Audio signal generator and (FFT)scope

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

PGDO

Member
Joined 2010
Hi all,
I am planning to design and build my first tube amplifier in 2011, probably with an EMS 300B-XLS tube.
I think that before I start this project I should first have the necessary test equipment. I currently own an old Fluke scopemeter (PM93).
What I think I need is a audio signal generator (distortion <= 0.05% ??) and a scope with FFT analyzing capabilities.
It should not be to difficult to buy a secondhand function generator online on one of the auction sites.

I am just not sure what the minimum requirements for a scope should be to be useful and still affordable.
Does anybody know if a TiePie Handyscope HS3 (5MHz) is sufficient, it has a build in function generator. USB oscilloscoop - Handyscope HS3 - TiePie engineering
Or the PicoScope 4226? PicoScope Precision USB Oscilloscopes from Pico
Would the AWG of above mentioned PC scopes be suitable for distortion measurements?
regards,
Peter
 
P

phaselag

Member
Joined 2010
Most scopes that have FFT capabilities, are not very good FFT analyzers. They will show basic spectra, but usually detailed measurements are not available. A good FFT analyzer will give you THD and accurate response data. As you noted, to go along, you will need a good audio generator with low distortion. Typically function generators are not good enough, so look for a good audio generator.

paul
 
P

pieter t

Disabled Account
Joined 2009
tribute-audio.nl
Hi Peter,

The world is small. I have lived in Zuidhorn for 15 years, now in Groningen since 2008.
Do you know Pete Millet's kit? Look at his website and see what you can do with it; I have the complete kit here unbuild (too busy winding transformers). When you're interested and have the skill you are kindly invited to assemble the unit, and use it of course.

Verslik je niet in de oliebollen!!
 
P

PGDO

Member
Joined 2010
nevermind said:
Scope fft's are useful, but a 12bit scope has about 72 db range, which is not so good for distortion measurements. I personally use a 96 khz sampling onboard sound card. The noise floor is at about -120, so distortion is easy to see.

Also found RMAA (http://audio.rightmark.org/download.shtml) very useful.
Click to expand...

Hi Nevermind,

TiePie claims 16 bit measurements (95db) which should be sufficient (have a look at their application note)

The thing with soundcards is that I don't thrust them that much as they are working in a noisy environment. Besides I read somewhere that their sampling rate is to low?

Also an concern is the high voltages used in tubeamps and the possibility of damaging the soundcard.

Thanks,
Peter
 
P

PGDO

Member
Joined 2010
mickeymoose said:
Check the free Visual Analyzer software at Visual Analyser 2011, combined with an E-MU0404 USB soundcard and a Behringer ECM8000 microphone (calibrated by Cross·Spectrum Labs - Sound | Vibration | Engineering) will get you a lot for little money! E
Click to expand...

Hi Mickymoose,

I had already installed the Visual anlyser software, nice program. Used it as an signal generator via the soundcard which is working for simple tests but not for serious distortion measurements.

As for the soundcard option, see my comments on the previous post from Nevermind.

I do already have an Behringer ECM 8000 mic with pre-amp but I want to be able to take measurements directly from various points in the amplifier..

Thanks,
Peter
 
P

PGDO

Member
Joined 2010
phaselag said:
Most scopes that have FFT capabilities, are not very good FFT analyzers. They will show basic spectra, but usually detailed measurements are not available. A good FFT analyzer will give you THD and accurate response data. As you noted, to go along, you will need a good audio generator with low distortion. Typically function generators are not good enough, so look for a good audio generator.

paul
Click to expand...

Hi Paul,

You're talking about a dedicated FFT analyzer, although probably the best solution it will likely be above my budget. I was hoping that a PC-scope with the right (FFT) software would suffice.

Thanks,
Peter
 
P

PGDO

Member
Joined 2010
pieter t said:
Hi Peter,

The world is small. I have lived in Zuidhorn for 15 years, now in Groningen since 2008.
Do you know Pete Millet's kit? Look at his website and see what you can do with it; I have the complete kit here unbuild (too busy winding transformers). When you're interested and have the skill you are kindly invited to assemble the unit, and use it of course.

Verslik je niet in de oliebollen!!
Click to expand...

Hoi Pieter,

Indeed a small world.
I had a look at Pete Millit's site and the thread about it on this forum, it sure looks interesting.

Please mail or PM me so that we can communicate more directly.

Jij ook voorzichtig met de oliebollen en vuurwerk, alvast een gelukkig nieuwjaar gewenst.

Peter
 
J

Johnny2Bad

Member
Joined 2010
PGDO said:
Hi Nevermind,

TiePie claims 16 bit measurements (95db) which should be sufficient (have a look at their application note)

The thing with soundcards is that I don't thrust them that much as they are working in a noisy environment. Besides I read somewhere that their sampling rate is to low?

Also an concern is the high voltages used in tubeamps and the possibility of damaging the soundcard.

Thanks,
Peter
Click to expand...

There are a number of issues with sound cards in a computer that simply do not exist with standalone, purpose-built test equipment. Not the least being computer companies and test equipment companies have very, very different attitudes when it comes to specification; if you assume something you may be very disappointed.

For example, there are many, many 24/192 sound cards, internal and external, that can only play back at 192Khz, record (ie your input for your test suite on the computer) is limited to 96Khz. Some have DC at the outputs. Some have poor IM distortion performance or the S/N figures are only valid with a single sine wave and fall with any other complex signal. Reducing level from 0dB means reducing bit depth. I honestly don't know of any 16-bit cards that can resolve more than 12-13 bits maximum; the rest is lost to noise or other issues. Some resample all data, even if you don't tell it to. And so on.

Broadly speaking you will not find good comprehensive specifications on the card from the manufacturer; the tendency is just to say nothing if there is any issue at all. You will have to hunt down tests online performed by competent, knowledgable users (which I encourage you to do) who are much rarer, whenever the word "computer" is involved, than many of us would like, to discover these limitations.
 
Last edited:
qusp

qusp

Member
Joined 2009
indeed sound cards can be an issue if PSRR and housing is an issue, but there are many external soundcards that will allow external power supply, as well as some computers and internal cards that have been well implemented, like my RME9632 for example. then you have the millet solution, which I have also had my eye on, looks very good
 
J

Johnny2Bad

Member
Joined 2010
qusp said:
indeed sound cards can be an issue if PSRR and housing is an issue, but there are many external soundcards that will allow external power supply, as well as some computers and internal cards that have been well implemented, like my RME9632 for example. then you have the millet solution, which I have also had my eye on, looks very good
Click to expand...

I agree; I'm an RME owner myself.

But with such a card we are now getting to a point where it's no longer a cheap option, or at least, where standalone equipment, carefully selected in the used market, should be reconsidered value-wise.

My comments were intended to make people aware that you need to proceed with caution and research your interfaces carefully. Many people in the computer field make too many assumptions and view the hardware as completely interchangeable when in fact it does matter. Perhaps worse, I would not want someone to buy something thinking it met their needs only to find later it did not.

I hate to say this, but I believe that sooner or later there comes a time where you need to use both. Purpose-built gear tackles measurement uncertainty with careful implementation of solutions and the confidence of repeatable measurement procedures ... there is a consistency that is valuable. Computer based solutions are complex and there is always the risk that some thing you didn't know about the setup can render previous measurements uncertain ... but the software is capable of tremendous flexibility.
 
Last edited:
P

PGDO

Member
Joined 2010
The Pete Millett soundcard interface

This is the final update to let you know what solution I have gone for.

Pieter from Tribute audiotransformers had a soundcard interface from Pete Millett which I assembled for him. He let me use the interface the past couple of weeks, thanks Pieter.

I have used the interface together with an E-MU 0404 USB soundcard and a couple of measurement programs via the ASIO interface. For most purposes I found the ARTA program the best.

It takes some practice to get good measurement, especially to have a setup and measurement which is reproducible.
All in all a good solution and I decided to buy the interface pcb from Pete Millett for myself.




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


The Pete Millett soundcard interface.

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


Arta loopback measurement via the soundcard interface.

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


A measurement at 1 kHz 1 watt of an amplifier.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom