|
Post by jimwilliams on Mar 19, 2014 11:40:05 GMT -6
I rebuild my bench and console Power One supplies with huge caps and film bypasses. I also add a time contraint mod to prevent power up latching. Add a 47 uf cap and a 220~270 ohm resistor from pin's 2 and 3 on the 739 regulator chips. Place the + end towards pin 2. That fixes a lot of power up issues.
|
|
|
Post by jimwilliams on Mar 19, 2014 11:33:13 GMT -6
Wouldn't a higher quality playback platform require higher quality recorded music to work? I don't see the need to reproduce an API, Neve mic preamp or a TL072 opamp.
Doesn't an Ipod accept 16 bit FLAC files as well?
|
|
|
Post by jimwilliams on Mar 19, 2014 11:26:46 GMT -6
The one with the broken power switch.
|
|
|
Post by jimwilliams on Mar 19, 2014 11:25:14 GMT -6
So if I buy a Strat and a Marshall I can sound like Hendrix?
|
|
|
Post by jimwilliams on Mar 5, 2014 10:01:06 GMT -6
I saw it at Namm. It's smoke and mirrors to me.
The LAST thing I ever think of is "did I pick the right mic". If it sounds good, I'm done.
I'm far more attentive to the talent on the other side of the mic. That is 99%, the mic is 1%.
If one considers the concept vs the execution, this is a failed concept. You have far too many variables to create a model that sells to me. There is the mic capsule, the head amp, the output circuits, all inside the mic. Does it have an output transformer or not? The cable, output impedance and mic preamp input impedance are new variables. Is the mic preamp tube, transistor or opamp based? Does it have an input transformer? Does the output stage have one? What cables are used? What converter is used? Does it have a transformer, direct coupled or? Sample rate, bit depth and converter input designs and filters are new variables. Then look at the output,what DAC is used? Sample rate? Bit depth? Opamp design? Filter design? Filter slopes? Filter pole numbers? Group delay or phase shift? Linear or non-linear phase shift?
Yes, I can go on and on.
Change just one of those and it's a new sound. Does that system/software compensate for all those options and variables? Can one adjust any or all of them? Can you use a different mic?
Yes, some simulators are very good, like flight simulators, but those are not a substitute for flying. It's only a training tool, which is something this design might be a good fit for, inside Full Compass.
|
|
|
Post by jimwilliams on Mar 2, 2014 12:53:32 GMT -6
It's not all that hard to do. You have +, - 15 volts and 48 for phantom. The Power One HDD-15-5AG runs about $160, the small 48 is $58. I would use a 5 pin XLR and run +, -, 48, power ground and seperate chassis ground lead. That would connect inside the rack to power ground and the chassis ground wire would connect to the AC ground on the power cable.
5 amps would run anything, nice and cool too.
|
|
|
Post by jimwilliams on Mar 2, 2014 12:46:34 GMT -6
Those are the Analog Devices opamps I specified when Ross built my DAC. He had not heard of them until I asked them to be used in mine. They are the ADA 4898-1. These slew at 60v/us, .9nv noise, 55 mhz gain bandwidth and very good low end output current. These have more energy in the bass than the other opamps as they have high current, almost class A output stages.
That is what makes them so great, the water clear top end of high speed, low THD parts with that missing depth in the low end. They have a more uniform sound than many fast chips only because the spectral balance is better.
I also did some consulting for Michal at Mytek. He does very good work and his low jitter clock designs are great.
|
|
|
Post by jimwilliams on Mar 2, 2014 12:39:34 GMT -6
If I had a 500 rack, I would rip the power supply out and wire a 4 or 5 pin connector. Add a Power One 5 amp dual supply and call it a day.
It's not worth playing with toys when you need tools.
|
|
|
Post by jimwilliams on Mar 2, 2014 12:35:21 GMT -6
Are people still doing this?
My solution is to adjust the volume control when listening.
|
|
|
Post by jimwilliams on Feb 27, 2014 10:11:49 GMT -6
#60 drill.
|
|
|
Post by jimwilliams on Feb 24, 2014 10:41:20 GMT -6
|
|
|
Post by jimwilliams on Feb 24, 2014 10:35:26 GMT -6
You need to tame the drummer.
|
|
|
Post by jimwilliams on Feb 24, 2014 10:27:08 GMT -6
Most of my mics have no transformer so I load them lightly at about 4.2k ohms through my preamps. Lower that impedance and THD creeps up as those curcuits have a harder time driving low impedance loads.
I prefer that higher impedance for my transformer coupled mics as well, they souud more open and airy. Loading down a transformer coupled mic is like loading down an electric guitar, you just roll off the top end.
If you want to optimize those loads, use a square wave oscillator at 2k hz and remove the primary lead in the mic and feed that square wave at around 30 mv into that transformer. Plug into the preamp/load under test and observe on a scope. Adjust the load until the top of the waveform is flat. That way you have found the sweet spot avoiding ringing and overshoot problems.
|
|
|
Post by jimwilliams on Feb 24, 2014 10:12:01 GMT -6
I never warmed up to the 500 format for several reasons.
First is the size, they are very crowded with tiny controls. Add to that the bundle of cables crowded into the rear, a ratsnest sized format to work with. Maybe it's ideal if you have small Japanese sized hands. 500 designs are the clown cars of audio. You never know how many knobs you can fit into them.
Next are the power supply issues. Power supplies are not that hard to do right. Somehow, the rack boys seem to always under power them. For me, it's no harder than building a power supply that will handle twice what anyone will throw at it. The obvious solution is to locate the power supply externally, like as done in an analog console. That way current limitations are avoided and that internal power transformer no longer generates hum fields into your audio path and audio transformers. It would also free up more rack module spaces as that power supply wouldn't eat up one end of the rack. You could then fit modules all across the 19" rack.
Last is interaction problems and crosstalk. Modules should not talk to each other. Again, power supply issues and ground distribution are key. I prefer my outboard to be self powered individually. The crosstalk is best that way. Someone should do a crosstalk sweep of common racks with common modules, but that info might get into the way of marketing the products.
|
|
|
Post by jimwilliams on Jan 23, 2014 11:57:11 GMT -6
Call me Odd Duck. I don't use global EQ on my mixes and never have. That tells me the mix has problems best addressed in the console, not outside of it.
|
|
|
Post by jimwilliams on Jan 23, 2014 11:52:58 GMT -6
I like my 1966 Fender Deluxe Reverb a bit more.
|
|
|
Post by jimwilliams on Jan 23, 2014 11:37:43 GMT -6
|
|
|
Post by jimwilliams on Jan 23, 2014 11:31:27 GMT -6
Must go active. All my guitars are and can drive 2000' cables without losses. Best if installed in the instrument before external losses kick in. That way the tone control roll-offs and volume pot loading can be avoided. You can get acoustic guitar zone harmonics if done that way.
Or use an external buffer box and put that in your pocket. Run a 1 foot cable to it.
|
|
|
Post by jimwilliams on Jan 22, 2014 11:11:20 GMT -6
Antares came out with their 'mic modeler' software years ago. They even included a modified Audio Upgrades AKG 414B-ULS in their samples.
Not considered is that all these emulations require digital conversion. A perfect simulation requires perfect conversion. Neither are available today.
|
|
|
Post by jimwilliams on Jan 22, 2014 10:54:21 GMT -6
I tried replacing the capsule in one of my original NT-2's I designed for Rode. I used a rk-12. It didn't work out well. I put the original capsule in.
The reason is the domed mount, it creates a mid bass resonance that you can't remove. It is a reflection. The original capsule with it's thinner low end and bumped upper end seems to mask most of that resonance.
|
|
|
Post by jimwilliams on Jan 22, 2014 10:41:37 GMT -6
500 modules have power rail limits, power current limits, a shared supply and ground increasing crosstalk. The format is also very crowded for larger hands.
Too bad the dbx/Aphex racks were not popularized as they had more room, a beefer power supply, + - 15 and + - 24 volt rails, better construction, etc.
When you have an analog console with -90 dbu stereo crosstalk at 10k hz, 500 modules in stereo will collapse that crosstalk spec.
One thing you will never see is a decent Audio Precision stereo crosstalk test done on a pair of 500 modules or a 500 series rack, the results would not look pretty.
Back in the 1980's when powered gear racks came out, they were marketed as a way to lower the costs of multiple processors and to save weight and space.
|
|
|
Post by jimwilliams on Nov 13, 2013 11:24:46 GMT -6
This web site shows -110 db EIN with a 300k measurement bandwidth and an unknown gain setting. That is confusing to most readers as many don't understand how noise is measured. Typically, audio noise measurements are done with a 20~22k hz measurement bandwidth to remove out of band noise (inaudible) from the readings. This is how Audio Precision, Rode and Schwartz and Prism audio test analysers do it. EIN means equivalent input noise, it's a sum of the gain plus the signal to noise ratio. 60 db of gain is also where mic preamps are measured, simply to equalise the test perameters for easy comparisons. Measureing EIN at higher gains does not improve the specs as preamp self noise begins to wash out the differences. Typically, a 150 ohm source impedance is also used to emulate the output impedance of many microphones and the source impedance of most mic input transformers. When I test EIN of mic preamps, I also test at 50 ohms input impedance to emulate active condenser mics. That will lower the EIN to below the -129.6 db EIN of a 150 ohm resistor and it's load, IF the preamp has less noise as well. Some designs will do -133 db EIN at 50 ohms. "A" weighting is also to be avoided as it skews the real readings and gives those devices that choose to use A weighting filters a gain in specs without showing the actual noise. I have seen in the last 15 years or so a complete lack of standardised audio tests and a lot of skewed, mis-applied test proceedures in order to either confuse the end user or to bump up their gear test measurements. I also realize many do not share my concepts of honest audio reproduction, but we are not talking about any of that, this is noise. Noise has nothing to do with sound or any audio influenced colors, it's noise, plain and simple. Either you accept a certain amount or flavor of noise or you do not. I'm in the second catagory as I find any sort or type of noise unacceptable. With modern design techniques and carefully choosen components and designs, very low noise specs are achievable. As I mentioned, an API opamp coupled to a proper turns ratio mic input transformer at +60 db of gain with a 20~22k measurement bandwidth will yield a -127 db EIN spec at 150 ohms impedance, as long as the DUT (device under test) is properly screened and the power supply is clean. Anything less than that is either a mistake or a problem. Many won't notice this problem using high output condenser mics, but connect a low output dynamic or ribbon mic and that hiss will eat you up. It's 2013 and there is no valid reason to print excess noise to your converters anymore, unless it is to be used as an effect, like emulating tape hiss. As long as buyers do not challenge the manufacturer's to present honest test results, don't expect them to suddenly become honest on their own. We have the manufacturers we deserve, sort of like government. It's up to the end users to demand honesty.
|
|
|
Post by jimwilliams on Nov 11, 2013 12:41:40 GMT -6
Just about any alnico 5 based 6k ohm DC resistance P-bass pickup will do. Even the low cost imports I get for $20 sound great, the quality of the bass makes up for the rest. My 1970's era P-bass pickup was rewound in 1979 by Seymour Duncan, before he manufactured pickups. It is wound to 1k ohms per coil. In series it has a 32k hz resonant peak so it's flat from 2~20k hz. It's very even sounding yet has an acoustic bandwidth and openess. That bass also has an active 2 band sweepable EQ circuit inside.
I have used copper foil to line the insides of the plastic covers and the rest of the innards, it's very quiet. A piece of black electric tape covers the magnet poles so your fingers don't create any buzz sounds while playing.
A great mod for stock 6k ohm P-bass coils is to add a series/parallel toggle switch. That gives you that 15k hz hi-fi bandwidth for an intimate, detailed option to the stock sounds. Add a "no load" Fender tone pot to disengage the tone circuit from the pickups when not in use. Then it does Alembic ranged top end as well as the regular sounds.
|
|
|
Post by jimwilliams on Nov 11, 2013 12:21:02 GMT -6
A -115 dbu EIN spec is about 12 db noisier than a decent API preamp loaded with a proper ratio input transformer, those can do about -127 db EIN at +60 db of gain. Neve-y designs reach -124 db EIN. Either this is a noisy design or something is amiss with the measurements.
Take 12 db's of hiss off and you will hear that.
|
|
|
Post by jimwilliams on Oct 15, 2013 9:05:03 GMT -6
Look up the apps notes for the Burrbrown buf634 at: www.ti.comBuild it as is, but use a quality low offset opamp like the LME49710. Add a servo if you use an opamp with DC offsets. You can use the buf634 as it's a DIP. I used the HA5002's mounted on a special Brown Dog adaptor. I also used the LME49713 CFA with a servo, no feedback caps on that opamp. Easy to make on a breadboard. The power supplies are more complicated than the headphone amp.
|
|