Converters

[gct]

HI

Anyone got a circuit for converters? Is it possible to upgrade Focusrite converters? Is it possible to build something? How complex is it? I need 24 bit 44.1KHz. I have 3 old ADATs at 16 bit.

Any thoughts?

Cheers

GCT

 

[JohnRoberts]

Most manufacturers offer comprehensive data sheets.

You need to look and old fashioned analog specs like S/N and THD, since 24b convertors are not all the same, or really true 24b resolution that in theory is something like 144dB dynamic range.

JR

 

[Rochey]

I'd suggest you start by specifying exactly what you want (channels, features, max sampling rate, interface, dynamic range, signal to noise ratio, maximum input voltage etc, word clock or spdif sync etc)

Then start digging through the schematics for evaluation boards for the ic's your thinking of using. Once you understand how each section of the board works, then you can decide which bits work, and which ones won't.

Get ready to be comfortable with pcb layout tools, as none of the converters on the market come in thru hole format. You'll need your own custom circuit board to support them.

 

[gct]

I'm getting the idea that this is very complex.

With the Focusrite converters is there a possibility that replacing components would resolve the problem. I've looked at the Voicmaster pro stereo out SPDIF outputs and for a 1k test tone I get spikes at roughly octave intervals above on the recordings. A consumer level Sony DAT recorder gives a recording with no spikes -same cables etc - but only 16 bit - can converters be swapped?


I would be interested in a stereo pair of clean inputs on SPDIF 24bit 44.1KHz or fixing my Voicemaster pro card and Octapre card to give good results.

Is there a thread here for anything related? - I didn't see one. I need more background info.

Cheers

GCT

 

[JohnRoberts]

You might want to re-test at 6dB less level just to make sure you are not saturating (clipping) the analog path around the convertor.

JR

 

[Rochey]

Yup, that really sounds like a clipping/thd issue.

You may find that if you run the converters with a little less signal (6dB below as John suggests) you'll get the clean output you were looking for.

 

[gct]

Ok will try but this has been a niggle for a while now - recording at all different levels- I could hear something was not right. Also Sony DAT gave excellent results -same settings/cables/digi in digi out. Thought it was the preamps or the mic circuits until I tried the Sony. On the Octapre also some channels were much worse than others.

I generated a clean test signal and recorded digi out to digi in from my RME sound card and this was perfect keeping in the digital domain. Then added the converter, no clipping shown on any metering. Then I analysed the result and all these peaks at regular intervals appear , but the Sony consumer DAT machine is totally free of peaks.

 

[JohnRoberts]

There could be something wrong or different with the gain structure on the analog side of the convertor interface.

If the distortion is still there after you drop the level fegetit...  It may be something you can fix easier than starting from scratch.

JR

 

[abechap024]

could bad clocking possibly be the issue? maybe try syncing the device to another clock if it has that option.

 

[Andy Peters]

HI

Anyone got a circuit for converters? Is it possible to upgrade Focusrite converters? Is it possible to build something? How complex is it? I need 24 bit 44.1KHz. I have 3 old ADATs at 16 bit.

Any thoughts?

Cheers

GCT

Upgrading the converter chips in a product is likely impossible. Differences in pinouts, power-supply requirements, reference requirements, initialization, etc,. all conspire against you.

-a

 

[gct]

Will clock from other source and try at lower levelsand post results.
Cheers
GCT

 

[JohnRoberts]

could bad clocking possibly be the issue? maybe try syncing the device to another clock if it has that option.

The symptom described sounds like harmonic spurs, generally caused by analog path saturation. Clock issues "might" cause noise, I doubt it would generate harmonics.

JR

 

[s2udio]

How complex is it?

As said by AP
E=MC"x SMD all now f"!*ed up .....how many light years have you to waste...lifes far to short !

 

[abechap024]

could bad clocking possibly be the issue? maybe try syncing the device to another clock if it has that option.

The symptom described sounds like harmonic spurs, generally caused by analog path saturation. Clock issues "might" cause noise, I doubt it would generate harmonics.

JR

Yes that is true, I agree. For some reason I read "spikes" and assumed there were little glitches in the audio. My mistake. If all you are seeing is good old Harmonic Distortion, maybe there is nothing wrong with your focusrite? Is it possible its had this "problem" all along and now its sounding worse after finding it? I've had that happen to me many times....ears will give you any chance to be biased.
Anyway how large are the spikes compared to the 1k test tone?

Cheers,
abe

 

[JohnRoberts]

How complex is it?

As said by AP
E=MC"x SMD all now f"!*ed up .....how many light years have you to waste...lifes far to short !

Different convertors will have very different clock requirements with some taking oversampling clocks at 256 Fs, 384 Fs, 512 Fs, and some even higher rates. Then the format of the serial data can vary in order of LSB/MSB, I2S, SPI, or dsp format, etc. I don't see the problem as SMD per se, but that will make it more difficult to tweak programming pins if the format is hard programmed. If the convertor expects soft programming, then there needs to be a brain somewhere send it serial data over another serial control port.

Not quite as easy as swapping out opamps, or moving the tubes around inside an old AM radio to get it maybe work, despite low gain tubes. 

JR

 

[s2udio]

Absolutly JR....As i said lifes to short to start modding SMD boards only to find out....
HNY