Frequency response of AVRs?

[Hakka]

MY equipment:

Denon 3803 receiver
Denon DVD 2200 dvd player
BFD1124p
Behringer A500 amp

I use a coax connection from the dvd player to the AVR.

If I connect REW directly to the BFD input I get a flat response down to 10hz.
If I connect through the 3803 I get a shrp rolloff at 15hz.

I was looking at the thread over at AVS with the Spectrum Labs waterfall plots of bass heavy scenes, these measurements are taken from the sub out jack on the receiver, they are all showing response down to single digit frequencies. There are a lot of different people doing these plots on lots of different sytems. I have also seen numerous posts here showing AVRs going well below 15hz.

I had a look at the specs of my Denon equipment. The DVD player freq response goes to 2hz, the receiver is listed at 10hz for the analog section (in direct mode). I tried direct mode (I normally use Stereo) and still got the sharp rolloff at 15hz.

Unfortunately I have no way of running a REW sweep in DD/DTS mode so I can't tell if it any extends lower, my soundcard does have an optical out so I could try a digital connection to the AVR. The only other thing I can think of is doing the same Spectrum Labs plots as the other guys and see what's missing down low.

Any other ideas on what could be causing this?

Hakka.

[brucek]

How have you taken the measurement with the receiver?

brucek

[Hakka]

Using REW, the soundcard is connected to the left and right analog CD input on the AVR, sub out to the BFD.
It makes no difference if the AVR is set to stereo (bass management) or direct (no BM).

When I measure the sub only I disconnect the soundcard input from the AVR and connect it to the BFD, so the only difference between the two measurement is the addition of the 3803 into the signal chain.

I use an ECM800 mic to take the meausement. With the 3803 in the loop it always rollsoff at 15hz, when I bypass the 3803 it extends to 10hz.

Hakka.

[JohnM]

The roll off you are seeing will most likely be in the analog input stage of the 3803.

[Hakka]

I tried some waterfall plots with Spectrum Labs and it showed output from the AVR sub terminal down to 1hz on DD and DTS DVDs.
I will try the optical connection with REW and see if that gives the same results.
It seems Denon's specs are wrong for the analog section of the AVR.


edit: we posted at the same time John, you are correct.

Hakka.

[MACCA350]

I Measured my Denon 3805 a while ago, but didn't use the mic, instead I hooked it up like this:

PC--(USB)-->Tascam 144--(RCA)-->line in,3805,preouts--(RCA)-->Tascam 144--(USB)-->PC

So it's basically a line level loop through the 3805. I loaded the previously measured soundcard.cal file for the Tascam 144 and ran the measurements. With the Eq set to None, this is what I came up with:

-1db points at 6.1Hz and 20.4kHz
-3db points at 2.9Hz and 20.7kHz


One thing I didn't do is set the 3805 to Direct or Pure Direct, I'll have a play around with it tomorrow.

Hakka, Try hooking your's up like this and see what measurements you get.

cheers

[brucek]

Yeah, I agree with MACCA350. You need to test the receiver by itself to get an accurate read on what is going on. I'm surprised that the analog input (or analog output) would have such a poor low end response.

Hook up a cable to the REW line-in and another cable to the REW line-out. Then connect those two cables together to create a loopback from line-in to line-out just as if you were doing a soundcard test. Now do a REW response measurement on that loop and ensure that it is perfectly flat from 2Hz to 200Hz. It should be, because the soundcard cal file should force it to be so when you are taking a measurement of a cable.

Now insert your receiver into the loopback and take an REW measurement. Since we know that the cable by itself is perfectly flat, then the response we get by simply measuring the receiver will be an accurate reading of its response. All you need to do is adjust the receivers volume control to get the REW levels set up. Use the CD or AUX input and the subwoofer output of the reciever.

I did the response of my processor and get the result below. You can see it tracks the REW crossover target and then drops off a little at the lowend. I did both the analog bypass plus the analog path with the ADC/DAC engaged. You can see the analog bypass doesn't provide any bass management, but gives a bit better low end response.





brucek

[MakeFlat]

Hakka wrote:

Using REW, the soundcard is connected to the left and right analog CD input on the AVR, sub out to the BFD.

Hakka.

If we don't asssume that all analog inputs were the same, we would certainly test the aux input just to find out if it differs from the CD input.

[brucek]

If we don't asssume that all analog inputs were the same

They would all be using the same input amplifier. The different jacks are simply switched to it..

brucek

[MACCA350]

Bruce, can we all set our graphs the same for this particular test to make comparisons easier? Also I used 1/24 smoothing.

I used:
top: 6
bottom: -6
left: 2
right: 30,000

Any other suggestions welcome. It'll be interesting to see if much variation between receivers.
I think I only ran the sweep up to 20kHz, I'll redo the test in the morning.

cheers

[brucek]

I only took my measurements up to 200hz, since I was interested in the subwoofer chain.

I wanted to see what the processor and BFD did to the signal before it got to the subwoofer. After I did the processor loop and BFD loop separately, I added the BFD and did the processor plus the BFD.

I adjusted my scale to -6 and +6, but again I only did the measurement to 200Hz. Some day when I get set all up again I'll extend it out to 30K. I also show the BFD and my processor plus BFD below. No smoothing..... Note the measurements of this kind are only accurate if the cable loop by itself is perefectly flat....



brucek

[Hakka]

Thanks guys, I will give this a try tomorrow.

[MACCA350]

Bruce, I found out that the drop off just after 20kHz was already there at the soundcard output. So there is no point testing above 20kHz. I also found that there was no difference between 'Stereo', 'Direct' and Pure Direct'

I started a thread here for comparing graphs since this has moved outside the scope of this thread

Hakka, let us know what you find.

cheers