IN13 Level Meter Linearity

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

thermionic

Well-known member
Joined
Jun 3, 2004
Messages
1,671
Hi,

I've built a two-channel (for stereo use) version of the IN13 40dB VU meter here: http://m.bareille.free.fr/vu-in13/vuin13_v21_schem.pdf

The issue I'm having is that the channel linearity doesn't match.

Let's say ch1 is @ -10, ch2 might be @ -8. But then, as you boost level, ch1 will out-accelerate ch2, so it now reads 2dB more than ch1 when @ +10...

Can anyone see an aspect to the schematic that might affect linearity in such a manner? It's easy to calibrate both IN13s to match at, say, 0dB - but they will not align together at lower or higher levels.

My initial thought was that the IN13 tubes needed 'matching' - but I now know the linearity was is due to the driver circuitry as I've swapped tubes between PCBs.

Which components do you think could explain the linearity issue?

Thanks in advance.
 
Assuming L/R same input current and frequency, parts tolerances setting Ibias and It at the 2252 could be the reason for different 2252-pin7 voltages.
R7 same value ?, R6/C2 same value ? C2 leaky?
A 50K rheostat in series to R7 and/or a 1M rheostat in series to R6 might be enough to catch integration time variation so both output voltages 2252-pin7 match over a wider level range.
Just a thought.
 
thermionic said:
Hi,

I've built a two-channel (for stereo use) version of the IN13 40dB VU meter here: http://m.bareille.free.fr/vu-in13/vuin13_v21_schem.pdf

The issue I'm having is that the channel linearity doesn't match.

Let's say ch1 is @ -10, ch2 might be @ -8. But then, as you boost level, ch1 will out-accelerate ch2, so it now reads 2dB more than ch1 when @ +10...

Can anyone see an aspect to the schematic that might affect linearity in such a manner? It's easy to calibrate both IN13s to match at, say, 0dB - but they will not align together at lower or higher levels.

My initial thought was that the IN13 tubes needed 'matching' - but I now know the linearity was is due to the driver circuitry as I've swapped tubes between PCBs.

Which components do you think could explain the linearity issue?

Thanks in advance.
The first thing is to determine the actual operating level of the 2252. This is governed by the input level, the gain of the second stage (adjustable) and the input resistor. Typically, with 30k input resistor, the 2252 is good for up to +20dBu.
I would recommend, for troubleshooting sake that you crank down the gain trimmers so the chain is unity-gain.

Now, there are a couple of things that you should check.
First the timing cap (4.7uF at pin 6) must be absolutely without leakage (use at least a 63V).
Then the resistor(s) between Pin 7 and Pin 1; their role is to avoid instability of the output voltage when the audio level is very small. But they impair linearity.

Here's my suggestion:
Disconnect these resistors and check the voltage at pin 7 without signal.
Try to find two 2252's with similar voltage.
Check linearity.
Reconnect resistors.
You may not find matched pairs of 2252's (particularly if you don't have a batch of them).  ;)

Alternatively, you may want to trim each detector with the attached mod.
Adjust trimmer for lowest voltage at Pin 7 with no signal.

 

Attachments

  • 2252 bias comp.jpg
    2252 bias comp.jpg
    48.8 KB · Views: 5
Without looking closely "what abbey said"..  Cheap log convertors are optimized to interface with VCAs so scale and offset error (not linearity) can creep in.

JR
 
Thanks. My first thought was to keep the i/p buffer at Unity. I've tried a few 2252s and it seems I'm unlikely to find a perfectly matched pair (although experimenting with them in sockets has gotten me better results since I last posted), so I may just implement Abbey's trimmer circuit.
 
Back
Top