Ok,
So I have found some LM3915's so i'm gonna get a go at using them.
The plan is building a VU/PPM meter using switching Between Bar/Dot.
So the principle is using 2 precision rectifiers with the correct timing to get VU (average) and PPM (peak hold) ballistics.
If I was using one LM3915 with a range of 30db and a maximum of around 24dbu this wouldn't be a big problem (-6db as lowest led is easy altough I would need attenuation as 24dbu is to large for feeding a precision rectifier with opamps)
Now The plan is on using a range of 90db from -66dbu to +24dbu and in theory it should be perfectly possible but I am a little worried for the precision rectifiers.
-66dbu is around 0.5mv peak and +24dbu is around 17,36v peak so to make sure the rectifiers don't clip I would have to attenuate the signal. Let say we take the signal down to 20dbu maximum (around 11v peak) leaving a little margin.
This means the lowest signal will be around 0.34mv.
I looked a little in the datasheet and they mention this as a problem for the lower leds as there could be 10mv offsets so it's better to keep the same reference and amplifiy the signal for the lower leds.
This would mean the first lm3915 get's it signal directly fro the recitifier then the second get's this signal but with a gain of 30db and the last get's this signal with a gain of 60db.
So far this looks possible altough I would need precision opamps with very low DC offset (ad707 or something like that) and for the lowest leds it would maybe need a zero trim.
Only problem remains the precision rectifier. Is it possible to rectify 0.34mv peak?
Using an amplifier before the rectifiers could be an idea but it also creates a problem. this would mean that each lm3915 has it's own rectifier and if using the PPM's slowly decay it would get weird with short transient signals.
The 90db range is mentioned in the datasheet but no schematics or info is giving. Google also gives a lot of cascading to 60db range but 90db is nothing I could find.
So I have found some LM3915's so i'm gonna get a go at using them.
The plan is building a VU/PPM meter using switching Between Bar/Dot.
So the principle is using 2 precision rectifiers with the correct timing to get VU (average) and PPM (peak hold) ballistics.
If I was using one LM3915 with a range of 30db and a maximum of around 24dbu this wouldn't be a big problem (-6db as lowest led is easy altough I would need attenuation as 24dbu is to large for feeding a precision rectifier with opamps)
Now The plan is on using a range of 90db from -66dbu to +24dbu and in theory it should be perfectly possible but I am a little worried for the precision rectifiers.
-66dbu is around 0.5mv peak and +24dbu is around 17,36v peak so to make sure the rectifiers don't clip I would have to attenuate the signal. Let say we take the signal down to 20dbu maximum (around 11v peak) leaving a little margin.
This means the lowest signal will be around 0.34mv.
I looked a little in the datasheet and they mention this as a problem for the lower leds as there could be 10mv offsets so it's better to keep the same reference and amplifiy the signal for the lower leds.
This would mean the first lm3915 get's it signal directly fro the recitifier then the second get's this signal but with a gain of 30db and the last get's this signal with a gain of 60db.
So far this looks possible altough I would need precision opamps with very low DC offset (ad707 or something like that) and for the lowest leds it would maybe need a zero trim.
Only problem remains the precision rectifier. Is it possible to rectify 0.34mv peak?
Using an amplifier before the rectifiers could be an idea but it also creates a problem. this would mean that each lm3915 has it's own rectifier and if using the PPM's slowly decay it would get weird with short transient signals.
The 90db range is mentioned in the datasheet but no schematics or info is giving. Google also gives a lot of cascading to 60db range but 90db is nothing I could find.
I think you also need a fairly fast op-amp. At the highest frequency of interest, it should still have enough open-loop gain to amplify that 0.34 mV to a voltage level high enough to turn on the diodes. It could get problematic if it overshoots: 1 % overshoot of 600 mV is already more than that 0.34 mV.
(By the way, see https://www.diyaudio.com/community/threads/log-conversion-for-vu-ppm-meter.410833/post-7752763 and https://www.diyaudio.com/community/threads/log-conversion-for-vu-ppm-meter.410833/post-7789886 for a different way to do the log conversion.)
(By the way, see https://www.diyaudio.com/community/threads/log-conversion-for-vu-ppm-meter.410833/post-7752763 and https://www.diyaudio.com/community/threads/log-conversion-for-vu-ppm-meter.410833/post-7789886 for a different way to do the log conversion.)
I used MCP 6294 quad opamps (datasheet 1) and BAT46 diodes (datasheet 2) when I built precision rectifiers ahead of VU meters. I chose to run from a single ended +5V supply, figuring that could be dropped into many different amplifiers. Those opamps are RRIO and 10 MHz bandwidth.
