Posted 12th July 2016 at 02:24 PM byTam Lin Updated 17th July 2016 at 08:03 PM byTam Lin(typo)
Theory tells us that each time the number of DAC chips is doubled the SNR increases by 3 dB. With 32 chips per channel, we should see a 15 dB increase. That is very good, but we can do better because there is no reason the paralleled DAC chips have to receive identical input data.
In essence, we have a 29-bit DAC for sample rates at or below 768K. For each 24-bit input sample value, we can provide a 29-bit value that produces an output current that is closest to the ideal. Accessing a 16MB look-up table 768K times per second is trivial for a modern 64-bit microprocessor. The table data comes from a one-time calibration procedure that analyses the DAC’s measured output performance for each possible input.
Above 768K, we are dealing with a delta that is obtained by scaling the difference between consecutive samples. Below 11.2896M, two or more chips are paralleled and a table lookup is used to improve accuracy.
I've not posted for a few months here as I've been involved in moving apartment which was quite a major project given the quantities of parts and assemblies I've accumulated. Even though I've been in my new place for over two weeks now, very little has been unpacked so far, but I have just yesterday rebooted my desktop active XO/amp system which had been quite literally assembled on my desktop with no casework whatsoever Its still without casework and survived the move with only a few wires falling off but now 'installed' in a drawer (pic attached). Its being fed from my 'Domino' balanced TDA1387 DAC and Taobao TFcard player and delivering bags of emotional satisfaction through stand mounted '3Nod' two-ways.
The amps are all LM4766 bridged running from 60V total but with heroic measures to keep supply noise under control, the central PSU has a CLC configuration and ferrite input transformers are used for coupling between AXO and amps. Output transformers (ferrite for...
To confirm the calibration of the sound card input and output gain. Also, to determine the relationship between the signal voltage, the recorded signal amplitude displayed in Audacity, and the signal peak and noise baseline levels in the FFT spectra.
* Setting the volume slider of the device output to 100 gives 1 V rms output for an amplitude 0.5 sine wave.
* Setting the volume slider of the device recording line input to 100 gives records a 1 V rms tone as an amplitude 0.5 sine wave, which is displayed in the frequency spectrum (FFT) as peak of magnitude 0 dB in Audacity when both channels are averaged.
* volume setting 100 needed for unity gain loopback.
* 0.5 amplitude sine wave = 0 dB FFT = 1 V rms.
* noise baseline in averaged stereo FFT is 3 dB lower than single channel measurement....
Posted 17th June 2016 at 02:44 PM byrjm (RJM Audio Blog)
Updated 20th December 2016 at 12:01 PM byrjm
I'm not totally sure this would work as advertised, but I can't see any obvious reason why it would not...
It's pretty much the same circuit as I used in the CrystalFET, which started out in a previous blog post in the Voltage Regulators for Line Level Audio series, but here I've replaced the MOSFETs with bipolars. It is shown configured to deliver 20 mA @ 12 V, split supply. Enough to power an op amp phono stage for example, or a preamp, or the voltage gain stage of a headphone amplifier.
Posted 4th June 2016 at 05:42 AM bygooglyone Updated 4th June 2016 at 06:18 AM bygooglyone
I finally got around to rolling out the distortion test set and the Amplifier of 100 Transistors to measure its performance.
I would like to say that I don't care - and that the whole thing is an engineering abortion. A complicated joke, and that the measurements don''t matter. The fact that I am making the measurement would however show me to be a liar - as if I didn't care, then why did I do this?
Anyway, with low distortion measurements, getting your head around the baseline of your test gear is key. With all gains / levels being equal, here is the loop-back distortion of the test system:
Which is fine, rolls along at about 0.0003% across the band.
Then I ran a sweep of the Amplifier of 100 Transistors with NO load at 3dB below clipping:
OK, this is saying the amplifier distortion raises it's head above the noise floor at 1KHz and is...
Posted 28th May 2016 at 03:18 AM bygooglyone Updated 28th May 2016 at 03:22 AM bygooglyone
In my previous post I presented an idea spawned from a very bad place, primarily boredom and probably too much alcohol. Plane trips from Australia do that...
The Amplifier of 100 Transistors was the result.
Between that posting and this a few things have happened. I finished the design - adding extra decoupling and 100 Ohm base resistors to all 104 output devices. I did this because I sincerely thought I was building more of an RF oscillator than amplifier.
And I built it.
And I got it working.
This is the beast from the back before I loaded the "output devices":
Note the ludicrous number of emitter and base resistors! Half way through I concluded that I was bonkers, and wasting several days hand building a complete folly.
Then again, even with the thing running, it is still a folly!
This circuit is to fulfill my curiosity of hearing the sonic differences between pentode, ultra-linear and triode gain stage. As with any line stage, this circuit can also be the input part of a power amplifier. I'm considering a single ended amplifier where both the input and output stages have adjustable ultralinear percentage.. I'll show the output stage design on another post.
Anyway, i'm not sure there will be an audible difference between each mode other than gain.. but until i actually build this circuit, i can't really be sure.
The idea is to use a pot to replace R5+R6. Dual 250K linear pot should do for both channel. You then adjust the screen feedback percentage by adjusting the ratio between R5 and R6. Here's how you obtain each mode:
1. Pentode mode is by making R5 = 200k and R6 = 0R. This will keep the screen grid at at fixed voltage regardless of plate swing.
2. Triode mode is...
Such mouthful blog title.. The idea is nothing new: to get the so-called "tube sound" from a hybrid amp. Tube sound for me comes from these two factors:
1. Triode-like linearity
2. Low to moderate damping factor
Personally, i believe number two is the main reason for tube sound. Any solid state amplifier manipulated to get a damping factor of around 1-2, instead of the typical SS amp DF that falls into the tens to hundreds, will sound like a tube amp. This is because with such low damping factor, the amp ceases to ignore the speaker impedance curve and start to become less of a voltage source and more of a current source where the bumps of speaker impedance curve will affect the amp's output.
Without further ado.. here it is:
1. Ignore the transistor and MOSFET type. I chose them because that's what i have in LTSpice. The small signal BJT can use BC547 for...