|
Post by tupisac on Mar 31, 2017 14:40:50 GMT
I don't care for scientifically correct numbers. I don't really understand what they mean. Are they even relevant in case of digital signal with 0dB as maximum?
Dancing needles is the main goal. Actually looking for the ballistics of a peak meter here, not slow and mellow VU movements. This driver theoretically gives me the possibility for some tweaking.
But I actually do care for some subjective reference points to be measured precisely. First, I would like to set 0dB point for my usual listening levels, to give the clear indication when I'm moving too much into the red area after some hours of listening. Second, I do want it to be precise enough to set the output levels of two amps visually with a 1kHz signal.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Apr 1, 2017 7:58:45 GMT
Will work well for dancing needles. When you want to read about dB's look HEREFor that it'll work fine, but for 1 headphone only because of differences in sensitivity. You may have to slow the meter response when looking for average values. That might be hard to do if you want the meters to show you the 0.1dB which is really needed when you want to A-B properly. The meter won't resolve this with the mentioned circuit. Also you would have to connect the meters to the output socket of the amp and would have to make a second input on the ensclosure that feeds the meter. A simple multimeter with a 2V AC range will be more accurate.
|
|
|
Post by tupisac on Apr 1, 2017 14:45:10 GMT
Will work well for dancing needles. When you want to read about dB's look HEREI've read the article. In fact, your site has been my go-to place for in-work procrastination for a while now. But I still have trouble grasping all the units and their meaning. It just doesn't ring a bell . Does VU stands for voltage times current? Why no Watt instead? No, no, no, we will use dB. The one describing the hearing thresholds? No silly you, it's just the logarithmic scale between voltage levels, nothing to do with your hearing. Yup, it gets confusing. Takes some time to sink in. It took me a while to grasp how the hell can audio guys describe power with just a voltage... No problem here. I love my HD-600 . I had some regrets not buying HD-650 initially because of better bass, but I've read that filters make them virtually the same, and I was stunned by the sonarworks results. Basically end game for me, especially after reading some of your remarks on HD-800. Next level is proper speakers. You're right, DMM is a way to go. Just have to whip out a simple banana-jack plug. What is more accurate in that matter - measuring RMS (my DMM can do that) or just simple AC voltage? Besides meters, there is another nice topic I would like to open. Grounding. In all of different meanings First, the grounding I understand. I plan to move all power supplies to separate brick. No way for me to build proper Class II chassis, and I imagine grounding it to mains earth could lead to all different problems. Also, proximity to Kameleon. KISS rule applied. But what about grounding that I don't really understand? Keep everything floating and isolated? In one case there would be a Polaris (which requires ground on pot) and VU driver. In the second, Kameleon + DAC (might be gutted iFi or xmos+AK4490, or just an ODAC, haven't decided yet). Any advices?
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Apr 1, 2017 18:18:25 GMT
I've read the article. In fact, your site has been my go-to place for in-work procrastination for a while now. But I still have trouble grasping all the units and their meaning. It just doesn't ring a bell . Does VU stands for voltage times current? Why no Watt instead? No, no, no, we will use dB. The one describing the hearing thresholds? No silly you, it's just the logarithmic scale between voltage levels, nothing to do with your hearing. Yup, it gets confusing. Takes some time to sink in. It took me a while to grasp how the hell can audio guys describe power with just a voltage... VU stands for Volume Units ... VA stands for Volt*Ampere = Watt. I tried to make some sense in all the dB things. But yes, all the different dB things will make things much more complicated. It is done because a dB in itself means nothing untill it is 'referenced' to some known 'value' which can be almost everything. I'll have a look at that whole article and maybe elaborate or streamline it a bit... You're right, DMM is a way to go. Just have to whip out a simple banana-jack plug. What is more accurate in that matter - measuring RMS (my DMM can do that) or just simple AC voltage? When the signal is a sinewave the RMS value and AC voltage would give the same readings. When the waveform differs the 2 values will differ as well. Besides meters, there is another nice topic I would like to open. Grounding. In all of different meanings First, the grounding I understand. I plan to move all power supplies to separate brick. No way for me to build proper Class II chassis, and I imagine grounding it to mains earth could lead to all different problems. Also, proximity to Kameleon. KISS rule applied. But what about grounding that I don't really understand? Keep everything floating and isolated? In one case there would be a Polaris (which requires ground on pot) and VU driver. In the second, Kameleon + DAC (might be gutted iFi or xmos+AK4490, or just an ODAC, haven't decided yet). Any advices? Ground can mean a few things in audio. 1: safety ground (3rd prong) which is there to avoid you getting an electric shock when metal parts can be touched and when connected to a 3 prong outlet AND when something went horribly wrong in that device and a live mains wire has gotten in contact with those touchable metal parts. 2: The metal enclosure of a device and or circuit. This is done to avoid hum and RF signals to disturb the audio signals. 3: The 'common' or 'reference' or 'ground' of a circuit which is generally 0V in a symmetrical power source. All signals going in and out are 'referenced' to that 'voltage level'. This can have a 'star' configuration where all signals end in one single point or can be (more than one) ground plane on a PCB. It is VERY important that this part is well designed. Failure there can lead to all kinds of weird sounds/oscillations and other problems. 4: shield, a wire mesh around a signal wire that actually acts as an extension of the metal enclosure of the amp. Designed to keep out hum and RF signals. 5: The ground you walk on may not be electrically very conductive but for radio signals it has a certain impedance (kind of AC resistance). When you have gotten a shock from touching a mains wire you have had a current running through you and the ground you walk on which shows that the ground you walk on also has some resistance. groundloops is the most difficult thing to understand and trouble shoot as signals can travel between appliances (through wires or wireless) but also between the devices and the ground we walk on. Different currents can take a second 'path' other than the one you thought it could only take and mess things up. Too complicated to write down in a few short sentences, took me quite a while and still sometimes have trouble 'seeing' it as it depends on many aspects. Many of which aren't obvious. Sometimes not even to trained eyes.
|
|
|
Post by tupisac on Apr 2, 2017 12:39:16 GMT
I'll have a look at that whole article and maybe elaborate or streamline it a bit... Don't mind me. Your articles are perfect, please do not dumb them down because of some noob showing his ignorance. Elaborate - yes What I do kindly ask for is some heavily commented math applied to practical examples, like you did with the o2 gain explanation. Something to tie everything together, and show where each abstraction go and why. 1. So, we have an EMU, with 3V RMS on the output with 1kHz 0dB signal (foobar on 0dB volume). We have o2 on gain=1, with potentiometer on full. We have HD-600. How many decibels (in terms of human hearing) will they generate? 2. Assuming 3V RMS on 0dB 1kHz signal on EMU output - how much negative dB I need to apply (with foobar volume control) to get 2V RMS on output? 3. How to calculate the position of the o2 potentiometer to give you 80 hearing decibels on HD-600 (3V RMS from EMU)?
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Apr 2, 2017 18:20:14 GMT
1: As the amp has 1x gain the output voltage is 3V max (vol pot max) The HD600 is specified at 105dB @ 1V But there is 3V ... 3/1 = 3 now we need the 20x LOG10 of 3 which is +9.5dB 105dB + 9.5dB = 114dB.... PEAK level. How loud the music will be now depends on the average recording levels assuming the peaks are at or just below 0dB in the recording. Assuming a DR of 10 is most common one can assume that average SPL levels of around 104dB dB may be present.
2: 20LOG10(2/3) = -3.5dB
3: For this you will need to know the attenuation of the used potmeter for various positions AND the DR rating of the recording AND peak levels present on the recording.
|
|
|
Post by tupisac on Apr 3, 2017 11:03:55 GMT
It just looks too simple.
So, with low gain of 4,5 on Polaris I get +22,6 dB 3V x 4,5 = 13,5; 20log10(13,5) = 22,6 dB
Which equals to 127,6 dB peaks on my senns.
To go from peaks to RMS I use the formula Vrms = 0.3535 * Vpp; So, 13,5 * 0,3535 = 4,77 V Which gives me +13,7 dB (118,7 dB on senns).
I've read that 20log is for waves, 10log is for power. Could you elaborate on that?
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Apr 3, 2017 17:33:52 GMT
It just looks too simple. So, with low gain of 4,5 on Polaris I get +22,6 dB 3V x 4,5 = 13,5; 20log10(13,5) = 22,6 dB Which equals to 127,6 dB peaks on my senns. Yes, correct. To go from peaks to RMS I use the formula Vrms = 0.3535 * Vpp; So, 13,5 * 0,3535 = 4,77 V Which gives me +13,7 dB (118,7 dB on senns). The mentioned formula is for calculating peak-peak voltages to RMS values in pure sinewaves. It is not applicable here as the peak in dB is already an RMS value. A sinewave of 1Vrms consists of a positive half and negative half. A 1Vrms sinewave thus has a positive and negative peak voltage of 1*SQRT2 = 1.414V = Vpeak (Vp) Thus Vrms = 0.707 Vp Because a sinewave has a negative and positive peak (Vpp) the total Vpp is 2x 1.414Vp = 2.828Vpp Thus Vrms = 0.3535Vpp That's what the formula is for. The answer you seek, however, is more complicated. You want the 'average' SPL but you only have the peak SPL. The problem here is Dynamic range and how loud the CD is recorded. Most CD's have peaks close to 0dB so that's easy. Have a look at DR ratings of albums and you will find many different DR values. Each song on that album even has different DR ratings. And to make things worse not each song may actually have peak values reaching 0dB. Certainly not older (not 'remastered' and 'loudness-war' victimised) recordings. Anyway... the DR rating of a song (assuming peaks are reaching near 0dB) will tell you the difference between 'average' SPL and peaks. Peak level-DR rating = average SPL. So a 127dB SPL peak may have an average SPL of 123dB SPL when the DR rating is only '4' a 127dB SPL peak may have an average SPL of 116dB SPL when the DR rating is '11' a 127dB SPL peak may have an average SPL of 108dB SPL when the DR rating is '19' I've read that 20log is for waves, 10log is for power. Could you elaborate on that? 20log is for currents, voltages and resistances 10log is for power The reason is power = voltage x current Current = voltage/resistance. Thus power = voltage² / resistance Note the square in voltage. take 3V for example (referenced to 1V) the difference between these values is a factor 3 (dB is about factors) 20*LOG 3 = 9.54dB3V in say.. an arbitrary value of 10 Ohm = 0.9W 1V in the same 10 Ohm = 0.1W Here there is a factor 9 difference in power with the same voltage difference of a factor 3. 9 = 3² 10log 9 = 9.54dBThe same dB values of course but a different formula simply because of the ² in the power calculation.
|
|
|
Post by tupisac on Apr 4, 2017 12:35:12 GMT
I'll just stop thanking at this level. Let's just say you have a patience level of a loving grandpa . This is indeed one of the best and friendliest forums I've participated in. More questions coming up, but the Polaris arrived so there is a slight interruption . Will write some proper remarks later, but for now I must say it is definitely an improvement over o2 and emu 404 amp.
|
|
|
Post by tupisac on Apr 4, 2017 20:08:31 GMT
I did some measurements. On 1kHz sine wave (0dB) and pot on full (0dB) EMU generates 2 V AC. On 3 o'clock it generates 1,7 V AC. On 12 o'clock around 0,4 V AC.
2 V seemed like a bit hot for Polaris. Besides volume being tad too loud (low gain, with attenuation, had to set the pot around 9 o'clock for usual listening level), the SQ was slightly worse than expected. Not as in O2, but it was harder to separate instruments than on EMU's amp. When I set output to 1 V AC, everything seem perfect.
What is the advisable voltage level for Polaris input - assuming measurements on 1 kHz sine wave?
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Apr 4, 2017 21:45:54 GMT
The advisable input voltage depends on attenuator settings and volpot settings. Unlike the O2 there is no input stage that can clip with a too high input voltage.
A higher input voltage means the volpot has to be set lower. The lower it is the worse the L-R tracking (true for most potmeters unless really expensive) So set the attenuator or DAC output voltage so that the amp has a pleasant working volume control.
|
|
|
Post by tupisac on Apr 4, 2017 22:40:00 GMT
Don't know the attenuation settings, I'm third-hand user: Is it possible for EMU to clip or compress at 100% level?
|
|
jello
extremely active
Posts: 1,569
|
Post by jello on Apr 4, 2017 23:26:47 GMT
As far as I know (the original owner didn't mention anything to the contrary) the attenuation module is the standard one, which halves the gain settings. I left the plastic jumpers in the same position in which they arrived and had always assumed the attenuation module was 'on'. However, looking at the 2nd last page of the Polaris manual it indicates that the default setting is with attenuation module 'on' but the image to the left of that text has the jumpers positioned differently (covering pins 1 & 2 rather than 2 & 3 if working from the back). Perhaps the text/image are not intended to correspond with one another but you could always try re-positioning the jumpers just to see (or wait for Frans to confirm)? www.garage1217.com/MANUALS/POLARIS%20MANUALS/POLARIS%201.2%20MANUAL%204-4-15.pdf
|
|
|
Post by tupisac on Apr 4, 2017 23:54:55 GMT
I played with those settings. I settled with attenuation on, gain low and EMU output set on 1:30 o'clock. This gives me comfortable listening levels between 11 - 1 o'clock on Polaris. Photo shows EMU and Polaris volume matched: And another shot
|
|
|
Post by tupisac on Apr 10, 2017 10:45:58 GMT
Got the boards for Kameleon in-line, all components ordered. Now waiting I've decided to ditch the idea of heavy black case and go Garage1217 style. I do like Jeremy's design a lot. Here is initial sketch after some measurements: Measuring wasn't easy, I've scratched my Polaris a bit with the caliper... :/ I'm thinking about small 3d-printed grilles on the front. Might add this bit of something while hiding that the pillars will be placed differently than on Polaris. EDIT: Already ordered laser-cut plexi boards. Cheaper than original Hammond enclosure
|
|