Onkyo TX-SR3100 AVR 5.1-Channel used as XLR preamp - measurements

[Will P]

Hi everyone!

This is my first post here.
Many thanks to Amir for the hard work and detailed measurements and analysis in each review.
Many thanks to the rest of the ASR community for the feedback and info provided.

Onkyo TX-SR3100​

It's advertised as 5.2 but in reality it is 5.1 with 2 subwoofer connectors.
It has full capabilities on the video side - DOLBY ATMOS AND DTS:X; HDMI 2.1A WITH 8K/4K Ultra HD; HDR10+, Dolby Vision, and HLG;
gaming-optimized video processing Variable Refresh Rate (VRR), Auto Low Latency Mode (ALLM), and Quick Frame Transport (QFT)
On the audio side has 5.1, 3.1.2 and DOLBY ATMOS HEIGHT VIRTUALIZER (5.1.2), Bluetooth, Zone B pre-out

Price on sale $249 (reg $399)
It's #2 on Amazon Best Sellers list. It's been on sale for the last 2 months.

I added DIY balanced attenuators (1:6) to the speaker outputs and I'm using it as balanced XLR preamp.
My main system is running Yamaha CX-A5000 pre-pro so that would be my benchmark for comparison.
I'll be using the Onkyo TX-SR3100 mainly for music (70%), BD concerts 5.1 (20%) and movies (10%).
I'll be looking at distortion levels, input signal clipping levels, output signal clipping levels, power protection levels and hopefully I can figure out the logic behind the well known power protection in Onkyo/ Pioneer brands.

Below are my notes I took for my personal use while I did the tests right out of the box after the latest firmware update was done.
I didn't do any room calibration, etc. I'll be using it as pre-pro to my active speakers through MiniDSP.

The Onkyo is in daily use for a month already. I'm really happy with it. For me the combined audio&video price/performance is unbelievable.

For the measurements I'm using FiiO X5 digital music player (max output at player level "120" gives 3.2VRMS (9.0VPP) no clipping, level steps 0.5dB), oscilloscope FNIRSI 2C23T, RTA Behringer DEQ2496

Let me know if you need more info on the notes below:

==========================
NOTE: Input Level Limit: 2VRMS (level 112 from Fiio X5) would start clipping the AVR if NOT IN DIRECT mode - the ONKIO Line (CD) inputs MAX LEVEL 2VRMS (2.8VPP).
In DIRECT mode - the MAX output from FiiO X5 is 3.2VRMS (9.0VPP) - still no clipping on FiiO or AVR ! ! !
=======================
At 0dB maintained signal (noise, sine wave) at AVR the output would switch (relay clicks) to 16VRMS and clips output signal - distortion starts
(rail voltage reduced 40% -7.5dB from 38VRMS to 16VRMS).

KEEP AVR VOLUME LEVEL at -10dB or below to avoid overheating protection rail voltage reduction.

At 2VRMS input (CD) at Music Direct setting and (-10dB) AVR volume level the output at speaker terminals is 12 VRMS (34VPP) and
never goes into reduced rail voltage since output voltage is already below it ! ! ! (that's 20W into 8Ohm load)

NOTE: For getting Balanced XLR signal out from the AVR ONKYO Speaker Terminals the balanced attenuator should be 1:6
so that at line 2VRMS and output level -10dB and attenuator (voltage divider) ratio 1:6 the signal at the XLR output would be 2VRMS.
2VRMS(XLR target voltage) : 12VRMS (speaker terminal output voltage at -10 Volume setting) = 1:6

NOTE: For regular Music material the unclipped output goes to 40VRMS (110VPP).
With regular music (not sustained sine waves) that unclipped peak output is possible at 2VRMS input and 0dB AVR Volume.
============
NOTE: The max speaker output with clipping is 38VRMS (110VPP). XLR attenuator should be kept at 1:8 or more to prevent XLR signal
going above 18dBU (6VRMS); 1:6 att = 6.3VRMS; 1:8 att = 4.7VRMS absolute max at clipping; 1:12 att = 3.1VRMS = 12dBu (standard level)
============
NOTE: Output terminal when loaded with 4 Ohm speaker vs no load - the voltage drops from 17.4 VRMS to 17.0 VRMS (70W output on 1ch)
when one front channel speaker gets connected and no other speakers connected - solid rail voltage
===========================================================================================================================
NOTE: Speaker Output MAX before clipping SIN wave = 40VRMS (110VPP). That's peak of close to 200W into 8 Ohm speaker (160W/8ohm ch rated)!!!
That is at 0dB Volume Level at MAX input Line (CD) 2VRMS.
Thermal protection would bring it back to 14VRMS (40VPP) after 10 sec or so if signal is maintained unless the volume level gets reduced below 14VRMS.


NOTE: The ONKYO doesn't lock itself at thermal protection. Just reduce the volume -10dB and it resets itself 10-15 secs later.
It allows for temporary peaks of (40VRMS/14VRMS) = 9db ! ! ! It's all good. You will not see it during regular music playback.

====================
Latest Measurements:
ONKYO Input Line (CD) - max levels before clipping
Movie TV / Direct: Max Line Input 3.1VRMS (MAX output 120 from FiiO X5) - still no input clipping
Music / Direct: Max Line Input 3.1VRMS (MAX output 120 from FiiO X5) - still no input clipping

All other Listening Modes - 2.0 VRMS Max Line (CD) Input Level before clipping

======================

Max output speaker terminals (no load; signal to XLR adaptor only) at max input 3.1VRMS (12dBu) sine wave:

3.1VRMS in Line Input CD in Music/ Direct-

Volume -10dB; output 21VRMS - no thermal switching
Volume -8dB; output 27VRMS - no thermal switching
Volume -7.5dB; output 29VRMS (2.5dB below output clipping) - no thermal switching
Volume -7.0dB; output 31VRMS (2dB below output clipping) - thermal switching after 3min to 16VRMS (-12.5dB volume needed for no clip);
It needs to get Volume reduced at or below the new lower rail voltage (-13dB volume for 15.5VRMS) and resets itself within 10sec and full 38VRMS (peak) output capability is restored
Volume -5.0dB; output 38VRMS (MAX output before clipping) - thermal switching after 3min to 16VRMS (-12.5dB volume needed for no clip); It needs to get Volume reduced
at or below the new lower rail voltage (-13dB volume for 15.5VRMS) and resets itself within 10sec and full 38VRMS (peak) output capability is restored

When playing MUSIC - protection becomes extremely tolerant - it engages when MUSIC volume output level is at +9dB (that's 14dB above clipping speaker output power)

It looks like the internal overload circuitry engages above 30VRMS output (volume -7.0dB) at speaker terminals even when no load (speaker) is connected.
Above that level rail voltage gets switched to provide only 16VRMS and distortion starts (-12.5dB volume level to get adjusted at Onkyo for no clip)
and it needs volume reduced below the new lower rail voltage (-13dB volume to match) and it resets rail voltage for full peak output of 38VRMS (-5dB volume level) within 5 sec.

NOTE: At MAX Input Line Level 3.1VRMS the ONKYO maintains 30VRMS Speaker Terminals output for sustained levels and 38VRMS for not-sustained (less than 20sec!!!)
signals. That's sustained signals at (-2dB) below clipping - more than generous - great;
If signal is sustained above 30VRMS then protection reduces rail voltage to output 16VRMS (-12.5dB output level) and distortion starts until volume is
reduced to -13dB (no clipping with the reduced rail voltage) and full rail voltage is resorted within 5sec.

================================

Max output speaker terminals (no load; signal to XLR adaptor only) at max input 1.7VRMS (7dBu) sine wave - At that level 0dB Volume Level is just below clipping output:

1.7VRMS in Line Input CD in Music/ Direct-

Volume 0.0dB; output 38VRMS - thermal switching on sine signal after 30 sec - that's full output below clipping - voltage drops to 17VRMS (
Volume +18.0dB; output 40VRMS, 112VPP - heavy clipping - thermal switching on sine signal after 30 sec - voltage drops to 17VRMS - volume reduced to -5dB - thermal
switching resets itself and capability for full sustained not-clipped output sine wave at 0dB volume level at 39VRMS

When playing MUSIC - protection becomes extremely tolerant - it engages when MUSIC volume output level is at +12dB (that's 12dB above clipping speaker output power)


NOTE: When Line Input level (CD) is 1.7VRMS (7dBu) the thermal protection engages when output clipping occurs for 30sec (above volume 0dB setting) and resets itself
when volume reduced to -5dB.


=========================

Bottom line:
1. Music playing - I never managed to trigger Power Protection in real life while using the speaker outputs as XLR preamp - that is Power Protection based on output voltages alone.
( I like listening at reference levels with 122dB peaks).
Power protection based on output signal voltage doesn't engage until severe clipping is maintained (+12dB above clipping)
2. Power protection usually restores itself when signal lowered 8dB below clipping. For severe power protection case turning AVR Off/On restores power. No need to unplug AVR.
3. Direct Mode Noise levels at 1:6 XLR attenuator and 2VRMS input - noise highest RTA band is 10k-20k with level -120dB, the rest of the noise is -125dB. RMS noise level around -96dB or better.
If you use XLR attenuator ratio 1:20 the noise will drop further -10dB.
4. Stereo Mode (internal DSP with bass management) Noise levels at 1:6 XLR attenuator and 2VRMS input = -105dB RTA levels. RMS noise level is around -85dB.
If you use XLR attenuator ratio 1:20 the noise will drop further -10dB.
5. Direct Mode Distortion levels at 1:6 XLR attenuator and 2VRMS input - 2-nd harmonic at -100dB, 3-rd harmonic at -105dB. SINAD looks like is around -95dB.
6. Stereo Mode (internal DSP with bass management) Distortion levels at 1:6 XLR attenuator and 2VRMS input - 2-nd harmonic at -90dB, 3-rd harmonic at -95dB. SINAD looks like is around 85dB.
7. Heat - I can barely tell the Onkyo is on when I touch it. Idling power is around 25W. It warms up just a little more when running at reference levels.
The Yamaha CX-A5000 on the other hand is warm even when idling and doesn't have any power amps inside, just pre-amps.
Long term heat exposure kills electronics. Low heat output is very important for me.
Onkyo performs really well in heat output too. I'm really surprised.

Noise observation "ear against the tweeter" - AVR level at MAX (0.0dB which is capable of 122dB peaks sound level), digital player plugged in "CD" input, player turned off: No noise can be heard from listening position.
Direct Mode - no noise coming from speakers while against the speaker
Stereo Mode - slight noise heard when against the speaker; noise goes away when AVR output level reduced to -10dB or below.

I'm really happy with the AVR and really surprised by the measurements and actual performance. I bought it mostly for the video capabilities, I'm keeping it for the audio capabilities. $249

 

[Will P]

=========================

Bottom line:
1. Music playing - I never managed to trigger Power Protection in real life while using the speaker outputs as XLR preamp - that is Power Protection based on output voltages alone.
( I like listening at reference levels with 122dB peaks).
Power protection based on output signal voltage doesn't engage until severe clipping is maintained (+12dB above clipping)

2. Power protection usually restores itself when signal lowered 8dB below clipping. For severe power protection case turning AVR Off/On restores power. No need to unplug AVR.

3. Direct Mode Noise levels at 1:6 XLR attenuator and 2VRMS input - noise highest RTA band is 10k-20k with level -120dB, the rest of the noise is -125dB. RMS noise level around -96dB or better.
If you use XLR attenuator ratio 1:20 the noise will drop further -10dB.

4. Stereo Mode (internal DSP with bass management) Noise levels at 1:6 XLR attenuator and 2VRMS input = -105dB RTA levels. RMS noise level is around -85dB.
If you use XLR attenuator ratio 1:20 the noise will drop further -10dB.

5. Direct Mode Distortion levels at 1:6 XLR attenuator and 2VRMS input - 2-nd harmonic at -100dB, 3-rd harmonic at -105dB. SINAD looks like is around -95dB.

6. Stereo Mode (internal DSP with bass management) Distortion levels at 1:6 XLR attenuator and 2VRMS input - 2-nd harmonic at -90dB, 3-rd harmonic at -95dB. SINAD looks like is around 85dB.

7. Heat - I can barely tell the Onkyo is on when I touch it. Idling power is around 25W. It warms up just a little more when running at reference levels.
The Yamaha CX-A5000 on the other hand is warm even when idling and doesn't have any power amps inside, just pre-amps.
Long term heat exposure kills electronics. Low heat output is very important for me.
Onkyo performs really well in heat output. I'm really surprised.

8. Noise observation "ear against the tweeter" - AVR level at MAX (0.0dB which is capable of 122dB peaks sound level), digital player plugged in "CD" input, player turned off: No noise can be heard from listening position.
Direct Mode - no noise coming from speakers while against the speaker
Stereo Mode - slight noise heard when against the speaker; noise goes away when AVR output level reduced to -10dB or below.


I'm really happy with the AVR and I'm really surprised by the measurements and actual performance.
I bought it mostly for the video capabilities, I'm keeping it for the audio capabilities.
$249. Wow.

 

[Trdat]

I'm confused, why are the XLR's connected to the speaker outputs? I am not doubting your expertise but is this a tried and tested method of adding XLR outputs?

And just to make sure I understand, the XLR are for Pre amp outs?

 

[Hayabusa]

I assume the speaker outputs are symmetrical ('bridged')? Other wise your XLR attenuator could have quite some common mode swing which some input have trouble with (I experienced this issue and had to adjust my attenuator schematic) Can you share your attenuator schematic?

 

[Will P]

What I'm doing here is a very unorthodox approach.

I'm using the output stage of the power amp (the speaker terminals), attenuating the speaker signal (from 40VRMS down to 6.5VRMS) so it gets back into pre-amp signal specs and wiring it to a XLR connector. I connect the XLR connector to the rest of my system.
The drawback is that the noise of the power amp is added to the signal chain if that noise is significant.
In my case I measured the noise and it's around -96dB for Direct mode and -86dB for Stereo mode (internal DSP is being used for sub signal separation, tone control, sound fields).
Now if I would increase the attenuator ratio from 1:6 to 1:20 it would reduce the max pre-amp signal to from 6.5VRMS to 2VRMS (still enough for most systems out there) but it would further reduce the noise by another -10dB. That would give Direct mode -105dB noise and Stereo mode -95dB noise which would be even better.
I my case 1:6 ratio is low-noise already and I'd rather have that extra headroom voltage on the pre-amp. My system runs on 12dBu which is 3VRMS.
I hope that makes more sense now.

What I'm getting here is fully loaded video processing 8k capable system with DOLBY ATMOS AND DTS:X; HDMI 2.1A WITH 8K/4K Ultra HD; HDR10+, Dolby Vision, and HLG;
gaming-optimized video processing Variable Refresh Rate (VRR), Auto Low Latency Mode (ALLM), and Quick Frame Transport (QFT).
On the audio side has 5.1, 3.1.2 and DOLBY ATMOS HEIGHT VIRTUALIZER (5.1.2), Bluetooth, Zone B pre-out.

I modified it and now it has XLR pre-outputs with 6.5VRMS while the speaker output capability remains. That is that I can choose XLR pre-out or Speaker Outputs for each of the 5 channels.

For me that's a great and "powerful" tool. My other AVR was the top of the line YAMAHA CX-A5000 12 years ago and it still works great but dealing with it's huge size gave me PTSD. I was trying to hide the 20in deep box in the living room under the flat screen TV. All my other components were 11" deep or less.
I had to customize the equipment cabinets below the TV to make the YAMAHA fit. It was the size of midsize suitcase. And it's just a pre-pro, no amp stages.

So for my second system I decided to find a way to keep it no deeper than 12 in.

I was thinking about getting an AVR with low noise and low distortion and having distinct layout separation between AV processing boards and power amp section - like in DENON AVR-X6800H.
It's relatively easily splitable right in the middle - front section power amps and power supply, rear section - AV processing.
I was planning finishing the 2 boxes missing sides after I cut the box in half, place both boxes side-by-side or on top of each other.

Many many years ago in tech school I built my own integrated amp - phono stage, mixer stages, preamp, tone control, LED level indicators (very rare back then), relay controlled power supply, speaker output overload protection, 2x50W output. So I have some knowledge just enough to give me false confidence to start modifications and eventually burn something up.

Here is how I arrived at the current ONKYO AVR - it's only 12in deep ! ! -
Specs wise I expected the ONKYO to have better than average noise and distortion similar to Pioneer VSX-LX505 since they are similar models of the same manufacturer.

I decided to jump the gun and see what I can get out of it. When testing it I found that the specs were equal or better than I expected. The ONKYO was the AVR I was looking for.

And it was only $249. Unbelievable. Those AVR are produced in very high volumes which allows the price to be kept really low.


So yes, it was a fun journey

 

[Will P]

 

[Will P]

That above ^^^^ was my plan on how to cut the DENON AVR-X6800H in half, have it in (2) separate boxes and complete the wiring and cabling connections between the (2) boxes. This way each box depth would be 12" or less to fit the depth of the rest of my components.

 

[GXAlan]

Nice work. Do you have graphs that more easily summarize the info? What would be interesting is to see how your approach compares to a transformer approach like this:

SP-2SX | Jensen Transformers

The design begins with a rugged flanged enclosure that comes standard with gold plated XLRs and a removable screw-down barrier strip for easy installation in NEMA enclosures and 19” racks. Plug and play easy to use, this passive interface does not require any power to work. Inside are two Jensen...

www.jensen-transformers.com

While it's really hard to beat the price of the onkyo, I've thought about using a Sonos Amp with of these transformers to have an eARC streaming DAC with wireless surround sound support and stuff like integrated Apple Music. I'm using the Bluesound Node Icon instead, which is a lot of money for a small form factor eARC HDMI-CEC DAC.

 

[Will P]

I assume the speaker outputs are symmetrical ('bridged')? Other wise your XLR attenuator could have quite some common mode swing which some input have trouble with (I experienced this issue and had to adjust my attenuator schematic) Can you share your attenuator schematic?

The speaker black terminals are grounded. They are not balanced.

The attenuator is balanced to the XLR pin 1 (GND)

 

[Will P]

 

[Will P]

Nice work. Do you have graphs that more easily summarize the info? What would be interesting is to see how your approach compares to a transformer approach like this:

SP-2SX | Jensen Transformers

The design begins with a rugged flanged enclosure that comes standard with gold plated XLRs and a removable screw-down barrier strip for easy installation in NEMA enclosures and 19” racks. Plug and play easy to use, this passive interface does not require any power to work. Inside are two Jensen...

www.jensen-transformers.com

While it's really hard to beat the price of the onkyo, I've thought about using a Sonos Amp with of these transformers to have an eARC streaming DAC with wireless surround sound support and stuff like integrated Apple Music. I'm using the Bluesound Node Icon instead, which is a lot of money for a small form factor eARC HDMI-CEC DAC.

Thank you.
I don't have software, etc. to do real analysis screenshots but I can take pictures of the DEQ2496 RTA graphs.

I would stay away from any transformers since they are not linear devices.

Using passive liner devices (resistors) keeps the signal clean of any extra distortion.

 

[Hayabusa]

View attachment 429054

This will indeed give you trouble depending on the type of XLR input and the amount of common voltage it can handle.
You better make an asymmetric divider to ground (black):


red -10k--- hot
|
2k
|
black-------cold

 

[Hayabusa]

This will indeed give you trouble depending on the type of XLR input and the amount of common voltage it can handle.
You better make an asymmetric divider to ground (black):


red -10k--- hot
|
2k
|
black-------cold

seems the spaces do not show correctly in ASCII 'art'
the 2k needs to be connected to hot and cold

 

[Will P]

Noise at 0.0dB max output in Direct mode
(18dBu = 6.5VRMS max output; RTA 0dB is 22dBU).
For actual values subtract 4dB. That is reading of 120dB is actually 116dB.

 

[Will P]

1kHz at 0.0dB max output in Direct mode.
(18dBu = 6.5VRMS max output; RTA 0dB is 22dBU).
For actual values subtract 4dB. That is reading of 120dB is actually 116dB.

 

[Will P]

This will indeed give you trouble depending on the type of XLR input and the amount of common voltage it can handle.
You better make an asymmetric divider to ground (black):


red -10k--- hot
|
2k
|
black-------cold

Could you please elaborate some more and give examples?

 

[Hayabusa]

Could you please elaborate some more and give examples?

in the way you made it there is common mode signal on both hot and cold of approximately Vpeak/2. (so if the amp outputs 20Vrms the common mode will be sqrt(2)*20/2= 14 volt). Some XLR inputs can only have a common mode signal not exceeding some value, depending in its design. For instance I used an E-MU 404 XLR in to measure my amp via a divider like yours and there was early clipping because of this common mode effect.
Changing it to asymmetrical solved the issue.

 

[peng]

View attachment 429049

The 6800 has full set of pre outs and flexible pre amp mode already, how and why you want to modify it? Would the goal be to have XLR pre out?

 

[Will P]

Splicing the 6800 would be just to reduce the size of the enclosure(s) to around 12" deep.

Otherwise 6800 is fully loaded unit and an awesome piece of equipment.

 

[Will P]

in the way you made it there is common mode signal on both hot and cold of approximately Vpeak/2. (so if the amp outputs 20Vrms the common mode will be sqrt(2)*20/2= 14 volt). Some XLR inputs can only have a common mode signal not exceeding some value, depending in its design. For instance I used an E-MU 404 XLR in to measure my amp via a divider like yours and there was early clipping because of this common mode effect.
Changing it to asymmetrical solved the issue.

Yes, you are correct about the common mode voltage limits on XLR inputs and the clipping.

I'm attaching a new picture of the same schematics but with resistors in line so it's easier to see and analyze the potential divider.

I my case the equipment inputs are rated for 22dBu ( 9.75VRMS) and I never had clipping problems feeding full AVR output into the components downstream.

In the RTA picture above I'm feeding the DEQ 2496 RTA with 6.5VRMS and still no clipping. The DEQ 2496 inputs are rated for 22dBu (9.75VRMS) and the ONKYO max XLR output after attenuation is 18dBu (6.5VRMS). At full output from ONKYO the max level on the RTA would be at -4dB. That's why I mentioned that the RTA charts above should have actual values reduced by 4dB. If noise level reads -120dB, the actual reading is -116dB

I case of clipping that you mentioned I'd suggest 1:21 attenuator (10k - 0.5k - 0.5k - 10k) which would provide 39VRMS / 21 = 1.85VRMS.
Most of the components out there would be rated for 2VRMS XLR so there should not be a problem with clipping.

In that case going from 1:6 to 1:21 attenuator would reduce the XLR voltage from 6.5VRMS to 1.85VRMS at full AVR output but it would also reduce the noise level by another -10dB which is great. Noise levels -95dB before now would be -105dB.

That 1:21 attenuator would be (10k - 0.5k - 0.5k - 10k).