Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Hilton]

Hi All,

One of my Velodyne CT100 Sub plate Amps (100W RMS / 250W peak) has been developing a more and more serious problem handling anything above -20db on my AVR.
It was making an occasional horrible clipping sound so I tried swapping the plate amps over and the problem moved to the other box with the amp so I was sure it was an amp problem.
I couldn't be bothered fixing it this time, as I replaced all CAPs 5 years ago and the subs are now 15 years old!
The drivers seem to be in good condition though.

The new AMP is a Jaycar 2 x 200W RMS into 4ohm Pro (cough) audio AMP designed for PA systems. If it proves itself I may not need to replace it.
I was going to go for a new Crown XLS-1502 if the Jaycar couldn't do the job.

I was able to calibrate my two 10" subs to exactly 75db to match the mains/surrounds pretty quickly with MC sound level calibration tones.
I've found my new SUB test track - Interstellar - where they are flying across the event horizon of the black whole near the end of the movie. The most awesome LFE I've ever heard!

With Interstellar - The new Sub AMP keeps up with my AVR set at -10db.

I measured in room sound level at over 105db during this section of the movie - which is almost too much for me and certainly way more than my wife would let me listen at if she was in the room.

I have the following setup:
Sub Amp input sensitivity set to .775v
Gain controls right up to max (0db) on the Jaycar Amp to get maximum output from it.
AVR set at -4db for LFE pre-out  (with +10 on the AVR LFE analog input)
Another -5db on SUB in MC21 room correction
-1db Sub Limiter in a PEQ
-48db High-Shelf @ 400hz Sub PEQ
 
I wanted some wise old JRiver advice on any particular setup precautions or advice based on the above settings as this is my first foray into active speaker amplification.

I haven't seen any clipping at -10db but the Sub amp clip lights start to flash at -7db.
The manual for my Sony STR-DA5200ES says the pre-outs are rated @ 2V which is either pretty rare or a miss-print if I understand correctly.
Most consumer AVRs only put out .775v on the pre-outs from my understanding.

Do you think I should switch the Sub AMP input sensitivity to 1.4V, looking at my settings above, I'm not sure I'd get enough output from the Subs with it set to 1.4V ?
I'll give it a try anyway.


The Amp Specs are reasonable for Sub duties:
2 channels of 200W RMS into 4 ohms Bridgeable to 400W RMS into 8 ohms
3 x switchable input sensitivities .775v/1v/1.4v
Ground switch - I needed to turn ground off to get a ground loop hum out the system - REALLY nice it has this option!
85db sig/noise (edit:errm not great but less important for Subs - (this is @ 1W so at rated output it's S/N Ratio is much better at around 105db - surely better than the Plate AMPs it replaced!)
1/4" Jack and XLR balanced inputs
Speakon connectors and binding posts
Toroid Transformer linear power supply with traditional A/B transistors as opposed to Class-D in the old plate Amps.
edit: oh and quiet fans! (its in my walk-in-robe with my gear and I cant hear it with the door open and it doesn't even get hot)

System sounds the best it ever has.

Thanks for any feedback!

[Hilton]

Just re-ran calibration @ 1.4v input sensitivity on the Sub AMP and was able to get same 75db output level on the Subs but with AVR LFE now set at 0db instead of -4db and MC is now at -2.5db on the Sub room correction instead of -5db.
So it would appear my AVR can drive the full 1.4V.
I guess that's the best gain setting to avoid clipping the input of the Sub Amp. Do you agree?

Amp closet by Hilton, on Flickr

Amp closet by Hilton, on Flickr

[mattkhan]

Modern AVRs/prepros can often put out a clean 7-8Vrms on the preout. To give an example, most Denon/Marantz kit is specified as 1.2V at "rated output" which actually means the voltage produced by a 0dBFS (sine wave) signal with master volume at 0 and the channel trim at 0. The LFE adds 10dB to this though so it becomes ~3.8V and then bass management can increase that further, potentially by another 10dB (which would take us to 12V and the analogue output into clipping) but more realistically by a maximum of ~7-8dB (which takes us to about 9.5V and still into clipping the analogue output). Hence, if using such an AVR for bass management, it will clip unless you run negative trims somewhere (either by not listening at reference or by dialling the SW down in the AVR and adjusting gain at the sub amp instead).

Whether your Sony has this issue is another question. I'm not sure whether the bass management bit applies either as you mention the use of jrmc room correction but also +10dB in the AVR. I thought using room correction in MC implied that it sorted out the +10 for the LFE so I'm not entirely sure what your signal chain is. Can you confirm?

What is this "-48db High-Shelf @ 400hz Sub PEQ" for btw?

[Hilton]

Whether your Sony has this issue is another question. I'm not sure whether the bass management bit applies either as you mention the use of jrmc room correction but also +10dB in the AVR. I thought using room correction in MC implied that it sorted out the +10 for the LFE so I'm not entirely sure what your signal chain is. Can you confirm?

What is this "-48db High-Shelf @ 400hz Sub PEQ" for btw?

Thanks for the feedback.
The +10LFE is done on the AVR LFE multichannel input. Normally this is done automatically for HDMI input but I'm using multi-channel analog into the AVR from the Asus Essence STX II 7.1 in the HTPC.

The AVR multichannel analog LFE pre-input requires the +10 to be added for Blu-ray and DVD sources.
The AVR can switch the analog LFE input between 0db or +10db. 10db is what's needed for the LFE from most sources.

MC obeys the industry standard +10db LFE rules by expecting the pre-amp to add the +10db for the LFE channel.  
MC also appears to sum the redirected bass from room correction to the SUB properly too as I've extensively tested the bass management in room correction with test-tones.

Regarding the high-shelf, I may not have done it correctly, but it's supposed to be a high pass shelf to filter anything above 400hz.
I noticed with a lot of blu-rays there's been a really slow roll-off filter applied to LFE or none at all. I don't want frequencies above 400hz getting to the Sub at all.
Let me know if this is not the right thing to do!

BTW I have no PEQ added to boost and cut any frequencies other than that high-shelf. jrmc Room correction passes mains and surrounds @ 60hz 12db/36db slopes to the sub, so that's summed with the LFE which is probably why I'm starting to clip when pushing the AVR to -7db. If I got a more powerful AMP for the Sub I'd get a little more head room or it might be the inputs clipping.
The manual doesn't say if they are input clip indicators.

The internal build quality of the AMP is pretty decent.
Untitled by Hilton, on Flickr

[Hilton]

Oh and there's no bass management or cross over available with multichannel analog input on my AVR. You have to do it all externally in the source. The only thing it will do with analog is distance and level and add the +10db for the LFE channel.
Hence doing bass management in MC.

[mattkhan]

On the high shelf, there was a change in a recent build that mentioned low pass filtering the lfe in some circumstances. It might be worth checking whether you still need it. If you do, why not just put a low pass filter in a peq stage after room correction to apply to the SW (output) channel? I agree with the need for it btw, there is certainly unwanted content on some LFE tracks that benefit from filtering away.

If you have a multimeter then it is easy to measure all this so you can see exactly what voltages are coming down the wire under different circumstances and hence be able to judge precisely whether (or when) you still clip the sub amp.

As an aside on bass management (just because i don't know exactly how jrmc does it as i don't use it)..... I have a similar setup except I do bass management manually. This means attenuating the mains by x db, the LFE by x+10, summing then adding x db back to the main channels. I then add the x+10 back to the SW feed in a downstream analogue stage. I am not sure how jrmc is doing this correctly if it leaves you to add 10dB back to the SW feed unless the mains are left x db down perhaps? Do you know how this is implemented exactly?

[Hilton]

The redirected bass from mains must be summed and then attenuated by 10db to work properly. Which I'm pretty sure is what mc does.
There's a diagram somewhere I saw recently that shows how it should be done.  I was doing extensive test-tones and cross-over slope testing with room correction and added a whole bunch of whacky filters to test if it was doing the right thing.  As far as I could tell it was, but the filters interact with one another in funny ways sometimes so it's hard to tell for sure.

filters by Hilton, on Flickr

Hopefully Hendrik can let us know!

I'll try find the diagram.

[Hilton]

I think MC is doing whats in the second bass management diagram in this link but I'm not sure.
http://forum.blu-ray.com/showthread.php?t=95817

[Hilton]

On the high shelf, there was a change in a recent build that mentioned low pass filtering the lfe in some circumstances. It might be worth checking whether you still need it. If you do, why not just put a low pass filter in a peq stage after room correction to apply to the SW (output) channel? I agree with the need for it btw, there is certainly unwanted content on some LFE tracks that benefit from filtering away.

This was only for down mixing I think in the output DSP.

[Hilton]

Just realised I think that MC should probably be attenuating LFE+Sub by 15db if bass management is active or only 10db if its just an LFE signal.
The further 5db attenuation is in the dolby specs for bass managed LFE+Sub.

Dolby Digital guidelines by Hilton, on Flickr

I'll play around with this tomorrow and see if I can cure the clipping on the Amp. In theory the Amp should easily have the power to run with 0db on the AVR with 2 10" Subs in my room. The original sub plate amps were only 100W RMS.  The LFE channel was only getting to -10db according to the MC analyser so I don't think MC is clipping the LFE output.  
I'll try dropping MC sub trim another 5db down and then add another 5db to the AVR Sub pre-out trim and see if that cures the clipping.

[mattkhan]

That is what I mean, you need headroom to sum all the main channels digitally (the x dB in earlier post). Typically this is 5dB  but you need a bit more to be completely safe with all content. There are some threads on data-bass that give examples, iirc the highest seen would require about 7-8dB headroom. Obviously such clips are pretty rare though so it can be valid choice to let clip protection do its thing instead (if you don't want to lose those few dB for some reason).

The clipping in the amp will likely be a gain structure issue though, do you know how much the asus puts out down the wire? https://www.pugetsystems.com/files/4962/parts/Sound-Card/ASUS-Xonar-Essence-STX-II-PCI-E-11566/E9122_Essence_STX_II_User_Manual.pdf says it put out 2V so if you had it peaking at -10 and you add 10dB in the sony then 2V would be arriving at the sub amp hence the clipping.

[Hilton]

That is what I mean, you need headroom to sum all the main channels digitally (the x dB in earlier post). Typically this is 5dB  but you need a bit more to be completely safe with all content. There are some threads on data-bass that give examples, iirc the highest seen would require about 7-8dB headroom. Obviously such clips are pretty rare though so it can be valid choice to let clip protection do its thing instead (if you don't want to lose those few dB for some reason).

The clipping in the amp will likely be a gain structure issue though, do you know how much the asus puts out down the wire? https://www.pugetsystems.com/files/4962/parts/Sound-Card/ASUS-Xonar-Essence-STX-II-PCI-E-11566/E9122_Essence_STX_II_User_Manual.pdf says it put out 2V so if you had it peaking at -10 and you add 10dB in the sony then 2V would be arriving at the sub amp hence the clipping.

Thanks I think you hit the nail on the head. Tomorrow I'll play with the trims as I mentioned above and throw a meter on to measure voltages to be sure.

[Hilton]

Just one further question for you boffins. Should the PEQ I'm using for high-shelf and Sub limiter be put in the chain after room correction or leave it where it is?

[mattkhan]

I would filter the lfe (input, i.e. before bass management) channel though it may not make any practical difference which way you do it as your proposed filter is operating well above your crossover.

I would put any limiter on the sw output channel (after bass management) as the total output of presumably what you want to limit not just one input channel.

[Hilton]

I've checked everything out and you were right there's a gain mismatch.

STX II outputs are rated at 2Vrms
Sony Pre-inputs are rated at 2Vrms
Sony Pre-outs are rated at 2Vrms
Jaycar AMP input set for max of 1.4Vrms to achieve AMPS max rated output with Gain set for 0db (max)


If I understand correctly, the Jaycar AMP doesn't have a high enough input gain or power output to get the Subs to a matched level with Sony AMP and the high sensitivity main speakers (96db) without clipping the output on the Jaycar.
I've discovered that most sub plate AMPS have input gain around 50-60db and the Pro Amps like the Jaycar often have much lower input gain.
So I'm clipping the Jaycar Sub AMP output??

I think that even if I dial-in some negative trim on all my mains in the Sony I'll still run out of Sub Amp output power at around 95db SPL in the room. Which still isn't too bad and as mentioned more than the misuss will let me listen at anyway.

What are your thoughts, anything else I can do?  I could bridge the Sub Amp and buy a second Amp and run it bridged with 400W RMS into each Sub?!!!??!  Or just live with a max of -10db on the Sony volume.

[mattkhan]

You don't need to match the input voltages across different amps, you just need to configure it so that it doesn't clip the input of the amp with whatever your peak source signal is. If that signal is 2V then you need to trim ~ -3dB from the SW somewhere along the line.

Once you've done this then that means you're getting maximum power out of the amp and that is going to result in a given SPL at the listening position. If this is less than what your mains can do then this will place a limit on the max clean output you can achieve and you will have to reduce the gain on the mains (either digitally or on the amps). If this is enough for your listening habits then you're all set, if you need more output then more subs and/or more amps are in your future Alternatively you dial in a limiter that compresses the sub output instead at the low frequencies to let the system as a whole go louder at the cost of some compression. The choice is yours!

[Hilton]

Here's the other problem!
This below is the same passage in Interstellar with bass management turned off. ie direct LFE signal.

Look at how much signal is going to the LFE.  Wasting all that Amp power on frequencies it just doesn't need to.
Interestingly the PEQ high-shelf doesn't seem to affect this??

LFE-no-room-correction by Hilton, on Flickr

[mattkhan]

I am still not sure why you're using a negative HS rather than a simple low pass filter. I've attached a comparison, green is the shape of your filter and magenta is a "traditional" LPF for the LFE channel (4th order at 120Hz). Ignore the noise at the end.

Your current filter rolls off the mid bass more than is arguably necessary and is a shallower slope too because jriver does not support a shelf with S>1 (and that would cause a ripple in the response anyway... note that what jriver calls Q for a shelf is really S but that's not really that important here, the important point is that it's not that steep a filter)

The amount of power used to deliver what you've shown is really v small in relation to the power required to drive the sub to significant excursion down low.

[Hilton]

I fixed my clipping by swapping the LFE high-shelf filter for a 48db low-pass @ 250hz.
Gained another 5db headroom on the LFE.
LFE signal dropped from -5db to -10 through that section of the movie and I can now turn upto -5db on the AVR before clipping the Sub amp. That's enough for me!

edit: snap..

[mojave]

Interestingly the PEQ high-shelf doesn't seem to affect this??

You have it turned off in your screen shot.

[Hilton]

You have it turned off in your screen shot.

Yes but it was making no difference on or off. I worked out it's just the way the built-in analyser rolls off after the cut-off frequency.
I installed SPAN analyser and it rolls off properly like Matts pic above. (edit: I realise it's a scaling issue)

Here's the analyser view in roughly the same point in the movie with a low pass 85hz 48db on the LFE and below that what it should look like in SPAN at about the same spot.
PS. I've also put a high-pass filter at 20hz w/ 48db on the LFE too now so I've hopefully got the full headroom to run at reference level now. (0db on the AVR and about 115db in room SPL)
Wont know till tomorrow when I can test again.


analyser-LFE by Hilton, on Flickr

SPAN-LFE by Hilton, on Flickr

[Hilton]

I can now run at 0 reference on the AVR with 110db avg and 120db peak SPL. That's plenty loud enough!  Damn it sounds good now!

I dialled in 5db negative trim for mains/surround in room correction.
I had to put a -3db cut at the crossover point of 60hz because i'm guessing the room correction crossovers are all butterworth. That returned a bit of head room too.

Are MC room-correction crossovers butterworth? If so, can I put it on the wish list to give us an option to tick a box to activate Linkwitz-Riley instead!

MC-crossover by Hilton, on Flickr

MC-PEQ by Hilton, on Flickr

[mattkhan]

Yes I believe they are butterworth filters (and that feature request is in https://yabb.jriver.com/interact/index.php?topic=99096.msg710907 fwiw), you can cascade 2 of them (in the PEQ block) to give an LR4 if you like

[Hilton]

Many thanks for your help Matt!!

Another AMP is in my future.  Then I can go with 400W RMS (bridged) into each Sub.

In thinking about it I decided I'd rather have neutral trims and turn the input gain on the AMP down, but this requires more powerful AMPs to reach the desired SPL.

The Amps are only $349 so it's a relatively small investment.

This way any signal source I connect will be set correctly.

The specs on the Amps are really quite good and they sound great. I'm impressed in the quality for the price.
Not the highest slew rate around but at 30V/us it provides pretty tight bass.
The damping factor is >300 which is pretty decent.  Certainly more than good enough for a Sub and way better than a lot of consumer gear.

[Hilton]

New extra Amp is in.

I took the opportunity to mount speakon connectors and run speakon cables instead of plain wire terminals.
Lots of head room now.    AVR and MC set for 0 trim allround and -7 clicks on the Amp input gains to attenuate the input.

Just checked the manual and the Amps are actually 450W RMS bridged into 8ohms.


Amps by Hilton, on Flickr


Back of an Amp
Amps by Hilton, on Flickr

Speakon on the back of Sub
Amps by Hilton, on Flickr

[Hilton]

I put a HPM master/slave (current sensing) powerboard in to slave the amps to the AVR. Amps turn on and off with AVR now.