NOT acceptable and any respectable audiophile would have to kill themselves. You know perfctly well that the performance HAS to be DC - light response with NO distortion or any other sources of error.
Seriously, I am not sure where the original poster of this thread was intending to use the VU meter but all that has been discussed so far will work just fine but I would err on the side of lighter loading as any imbalance of the resistors will affect the common mode balance less and if you are using capacitors on the input (which is probably a good idea) then they can be smaller and non electrolytic.
Happy days!
Matt Syson
I am sorry but to load it by this much, nearly 0.25 dB if the source is 600 ohm is NOT acceptable and any respectable audiophile would have to kill themselves. You know perfctly well that the performance HAS to be DC - light response with NO distortion or any other sources of error.
Seriously, I am not sure where the original poster of this thread was intending to use the VU meter but all that has been discussed so far will work just fine but I would err on the side of lighter loading as any imbalance of the resistors will affect the common mode balance less and if you are using capacitors on the input (which is probably a good idea) then they can be smaller and non electrolytic.
Happy days!
Matt Syson
enthalpystudios
Well-known member
If I wanted to drive 2 meters for a single mono source, would I need two of these, or could I run 2 meters in parallel from the output of a single buffer?
I'm trying to make a buffer for a mic amp that will drive both a peak and a vu.
oh yeah, could I use the 48v with a voltage divider to run 32v single ended? Or maybe is there a similar chip that will run on 24v single ended? Its for a 1272, so I've got 24/0/48 in there.
thanks for the help and designs, dave and brian both!
billy
I'm trying to make a buffer for a mic amp that will drive both a peak and a vu.
oh yeah, could I use the 48v with a voltage divider to run 32v single ended? Or maybe is there a similar chip that will run on 24v single ended? Its for a 1272, so I've got 24/0/48 in there.
thanks for the help and designs, dave and brian both!
billy
