How do I get these old analog VU meters working? #2

Thread Starter

Hunteru

Joined Feb 7, 2021
11
Back in another life I helped built a TV station and designed/built/rebuilt several remote trucks.

A VU meter has a defined set of ballistic parameters. That is, the amount of needle overshoot that happens at a sharp transient increase in volume is standardized so all VU meters show the same effective volume level. The ballistics and damping are calibrated when the meter coil is loaded with a specific non-zero impedance. My guess is that it is 3.6K ohms for this meter, so you want the coil to see 3.6K ohms as the source impedance no matter what the input signal level is. A common way to do this is a variable gain opamp stage followed by a 3.6K resistor in series with the meter.

Alternatively, 3.6K might be the impedance of the meter when it is applied to an external circuit. However, the 3.6K loading impedance rings a faint bell...

I don't see any diodes on the on board, so the meter must have it built in. This means it is a true VU meter, not just a meter with a VU scale.

1.734 V = 1.228 V (0 dBu) x 1.414 (sqrt 2). It is the peak value of an input sine wave that will produce an indication of 0 dB.

https://en.wikipedia.org/wiki/VU_meter

ak

Hi Ak ,

I currently have 'Vu meters' that can be found everywhere on eBay like the Nissei TN-73.

I don't see any diodes or 3.6k resistors inside the VU, the meter doesn't have it built-in. this means they are not a true VU meter, just meters with a VU scale like you said.
And I wonder how can we turn them into real VUmeter with its true law.

Do you have any idea how to do this step?

Thank you very much

Moderator edit: New thread created from here.
 

Thread Starter

Hunteru

Joined Feb 7, 2021
11
You can build a precision rectifier using opamps and no diodes.
Do you have a link or a circuit to share to achieve this? My knowledge is limited, but I know how to combine this easily.

Also, if possible I would like to get closer to the values commonly used on the Sifam al-29 which are real VU meter. I know there is a 3.6k resistor on the hot pole but is there something else with it internally? A germanium diode or something?
 
Last edited:

Thread Starter

Hunteru

Joined Feb 7, 2021
11
Link to the actual meter you have?

Also, what is your skill set for designing / building a small analog circuit?

ak
Hi Ak,

They are similar to these ones:
https://www.ebay.com/itm/2PCS-VU-Panel-Meter-TN-73-High-Precision-VU-Meters-Header-DB-Meter-Level-Audio/383765675810?_trkparms=aid=1110006&algo=HOMESPLICE.SIM&ao=1&asc=20201210111314&meid=4dac5393fa504c9aac9d1b1335754cca&pid=101195&rk=3&rkt=12&mehot=lo&sd=254503943726&itm=383765675810&pmt=1&noa=0&pg=2047675&algv=SimplAMLv5PairwiseWebWithDarwoV3BBEV2b&brand=Unbranded&_trksid=p2047675.c101195.m1851

I measured the external impedance and I had 55ohm as a result.
With a AA battery and a 10k pot, I had:
1.180 V - 8kohm at "0"
1.325 V - 9k ohm at "+3"

My ability to design is limited but I have good soldering skills. I can light something as long as I have the correct component in front of me.
 

Attachments

Thread Starter

Hunteru

Joined Feb 7, 2021
11
I just received this from an old friend and he can certainly build it for me. It's an old circuit driver from Neve. Do you think it's worth the shot?

NEVE B-669s0.jpgNEVE B-669s1.jpg
 

AnalogKid

Joined Aug 1, 2013
10,986
There are two adjustments, but neither one affects damping. Also, there is no rectification. My guess is that this circuit assumes a true VU meter with rectification and proper ballistics built in.

For your use, are you working balanced or unbalanced audio sources? The left amp is a low-gain balanced to single-ended converter, so you might be able to eliminate it.

For your meter, you might want to make the 2.2K output resistor a 5K pot. As the resistance increases, the damping (overshoot) increases. But also, the signal indication decreases because the output resistor and the meter impedance form a two-resistor attenuator. This is why you need a gain stage.

On second thought, the 20K pot might affect damping. It changes the gain of the stage, but it also change the low frequency roll-off. Hmmm...

ak
 
There are many terrible VU meter driver circuits out there.
First you have to know the meter, there's four possibilities:
internal bridge rectifier, no rectifiers, resistor built-in, or none. Just use an ohmeter with both polarity to see what you've got.
There seems to be many different TN-73's out there, with no datasheets. I've seen specs for it:
50ohm 1mA FSD
600ohm 1mA FSD
500ohm 0.65mA FSD (TN-73-HS-9213)
0dB 0.8VAC, or w/ext 3.6k resistor 1.228V for 0dB (TN-73-HS-S0337).

I think for you to read 55R ohms, it's just the coil. But 1.325V in to read "+3VU" fullscale ? 1.325/(55+9k) = 0.146mA FSD so still seems odd. 1.118V/10k is 80% down to 0dB.

That Neve-copy circuit, I don't see a rectifier there. Most people want single-supply operation which makes the circuit a bit harder to get a good quality signal driving the meter.
 

Thread Starter

Hunteru

Joined Feb 7, 2021
11
There are two adjustments, but neither one affects damping. Also, there is no rectification. My guess is that this circuit assumes a true VU meter with rectification and proper ballistics built in.

For your use, are you working balanced or unbalanced audio sources? The left amp is a low-gain balanced to single-ended converter, so you might be able to eliminate it.

For your meter, you might want to make the 2.2K output resistor a 5K pot. As the resistance increases, the damping (overshoot) increases. But also, the signal indication decreases because the output resistor and the meter impedance form a two-resistor attenuator. This is why you need a gain stage.

On second thought, the 20K pot might affect damping. It changes the gain of the stage, but it also change the low frequency roll-off. Hmmm...

ak
Yup , let's just forget the Neve. I guess this is not suited for this particular case.

Thanks for your input though. the entire chain is Balanced.

From what I understand now. Is that I need a rectifier circuit and a driver. If that's it, I found a circuit that could go :here: (from Karl Adams Youtube Channel)

Schematic here: DropBox
 

Thread Starter

Hunteru

Joined Feb 7, 2021
11
There are many terrible VU meter driver circuits out there.
First you have to know the meter, there's four possibilities:
internal bridge rectifier, no rectifiers, resistor built-in, or none. Just use an ohmeter with both polarity to see what you've got.
There seems to be many different TN-73's out there, with no datasheets. I've seen specs for it:
50ohm 1mA FSD
600ohm 1mA FSD
500ohm 0.65mA FSD (TN-73-HS-9213)
0dB 0.8VAC, or w/ext 3.6k resistor 1.228V for 0dB (TN-73-HS-S0337).

I think for you to read 55R ohms, it's just the coil. But 1.325V in to read "+3VU" fullscale ? 1.325/(55+9k) = 0.146mA FSD so still seems odd. 1.118V/10k is 80% down to 0dB.

That Neve-copy circuit, I don't see a rectifier there. Most people want single-supply operation which makes the circuit a bit harder to get a good quality signal driving the meter.
Hi,

I guess I have a pair of 50ohm 1mA FSD. There is no internal bridge rectifier, no resistor built-in, and this is what I would love to achieve and easily wire them to my chain.

I do not exclude that I may have been wrong on the statement. On the other hand, I am sure that it is indeed a 50ohm when I measure from the terminals.

A bit like what is in the Sifam AL29WF or other good copy Input:
The VU meter reads 0db at 1.228V +/- 10% trough an external 3.6k resistor
Internal Resistance: 3.9K ohms +/- 10% Insulation
Resistance:> 50MO
Model: DA-WF1029X

Cheers
 

AnalogKid

Joined Aug 1, 2013
10,986
Neve makes very high-end recording studio audio consoles. As such, their meters are almost certainly true VU types, with the rectifier and part of the back-impedance built in. This is consistent with their meter driver circuit, which is why it does not perform rectification and has a build-out resistance.

I was going to dig up Howard and ask him, but I can't find my shovel.

Before he literally wrote the book on broadcast engineering, Howard Chin wrote a paper describing his invention - the VU meter. Warning - it's a bit deeper than you think.

https://www.aes.org/aeshc/pdf/chinn_a-new-svi.pdf

For this thread, the fun starts on page 16. Note that a true VU meter has defined ballistics. That is, it overshoots intentionally when a tone burst is applied at full volume. Howard had no formal training in psychoacoustics, but he was a pretty smart puppy.

A "correct" meter driver circuit to make a generic meter perform as a VU meter has an optional balanced-to single-ended circuit front end, followed by a precision full-wave rectifier circuit and a gain stage driving an adjustable output impedance. The idea here is that you test the meter ballistics with a tone burst generator, going back and forth between adjusting the output resistor to get the right ballistics, and adjusting the gain to get a correct reading for a steady-state tone. It is an iterative process, because adjust the output resistance affects both the transient and steady-state indications.

This might be more than you want to mess with, in which case one dual opamp per meter will work - 1/2 for the rectifier and 1/2 for a driver, ignoring the output impedance and ballistics.

ak
 

Thread Starter

Hunteru

Joined Feb 7, 2021
11
Neve makes very high-end recording studio audio consoles. As such, their meters are almost certainly true VU types, with the rectifier and part of the back-impedance built in. This is consistent with their meter driver circuit, which is why it does not perform rectification and has a build-out resistance.

I was going to dig up Howard and ask him, but I can't find my shovel.

Before he literally wrote the book on broadcast engineering, Howard Chin wrote a paper describing his invention - the VU meter. Warning - it's a bit deeper than you think.

https://www.aes.org/aeshc/pdf/chinn_a-new-svi.pdf

For this thread, the fun starts on page 16. Note that a true VU meter has defined ballistics. That is, it overshoots intentionally when a tone burst is applied at full volume. Howard had no formal training in psychoacoustics, but he was a pretty smart puppy.

A "correct" meter driver circuit to make a generic meter perform as a VU meter has an optional balanced-to single-ended circuit front end, followed by a precision full-wave rectifier circuit and a gain stage driving an adjustable output impedance. The idea here is that you test the meter ballistics with a tone burst generator, going back and forth between adjusting the output resistor to get the right ballistics, and adjusting the gain to get a correct reading for a steady-state tone. It is an iterative process, because adjust the output resistance affects both the transient and steady-state indications.

This might be more than you want to mess with, in which case one dual opamp per meter will work - 1/2 for the rectifier and 1/2 for a driver, ignoring the output impedance and ballistics.

ak

Yes, Rupert Neve is the G.O.A.T. 90+ years old and still at work. The VUs on their consoles at the time like the 8058 were true VUs with the rectifier built-in. I read from another legend that worked at Neve (when Mr. Rupert was still the owner) they used to put just a 3.6K ohm on the hot phase.(link)

The rabbit hole was amazing to explore. I liked it. Yes, Howard Chin knows a lot, very nice. respect.

Page 16 is gold, all the requirements are exactly what we need. 300ms of dynamics. The sensitivity of 1.228v at 0-Vu and impedance 3k9 in the meter with 3k6 outside the meter.

I guess the only piece missing is an instruction to a) connect a diode and resistor inside the 1mA FSD 50ohm meter OR b) Use a dual opamp - 1/2 for the rectifier and 1/2 for a driver.

I'd rather grab the a) :).
 
Top