Homebrew Digital Sampling Audio spectrum analyser.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

GK

Disabled Account
Joined 2006
infinia said:
should U5 pin 4 be on digital split?

What about ADC chips that have internal ref. and differential inputs, would save a lot of parts.


According to the datasheet the device is intolerant (even on a transient basis) of any potential difference between AVDD and DVDD. Also, I'm only running a 2MHz clock (the part is specified at 26MHz for max. throughput), so in this much less demanding application the pins should be just strapped together, but heavily bypassed.

An ADC with differential inputs would not allow a simpler circuit. The VIN’s still need to be buffered to provide a high input impedance (and to charge the internal switched sampling capacitance) and protected from excessive (generally 300mV max. beyond the rails) input voltage swing. CMRR would be inferior to that of the dedicated in-amp input stage and there would be little allowable common mode input signal (as opposed to the current ~26Vp-p).

And the unipolar-input option would still need the additional circuitry, with access to the ADC reference (and most likely an external op-amp buffer for internal Vref ADC's too).

Cheers,
Glen


EDIT: Pin 4 (AVDD) is also on the "analogue" side of the SOIC package. In other words, it is directly decoupled to the analogue ground plane side with it's own local by-passing. That is another reason why I selected this ADC - a nice big SOIC package with a pin-out that is home-made PCB friendly.
 

GK

Disabled Account
Joined 2006
The display module.

All that is now left for this part of the analyser is the USB/RS232 module and the mother/memory/recorder board.


An externally hosted image should be here but it was not working when we last tested it.
 

Attachments

  • dism.jpg
    dism.jpg
    88.7 KB · Views: 228

GK

Disabled Account
Joined 2006
And here is the control panel for this part of the instrument. As stated previously, this is essentially a low frequency digital storage oscilloscope adaptor that can be used (with external equipment and modules) as either a digital CRO for observing/recording low frequency signals (such as DC bias drift in a class AB power output stage of an audio amplifier) or mated with a superheterodyne receiver to make a narrow bandwidth spectrum analyser.

To summarise the controls and indicators:

Timebase:
The timebase sweep duration of the digital recorder will range from 0.01s/Div to 500s/Div in 15 1-2-5 steps.
The timebase duration being indicated on the front panel on three 7-segment LED displays and set with a pair of pushbuttons. A 10 segment LED bargraph display shows the sweep progress (to suppress the suspense when waiting for a long duration sweep).

Trigger:
The timebase can be set to either free run, trigger on an external input or record a single sweep on the pressing of the single shot button.

Comms:
The unit can be run either as a stand alone unit with an external analogue oscilloscope with X/Y inputs acting as the display, or with a PC, primarily for storage of recordings. A toggle switch selects either local or computer control and both RS-232 and USB interfaces are provided. The comms module will hardware default to RS-232, but automatically switch to USB when a USB connection is detected (+5V on line 1 of the USB connector).

Display:
The Vertical and Horizontal sweep outputs for driving an external analogue oscilloscope for the display.

Recorder sweep:
These are the outputs necessary for synchronising the internal recorder with a local oscillator VCO in spectrum analyser applications. There are many other possible applications also, such as using the 0-10V ramp output to drive a V-to-F convertor, used in conjunction with a log detector or an RMS convertor to plot and record the frequency response of audio filters, etc.
The Clock (4096 pulses per sweep - 12 bits)and Sync outputs will be used to sync the recorder with the DDS based, uC controlled sweeping local oscillator which I have on the drawing board for the audio spectrum analyser unit.

Input:
The differential analogue input to the ADC board as described previously.
 

Attachments

  • dsomk1.jpg
    dsomk1.jpg
    39.3 KB · Views: 230

GK

Disabled Account
Joined 2006
Continuing to talk to myself for some reason……...

The Mother Board schematic is completed; will have the hi-res version up in a day or two along with the PCB layout.

Interesting how this place works. Numerous threads on measurement equipment and the supposed importance of measuring the THD spectra of audio equipment. Come up with and report on a design to do such instead of just yakking and once things get a bit technical everybody shuts up.

:violin:
 

Attachments

  • mb.gif
    mb.gif
    47 KB · Views: 328
AX tech editor
Joined 2002
Paid Member
G.Kleinschmidt said:
[snip]Interesting how this place works. Numerous threads on measurement equipment and the supposed importance of measuring the THD spectra of audio equipment. Come up with and report on a design to do such instead of just yakking and once things get a bit technical everybody shuts up.

:violin:


Welcome to The Real World (tm) :D

Jan Didden
 
Hi Glen

I doubt that there are more than a handful of people, certainly percentage wise, on the entire forum that actually build anything. As well, instrumentation has no associated religious war.

What you really need to do is insist that a fine drizzle of extra virgin olive oil, from just the right valley in Tuscany, improves the performance.

There are other threads where you can succumb to self-gratification if you need it - which I doubt. So chin up and keep making progress.

I see you use Altium, so are you looking for some grunt help with the PCB design?

I am a Tek man myself ( so I was wrong - Tek vs HP is as good a religious war as anything out there! ). My newer GPIB-equiped test gear has gone the route of pushbuttons for the HMI - really painful and confusing at times, unless you have the manual open on your lap. So I would suggest rotary controls on the front panel rather than push buttons e.g. your timebase selection. They are much easier to grab/manipulate while you are looking at the screen.
 
G.Kleinschmidt said:
Continuing to talk to myself for some reason……...

The Mother Board schematic is completed; will have the hi-res version up in a day or two along with the PCB layout.

Interesting how this place works. Numerous threads on measurement equipment and the supposed importance of measuring the THD spectra of audio equipment. Come up with and report on a design to do such instead of just yakking and once things get a bit technical everybody shuts up.

:violin:


whatya want us to say?

ch.gif

good job
 
Looking at your ADC module.

Depending on the likely source impedace (which would be a fractional Ohm for power amps and a few Ohms for preamp outputs - ignoring tube based units...), you might be better looking at the INA103 which has lower noise for impedance less than 1k.

Sadly it is not pin compatible.

I can't make out the Rg value on the jpg, but if your gain is high then I would recommend more serious RFI protection at the front end. You can always elect to not fit those components at a later date. See Analog Devices Application note AN-671.
 

GK

Disabled Account
Joined 2006
Hi Alan.

The boards are 90% complete. I was going to use a rotary switch for the time base but the decoding is too much of a hassle and I've got no spare I/O pins on the uC to take 15 direct inputs.

The INA118 gain is set to only 1.25 (2V FSD). An external interface will be needed to interface with the outside world.
In other words, the ADC module input is not designed to connect to devices under test.

Because I want to use this unit as either a LF DSO or the recorder/display end for a sweeping frequency spectrum analyser, I'm keeping it modular. As it is, the instrument will be pretty useless on its own.

I like the old school Tek analogue scopes, but never prefered their DSO's over the competition. Opted for a LEcroy for my bench at work :)

Cheers,
Glen
 

GK

Disabled Account
Joined 2006
Here is the completed motherboard. I used DIP'ed devices for the IC's and SMD for most of the passives. I could have easily made it 1/4 the size if designed to be manufactured at a board house, but I had to keep it homebrew friendly with tinned wire feed-throughs instead of plated through holes.
There are so many parallel address and data lines that you have to keep to a minimum via size and track width / clearance to ensure success. You also have to return tracks to component pads on the layer that a soldering iron can get too. Anyway, like with the VGAASAMK1 project I’ve erred on the conservative side and this board will be a piece of cake to make / hand-drill and load.

I made a slight modification to the motherboard schematic also, in that the SMD memory chips have been put onto a pair of “memory module” daughter boards that plug in to the motherboard vertically via 90-degree SIL headers. The paralleled address line routing was getting too much of a mess.

There is also another daughter board which I forgot to mention before – the “DAC module” which provides the 0-10V ramp output, synchronised to the recorder section timebase for controlling an external analogue VCO for spectrum analyser applications.

The RS232/comms module and the timebase display module are separate from the motherboard and will mount behind the instruments front panel.

The design work is complete but it will take at least a couple of week before I have this thing up and running as 50% of the parts I need are on back order. Farnell Aust. seems to be keeping an ever-dwindling inventory of in-stock items.

In the meanwhile I will start up loading the final schematics onto my website under test equipment, DSOMK1.

When it’s running I’ll post some RF spectrum analyser displays that I’ll be able to immediately generate with some existing equipment.

Cheers,
Glen
 

Attachments

  • mb.jpg
    mb.jpg
    91 KB · Views: 469
I don't think glen is with us anymore - his account has been disabled.

Although occasionaly a bit direct and absolutely dissmissive of subjectivist b.s., he did bring some great technical insights to the forum.

This is a deep loss for DIYaudio. Glenn is top-notch. None of us is perfect, and Glenn's contributions far outweigh any controversy or directness he brought. A bit of conflict and disagreement is good. We all learn from the back-and-forth. I hope Glenn will return.

Cheers,
Bob
 
This is a deep loss for DIYaudio. Glenn is top-notch. None of us is perfect, and Glenn's contributions far outweigh any controversy or directness he brought. A bit of conflict and disagreement is good. We all learn from the back-and-forth. I hope Glenn will return.

Cheers,
Bob

I agree Bob, it is also unfortunate to see Syn08 Binned again. Seems they don't want real competent engineers here. I have told other friends that people don't appreciate real work unless they have to pay for it. Imagine what one would have to pay to get Syn08's designs under contract ... LOL! I believe that we're seeing bias against real degreed engineers here and that is unfortunate.

I have considered putting some of my work on forums in the past, but I hold back and do very little since it is not often appreciated.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.