CE Amp to Measure Voltage

[NotaBene]

Hi, I'm a bit of an electronics noob, be gentle

Trying to make a circuit to measure the signal level (voltage).  Ideally this would be boosted/cut to 0-5V.  I've been studying simple common emitter amps and even virtually tested some.

They work, but the voltage seems to be all over in my actual effects chain.  From the guitar it's < 100mV, and my preamp out has a huge range depending on where the volume is set.  I haven't tested elsewhere in the chain, but cursory research indicates there IS no accepted level.  Up to 9V, I guess?

I'd like to reliably have some kind of amp that doesn't mess with the normal signal but boosts or cuts to 0-5V to make measuring easier.  Is this possible?

If so, ideally it would be totally transparent and not mess with the actual base tone at all.

[R.G.]

Your results are what I would expect. The signal levels do vary a great deal. The level of the guitar's signal varies depending on the guitar volume setting, of course, but the maximum output from a guitar may be as little as 10-20mV from older low output pickups to as much as a volt or two with specially designed high output humbucker types.

And pedals do amplify the guitar signal by varying amounts, and the pedal volume knobs do make it vary some more.

I'm a little confused about that you're trying to measure. Can you tell us a little more? Is it just how big your signal is compared to the guitar, or something else?

[NotaBene]

Thanks fer the help.  I'm making a signal meter to display the relative strength of a signal using an LM3914.

I can adjust the ranges with a pot, not a problem, but since the signal is going to vary so much I need a way to make the maximum around 1.5 to 5V to accommodate the IC.  It's looking like this is some kind of amplifier?

[PRR]

I hope you can work in dBs. The range of sounds is way too large to express in linear numbers.

Live sounds vary from under 20dB SPL to over 120dB SPL, a 100 dB range. (Why CDs were designed for 96dB, which happens to be a round number in binary, 2^16 or 65,536:1.)

LM3914 covers a 10:1 range, 20dB. This is kinda-OK for reading the maximum peaks of audio signals.

LM3915 covers about 30:1 range, 30dB. And smaller steps at the bottom where you want them. It can be cascaded for 60dB range; this is in the datasheet.

To actually measure "any" audio voltage, from the softest whisp to the Marshall Stack, you need more than 60dB. Fortunately we almost never need that much.

$40 SPL meters cover 76dB range by using a 7-position switch, manually dialed up/down. Fancier ones (some at lower price) auto-range.

For the specific purpose of guitar-cord measurements: a soft strum may start 20mV and fade to infinity, which is more like 0.01mV (where we run into universal hiss), and we often do not need to know how far down it goes. The upper limit is surely 300mV or more. Some reports of well over 1V from hot pickups, and a 9V pedal can be capable of 3V RMS output; hot-rods more.

Taking 20mV-200mV as normal, and another 200mV-2V as "hot", gives two ranges which could be displayed on '3914 (however use '3915).

The '391* needs a rectifier, and also pre-gain to get into the 1.2V range. (It can be worked lower but errors accumulate.) Basic rectifier plans are in the datasheet. Pre-gain is just a "booster".

[NotaBene]

Thanks!  I actually understood most of that!  Not really because I'm smart, because you wrote it at my level

The range thing is problematic because I kind of wanted the meter not directly after the guitar, but maybe somewhere in the signal chain of pedals... like in my normal FX loop of my amp.  Would your second range, the 200mV - 2V, be good for that?

Better yet, I could make the low AND high ranges adjustable with pots, to compensate for placement and wacky guitars/pedals?

Low: 20mV - 1.5V
High: 200mV - 3V

I'm not sure what you mean by adding a rectifier and/or pre-gain, other than what I've already been doing for the basic layout of the IC.

One more weird thing.  I was getting weird flickering of the LEDs until I followed the left side of this circuit I found somewhere.  Diode, 3.3uF cap, and 120k resistor.  Then everything worked fine!  My guess is that the electrolytic cap is removing half the AC waveform?  Not sure what the other stuff is for, but it worked somehow.

[antonis]

Diode, 3.3uF cap, and 120k resistor.  Then everything worked fine!  My guess is that the electrolytic cap is removing half the AC waveform?

Diode is half-wave rectifying input signal..!!
(it only lets more positive than 0V + diode forward voltage drop signal curves to pass through it..)

3.3μF cap is there to "recover" the voltage/current needed during the negative signal cycle..
(when signal voltage is less than about 600mV, for 1N4148..)

[NotaBene]

Diodes do that?!  Sweet, this stuff is so exciting

[PRR]

Diode is your rectifier. BUT! A simple diode is deaf below about 0.5V 500mV.

The diode and cap is a peak catcher. Audio is very dynamic. Mostly we want to know what the peaks are. The cap holds the peak long enough for the eye to see.

Putting the plain diode-cap directly across a guitar or many pedals will distort the signal badly. (It is a heavy load to charge-up a cap.)

Google "lm3915 application notes". I did not read all the pages found. Some are from lightweight sites hardly worth reading. You would like to find the *original* app-not from the 1980s, as well as browse the modern writings for nuggets.

This extended data-sheet]http://kontel.hu/shop_ordered/67768/pic/pdf/LM3915N-1.pdf]This extended data-sheet in Fig. 1 on page 10 shows an improved rectifier which will work below 0.1V and with less loading of the source.

"Precision Rectifiers" are a thing, and you can get lost forever trading-off cost and performance.

In general I would be thinking the semi-precision rectifier of Fig 1 with a booster in front.

FWIW: this is how Neve (giant broadcast/recording consoles) did wide-range metering in days before chips or LEDs. There are several very subtle elegant tricks in here. The first two stages are a booster. The rectifier is full-wave(!). The output is linear in dB over about 50dB. The meter is a standard linear meter except the face is black (BBC PPM standard).

[NotaBene]

Bleh, I replied earlier but it got erased somehow

- I did notice that changing the 3.3uF cap value changed the "hysteresis", I think I'm using that word right?

- I read the "improved" spec sheet, that's very helpful.  I think I sort of get the need for the Fig.1 peak detector.  But it seems like the need for that AND the need for another boost/amp circuit depends a lot on where the thing is in the chain.  If it's right after the guitar, it might need a booster + half wave peak detector for the <50mV signals.  But if it's in the chain, it might not need anything special - 0-4V adjustable at pins 4 and 6 seem like enough.  Maybe... a switch for the boost/buffer?

- If there is a boost of some kind right before the peak detector, which kind should I make?  I've researched a few, op-amp and transistor and JFET and MOSFET.  I tried a simple transistor one today, but it didn't work well.

- I'm sort of starting to grasp why I need this stuff, but still a bit clueless on how it hooks up to the ins and outs.  I think i can wire it true bypass with a DPDT switch, but is that enough?  I'm assuming the meter + extras goes in series with the "main" input signal?

- That schematic is a bit above my pay grade, but I would guess they are using zeners for comparators instead of the IC.