I plan on doing some testing this weekend using 'trueRTA' based on a laptop....
(Yes, I know the soundcard is VERY CRAP but just experimenting/fooling round anyway....)
Never done this using a PC, what are the normal limits of signal input of a sound-card, do I need to make up an attenuator...?
Guess a soundcard wouldnt like tens of volts shoved up its mic socket...?
Any help appreciated....
(Yes, I know the soundcard is VERY CRAP but just experimenting/fooling round anyway....)
Never done this using a PC, what are the normal limits of signal input of a sound-card, do I need to make up an attenuator...?
Guess a soundcard wouldnt like tens of volts shoved up its mic socket...?
Any help appreciated....
Unless you're using a microphone, use the Line IN input. That should be able to handle a few volts (probably 3.3 Vpp). If your amplifier outputs more than that, you'll need an attenuator.
If you want to get more serious with the computer measurement thing, you'll likely need a better sound card (better dynamic range, lower noise, better distortion, etc.) But the built-in sound card is a fine place to start to get a feel for the process.
~Tom
If you want to get more serious with the computer measurement thing, you'll likely need a better sound card (better dynamic range, lower noise, better distortion, etc.) But the built-in sound card is a fine place to start to get a feel for the process.
~Tom
The normal limits are around 2v RMS for signal maximum on the digital scale. Typically there will be an opamp straight after the line input so the maximum theoretical voltage you could apply would the the maximum that the opamp can survive on its input pins.
Usually a single resistor placed in series with the output of the amplifier will work, this creates a resistor divider with the input impedance of the input stage and scales down the voltage output of the amplifier. The best thing to do is experiment with resistor values and see what works best, just remember to keep it turned down and turn things up slowly.
Also it bears mentioning that, even if plugged in, the laptop will have a floating ground. Make sure that a ground connection on the amplifier is also connected to the laptops ground.
A mic socket will also not really work for this type of measurement, you want a line input. If they share a common socket you can usually switch between mic/line in the sound setup somewhere, or specify which you want to use in the measuring program.
ARTA is another program suite worth considering for measurements, it's almost fully functional when freeware and comes with a lot of flexibility/options for measurements.
Usually a single resistor placed in series with the output of the amplifier will work, this creates a resistor divider with the input impedance of the input stage and scales down the voltage output of the amplifier. The best thing to do is experiment with resistor values and see what works best, just remember to keep it turned down and turn things up slowly.
Also it bears mentioning that, even if plugged in, the laptop will have a floating ground. Make sure that a ground connection on the amplifier is also connected to the laptops ground.
A mic socket will also not really work for this type of measurement, you want a line input. If they share a common socket you can usually switch between mic/line in the sound setup somewhere, or specify which you want to use in the measuring program.
ARTA is another program suite worth considering for measurements, it's almost fully functional when freeware and comes with a lot of flexibility/options for measurements.
I used ESI Juli for this purpose, input signal limit is around 5V if I remember correctly.
You have to be very careful not to damage your sound card input. Just build a 10:1 resistor attenuator, it will lower output voltage to safe level. Additionally, you will need 25 - 100W dummy load - wirewound resistor or potentiometer.
You have to be very careful not to damage your sound card input. Just build a 10:1 resistor attenuator, it will lower output voltage to safe level. Additionally, you will need 25 - 100W dummy load - wirewound resistor or potentiometer.
I tried this on an old IBM Thinkpad pIII 450Mhz. The built in sound was unusable. Looping line out to line in showed a bandwidth of only 8Khz. After that it looked like the ADC just stopped sampling.
If the laptop has USB 2 ports then I recommend looking at a cheap external soundcard. I bought a refurbished Creative card for under $50, and it does 96Khz 24 bit recording. I planned on breaking out the power wires in the cable and running the card from a linear supply but the noise floor is pretty quiet already.
As far as interfacing to the line in jack, Tubelab said he got the best results using a 500 ohm pot across the load resistor. Wish I could recall what thread it was in...
If the laptop has USB 2 ports then I recommend looking at a cheap external soundcard. I bought a refurbished Creative card for under $50, and it does 96Khz 24 bit recording. I planned on breaking out the power wires in the cable and running the card from a linear supply but the noise floor is pretty quiet already.
As far as interfacing to the line in jack, Tubelab said he got the best results using a 500 ohm pot across the load resistor. Wish I could recall what thread it was in...
I recommend making a dummy load with voltage adaptation to the sound card:
Described in this thread:
http://www.diyaudio.com/forums/tubes-valves/162214-distortion-measurements.html#post2101893
I also use ARTA.
Svein_B
Described in this thread:
http://www.diyaudio.com/forums/tubes-valves/162214-distortion-measurements.html#post2101893
I also use ARTA.
Svein_B
I think that level is more like 2 Vp-p. Sound cards can possibly do that but the on-board audio codecs run 5 V power. 2 V RMS is 5.66 Vp-p which impossible on a 5 V supply.
Also, be very careful about excessive input levels as the overvoltage is conducted out through the protection diodes. I had some load hum into an onboard sound system destroy the codec. I strongly recommend external clamp diode clippers to prevent destructive accidents.
Your resistive tree idea is good but I'd use my own second resistor to ground rather than relying on the input impedance of the codec. To keep noise down use the lowest practical impedance. For solid state I use 5-25 K loads. Hollow state glassfets will probably need higher impedance but that's up to you. I avoid toobs.
G²
Also, be very careful about excessive input levels as the overvoltage is conducted out through the protection diodes. I had some load hum into an onboard sound system destroy the codec. I strongly recommend external clamp diode clippers to prevent destructive accidents.
Your resistive tree idea is good but I'd use my own second resistor to ground rather than relying on the input impedance of the codec. To keep noise down use the lowest practical impedance. For solid state I use 5-25 K loads. Hollow state glassfets will probably need higher impedance but that's up to you. I avoid toobs.
G²
I saw this in Sci.Electronics.Design and it makes an excellent point.
The built-in audio systems are unpredictable; especially in the laptops.
Use an external USB sound thing; get a reasonably good one. The other
reason for using an external sound is that someone will eventually apply
full scale AC into it. "Psst-bam!" as the zurgs say. It is better to
"psst-bam" just an external card then (sic) entire laptop.
Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
AbVolt - DSP - Embedded - Consulting
G²
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.