G9 and its frequency response ?? (impedance matching?)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

matthias

Well-known member
Joined
Jun 4, 2004
Messages
768
Location
germany / frankfurt
Hi,

I finished my G9 a while ago and today I was coming back to it...

I made some tests with the rightmark audioanalyzer tool
(http://audio.rightmark.org/products/rmaa.shtml)
regarding the frequency range...

I have used OEP-trafos at the input and lundahl at the output..


here's a test using the xlr-in (signal gets thru the oep input transformer)
it seems that OEP causes a slight shelf to the high frequencies


edit : sorry I don't have this pic anymore...

and here one test without the oep, thru the jack at the front...

edit : sorry I don't have this pic anymore...



of course the low cut switch is in its off-position
and I did the test on both channels, they behave exactly the same...

maybe someone of you with a G9 could do this test...



thanks,
mat
 
you test may be correct

BUT
before making any comments and as this is a Software/Blaster Tester
it would be nice to see a couple of basic self tests first.
Feq plots for just the tester and cables at a couple of levels AND loads.

The G9 test didn't say at what level or if there was a load on the G9.
 
whoa!

That low end roll off certainly looks like something is wrong..

dosent it !

Surely, it must be a problem with your particular unit?

I cant imagine it being correct for the G9, with all the people that have said how much they love it for bass ! :shock:
 
How are you driving the input transformers? Your soundcard sure can´t drive it directly. You have to use a pad network. One that will both attenuate the signal AND make it appear as a 150ohms source to the input transformer primary...

Try two matched 22k resistors in series with the signal lines (+and-) and a 150ohms resistor shunting across...
 
for testing I only connected the input and outputs to my rme multiface soundcard... its output impedance is 47r... input is greater than 5k

the input level for testing was -8db and the test signal a frequency sweep
 
I've done the same tests with my G9... and I achieved almost the same results...

I've used Lundahls on input and output, and i don't have that ramp on high frequences, but almost the same roll-off starting at around 300Hz.

I'll try to test it again following rafafredd suggestion...
 
ok... I need to match the impedance of the soundcard output to the impedance of the input of the G9

1) I always thought, if the output impedance is lower than the input impedance, everything should work allright...


2) is there a simple way to measure the real impedance of the unit??

3) is there a formula, to calculate the needed resitors for the pad..??



??
 
but what about the deficit in the bass... that's not ok, isn't it??

Hmm You the user has to be the judge of that I mean ! What is the "norm" or the "standart"..does it sound good to your ears or not..that is the real question to be asked imho *S*

If you think about rulerflat freq responce 20Hz-20Khz or more..then it would not be okay, but again as I said..how does it work and sound to you in your studio..??? does it make you and your clients happy or sad ???
in the end its all that matters...not freq responce chats, THD+N..IM etc..it´s funny to look at..though *S*

Kind regards

Peter
 
I want to use the G9 for my passive pultec to make-up the lost gain...

When I had run some tests, the bass boost pot of the pultec was always at its full clockwise position... this made me thinking that there must be something wrong...


Viitalahde said, I don't have to match the Z,
but what's the technical premise to get a nearly flat response??


thanks,
mat
 
no... forget about the pultec...
at first I wanna get sure, that the G9 works properly...

for the tests above, I connected the G9 only to my RME converter...


btw: my pultec has a 600:600 input transformer and no output transformer.. I want to go unbalanced into the G9 ... I want to use both as an
"external instrument" in cubase sx, so I have to connect them to my rem converter...
 
Have you tried the tests without the input transfomer, using the instrument input?
 
I missed the note on the 2nd diagram.

The curve is kind of hard to read but it looks like more than one roll off because it looks to be greater than a 6 Db per octave.
 
What's the big deal? Looks like every step up transformer curve I have ever seen. Bass rolling off badly delow 10 hz, resonant peaks from 10K and up.
Do you have a singer ( or a mic) that can hit 10 hz? If so, sign him up! (or her, God forbid!)

You might try bridging your coupling caps with something larger, if you really want a mic pre trhat sounds like a Marantz stereo system, but is that what you want?
 
It is sometimes hard to get an accuate sweep at the extremeties of the audio range due to anomalies associated with the test equipment.. At the high end, the scope probe is important. For instance, if you set your scope probe to 1X and do a sweep, then set it on 10X and do the same sweep, you will sometimes get two different sweep charts.

With the low end, you have two problems. The generating source, and the equipment looking at the output. A signal generator might go down to 1 hz, but it's waveform might suffer dramatically depending on what it is driving and what kind of generator it is. This is why Jensen and others use a synthesized waveform to measure low end response.

You might have capacitors in the signal path of a generator that can form resonant circuits with the transformer primary of the piece of equipment you are measuring. This can cause a rise in the freq chart at the low end, which in most cases should be a sign that something is awry. I doubt that there are transformers out there with enough capacitance to form a 10 hz resonance with the inductance.

Most scopes have an option for AC or DC coupling. With AC, you are obviously putting a large capacitor in series with the scope probe in order to block out any DC offset. This cap can roll off your low end when you are doing sweep tests.

A mic pre is actually a musical instrument amplifier. Sometimes this means that better results can be had with a singer if the frequency response is limited at both ends, and perhaps tilted a bit in the middle. If you look at the frequency response of microphones, you will soon realize that the microphone is rolling off both ends way before the 20-20K endpoints, and with other dips and valleys in between.

OK, so lets say your charts are perfect. Just start at the beginning of the circuit and work your way forward until you find the culprit who is stealing bass. A transformer that is built for a mic level signal at low frequencies will probably crap out if you hit it with today's CD levels.

I would try sweeping the G9 at different gain settings to see if your charts stay the same. Maybe playing with some of those feedback caps will help out.
 
[quote author="CJ"]I would try sweeping the G9 at different gain settings to see if your charts stay the same. Maybe playing with some of those feedback caps will help out.[/quote]

correct
BUT
as I said in my first reply

test the test gear first
and
indicate what level the test is at
:shock:

say -60dBu, -40dBu, -20dBu and 0dBu
use just the TWO balanced connecting cables ... with and without a load ...
change the load and see how it effects the test gear.
 
Back
Top