MiniDSP + Denon x3700H signal clipping.

[KurantAK]

Something I'm a little confused about.

I followed HTG MiniDSP tutorial as he said. Mainly, the idea that level matching the subs to 75db is what throws me off. Starting, I set the sub gain in the AVR itself to 0, turned Audyessy off, and then to start calibrating with REW he says to adjust the volume with the main AVR volume to 75db on the SPL meter in REW, and then start your sweeps to calibrate with REW.

I did that, everything went as it should. But my room is 16x24. I like to listen to my movies loud. I recently build 2 18" LaVoce Marty's with an NX6000D. I went thru the calibration, but when I turned it up unknowingly at that point, at high volumes it reached a point where I'm guessing it was input clipping, or perhaps even just reaching a point of distortion it made a awful sound in the woofer itself. I thought they were blown or ruined. I really had no idea.

I unhooked the DSP and went straight to Audyessey, loaded my old calibration, and the noise went away at high volumes. Hooked the DSP back up, the noise came back. I started reading a bit and noticed people talking about input clipping, so I started lowering the gain in the AVR itself.. Went to -8, turned up the gain on the amp and in the miniDSP software - I can now play at any volume I want, turn the subs up to any level I want and the noise is gone.

What causes what I'm guessing, is the input's clipping? Is it not enough voltage in the signal? Just a point where is distorts?

 

[lovinthehd]

Who is HTG? Sounds like in the end you lowered the signal, so it was too much. My older miniDSP 2x4s (unbalanced/balanced, pre-HD models) you had an input voltage sensitivity adjustment you could make by flipping a pin switch/jumper. What's your miniDSP model and what setting for input voltage is it set to (if it has such)? IIRC on my old one it was a choice between max input of .9V or 2V on the unbalanced and a choice of 2 or 4V on the balanced model....

 

[TLS Guy]

Something I'm a little confused about.

I followed HTG MiniDSP tutorial as he said. Mainly, the idea that level matching the subs to 75db is what throws me off. Starting, I set the sub gain in the AVR itself to 0, turned Audyessy off, and then to start calibrating with REW he says to adjust the volume with the main AVR volume to 75db on the SPL meter in REW, and then start your sweeps to calibrate with REW.

I did that, everything went as it should. But my room is 16x24. I like to listen to my movies loud. I recently build 2 18" LaVoce Marty's with an NX6000D. I went thru the calibration, but when I turned it up unknowingly at that point, at high volumes it reached a point where I'm guessing it was input clipping, or perhaps even just reaching a point of distortion it made a awful sound in the woofer itself. I thought they were blown or ruined. I really had no idea.

I unhooked the DSP and went straight to Audyessey, loaded my old calibration, and the noise went away at high volumes. Hooked the DSP back up, the noise came back. I started reading a bit and noticed people talking about input clipping, so I started lowering the gain in the AVR itself.. Went to -8, turned up the gain on the amp and in the miniDSP software - I can now play at any volume I want, turn the subs up to any level I want and the noise is gone.

What causes what I'm guessing, is the input's clipping? Is it not enough voltage in the signal? Just a point where is distorts?

No, I suspect this is an error due to inexperience. Those LaVoce 18" sub drivers are high efficiency pro sub drivers. They are NOT home HT drivers.

They seem a poor outfit, who only show very limited specifications. However, they are high efficiency driver with 96 db. sensitivity. Now there is an axiom in speaker design, that you can have sensitivity or bass extension but not both. For home HT people look for bass extension and in the pro world sensitivity.

Now, what I suspect is happening is that the DSP is trying to extend bass extension. In a tuned ported enclosure, if you try and extend the bass by adding Eq, you can't as the driver decouples from the box, and useless over excursion of the driver results. Then you get bottoming if the VC which is the sound you heard, and if you keep it up you will definitely ruin the drivers.

Unfortunately you went ahead without posting here and taking advice, and now you have gone down the wrong rabbit hole and have the wrong equipment for your application. La Voce only quote specs for those drivers to 35 Hz, which is actually typical for pro sub drivers. Lastly their web site does not inspire confidence, their spec. sheets are incomplete and totally inadequate. I personally would never source one of their drivers.

 

[KurantAK]

Who is HTG? Sounds like in the end you lowered the signal, so it was too much. My older miniDSP 2x4s (unbalanced/balanced, pre-HD models) you had an input voltage sensitivity adjustment you could make by flipping a pin switch/jumper. What's your miniDSP model and what setting for input voltage is it set to (if it has such)? IIRC on my old one it was a choice between max input of .9V or 2V on the unbalanced and a choice of 2 or 4V on the balanced model....

2x4 HD.

The clipping is now gone at any volume level. I was just curious as to why it happens. I believe the HD also has the jumper after some research yesterday. I’ll mess with that when I return home Wed.

 

[KurantAK]

No, I suspect this is an error due to inexperience. Those LaVoce 18" sub drivers are high efficiency pro sub drivers. They are NOT home HT drivers.

They seem a poor outfit, who only show very limited specifications. However, they are high efficiency driver with 96 db. sensitivity. Now there is an axiom in speaker design, that you can have sensitivity or bass extension but not both. For home HT people look for bass extension and in the pro world sensitivity.

Now, what I suspect is happening is that the DSP is trying to extend bass extension. In a tuned ported enclosure, if you try and extend the bass by adding Eq, you can't as the driver decouples from the box, and useless over excursion of the driver results. Then you get bottoming if the VC which is the sound you heard, and if you keep it up you will definitely ruin the drivers.

Unfortunately you went ahead without posting here and taking advice, and now you have gone down the wrong rabbit hole and have the wrong equipment for your application. La Voce only quote specs for those drivers to 35 Hz, which is actually typical for pro sub drivers. Lastly their web site does not inspire confidence, their spec. sheets are incomplete and totally inadequate. I personally would never source one of their drivers.

After weeks of discussions about their products and my application, GSG suggested that driver. It’s also a very popular driver in their products. I trust that GSG made the correct suggestion when helping me. On top of that, they sound phenomenal behind my screen now that they are calibrated correctly.

Secondly it is not extension. After more research yesterday, it was the input signal clipping, I am 100% confident of that. The driver itself is fine - and since the issue has been fixed.

 

[TLS Guy]

After weeks of discussions about their products and my application, GSG suggested that driver. It’s also a very popular driver in their products. I trust that GSG made the correct suggestion when helping me. On top of that, they sound phenomenal behind my screen now that they are calibrated correctly.

Secondly it is not extension. After more research yesterday, it was the input signal clipping, I am 100% confident of that. The driver itself is fine - and since the issue has been fixed.

Clipping or over extension the underlying cause will be the same, which is trying to get deep bass out of a pro driver it can't produce.

If you can find, or link me, to a full set of the Thiel/Small parameters, I will do an accurate model of what that driver is capable of, as well as the dimensions and porting of the optimal enclosure. I have almost 70 years experience in speaker design now, and up to date modelling software.

 

[KurantAK]

Clipping or over extension the underlying cause will be the same, which is trying to get deep bass out of a pro driver it can't produce.

If you can find, or link me, to a full set of the Thiel/Small parameters, I will do an accurate model of what that driver is capable of, as well as the dimensions and porting of the optimal enclosure. I have almost 70 years experience in speaker design now, and up to date modelling software.

Input clipping has nothing to do with extension. There are loads of threads about it in the communities on the MiniDSP website. Even discussions that include their staff.

 

[Verdinut]

Clipping or over extension the underlying cause will be the same, which is trying to get deep bass out of a pro driver it can't produce.

If you can find, or link me, to a full set of the Thiel/Small parameters, I will do an accurate model of what that driver is capable of, as well as the dimensions and porting of the optimal enclosure. I have almost 70 years experience in speaker design now, and up to date modelling software.

Here you will find the Thiele/Small parameters for it:

https://www.parts-express.com/LaVoce-SAF184.02-18-Subwoofer-8-Ohm-293-728?quantity=1

 

[TLS Guy]

Here you will find the Thiele/Small parameters for it:

https://www.parts-express.com/LaVoce-SAF184.02-18-Subwoofer-8-Ohm-293-728?quantity=1

Thanks for that. In my quick search this morning it did not list Part-express as a vendor. However F3 is 38 Hz and x-max is only 8.4mm. Sensitivity is 97 db. So it is a classic high output pro sub driver. I will model it, but the parts x-press model does not extend Fs below Fs, which of course is impossible. So this driver is not a deep bass unit, and xmax does not allow room for Eq. I will model it and see how the graphs look. I will model it vented and sealed. But I doubt it has the x-max to make a decent sealed sub. This has the stamp of a driver to make lots of noise in a big space, but not deep bass.

 

[TLS Guy]

Input clipping has nothing to do with extension. There are loads of threads about it in the communities on the MiniDSP website. Even discussions that include their staff.

I don't understand what you are saying, if you use too much Eq, then a driver can run out of excursion, you can clip the amp or both.

Here are two designs I have done. I have modelled the optimal box, and mine agrees with the Parts x-press model. I have also done a sealed alignment. The driver is too big to fit in the optimal box, so I made it large enough for the driver to fit. In any case this driver is totally unsuitable for sealed alignment.

I can not sugar coat this as this is not a good driver, and is not really a sub, as it can not reproduce the last octave. In addition the specs have embellished the truth in that this is NOT a 1200 watt driver at bass frequencies as you will see in the attached graphs.

I will only go over the vented alignment as the sealed alignment is not worthy of discussion as F3 is 70 Hz, and there is no room for any Eq, as the driver quickly runs out of cone excursion and the response shows that in that application the driver is a classic "thumper" driver beloved of car audio enthusiasts.

So lets return to the vented alignment. As I stated before, the free air resonance of a driver Fs absolutely sets the lower limit for a vented alignment. Attempts to Eq below that bottom the driver. The driver decouples from the box below that point and large stroke excursions of the cone take place with output falling at 24 db. per octave, resulting in fatal driver damage.

The first and second graph NA and CA confirm what I have just told you. AP shows you that available output is limited below 100 Hz or so. EP shows the power the driver can handle plotted against frequency.

CD the cone displacement graph shows that the driver is significantly power restricted below 80 Hz.

VV the vent air velocity shows that port is correctly configured with the vent air velocity below turbulence at 20 msec.

I shows the impedance curve and that optimal box tuning is 35 Hz.

The last two graphs show phase response and group delay.

From the description of the sound you describe in your first post, I can be certain that what you heard was driver bottoming. Your description is classic and the model shows that this speaker is highly susceptible to driver bottoming.

Now I understand that this data will be very disappointing to you, as you have put a lot of work into this, to say nothing of the expense. However all of us DIYers have had failures, including myself, especially in the early years. You have to regard these episodes as learning experiences. In fact I think failures teach you more then success.

Certainly I and others remain ready to assist you in your next project. Do not be disheartened.

 

[lovinthehd]

Interesting to get @JasonGSG 's opinion here perhaps.

 

[Verdinut]

Thanks for that. In my quick search this morning it did not list Part-express as a vendor. However F3 is 38 Hz and x-max is only 8.4mm. Sensitivity is 97 db. So it is a classic high output pro sub driver. I will model it, but the parts x-press model does not extend Fs below Fs, which of course is impossible. So this driver is not a deep bass unit, and xmax does not allow room for Eq. I will model it and see how the graphs look. I will model it vented and sealed. But I doubt it has the x-max to make a decent sealed sub. This has the stamp of a driver to make lots of noise in a big space, but not deep bass.

Yeah! Over 99% of Pro sub drivers are not real subwoofers as none of them can reproduce any bottom infrasonic frequencies. Perhaps the only exception is for one of the huge Danley Sound Labs tapped horn sub driver design, the DTS10 with useful response down to 12 Hz:

https://www.danleysoundlabs.com/wp-content/uploads/2022/02/DTS-10-spec-sheet.pdf

 

[KurantAK]

I don't understand what you are saying, if you use too much Eq, then a driver can run out of excursion, you can clip the amp or both.

Here are two designs I have done. I have modelled the optimal box, and mine agrees with the Parts x-press model. I have also done a sealed alignment. The driver is too big to fit in the optimal box, so I made it large enough for the driver to fit. In any case this driver is totally unsuitable for sealed alignment.

I can not sugar coat this as this is not a good driver, and is not really a sub, as it can not reproduce the last octave. In addition the specs have embellished the truth in that this is NOT a 1200 watt driver at bass frequencies as you will see in the attached graphs.

I will only go over the vented alignment as the sealed alignment is not worthy of discussion as F3 is 70 Hz, and there is no room for any Eq, as the driver quickly runs out of cone excursion and the response shows that in that application the driver is a classic "thumper" driver beloved of car audio enthusiasts.

So lets return to the vented alignment. As I stated before, the free air resonance of a driver Fs absolutely sets the lower limit for a vented alignment. Attempts to Eq below that bottom the driver. The driver decouples from the box below that point and large stroke excursions of the cone take place with output falling at 24 db. per octave, resulting in fatal driver damage.

The first and second graph NA and CA confirm what I have just told you. AP shows you that available output is limited below 100 Hz or so. EP shows the power the driver can handle plotted against frequency.

CD the cone displacement graph shows that the driver is significantly power restricted below 80 Hz.

VV the vent air velocity shows that port is correctly configured with the vent air velocity below turbulence at 20 msec.

I shows the impedance curve and that optimal box tuning is 35 Hz.

The last two graphs show phase response and group delay.

From the description of the sound you describe in your first post, I can be certain that what you heard was driver bottoming. Your description is classic and the model shows that this speaker is highly susceptible to driver bottoming.

Now I understand that this data will be very disappointing to you, as you have put a lot of work into this, to say nothing of the expense. However all of us DIYers have had failures, including myself, especially in the early years. You have to regard these episodes as learning experiences. In fact I think failures teach you more then success.

Certainly I and others remain ready to assist you in your next project. Do not be disheartened.

It is not an excursion issue. I don’t know how many times I have to say this. The problem ONLY arises when the MiniDSP is being used. If I run Audyessey, play anything I want at any volume, the clipping is not present. Put the MiniDSP with a fresh factory reset into the equation , and the sound shows up. This has nothing to do with frequency response as much as you want it to.

Again, this is not an excursion issue. It is a voltage issue from the Denon into the input of the MiniDSP. A engineer responded to my message and he agreed that the incoming voltage was likely the issue. There are loads of threads over there on this topic alone.

I know what you’re saying, but you refuse to listen.

 

[TLS Guy]

It is not an excursion issue. I don’t know how many times I have to say this. The problem ONLY arises when the MiniDSP is being used. If I run Audyessey, play anything I want at any volume, the clipping is not present. Put the MiniDSP with a fresh factory reset into the equation , and the sound shows up. This has nothing to do with frequency response as much as you want it to.

Again, this is not an excursion issue. It is a voltage issue from the Denon into the input of the MiniDSP. A engineer responded to my message and he agreed that the incoming voltage was likely the issue. There are loads of threads over there on this topic alone.

I know what you’re saying, but you refuse to listen.

If you want us to believe you, then you need to tell us what you are doing with the DSP. If you are boosting the frequencies below 40 Hz, it is definitely an excursion issue. That design would preclude any boost below 40 Hz, and if you did the woofer would bottom in a hurry, of that I can be absolutely certain.

 

[KurantAK]

If you want us to believe you, then you need to tell us what you are doing with the DSP. If you are boosting the frequencies below 40 Hz, it is definitely an excursion issue. That design would preclude any boost below 40 Hz, and if you did the woofer would bottom in a hurry, of that I can be absolutely certain.

No offense, but did you bother to actually read what I wrote? I'll repeat it again.

The issue arises ONLY when the MiniDSP is in play.
Audyessy calibration does not cause the subs to clip. I can use LFE 1 and 2 on the AVR. I can play it at any volume I want, and the clipping doesn't happen. If you want to tell me Audyessy isn't applying EQ below 40hz, I have a bridge to sell you.

I can take a freshly factory reset miniDSP and hook it up, and the clipping happens. Once again, that means NO EQ. I can assure you there is no issue with the woofer itself. It is the gain structure and the voltages into the inputs.

I don't even know why I'm going back and forth with you. I can appreciate you understand speaker design and have lots of experience in it, but you refuse to read and comprehend and that is very frustrating. Good luck to you.

 

[TLS Guy]

No offense, but did you bother to actually read what I wrote? I'll repeat it again.

The issue arises ONLY when the MiniDSP is in play.
Audyessy calibration does not cause the subs to clip. I can use LFE 1 and 2 on the AVR. I can play it at any volume I want, and the clipping doesn't happen. If you want to tell me Audyessy isn't applying EQ below 40hz, I have a bridge to sell you.

I can take a freshly factory reset miniDSP and hook it up, and the clipping happens. Once again, that means NO EQ. I can assure you there is no issue with the woofer itself. It is the gain structure and the voltages into the inputs.

I don't even know why I'm going back and forth with you. I can appreciate you understand speaker design and have lots of experience in it, but you refuse to read and comprehend and that is very frustrating. Good luck to you.

I really can't answer your question. I understand what you are doing, but have no idea of the actual gains involved. You would need to provide actual voltage readings.

What I can tell you is, that you absolutely can NOT provide any significant Eq below 40 Hz with those drivers. The analysis I did, especially the cone displacement graphs, tell the whole story. I used excellent and highly respected up to date software. In any event it is a tenet of good practice that you do NOT boost ported enclosures below F3, as it is ineffective and damaging.

The sound you describe is classic driver bottoming from exceeding not only driver linear xmax, but also mech. xmax. What you describe is precisely how it sounds. So I have to conclude that Audyssey is providing minimal Eq. if any, below F3, and that mini DSP is providing significant and inappropriate Eq.

The other issue is that it is clear those drivers are over specked in terms of power handling, by a significant power margin. For any 18" sub that xmax spec. is paltry.

The fact is that for a high spl. design for that room, I would never have selected those drivers for my subs, or for that matter any sub.

You have to face the fact that you have a design well wide of requirements.

I have designed a significant number of subs for members here over the years, and all have been very pleased with them. I have a number of good designs on my website. I make no money from this, and it is all done for the good of the order.

You can check any source on the Internet to check that my assertion that you can not Eq a ported sub below driver Fs is absolutely correct, and disastrous if you do.

Unfortunately those subs are way short of your desires. That is the honest truth. So you need to cut your losses and build a new and more effective design. Certainly I will offer you any assistance you wish, as I am a keen proponent of DIY speaker builds. I have always designed and built my own speakers.

 

[TLS Guy]

Interesting to get @JasonGSG 's opinion here perhaps.

I sure would like to know Lovin! I have not looked into GSC until today, and I'm appalled at what I found. He is selling flat back cabinets purely on driver size. He has selection of drivers with totally different T/S parameters that he says can be put on the cabinets. He is selling cabinets based on what the cabinets are tuned to.
So the unwary will think they will get deep bass, on account of what the cabinet is tuned to, when as in the case here it is impossible just based on the Fs of some of the drivers.

I find this very distressing, as it gives DIY a bad name.

If that Jason chap ever shows even the top of his head above the parapets here, he had better be prepared for the dressing down of his life from me.

 

[lovinthehd]

I sure would like to know Lovin! I have not looked into GSC until today, and I'm appalled at what I found. He is selling flat back cabinets purely on driver size. He has selection of drivers with totally different T/S parameters that he says can be put on the cabinets. He is selling cabinets based on what the cabinets are tuned to.
So the unwary will think they will get deep bass, on account of what the cabinet is tuned to, when as in the case here it is impossible just based on the Fs of some of the drivers.

I find this very distressing, as it gives DIY a bad name.

If that Jason chap ever shows even the top of his head above the parapets here, he had better be prepared for the dressing down of his life from me.

I'm curious on that aspect as well. I have several of the drivers the Marty was originally designed for (the Stereo Integrity HT18D2) and have wondered how some other drivers, like this one, fit the design and how they particularly perform as a result.

 

[TLS Guy]

I'm curious on that aspect as well. I have several of the drivers the Marty was originally designed for (the Stereo Integrity HT18D2) and have wondered how some other drivers, like this one, fit the design and how they particularly perform as a result.

I wasted some time playing with this. If you model and extended bass alignment, the box size goes up to 9 Cu.ft. and you gain 0.8 Hz in bass extension!

Just to make it clear, the driver Fs pretty much does set the F3 of total system, now and again it may be 3 to 4 Hz lower, but more often then not f3 is a few Hz higher than Fs.

The OP is pretty upset and in strong denial of the physics of the situation.

I now know to steer members away from GSC.

 

[KurantAK]

I wasted some time playing with this. If you model and extended bass alignment, the box size goes up to 9 Cu.ft. and you gain 0.8 Hz in bass extension!

Just to make it clear, the driver Fs pretty much does set the F3 of total system, now and again it may be 3 to 4 Hz lower, but more often then not f3 is a few Hz higher than Fs.

The OP is pretty upset and in strong denial of the physics of the situation.

I now know to steer members away from GSC.

First off, no one is upset. I'm perfectly happy with what I purchased. You just assumed it was specifically for home theater, and you would be entirely wrong.

Secondly, switching the jumper from 0.9v to 1.2v fixed the input clipping entirely. Now, it's quite clear you don't' have a clue about the DSP I'm using, and that's fine. I'm not going to explain it again, because you can't comprehend it anyway, obviously. Just be assured, it was NEVER a driver problem.