My daughter has been getting into vinyl recently, so I've decided to get her a decent-ish turntable (Fluance RT82) and pair of Edifier S2000MKIII speakers. Right now she's destroying her vinyl on a cheap plastic piece of junk turntable from Amazon and migraine inducing soundbar.
The turntable will inconveniently need to be placed on the opposite side of the room as the speakers, so I plan on pairing a WM8782 ADC to an AptX HD I2S input BT transmitter (24bit 48kHz). Not ideal, but it will be infinitely better than what she has now.
Vinyl and turntables are somewhat foreign to me, but I guess the recording levels can be all over the place? I'm kind of lost on how much headroom I need for the ADC to optimize for noise and potential clipping.
The stylus is an Ortofon OM 10 (4mV @ 1kHz 5cm/s) and the preamp I purchased has a gain of 40dB. Right now the WM8782 board is set for 2Vrms full scale, but it can easily be changed. I don't have a clue if that's too little or too much.
The turntable will inconveniently need to be placed on the opposite side of the room as the speakers, so I plan on pairing a WM8782 ADC to an AptX HD I2S input BT transmitter (24bit 48kHz). Not ideal, but it will be infinitely better than what she has now.
Vinyl and turntables are somewhat foreign to me, but I guess the recording levels can be all over the place? I'm kind of lost on how much headroom I need for the ADC to optimize for noise and potential clipping.
The stylus is an Ortofon OM 10 (4mV @ 1kHz 5cm/s) and the preamp I purchased has a gain of 40dB. Right now the WM8782 board is set for 2Vrms full scale, but it can easily be changed. I don't have a clue if that's too little or too much.
When I designed a phono EQ amplifier a few years ago, I considered how much gain should be set to even out the difference with the digital source. Generally, the standard is 40dB (@1kHz) for MM 5mV. My DAC is 2.5Vrms(0dBFS), MM cartridge is 4mV(5cm/sec) and started with a setting of 43dB. However, this was insufficient, so we decided to increase it by 4dB to 47dB. Of course, this is the average (perceptual) of the vinyl I own, so there are many that exceed it. I was concerned about the maximum allowable input, so I selected a vinyl that seemed to have a high level (including one specially designed for audiophiles, 45rpm, about 8 minutes per side) and observed the output. However, the peak voltage was about ±7V. The peak 7V is equivalent to 5V when converted to rms of a sine wave.
If you convert from that to your (daughter's) case, the peak will be 2.2Vrms.
Shure has published a graph that measured the peak level of vinyl in the past (attached file is of very low resolution), and according to it, it was
35cm/sec (@1kHz) 80cm/sec (@4kHz)
Considering RIAA characteristics, 80cm/sec at 4kHz is the same level as 1kHz, so when considering 1kHz, in your case it is 2.8Vrms.
Based on the above, we recommend that you first try it as is, and if you feel that the headroom is insufficient, adjust it by about 3dB.
If you convert from that to your (daughter's) case, the peak will be 2.2Vrms.
Shure has published a graph that measured the peak level of vinyl in the past (attached file is of very low resolution), and according to it, it was
35cm/sec (@1kHz) 80cm/sec (@4kHz)
Considering RIAA characteristics, 80cm/sec at 4kHz is the same level as 1kHz, so when considering 1kHz, in your case it is 2.8Vrms.
Based on the above, we recommend that you first try it as is, and if you feel that the headroom is insufficient, adjust it by about 3dB.
Attachments
Thank you very much for your reply. I'll change the ADC to 2.8Vrms full scale and see if the noise level is acceptable. The powered speakers should have enough headroom since she lives in an apartment and has to keep them at "neighbor friendly" levels anyway.
I also plan on putting the ADC and BT transmitter in a nice little case with VU meters. My original plan was to get a test record with a 1kHz sine to calibrate the VU meters to 0dB, but such a record costs more than I was expecting, especially for something I'll use once. If I use a signal generator instead, can you recommend what voltage I should use for 0dB? They are mostly just for aesthetics, so they don't need to be perfect.
I also plan on putting the ADC and BT transmitter in a nice little case with VU meters. My original plan was to get a test record with a 1kHz sine to calibrate the VU meters to 0dB, but such a record costs more than I was expecting, especially for something I'll use once. If I use a signal generator instead, can you recommend what voltage I should use for 0dB? They are mostly just for aesthetics, so they don't need to be perfect.
Hot pressings, particularly dance 12" recordings, can be well above the 5cm/c^2 level. Its best to route a MM preamp into a preamplifier so levels can be adjusted before an ADC, since levels vary significantly between different kinds of vinyl recording.
Ok, I have a plan. Feel free to correct me if this doesn't make sense.
Phono preamp -> potentiometer -> unity gain op-amp buffer -> ADC
4mV @ 1kHz 5cm/s should give 400mV out of the preamp. With the ADC set for 1Vrms 0dBFS, that gives roughly 8dB of headroom with the pot set at max. If more headroom is needed, the pot can be turned down.
For the VU meters, I can connect them after the buffer and calibrate them to 1Vrms, or maybe even 0.9Vrms. I'll tell my daughter if she hears a bunch of static/crackle/popping on music peaks, to make sure the needles aren't going over 0dB and to turn it down if they do.
Phono preamp -> potentiometer -> unity gain op-amp buffer -> ADC
4mV @ 1kHz 5cm/s should give 400mV out of the preamp. With the ADC set for 1Vrms 0dBFS, that gives roughly 8dB of headroom with the pot set at max. If more headroom is needed, the pot can be turned down.
For the VU meters, I can connect them after the buffer and calibrate them to 1Vrms, or maybe even 0.9Vrms. I'll tell my daughter if she hears a bunch of static/crackle/popping on music peaks, to make sure the needles aren't going over 0dB and to turn it down if they do.
I think the signal used for calibration is a continuous sine wave, but the VU meter does not indicate the peak, so with that method, when the VU meter shows 1Vrms, or 0.9Vrms with a music signal, the instantaneous peak value may already exceed the ADC's permissible input (1Vrms, in other words -1.4 to +1.4 range).
Last edited:
Thanks. I don't think it matters now anyway. Received the meters yesterday and they are not accurate or sensitive enough. With sensitivity set to max, they require around 1.4Vrms to reach 0dB. If I reduce the test signal by -3dB, the needle only drops to -1dB, and with -9dB it drops to -5dB.
Didn't realize that they're passively driven. They are little 34mm round meters with the PCB mounted directly on the rear. Made the wrong assumption that there were other components on the rear of the PCB. The "AC" connection is simply to power the LED.
Didn't realize that they're passively driven. They are little 34mm round meters with the PCB mounted directly on the rear. Made the wrong assumption that there were other components on the rear of the PCB. The "AC" connection is simply to power the LED.
Phono pre-amplifiers are designed to have >20dB of headroom because of dust (and the resulting pop!).
Overloading a capacitively-coupled amplifier can shift the bias so much that the signal will drop out until the bias has re-stabilized. That can take a fraction of a second.
All capacitively-coupled stages prior to the level control need headroom. You will need to experiment. Cleaning the records is a good way to eliminate pops.
Ed
Overloading a capacitively-coupled amplifier can shift the bias so much that the signal will drop out until the bias has re-stabilized. That can take a fraction of a second.
All capacitively-coupled stages prior to the level control need headroom. You will need to experiment. Cleaning the records is a good way to eliminate pops.
Ed
Yes, they are passive circuits that are just rectified with diodes.
Because of this, it is affected by the nonlinearity of the diode, so the scale is scaled so that it displays correctly when driven with an impedance of 3.9kΩ. Other drive impedances will result in larger errors.
@EdGr I'm not to worried about any catastrophic effects of clipping. The WM8782 has a DC filter and the signal will also be filtered by the inherent BW limitation of bluetooth . I'm guessing that worst case will just be a lot of nasty noise.
Now I'm wondering how something like the Pro-ject Phono Box E BT 5 (bluetooth phono preamp/transmitter) is set up. It doesn't have a level adjustment, so maybe they gave it a ton of headroom? Not paying $190 to find out though. I'm just worried too much headroom will cause excessive noise.
@mason_f8 Most of these cheap "VU" meter kits have an active driver board. This is the first one I've seen that is passive, and it was obviously done just to save money. I don't think the meters are true VU meters.
Now I'm wondering how something like the Pro-ject Phono Box E BT 5 (bluetooth phono preamp/transmitter) is set up. It doesn't have a level adjustment, so maybe they gave it a ton of headroom? Not paying $190 to find out though. I'm just worried too much headroom will cause excessive noise.
@mason_f8 Most of these cheap "VU" meter kits have an active driver board. This is the first one I've seen that is passive, and it was obviously done just to save money. I don't think the meters are true VU meters.
The clipping can be followed by a momentary dropout. I expect the phono pre-amps with built-in ADCs to have much less than 20dB of headroom.
Clean records won't need that anyway.
Ed
Clean records won't need that anyway.
Ed
Received a proper driver board for the meters. They scale perfectly now. Still waiting for the enclosure.
I suggest this: Set the level so that your calculated peaks just reach say -24dbFs. The just do the recording at 24 bits. Now when you have the file, use a sound program (Audiacity?) to do an analysis of the recording - it will tell you peak values for both channels. So if you want to keep these unclipped, you just gain the whole file to end up at -1 dBFs. Finished. But you say - didn't we lose 24 dB in Noise or/and headroom. No, because the LP has about 50-70 dB SNR but the ADC has perhaps 120 dB so you still have 120-24-70=26 dB left as a safety margin. Dont try to hit 0 while recording - it will just clip as VU meters are way to slow and un-precise. If you do it like this you can skip the meters. Period.
Good luck!
Good luck!
Trick I found with audacity, use the amplify effect and it defaults to scale to max volume. That number is the headroom. There are probably other ways as well. Audacity has more features than I'll explore in my lifetime.
I'm not ripping vinyl though. I've already mentioned that the ADC will be connected to a bluetooth transmitter.
Understand, I'm suggesting running a couple of albums thru this to get an idea of the output from your cart/preamp to see what actual values look like.
I would use a -20db inline RCA attenuator per channel.
The best approach to look at setting up ADC circuits is to provide a load for the signal, then sample the signal to the ADC from the load point.
