Autoranger for soundcards

Jan,

Thanks for the additional information on Pilgham's experiences.

I was having issues getting any of the frequency response calibrations completed. After adjusting the capacitor values to the new recommended values and not noticing any improvements in the calibration, I took a step back, disassembled, and reviewed each component. Turns out I had swapped the attenuator's RV1 and the display/control RV1! I was wondering why the display pot was getting so hot! Unfortunately, the pins for the display RV1 were right underneath the perfectly soldered and aligned LCD module :eek:

It took me a few days of walking into the workshop, sighing, and then moving on to something else until I mustered up the strength to rework all 32 pins of the LCD (yes, I had even soldered the entire bottom row). Between a hot air rework station, solder wick, a solder sucker, and lots of flux, I managed to remove the LCD, swap the pots, and successfully calibrate the unit.

I have shared the BOM I purchased from Mouser USA at Mouser Electronics if it is use to others (it does not include the LEDs). Unfortunately, it seems that the Coto relays (with diodes) I used and the AD536AJQ are currently backordered.
 
I just about have mine wrapped up but have no idea what connectors to order for the board-to-board interconnects. These would be, for example, the female connectors on the wires from the power board. (Credit to ergo for the picture in post 384). Can someone point me in the right direction?

644167d1509954183-autoranger-soundcards-ss_view-jpg


Thanks!
Phil
 
Last edited:
These would be, for example, the female connectors on the wires from the power board. (Credit to ergo for the picture in post 384). Can someone point me in the right direction?

I used pre-terminated cables like the following: GenBasic 80 Piece Female to Female Solderless Ribbon Dupont Jumper Wires (4 and 8 Inch) for Breadboard Prototyping Amazon.com: GenBasic 80 Piece Female to Female Solderless Ribbon Dupont Jumper Wires (4 and 8 Inch) for Breadboard Prototyping: Computers & Accessories.
 
AX tech editor
Joined 2002
Paid Member
Good evening. After having spent some time with the AutoRanger I would like to suggest a software-modifcation.

Here is why:
# In idle (small or no signal present at input) the AR switches to highest possible gain.
# If there is is a sudden increase of input level, the the AR clips and overdrives the soundcard/ADC .
# after some delay the AR ranges up in attenuation to the target range
=> performing stepped level sweeps, this overdrive-situation happens every single step.

Auto-ranging circuits in Audio Analyzers normally work the other way round, don't they? i.e.:
# idle = highest attenuation available
# signal present -> ranging down to the target range
 
Disabled Account
Joined 2012
Good evening. After having spent some time with the AutoRanger I would like to suggest a software-modifcation.

Auto-ranging circuits in Audio Analyzers normally work the other way round, don't they? i.e.:
# idle = highest attenuation available
# signal present -> ranging down to the target range

yes, that is the norm... highest to lowest range. Default and at turn-on is to highest range.


THx-RNMarsh
 
AX tech editor
Joined 2002
Paid Member
Need to think this through. If it just sits there, how does it know that there is not a small input signal that has to be amplified to the set value, rather than the user just wants it to wait for a bigger signal?

I'll check how that works with my dScope.

IIRC it starts up at the highest attenuation for safety, but eventually it looks at the output signal and if lower than the set value it cannot do much else than ramp up the gain, unless you select Hold mode and/or set a specific attenuation.

At any rate, the output level is limited to about 5V peak max to avoid damaging the sound card.

Jan
 
Need to think this through. If it just sits there, how does it know that there is not a small input signal that has to be amplified to the set value, rather than the user just wants it to wait for a bigger signal?

I'll check how that works with my dScope.

IIRC it starts up at the highest attenuation for safety, but eventually it looks at the output signal and if lower than the set value it cannot do much else than ramp up the gain, unless you select Hold mode and/or set a specific attenuation.

At any rate, the output level is limited to about 5V peak max to avoid damaging the sound card.

Jan

Thanks for your input.
 
A second thought on the topic:

In the current circuit implementation the AR has no chance to "react" to a small perturbation at the input, if attenuation is max in idle state. The AR needs a trigger, which could be derived from:

(1) a "signal-present" detection circuit, a comparator or similar, that monitors the Input voltage on a parallel path.

(2) If there would be a connection between AR and PC, the AR could be trigered at the moment the acquisition is started (i.e. in a playback + record situation).

(3) Same as before, but the soundcard output signal could be monitored in order to detect the acquisition start.

To conclude my proposal from post #428 was wrong.
 
Last edited:
AX tech editor
Joined 2002
Paid Member
OK, a few points. The AR does not have an idle mode as such. If there is no input signal (or very low), it ranges up the gain to get the nominal output, say 1V. With no input that means it goes to +12dB, not max atten.

The ranging slew rate is about 50dB/sec. If you change the input signal from say 1V suddenly to 20V, it takes a finite time to get to the new situation. It is possible that for a brief instant, the output is higher than the 1V, up to a (limited) max of 5V peak.

Note that the issues related to the max ranging slew rate are exactly the same as with any feedback control system. You want to get there as fast as possible, but without (too much) overshoot and certainly no ringing. The time constants involved are the ADC conversion time, the relay switching time and especially the RMS to DC converter filtering.

Jan
 
OK, a few points. The AR does not have an idle mode as such. If there is no input signal (or very low), it ranges up the gain to get the nominal output, say 1V. With no input that means it goes to +12dB, not max atten.

Yes, of course. I was referring to my previous post, where I was proposing "to do it the other way round".

The ranging slew rate is about 50dB/sec. If you change the input signal from say 1V suddenly to 20V, it takes a finite time to get to the new situation. It is possible that for a brief instant, the output is higher than the 1V, up to a (limited) max of 5V peak.

Note that the issues related to the max ranging slew rate are exactly the same as with any feedback control system. You want to get there as fast as possible, but without (too much) overshoot and certainly no ringing. The time constants involved are the ADC conversion time, the relay switching time and especially the RMS to DC converter filtering.

Jan

Fully understood. You did it perfectly right. From a system-level point of view its the right decision. (Just wanted to make this clear).
 
Last edited:
I have mine socketed but that was more for convenience using different opamps during development. If you use any, use good quality with turned socket pins. Some have internal pin gold plating but that is only valid for just a few insertions as it is very thin. But they are not expensive.

Jan



Understood and thanks. Also no need for the display/controller board U2 microcontroller socket? The construction guide refers to installing the socket first, but a socket is not on the BOM.


BK