Old Siemens meter?

Well, a friend of mine came to my workshop with a pair of those:




They seem to be in good condition, they light up, and the mirror moves when applying current.

I have looked for infos, and so far found the manual on Kubarth's site, look for: NWDR C-B-J47B
at: http://audio.kubarth.com/rundfunk/index.cgi

I don't speak German, but I guess that those need an external amplifier named B-U 21.
I couldn't find anything about those B-U 21...

EDIT: Following Holger's suggestion, I've googled a bit and it appears that the scale -50/+5dB is typical from the German DIN PPM standard.

Anybody could point me in the right direction?

Axel

and you wanna let us know or need to know something ??

Sorry, I'm still trying to edit my first post!
And yes I would like to know something...

Looks like an old Siemens meter with the DIN scale.
Edited.
-50=3.450mA
-10=2.102mA
0=1.445mA
+5=1.033mA

Hi Holger,
makes sense wrt the age of the thing, though I guess that the scale could be adjusted with the dedicated amp...

It sure doesn't act like a regular VUmeter, as it needs more current to show -40dB than to show 0dB!

mad.ax, you're right. Edited my post. Do you need special information? Maybe I can help you.

Well, I'd like to have a look at that BU21 schematic of course, in order to figure out a way to make these puppies running again.
I guess there are some tubes and transformers involved in the original unit, but it should be possible to build something with opamps.

Perhaps somebody has allready designed a modern ckt for that purpose?

> it needs more current to show -40dB than to show 0dB!

Odd, yet logical.

Build an AM radio with AVC (Auomatic Volume Control). The AM detector DC voltage is filtered and fed back to IF stage grids to reduce gain on strong signals. The IF stage may run 10mA at no-signal and 1mA for strong-signal. The relationship is quasi-exponential.

Replace radio IF transformers with audio couplings. Fiddle the current. The "IF stage" current is the correct function for this meter.

If you know the original tube.

Sure would be easier if the schematic were available.

Your unit is jus a DC moving coil meter. So, you need a logarithmic meter amplifier to use it as a PPM. I have one made by Disa but there were many other (mostly German) manufacturers too. They should all be interchangeable.

Of cource you can do it in more modern way too (with op-amps). A good starting point is the peak programme meter circuit of the Studer 069 (or almost any other) console.

> PRR,
I didn't tought of this "radio" approach. Thanks! If I understand right, this would deal with the quasi log conversion.

I have seen a much more modern English PPM meter which use a conventional needle. When not powered, the needle is on the left(-40dB), when powered it goes full scale on the right (+6dB), and when applying signal, it behave the normal way...  So my guess is : Since we need a cap to deal with the time constants of the PPM specs, it's easier to quickly short this cap than to charge it. With an inverted moving coil, shorting gives us fast attack time, and charging the cap a slow delay...
Could I be on the right track?

>Audiox
I guess that the B-U 21 is indeed a logaritmic amplifier. You don't have the schemo for your Disa unit, don't you? Or maybe you could have a look inside it and tell me how much complex it looks like?
I'm not sure they are all interchangeable though, unless they have the same name (B-U 21), because looking at Kubarth's site, I have seen a previous version of this meter, the BJ 45 A (mine is a BJ 47 B), and the manual shows a moving coil drawing 7.55mA instead of 3.45...

Thanks for the suggestion! There's not many good PPM meter schematics around, I will google for the Studer...

Axel

So, I downloaded the 069 service manual from the Studer website :
ftp://ftp.studer.ch/Public/Products/Mixing_Analog/069/Manuals/

The PPM description starts page 101, and the schemo is page 141...

Obviously this one use another different type of DC moving coil meter! From the Studer manual:
"The mechanical zero position of the movement corresponds to the reference indication 0dB. For inputs levels which gives a deflection between 0... +6dB on the meter, the output of amplifier 4 changes the polarity of the voltage."

Axel

Those old German logaritmic meter amplifiers are built using valves. So it makes no sense to clone them. If you want those valve units to be accurate they need calibration almost every month. It is so much easier to do it with a few op-amps.

You can change the reference point of the Studer circuit to make it work with your meter.

Totaly agreed! My interest for the original log amp is mainly curiosity...

I will probably build something with opamps, and while going this "easier" route, I will try to make it run with a single supply in order to use the 12V power of the internal bulb...

Axel

I found something to feed my curiosity... Looks like the B-U 21 has been replaced by U70 and U71 and those are well documented on Kubi's site.
4 tubes, input and output transformers, a 60mH coil, and even an integrated meter for calibration!! Makes opamps definitely tempting...

There is also the TAB U370a, a transistorized version which runs on 24V DC, but needs 3 transformer plus the coil, and it looks like the pseudo log conversion is made by a bunch of diodes and adjustables...

So far, the Studer 069 seems to be one of the best way to make a simple but accurate PPM.

But I'm also very interested in PRR's favorite B192

Axel

> much more modern English PPM meter which use a conventional needle. When not powered, the needle is on the left(-40dB)

Yes, there are systems like that. The meter is conventional.

When yours is idle, does the needle sit at "100%/0dB", or at "180%/+5dB"?

If it sits at FULL scale, my "AM radio AVC" plan can work. If it sits at some nominal level less than MAX, it won't.

> Since we need a cap to deal with the time constants of the PPM specs, it's easier to quickly short this cap than to charge it. With an inverted moving coil, shorting gives us fast attack time, and charging the cap a slow delay...
Could I be on the right track?

No. The rectifier is conventional. And it is not necessarily "easier to quickly short this cap" unless you only use a screw-driver. If you use a tube/transistor, you can as-easily charge to a voltage.

> "The mechanical zero position of the movement corresponds to the reference indication 0dB.

If your meter is like this, you can use this plan, with maybe some resistor mods. The classic/defunct 4136 chip can be TL074 if you adjust the pinout.

This scheme logs the signal in a diode D7. The diode voltage is temperature sensitive. They balance this with another diode D8 working at fixed reference current. The correction is perfect only when input signal equals reference signal. They set the reference to "0dB", so that it will be spot-on at lineup level, very-close at normal program levels. The temp-error falls to the -50dB end of the scale, where nobody is too fussy about "-50dB" really being -52dB or -48dB.

> "....the output of amplifier 4 changes the polarity of the voltage."

That's said un-clearly. This is a Log-Difference scheme. See Walt Jung's books. The output of the log-diff is zero when input current equals reference current, and the polarity indicates above/below reference; amp 4 is not doing anything tricky about the polarity.

> I will try to make it run with a single supply in order to use the 12V power

That is possible; Neve has one. However it will be far more dificult. The 2-diode log-difference scheme works, well, elegantly, but has to be butchered-up to work on a unipolar supply.

What is the no-power idle indication? What current causes it to read how much of the scale? What is the resistance of the meter movement? Be VERY careful: meters are easily blown or bent.

Thanks for your answer PRR.

When idle, the Siemens "needle" sit out of the range.

>"And it is not necessarily "easier to quickly short this cap" unless you only use a screw-driver"
Sure! I know that I'm much slower than the laziest electron
By easier I meant using less power. My reasoning was: This movement is not a 50µA, but more something like 4mA (and the first version BJ45A was around 8mA), as we want a slow decay time in this application, we need a big cap to keep the current for 1.5 sec. It would take power to charge this big cap fast enough to get a few miliseconds rise time...
Bah, you are right, it can be done... I was just overthinking and confusing a PPM driver with a 660 sidechain!

>"What is the no-power idle indication?"

Out of range, but I will check wich side...

>" What current causes it to read how much of the scale?"

From the BJ47 manual:
-50=3.450mA
-10=2.102mA
  0=1.445mA
+5=1.033mA

>" What is the resistance of the meter movement?"

I have to check this, very carrefully
Unfortunately, I have no internet in my workshop, so it will take a few hours before I post the results.

Axel

> From the BJ47 manual:
> -50=3.450mA
> -10=2.102mA
>    0=1.445mA
> +5=1.033mA

This is annoying. I'm sure there is some way these values make perfect sense, and I'm inclined to think something like the attached plan, but reverse engineering it is just too much work.

There's indeed a lot of similarities between your plan and the original one!

The meter is connected between pins 3a & 3b, so between +150 and the ECC81 plate which forms an output buffer.
Your plan is centered around the EBF89 with the transformer in the spliter's role...

Axel

So I made some measurements on the BJ47B...

With no current applied, the "needle" sit on the right side.
Any current inferior to 1mA and it stay out of range (at the right of +5dB).
1ma gives a +5dB reading
2mA = -10dB
...and so on according to the manual, seems to work perfectly.

DC resistance of the coil is 90 Ohms

Axel

If it's just about getting the meters to work, you could try to get hold of some transistorized versions of the meter amps. The naming of the german broadcast technology has a system, the first character shows the type of amplifier, with the tube stuff it is followed by a 2 digit number which specifies the device. Later transistorized versions add a third digit in front of that number, which indicates the series / manufacturer. So if the tube version is a U70, you're looking for a U270, U370 etc. I don't know when this meter technology was abandoned, so maybe there is no later version.

You could try ebay or maybe Schmittronic in Cologne, these amps are not sought after and might go for little money.

Michael