Strange frequency behavior of new installation

[cartraveller]

Hi,

I'm new on this forum. I'm from The Netherlands and have a Toyota Landcruiser 150 from 2018. The original speakers are removed and replaced by an installer. My new gear is:
- Exact HX20 tweeters for the highs
- Gladen Zero 130 for the mids
- German Meastro EV6908 woofers for the lows
- A Mosconi 130.4 DSP amplifier / processor is installed. Has 4 channels. 2 channels for the lows and two channels for the combined mid & highs.

When I measure the frequency curve of the German Meastro and the mid/highs separately (without EQ, filtering etc.), the result is very strange. See attached picture from REW.

I know that the original setup has a HUGE emphasis on bass (as you can see), so that doesn’t surprise me. But the German Meastro should be more or less flat untill 1 or 2 Khz at a minimum I guess (Can’t find in in the specs). The Gladen Zero 130 mid unit has a flat curve until 125 Hz according to the specs. The measurement doesn’t even come close! I have a huge dip in the mid area!

First I thought the mids were not connected (the frequency curve of the mid/high combined is almost the same as the natural slope of the tweeters alone ), but it seems they are connected. What could be reason for this strange behaviour? I can think of two options:
1) The original crossover filters have not been removed by the installer (although, the installers states that they always remove the crossover filters) à is there an easy way to check this? (I’m not comfortable to remove side panels in my car etc.)
2) The head unit is already filtering the signal à Is that ‘normal’ form head units? Apparently, it is also boosting the bass, so some kind of DPS processing must be going on inside I guess
3) ..... any other explanation that I didn't think off?

Hope you can help me out on this.

Click to read more...

 

[1aespinoza Premium Member]

Welcome cartraveller. From what you have already stated, your factory radio has pre selected frequency curves designed to work in conjunction with your car's interior/speaker quality. You may need a dsp to flatten your signal before it reaches your amplifier. Your amp has built in dsp but I was unable to find if it corrects the signal before going to the filters stage. So the issue may be correcting the curves via computer download or presignal adjustment.

 

[cartraveller]

Okay, so probably the signal is already filtered when it leaves the head unit? So the filtered signal enters the DSP amplifier. That is bad luck...... the amplifier cannot 'add' something to the signal that was already filtered out :-(. I guess that would mean that I cannot use the filter options in my DSP amplifier and will have to live with the crossover filter settings from the head unit? Btw, is this in particular the case for the head unit in my Toyota, er is this the case in most head modern pre installed head units?

 

[1aespinoza Premium Member]

Btw, is this in particular the case for the head unit in my Toyota, er is this the case in most head modern pre installed head units?

Yes, it. There actually exist dsp units that correct the signal. Audiocontrol, JL Audio are the ones off the top of my head that make such units. Most modern cars have nice radios with features to spare, but coupled to mediocre speakers. So the radios are programmed to soften the signal for speaker protection. So companies came up with the Active Line Output Converters to "fix" the source signal, to a flat clean usable signal. Look for Active Line Output Converter.

 

[cartraveller]

Sounds very mystic to me . The 'active line output converter' can not know what signal is lost or filter out. So how does it know what to restore? Seems mystic that ik can 'add' something to the signal that was already filtered at en earlier stage.

I guess it comes down to the questions: does the car has the crossover filter build into the head unit, or did it feature separate passive crossover installed between the head unit and the speaker. If I understand you correctly you say 'most units have the crossover filter build into the head unit'?

 

[Jeffdachef 5,000+ posts]

your signal is garbage because you are using the stock head unit which alters the signal to make the stock speakers sound better. There's no way of getting out of this unless you have a DSP that can signal sum then de equalized the signal recreating a flat signal, the dsp has to have input EQ capabilities which makes this possible. Also you might want to do some door treatments as well. Dips happen from vehicle acoustics, thats your next step in improving the sound.

You can also just get a brand new head unit and ditch the factory one which is the best option for pure sound quality.

 

[1aespinoza Premium Member]

Sounds very mystic to me

I am on the same boat buddy. I will not pretend to know how it is accomplished, but it is used by a lot of audiophiles with positive results. But like Jeffdachef said, it might be easier to spring for an aftermarket radio. I guess it will also depend on the features connected to your radio. Steering wheel, alarm, warning chimes. Maybe you will end up paying the same whichever route you choose.

 

[cartraveller]

Okay, thanks! Ditching the head unit is not an option for several reasons. I see two possible ways for going forward:

1) combine the inputs and let the amplifier do the filtering

The amplifier has 4 inputs and 4 outputs. I'm able to decide the mapping. Currently input 1/2 are mapped on output 1/2 and input 3/4 are mapped on output 3/4. Would it be better if I would combine input 1 and 3 and fix both on input 1. And the same for input 2 and 4 and attach both to input 2. This way the input signal is combined and looks 'reasonably okay'. I can map the input signal on output 1/2 and 3/4. I will use 1/2 for the lows (with a low pass filter) and 3/4 for the mid/highs (with a high pass filter), so I can decide on the filtering myself. I would like to get more lows out of the mids (up until approx. 150 Hz) and use the lows purely for the lows (>150 Hz). Would that be possible and would that help to improve the sound? I'm not sure whether it is 'technically' possible to combine two outputs channels from the head unit on one input of the amplifier.

2) Use the filtering from the head unit

Don't change anything in the wiring, and don't set any filtering in the amplifier and basically accept the current filtering that is coming from the head unit. I can still correct the frequency curve with the DSP functionality from the amplifier, but there will be a huge overlap in frequencies that come out of the lows and out of the mids as you can see in the graphs that I posted.

Not sure whether I managed explain my self (sorry for my English...). What would be your advice? Option 1 or option 2?

 

[1aespinoza Premium Member]

Not sure whether I managed explain my self (sorry for my English...). What would be your advice? Option 1 or option 2?

Your English is fine. Do not adjust any wiring, everything is as it should be. What we are trying to convey is that the altered signal from the head unit is not adjustable. The adjustable eq settings in the radio receive the altered signal, so those eq settings will not be able to flatten what comes out. This video better explains it.

 

[cartraveller]

Thanks for clarifying. Now I understand it: it works with a calibration file and restores the original signal. Changing the wiring is indeed not a good idea because there might be phase differences between the channels (I missed that possibility). So the best thing I can do is live with the current signal, don't use the cross over settings in the DSP amplifier, but accept the cross over settings as implemtned in the head unit and try to flatten the signal to some extend with the DSP capability of the amplifier. If that would not be enough, I have to buy such a Fix86 to restore the signal before feeding it to the DSP amplifier. Thanks for the help.

 

[cartraveller]

Okay, I have made some progress! And some new questions

The progress: I can do input summing!
I'm a lucky guy! My Mosconi has also the capability to sum signals. They call it 'input mixer'. Unfortunately, no input EQ (I should have bought the more advanced model for that). I have tested it, and it works!

The new problem
But it seems that I have also found a new problem. The original tweeters had a very simple filter attached to it (a single capacitor). It seems that the new tweeters have no filter at all. So the tweeters are now without any form of filtering / protection. After summing the two input signals, more bass will be directed to the mid unit (and thus also to the tweeter). As a short recap: my system is a combi of active / passive. The lows have separate amplifier and the mid/high have a separate amplifier (4 channels in total):

• Low signal from head unit >>>>> Mosconi input channel 3+4 >>>> output channel 1+2+3+4 (with input summing)
• Mid/low signal from head unit >>>>> Mosconi input channel 1+2 >>>>> output channel 1+2+3+4 (with input summing)

So my questions:

• I know that exposing the tweeters to a full range signal can result in blown up tweeters. If I would filter the mid/high signal in the amplifier and set the crossover to 250 Hz, would that be safe for the tweeters?
• I think not having a filter between mid range and tweeter will not result in good sound quality?

I have attached the original input signal after summing. The mid range is Gladen Zero Pro 80 - not the 130 as I thought! - (https://www.hifimotive.nl/files/products/f85454e8279be180185cac7d243c5eb3.pdf) and the tweeters are Exact HX 20 (https://www.hifimotive.nl/files/products/f73b76ce8949fe29bf2a537cfa420e8f.pdf .... unfortunatly, not in English)

My last questions:

• Two very strange dips in the profile. Is this also filtering from the head unit, the acoustics of the car of something else that is going on?
• What would be the logical crossover point between low and mid/high in your view?

Thanks for your support and knowledge sharing!!! So much appreciated!!

 

[hispls 5,000+ posts]

I know that exposing the tweeters to a full range signal can result in blown up tweeters.

Can and will very quickly. Be extremely careful you keep it to very low power if you're trying to test stuff now.

What's jumping out at me looking at your curve is where your nulls are. 150, 300, 1500, and 3000. This leads me to suspect some standing waves being created at those lengths and I'd wager if you could graph up to 15,000 you would see a dip there.

As far as "fixing" your response let me first say "if it sounds good it is good". Most people find deal flat RTA not pleasing for listening so if you're not trying to do that for the sake of a competition don't get hung up on it. Now if something sounds wrong it can be a good tool to try to pinpoint the trouble.

That being said, you MUST select crossover points before you try to chase down peaks and nulls, and either figure out how to make your amp/source unit do the job or build your own passive crossover network; you can find passive crossover calculators online and buy components at mouser.com (and possibly somewhere in Europe as well). If you want to be snobbish about components chose air core ribbon inductors and poly caps. Otherwise iron core round wire and electrolytic will get the job done just as well. Use ceramic power resistors either way.

Peaks can and should be tamed with EQ

Nulls in your response may be happening at or around crossover points. If overlapping more causes the null to get worse you can try to move the crossover point and see what that does, overlap less, or start experimenting running things out of phase. Try every single combination, one mid only, one tweet only, both mids, both tweets. It doesn't have to make any sense but I have personally seen this fix things. Location and aiming of drivers is a big limitation in a car and often the most difficult to easily overcome. Tweeters can be mounted in pods which can be swiveled and re-positioned easily mids not so much. Ideally all sound sources come from exactly the same point and left, right and sub sources should all ideally be equal distance from each other and the listener. This is nearly always impractical in a car so most of this other stuff we do is just trying to work around not being able to do it. Lastly I've seen one guy who had to keep a sweater rolled up on the hump between the driver and passenger's legs by where a shift knob would be on a manual. This inexplicably tamed a nasty peak he would get otherwise.

I really don't know much about the gear you're running nor your source unit so I can't really help you plan around that, but keep in mind these general tips. Always remember the "perfect" sound people want to experience is what you would get in an anechoic chamber (or in the middle of a wide open outdoor area) sitting at the vertex of a perfect equilateral triangle with two full range tower speakers. Anything you're missing is from some limitation of not having an anechoic listening space and not having all the waves from each side and each driver hitting your ears at precisely the same instant.

 

[1aespinoza Premium Member]

• Two very strange dips in the profile. Is this also filtering from the head unit, the acoustics of the car of something else that is going on?
• What would be the logical crossover point between low and mid/high in your view?

Those dips seem to be tuned to speakers that got loud in the 300hz and 1200hz range. So it does seem to be direct from the head unit. Your tweeter is designed to play from 1500hz, so I would cross it at 2000hz. Below is a site to figure out capacitance and inductance.
Also, your new chart corresponds with your first two charts.

Crossover Calculator

This crossover calculator will help you design amazing sounding speaker units.

www.omnicalculator.com