REW Update

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

alexc

Well-known member
Joined
Sep 21, 2004
Messages
2,571
Location
Hobart
The fabulous Room EQ Wizard (REW) application from John Mulcahy and team  has a new update!

[  REW is the business for audio real time analysis, especially distortion measurement  ]

https://www.roomeqwizard.com

SOme interesting improvements and extras, and so far everything I have  is working great with it.

Thanks REW for services to audio measurement-kind  :)



 
+1 for REW as an audio measurement tool. The correlated white noise source is a boon for measuring frequency response. The distortion pop up show THD, THD+N and harmonic levels up to about 7th. Noise measurements are more or less direct and include A weighting if you want to indulge in specmanship.

Cheers

Ian
 
Of most  interest to me - the main RTA window  THD and Noise measures are more detailed,  and one can now set the BW of the thd measures.

Also, there  is now control of 'stepped sweep' from the RTA window, and the option of 'dBc' amplitude unit.

Plus much wider basic bandwidth .. up to 384K sampling rate - I think with the current generation of 120dB SNR audio interfaces out there, one can really go to town !

One day I hope REW adds :

- calibrate-able dBu unit for RTA amplitude
-  intermodulation distortion test measures
- a second 'measurement cursor' with the difference between them displayed ... like my Tektronix CRO  :)

Just a fantastic piece of software engineering.

 
I spent another half-day noodling with REW 

..  trimming up a simple 6SL7 in a variety of ways  ..  single gain stage with cathode follower, parallel'd gain stage, cascaded gain stages with/without nfb  etc.

You can really dial in the desired perf with quite a lot of dexterity ..  and compare with the data from the 'LoadSim' calc-er.

btw, I finally tested some of the russkie sl7 type against some modern tung-sol  ...  the modern tung-sol was streets ahead in hum performance  ...  didn't bother testing further  .. something like 10dB diff in hum , albeit at some little more dB gain.

I was surprised .. maybe I have some poor examples ... Also tested up some '6L6 equivalents' from the old soviet days  ... against some modern TungSol types.

Once again, no contest - this time mostly due to really poor match in so called 'matched pair'  ..  10mA diff!

A pair of modern TungSol 7581a  basic-match were less than 1mA different (single spud SE amp, 6K-8ohm 50mA output traffo).

The crapola sino SE traffos in my 'spudster test platform' crap out way before the finals, tho', so the poorer performers don't make much difference at the end of the chain. 'Its all good!'

....

And what a fab tube the  6SL7 is.

One of the greats.  I have a few variants and they can be really stellar, both vintage and modern (even more so).

I want to combine it with a modern hi-perf power amp module, just to see if I can preserve that beautiful spectra to a pair of 30W speakers.  Sort of a  6SL7-with-straight-wire-power kind of thing.

I have a tiny pcb module with that  super hi tech Texas Instruments 'amp driver chip+heatsink'  and a pair of power bjts onboard.

Just add a large heatsink, even a little fan, plus some high current bi-polar power rails and voila! Award winning audio  ;D

And just for kicks, a 6SL7 driving a JLH 10W bjt class A module  :)

. ..

Have to find how close I can get to that pure 6SL7 spectra at decent levels!

..

REW really shines a light on audio stuff for those with only half-days to noodle away!
 
Higher hum level is usually a sign of poor heater cathode leakage. Often occurs with age in tubes. One reason I do not buy NOS tubes any more. Sometimes it gets a little better if you run the tube for a day or so. Sometimes heater elevation works.

Regarding REW the one feature I want is to be able to set the zero reference level on the RTA graph to an arbitrary input level so I don't have to keep remembering to add 7.7dB or some similar figure to the noise measurements to get absolute values.

Oh, and automatic Q measurement of EQ plots - and Bode plots with phase response as well as amplitude.

Cheers

Ian
 
ruffrecords said:
Higher hum level is usually a sign of poor heater cathode leakage. Often occurs with age in tubes. One reason I do not buy NOS tubes any more. Sometimes it gets a little better if you run the tube for a day or so. Sometimes heater elevation works.

Good points Ian, thanks.

I had quite good results with the soviet era el84 equivalents, which did perform very well.

I guess, as always, one has to hunt for the perf.  I just can't resist the Russian tubes  :)

Before REW, the lens for selection was much coarser.
 
A key aspect is the on-going development and bug fixing - amazing for such a detailed app.

It is also worth checking on the rew forum about wish list improvements - it may seed a future addition, or get an informative response.

Btw, what intermodulation distortion function is lacking?  From a vintage perspective it provides the common standard techniques that were in use.
 
There is a facility in the sig-gen part of REW for multi tone generation, but as far as I know, there are not Intermodulation
Distortion calculations anywhere - IMD measurements are the only thing I still do in Rightmark.

Another fairly advanced thing I'd like to see are 'distortion residuals' ... 

There are a ton of features that I don't use - I most only use the sig-gen/RTA/sweeps.

 
IM is reported in the distortion window (for the given test signal configuration chosen in the tone genertor).  The distortion window identifies the harmonic residuals and there levels, and automatically identifies them on the spectrum plot.
 
trobbins said:
IM is reported in the distortion window (for the given test signal configuration chosen in the tone genertor).  The distortion window identifies the harmonic residuals and there levels, and automatically identifies them on the spectrum plot.

I didn't know that.Must give it a try.

Cheers

ian
 
Oh snap!

Thanks for the tip !

With the 'Dual Tone' signal selected in the Generator window, the RTA window's Distortion panel does indeed show Intermodulation Measures  :)

Exactly what I was looking for!


 
I ran up REW yesterday for the first time , superb for measuring the dist on my Hp200cd sine gen , with a few replacement components, Im now able to get  a thd figure of less than .5% which is on spec , frequency is rock solid and settles in 3 instead of the allowed 5 seconds . Amplitude was a bit up and down also , a clean of the output att pot seems to have improved matters a bit here .

I used the standard soundcard mic/line in , it was very easy and intuitive to set up ,  overdriving the sound card input with the oscillator harmonics up to about 9th appeared in the window ,as expected the noise levels were poor as the sound card can only accept a few hundred millivolts .

For anyone who wants too but hasnt tried it yet , get stuck in , you should be up an measuring stuff in no more than a few minutes ,between download ,install and set up . It reminds me a bit of Reaper in terms of intuitiveness and ease of use .
Only thing missing for me in it was no scope ,or at least I didnt find one . I reverted to YMEC audio tools for its scope abillity and sure enough the 200cd is cranking out a nice clean sine , the limitations of the inbuilt soundcard I used were very plain to see with anything over about half a volt causing the waveform to 'flat top '

I feel a bit stupid now for procrastinating so long before trying REW , Ive barely scratched the surface but the abillity to see how the proportions of various harmonics change with level is fantastic to have .

 
I used a scope app for a while, but a spectrum plot tells me so much more, that I don't get the urge to swap over much nowadays.
 
Theres no doubt about it the distortion panel in Rew gives great indication whats going on in your circuit , much more than a simple waveform on screen can say . the whole thing has me wondering if a digital scope is even nessesary , all the math functions you need are built into Rew ,  a simple old fashioned analog scope would do for waveform indication .
It seems to me the REW is much more versatile than just a speaker test program , its just as good with line inputs and using external signal sources ,as opposed to mic input , I cant help thinking the scope of this program is somewhat limited by its name, functionally its much much more .

Anyway big thanks to John Mulcahy , quite an Irish sounding name , I do hope the free(donation) business model is giving a return for all the work put in .

The other thing is I ran the program up on  single core 32 bit based windows 3 Gig ram , its works great despite the lack of number crunching power , Mac, Windows or Linux platforms means just about anyone can run it as well.
 
A scope will tell you things REW won't, like if your circuit is oscillating in the MHz region, what your square wave performance is like, if there is a dc offset on your output and so on. They are both tools that help you design and debug your circuits.

Cheers

Ian
 
Yes it is "horses for courses".  For valve amps, it can show parasitic feedback resonances coming out of the noise floor well before they would ever show up in a waveform.  And with 96kHz bandwidth, i reckon i would see anything amiss from an 807 acting as a radio tx, as something would fold back in to tha audio band.

But i get the scope out for X-Y low frequency phase shift and resonance effects - it can be fun trying to use various tools to gain a better awareness of some topic.
 
Tubetec said:
The other thing is I ran the program up on  single core 32 bit based windows 3 Gig ram , its works great despite the lack of number crunching power , Mac, Windows or Linux platforms means just about anyone can run it as well.

What OS are you running?

Bri
 
I routinely run a 'high voltage' probe as an input to REW ....

I did up a small alu box with a connection for the probe, and a parallel BNC connector to feed to my CRO. The box has a dc blocking cap feeding a high impedance network with a  6 way rotary switch to vary the attenuation of the probe ac signal.

So I can measure (unbalanced) ac signals of up to 450Vp-p both on the CRO directly and/or  in REW, with that signal suitably divided down so as to suit my soundcard's balanced input, a range of 0 to 10Vp-p or so corresponding to -130 to -3 dBFS in REW.

There is some loading effect by the probe of course, but it is reasonably minimal ..  I can check a signal's true value on the CRO by simply disconnecting the ac probe and using the standard CRO probe alone.

ANYway, the point is one really needs *both* a CRO and REW at all times!

The best thing about my modest Tektronix is the pair of measurement cursors ...  I like to set one cursor  to the bottom div of the graticule, the other cursor about half way up ...  dial the CRO sensitivity to approx lie within the two cursors and then quickly adjust to get an accurate reading of the CRO screen. 

Beats counting divs any day!

I'd like similar on the REW RTA .... one cursor is a line at the 'hum floor'  and the other at the 'return signal peak' with each level displayed in a 'cursor window' as well as the difference between them. 

I do it manually when I do my 'snapshots' of the REW RTA window ...  my app is 'SnagIT' for Windows  .... lets me annotate and draw lines and what not in it's editor ...

Like this ....
 

Attachments

  • API Sum Bus  - 1ch Routed SS balancing Out Rch baseline01.png
    API Sum Bus - 1ch Routed SS balancing Out Rch baseline01.png
    169.7 KB · Views: 48
Given the typical bandwidth limitation of soundcards, it does relax the need for top of the line probes. 
Even cheap 10:1 and 100:1 from fleabay perform nicely if their compensation range is acceptable. And even if they don't, it isn't too daunting to swap out the bulk cap typically hidden away inside the bnc end for a value that does allow flat response.  And even then, a simple calibration sweep could be used to 'iron out' grossly poor response as an easy way out.
 

Latest posts

Back
Top