Any secrets to good car speakers?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi ,

I've got a car system that is good, but not as good as I was hoping - maybe someone could offer some insight as to why?

It sounds great for low frequencies - low pitch male voices sound perfect, however higher female voices seem 'shrill' and seem to 'scream' when the volume level is turned up to a level that would be if the singer was sitting in the car.
(at low levels it still does this but not as much).

I've bought some not-cheap 2-way car speakers, placed them in 4L sealed boxes (which makes a Q just >1) which are stuffed with accoutic foam. The speakers are on the back shelf of the sedan, facing forward.

The subwoofer is a cheap 12" sealed box unit in the boot. With the backseat armrest down there is a hole straight thru to the boot, the subwoofer is up agaist this.

The amp is a cheap ebay 4-way Gazillion watt amp (so they say) - dont know how powerful it really is - it has two 20A fuses.

What makes it all great is the MiniDSP equaliser which has been setup using REW. The speakers are crossed over to the sub at 200Hz to avoid the speakers entering the region where the high Q (from the small boxes) boosts the levels.

What do I do now?
I'm thinking of checking the signal levels through everything are good, and finding out the max output power of the amp by testing.
If I can't find anything wrong there all I can do is replace the speakers, but how can I know before buying which ones will be any good?
 
So you did come up with a system Qtc........Check your drivers for EBP, Efficiency Bandwidth Product, EBP=Fs/Qes, The values less than 50 tend to function best as sealed, over 50 towards 100 puts you in the ported camp. I would bet your drivers will run best ported..........as in, Think about the environment car door drivers are designed for ....un-sealed doors with little to none acoustic suspension.....& yes the 'woofer' extends up into the vocal range easily perhaps breaking up & messing up your vocals.....Just some thoughts to mull over.

____________________________________________________Rick..........
 
If you can find speakers that have a flat frequency response, that will make getting good sound quality much easier. Many times, the frequency response graphs provided by the high-end manufacturers are the only reliable graphs. If you can find an independent review with graphs and they agree with the manufacturer's graphs, then you can probably rely on those.

Too many manufacturers try to get high efficiency and high power handling and often those speakers have terrible frequency response. The least efficient often end up being the best choice.

I don't know if it's still the case but a few years back, it seemed like many of the speakers with exotic cones (not poly or paper) had a frequency response that looked more like a pyramid than a flat line. I saw a lot of people try to make them sound good but it was difficult and it required either a very complex passive crossover or a lot of equalization.
 
Ugh, … keep it in perspective. Good sound in a car is nice but I’m not sure it’s an environment that takes to perfection. Besides engine/exhaust/road noise there’s always issues with early reflections, lensing, absorption, etc. Not to mention obstructions like headrests and dash boards. These factors can completely alter the sound of any speaker, so knowing a speaker’s response ahead of time will tell you nothing about how it sounds installed in a specific application. Custom filters are sometimes necessary to correct for these variables, and may be what you need. Check to see if there’s some local car clubs in your area for your car. If there is you can listen first hand to what they have and maybe make some decisions based on that. You could even check online for some forums related to your specific car and ask what other people have found to sound decent. Not the same as hearing it for yourself, but may serve as a starting point. As a last resort you could always try to find a guru in a local automotive sales/installation shop and talk to them to see if they’ve done installations on your type of vehicle and see what they feel works well.
 
As other posts have pointed out, there are several factors that could cause the sound to be less than desirable in a car audio installation. The fact that the highs get more shrill when you turn it up, suggests that the tweeter is crossed over too low or the power amp has non-linear distortion that increases at higher power levels. Why did you cheap out on the amp after spending money on miniDSP? Amps that are clipping are much more likely to damage speakers than a high quality power amp putting out a clean signal at higher power. What you are hearing may be distortion in the amplifier output signal that is punishing your tweeters and your ears. You should also verify that the output signal voltage of the miniDSP is properly gain matched to the input capability on your amplifier and the input of the miniDSP is not being overdriven by your head unit. Any of these could be the source of the problem. I would not rush to replace the speakers until you know the gain levels are set properly for each stage of the system. If you don't understand these terms, you probably should be taking your car to an expert.
 
Last edited:
Thanks to all for your comments.

...Why did you cheap out on the amp after spending money on miniDSP?...
If I could buy a decent amp for a decent price that would be good, but the problem is how do you know the amp you are buying is not just a rebranded cheapie anyway. As an electronic engineer I look into the guts of most everything I buy and quite often that is the case.
I will look into the distortion under load (I have a dummy load). I was thinking if the amp is bad, just so long as the DC power supply is good enough I might buy some IC based amps and replace the amps in the amplifier case.
Do you know of powerful amps, 3 or 4 channel with a reasonable price and a good audio quality spec?
 
What happens if you lower 2,500hz on your EQ by about 2~6db?

I also second getting a good amp. From an electronics standpoint you are mostly right, only you should start treating the actual sound output like a musical instrument. i cant recommend any specific amp for you as only your ears will tell you that, but i can tell you I've been through about 5 different amps on the same door speakers and each sound different - from harsh to mellow.
 
Thanks to all for your comments.


If I could buy a decent amp for a decent price that would be good, but the problem is how do you know the amp you are buying is not just a rebranded cheapie anyway. As an electronic engineer I look into the guts of most everything I buy and quite often that is the case.
I will look into the distortion under load (I have a dummy load). I was thinking if the amp is bad, just so long as the DC power supply is good enough I might buy some IC based amps and replace the amps in the amplifier case.
Do you know of powerful amps, 3 or 4 channel with a reasonable price and a good audio quality spec?

I have been experimenting with a Soundstream STL4.500 4-Channel amp that I picked up from Amazon and I am impressed with the performance and value. It uses surface mount components for most of the circuitry other than the output transistors and power supply switching transistors. This results in a more compact size while still being a Class A/B analog amp. (I don't like the sound of most of the digital switching amps I have listened to.) Power rating is 80W X 4 into 4 ohms; 125W X 4 into 2 ohms with low distortion. I don't remember exact distortion values but it is <<< 1% at these power ratings. Most cheap amps give you max power rating at something like 10% distortion which is not something you want to listen to. Current cost is $131 from Amazon.
 
A shrill female vocal is often associated with to much energy around 5KHz. This could be caused be either a peak in the response or by higher order distortion from the mid range. It is less likely to be distortion in the tweeter but could be a peak in its response. If you have measurement equipment you could look for this but a good test would be to try putting a band reject of about 3 - 6dB at around 5KHz.

If you find this makes the system sound lifeless, increase the energy around 8 - 10KHz this tends to give the sound a crisp bright sound without causing it to sound shrill or shouting.

Regards,
Andrew
 
1. Forget rear speakers, concentrate on the fronts
2. Car drivers can be high Q and not always well suited to enclosures-try IB mounting them in the rear deck, or get them up front in your doors or kick panels
3. The amp is probably around 240wrms, so 80wrms x 4 (20a+20a=40a, 40a x 14.4v = 576w, 576 x 60% (most efficient the amp is likely to be)= 345.6w, 345.6 x0.707=244.34wrms) but that doesn't mean it will give you that power cleanly-how have you set the gains?
4. Tell us what speakers they are so we can find out more about them
5. What car is it? What size speakers will the front doors accept?
 
Thanks for all your comments.
I got around to testing the amp, it gives 28W RMS into 4ohm per channel before clipping starts. Only tested 1 channel at a time. (writing on amp says 5400W!!!)
The distortion measured just a little over what I would get with a loopback cable, so it wasn't that bad.
I'm thinking 28W isn't enough though for realistic sound levels - any comments?
Do any car amp manufacturers publish real data on their amps?
I'm thinking of building my own from IC amp boards from ebay, not sure how to get a power supply for them though.

On the frequency response issue, as some of you have eluded to, it may be a case of trying to get a good "room response" rather than flat. I've found a few I might try once the amp is sorted.

Thinking about mounting the speaker into the back shelf too.
 
Wow, that's some discrepency! 28wrms is around the output of most stereos, so good for about 100dB on "most" car speakers. If the sub is prety senistive then that should be fine too. Mounting them in the shelf would be better I'm sure, though i'd still say to concentrate on the fronts and forget the rears until the front is sorted-and even them I personally wouldn't bother with them, can be a lot more trouble than good.

All amps that are "CEA Approved" detail theair actual outputs and don't make false claims. Any decent brand amp should be fairly accurate on the RMS out put of their amp-if you're looking at new amps post them up here (wih links) and I'll tell you what I think/know about them.

I found with my car that the best way to get a "flat" response was to do the opposite of most and go for an undersized bass enclosure-cabin gain makes a huge impact on bass freqs, so you can either build big, tune low and EQ down the hump or make a Linkwitz Transform enlosure and use the gain of the vehicle for the EQ-this has worked very well in my car, but then I had the cabin gain mapped out and have tiny car which gives a huge gain!
 
Have you listened to the same track with "shrill voices" in different positions of the car? Your driver's position could just be a "sweet spot", or not so sweet spot for those particular shrill frequencies. If they go away when you move around, you might just have to "re-angle" the mids/tweets. Trial and error works, but it time consuming.

28 "real" watts should be plenty for a moderate system in terms of loudness. However the cheap factor could mean that there in an inherent nonlinearlity to the amp. Only real way around that is EQ, or simply a higher quality amp. If its possible you could keep the cheap amp for the sub(s), and get a good amp for the highs. There is more tuning to a multiamp system, but with that alot more flexibility.

...and yes....forget about vocal speakers behind you. Gotta get them up front for any sort of imaging and soundstaging. I dont even use "rear fill". I also try to crossover the sub/mids as low as possible...usually around 80-100Hz. Every car is different though too...as is the listener's perception. What sounds good to one, may sound horrible to another. Just my two cents.....
 
Just curious, but what parameters did you use to test the power output of the amplifier?
DC Voltage; frequency; load impedance; distortion threshold (or, if no measured threshold, then what; ie viewing onset of clipping via scope, etc).

Also, did you measure the one channel with both channels driven into your dummy load?

Although there is such a thing as a standard car audio measurement criteria, it's not as stringent as the FTC method for home audio (28 Watts would be pretty decent if specified as continuous per channel from 20~20,000 Hz, both channels driven into [specified] impedance, at less than [specified] total harmonic distortion, measured after a rather stringent pre-conditioning of 1/3 rated power @ 1 Khz for 60 minutes under the same conditions and load).

It might be worth mentioning that most multi-channel home amps and especially receivers don't follow the FTC method either since manufacturers successfully argued that the FTC method, mandated in the US (only) in 1974, was only meant for mono and stereo amplifiers.

CEA compliance for mobile amplifiers can be attained by displaying just two criteria: Output Power and Signal-To-Noise ratio. Output Power must be specified at 14.4V into a 4 ohm load from 20~20 Khz @ 1% or less THD + N, while S/N must be specified at 1 watt/4 ohms output level.

Specify those two and you can display the sticker on the box ... there are other CEA-compliant specs you can voluntarily add, but they're not mandatory. Compliance is voluntary.

CEA-mobile compliance does not prevent amp makers from using "other" methods in advertising, on the amp itself, etc. Nor are any power claims into any load other than 4 ohms covered by CEA. The CEA compliant data must appear next to the compliance logo on the product box, in rather small type, and nowhere else.

So, you will still see ... shall we say "interesting" ... power output ratings in much larger type from the compliant manufacturers as well as the non-compliant ones, on the product packaging, the amp itself, etc.

Also, it's unlikely that you will be able to see 14.4V at the DC power input of any car amplifier under most conditions, so CEA-mobile isn't exactly a true "real-world" rating. (The S/N spec is, however, rather stringent since noise will dominate at 1w 4 ohms output, versus rated power).

As installed in a working car, at typical alternator RPMs and typical electrical loads, power output will likely be less unless the amp is deliberately under-rated by the manufacturer.

Some car audio manufacturers will get creative with the measured performance to give the "bold type" power rating. They measure at some relatively real-world criteria, then use the math (rather than actually driving and measuring the amp) to get the rest.

So, they start with a measured 10% THD rating at 1 KHz one channel driven, then get out pen and paper for the rest rather than actual measurement, and taking advantage of rounding up to some "whole value" at each stage, grossing up the power by assuming some high alternator charging voltage, then calculate peak power (not RMS), then double it for peak-to-peak, then double it again for a theoretical 2 ohm load, or maybe twice, for a theoretical 1 ohm load, and so on.

Reputable manufacturers are more stringent but you have to realize they are competing with lower-priced alternatives so they also have a bold-print number that is not comparable to a home amplifier. However, if they say, for example, x watts into 1 ohm, they probably actually ran the amp at 1 ohm.
 
Last edited:
Just curious, but what parameters did you use to test the power output of the amplifier?
DC Voltage; frequency; load impedance; distortion threshold (or, if no measured threshold, then what; ie viewing onset of clipping via scope, etc).

Also, did you measure the one channel with both channels driven into your dummy load?

I have loaded the amp with 4ohms (one one channel - only have 1 dummy load), 1khz signal and increased signal until clipping is observed on the scope. At the same time the THD+N greatly increases. THD+N with a loopback cable reads 0.1% while the amp reads ~0.15%, so the exact figure is unknown but does not seem to be a problem.
Quickly tried a few other frequencies with similar results.
NB also tested it with 2 channels bridged into 4ohm and it supplied virtually the same voltage, so this equates to a lot more power.

Just realised though I did this with the car engine off, so the voltage reaching the amp would be a lot lower than normal - might need to re-test.
 
I have loaded the amp with 4ohms (one one channel - only have 1 dummy load), 1khz signal and increased signal until clipping is observed on the scope. At the same time the THD+N greatly increases. THD+N with a loopback cable reads 0.1% while the amp reads ~0.15%, so the exact figure is unknown but does not seem to be a problem.
Quickly tried a few other frequencies with similar results.
NB also tested it with 2 channels bridged into 4ohm and it supplied virtually the same voltage, so this equates to a lot more power.

Just realised though I did this with the car engine off, so the voltage reaching the amp would be a lot lower than normal - might need to re-test.

I don't think it would change significantly if the car were idling under alternator power so I don't think you *need* to re-test for output unless you want to satisfy your own curiosity. More importantly, I think 25~30 wpc observed via the method you used is adequate to power a pair of car speakers to decent levels.

If the amp is the problem, I would expect the reason would lie more in the distortion spectrum (high order odd harmonics, IMD) rather than output power itself. Perhaps also there are noise issues or other bad behaviour that arise under alternator power.
 
Last edited:
I have loaded the amp with 4ohms (one one channel - only have 1 dummy load), 1khz signal and increased signal until clipping is observed on the scope. At the same time the THD+N greatly increases. THD+N with a loopback cable reads 0.1% while the amp reads ~0.15%, so the exact figure is unknown but does not seem to be a problem.
Quickly tried a few other frequencies with similar results.
NB also tested it with 2 channels bridged into 4ohm and it supplied virtually the same voltage, so this equates to a lot more power.

Just realised though I did this with the car engine off, so the voltage reaching the amp would be a lot lower than normal - might need to re-test.

I don't think it would change significantly if the car were idling under alternator power so I don't think you *need* to re-test unless you need to satisfy your own curiosity. More importantly, I think 25~30 wpc observed via the method you used is adequate to power a pair of car speakers to decent levels.
 
You couldtry this test. get a singer to bellow it out in your car . then do an ab test with a studio recording of said singer. . i guess car audio is like sitting inside a very poorly made loudspeaker enclosure made out of the worst materals you can lay your hands on. All them reflections resonances rattles and not forgetting the electrical operated adjustable ports.

Seasons greetings too all lol
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.