| Squeeky Voice-coil,Linkwitz Transform causing it? - Click HERE for Original Thread |
| mikee55 |
Hello Y'all,
As you may or may not know, I've built a Linkwitz Transform sub recently, and have a new problem today. This time its not my electronics, but my driver. My cheapo SPLX 12" car subwoofer occasionaly makes a squeek in time with the bass, that seems to eminate from the centre of the cone. Could be something to do with the voice-coil stressing perhaps?
Can anyone tell me whats going on?
Cheers Mikee55 :) |
|
|
| sek |
Hi Mike,
could likely be the voice coil former touching the back plate of the magnet assembly.
This is nothing to worry about as such, you just have reached your mechanical excursion limit for that driver. The likelihood of destruction is small if it doesn't happen too often, but don't challenge your voice coil former too much, though - the back plate is stronger. :cannotbe:
Could also be something else, e.g. clipping of an amplifier stage. This would have to be determined before a solution can be found, I'm afraid. ;)
One question: by writing "occasionally", do you mean something like: at exceptional excursions due to high SPL and/or low bass content?
Cheers,
Sebastian. |
|
|
| mikee55 |
Hello Sebastian,
I think an op-amp could be clipping. Because I'm using a car amplifier running off an old AT powersupply, and the fact that the protector light goes out at the same time, suggests clipping. The AT supply is only putting out 8amps max. Also I'm not sure how powerfull this car amp is. I also think that if the coil former is hitting the back plate, it would be a knocking sound, more than a squeek.
What do you think?:)
Cheers Mike :) |
|
|
| pinkmouse |
| Whatever the exact cause, mechanical noise from the driver means you're pushing it too hard. I suspect the LT circuit has too much gain, pushing your driver to its limits and overloading the amp and power supply at the same time. LTs put huge demands on the system, and if it can't cope, it will go bang eventually. |
|
|
| mikee55 |
If I reduce the gain, will I loose the ability for the system to go as low as I designed for,(20hz)? I was playing some drum and bass at the time. Maybe I need to filter off the cone flap.
Cheers Mike:D |
|
|
| pinkmouse |
| Yes. For LT to work at low frequencies, you need a good driver, and a powerful amp with a good PSU. With respect, it sounds like you have neither. |
|
|
| mikee55 |
I wanted to find how an uncomfortable drive felt before I bought something flashy.
I had bought the driver 'cause I got it cheap, I was given the box and amp, and the powersupply wasn't bin-worthy. The whole project is for education and fun and to see how the ride could go if I were to spend some cash on it. Its a compact way to get some serious bass. I can talk on the phone while it shakes my kitchen, mad, its doing its job and I ain't turned the volume up yet.:D
Cheers Mike |
|
|
| pinkmouse |
Hey, don't get me wrong, I'm all for fun! :)
Okay, you can run it like that, and it will have a short, but happy life. Or, if you really want to use this as a learning tool, then you could post your simulation graphs and models, and we might be able to help you make it better! |
|
|
| 4fun |
Hi mikee55,
Your AT supply is 8A 12V that makes 96W. How much power does your amp draw, or supposed to give?
A voltmeter or much better a scope will tell if your 12V is up to the task.
pinkmouse is right, LT is going to put severe stress on amp and driver. How much gain diff from LT do you have at most within your wanted sub bandwidth?
Have you done any driver excursion analysis with your LT applied? |
|
|
| mikee55 |
Hello, pinkmouse I don't wish for you to take me the wrong way either, I just felt I'd be laughed at for throwing together a load of cheap parts in a design like this. I can't afford 1000 watts of muscle to drive a Tempest or Peerless XXLS, I'm just getting along the best I can. I've had too many people laugh at me in my life it makes it difficult to ask for help without feeling small or I shouldn't be asking. I've had a lousy childhood, and that effected my education. Everything I have learnt has been through asking people like on here. My confidence is low, so I feel inferior.
Sorry for waffling...
I was originaly using a Technics amp, that was dead on one channel it was a Technics SU-V2 and I'm now using this car amp. A post on an ICE forum has just thrown up that its a cheapy amp of which lots were made.Its 4 Channel Mosfet job so I assumed it would have been advertised as 4 x 100 watts. Its chassis isn't as long as the 600W version.I'm using just one channel. The only graph I can show is of the Transform spreadsheet. I have the T/S parameters written down too.
Cheers Mike :) |
|
|
| mikee55 |
I have a volt meter but no scop, and wouldn't know how to do an excursion test with LT applied.
Could you please explain what you mean, further.
If I connected my DMM to measure current drawn would it help?
Cheers Mike:) |
|
|
| pinkmouse |
| As I thought, too much gain. Try knocking the gain down to 10dB. With the small size of house in this country, room gain will pick up the difference. One of the other sim programs has a graphical indication of required power and excursions for a given filter. I can't remember what its called offhand though. Have a google. |
|
|
| mikee55 |
Hello pinkmouse, Do you mean reduce the input, output or the gain of the Transform?
Would I get way with lowering the gain on the Input Mixer?
Cheers Mike:) |
|
|
| pinkmouse |
| The gain of the LT. Try and find that sim program, it makes things so much easier to explain. |
|
|
| mikee55 |
Al, can I have an invite to the meet, please? Can you post me some details. I'd love to shake the hands of the people who have been helping me with this project.
Thankyou, Mike:) |
|
|
| pinkmouse |
| Details still TBA. Anyone is welcome as long as they bring something along, ( hifi related preferably, but booze will do ;) ) |
|
|
| mikee55 |
....meanwhile, still trying to find sim.
:D |
|
|
| mikee55 |
Hi Al, Would it be the software from FRD Consortium, by chance?
Cheers Mike :) |
|
|
| 4fun |
Hi,| quote: | Originally posted by mikee55
I have a volt meter but no scop, and wouldn't know how to do an excursion test with LT applied.
Could you please explain what you mean, further.
If I connected my DMM to measure current drawn would it help?
Cheers Mike:) |
I mean that if you have a voltmeter, connect it to your supply while playing loud. Your supply probably have current limit close to 8A and if exceeding that it is not going to supply further current so your voltage will drop, easily seen on a voltmeter.
Yes you could connect an amp meter (if it's up to the task) in series and if current approaching 8A (probably) you are in trouble.
Excursion test, in a sim program for example the popular WinISD Pro program there it is possible to add your LT and watch power and excursion plots.
One example with values from your spreadsheet:
We apply 10W, at 50 Hz there is no gain from LT, so therefore 10W to speaker. If we just lower frequency we get:
At 30Hz there is approx 10dB gain from LT so 10*(10^(10/10))=100W
At 20Hz there is approx 15dB gain from LT so 10*(10^(15/10))=316W
If I got this right you will see the rapid change in required power in order to keep response flat. So reducing gain a bit is a good idea. As mentioned before, room gain is going to fill in. |
|
|
| pinkmouse |
| quote: | Originally posted by mikee55
Hi Al, Would it be the software from FRD Consortium, by chance? |
Could well be! |
|
|
| mikee55 |
Hi,
I've downloaded WinLsd 0.50a7. Er, does anybody want to tell me what to do, it looks even more complicated when I know I got to use it properly. I downloaded it ages ago, but only played with it. I have a list of parameters and I'm going to assume I've got to put the driver into the database. So I'll get on with that.
Cheers Mike:) |
|
|
| mikee55 |
Hello again. I don't think I can give it enough info. None really made sense. I have attached my driver Mine.wdr
Do I have to measure driver physical dimensions?
Cheers Mike :confused: |
|
|
| 4fun |
Hi mikee55,
You also need to input:
Xmax, peak one way linear excursion.
Sd, displacement area, includes half surround.
Pe, maximum power
Help function is good in WinIsd ;)
There is info on how to input parameters. |
|
|
| richie00boy |
Back home after a weekend away. Mike I think your power supply and amp is not up to the job and it's clipping, making the noise. An advertised 100w amp is more like 50w in reality and your power supply will be struggling to keep up with even that.
When Pinkmouse was on about the gain of the LT he means don't try to EQ down so far in the frequency range, e.g. just try for 30Hz. Like I suggested in the first place :angel: |
|
|
| mikee55 |
...OH WHAT AM I TO DO WITH ME??
I just wanted to get down low with my audio, I get such a buzz from a real deep bass. I was greedy and went for 20Hz. What can I say, I'm a Bass Head. If I had a big enough amp/powersupply can I stay at 20Hz? I don't want to dismantle my new toy. I just love its potentcy.
richie000boy, bless you for helping me, I want to build an amp for it and that will be my next project. Does this mean I can't find out how to measure what I have created?
Anders, I don't have these parameters, so I guess measuring is outta the question?
Cheers Mike :)
Was Big Brother fixed?(Sorry):) |
|
|
| pinkmouse |
| Possibly the only thing saving your driver from explosive disassembly is the weakness of your amp. ;) |
|
|
| mikee55 |
Yeah, but I've got the gain turned right down on the amp. If I turn it up, things get really physical. I've got the little screwdriver pot turned half a notch, I'm playing Leftfield at the mo and its nice. But then, its late at night, earlier I was playing some and that squeaky noise was there again!
I just hope the system lasts long enough while I have a go at amp building.
Cheers Mike:D |
|
|
| pinkmouse |
| Just don't play any Prodigy! :) |
|
|
| 4fun |
mikee55,
| quote: | | Anders, I don't have these parameters, so I guess measuring is outta the question? |
Not necessary so :)
Sd, the most important here is easy.
Sd=(d^2*pi)/4 where d is diameter of the cone, half of surround included.
X-max, if you can see the coil through the vents in your pic #2, just gently press the cone until coil just dissapear in gap, required travel from resting position is approx X-max.
Pe, max thermal power limit, important if you want to know just that. Usually grossly exagurrated on cheap drivers. |
|
|
| mikee55 |
The missus starts her new job tommorrow so I'll be playing with various styles. At the mo I'm playing my k.d.lang invincible summer DVD Audio. I have Intervideo WinDvd Platinum version 6, it can play Dvd Audio if you have a Creative soundcard.
So 24bit/96 sounds good, the missus told me to turn it down but I argued that we are speaking, not shouting and so I havn't turned it down.
Its mad.
Could you explain what would happen if I installed a bigger amp, not technicaly but acousticly? I've only ever used these sized amps.Obviously I don't want to be blasting it out, I have a 24 db crossover inbetween the Input mixer and Transform, richie000boy told me to make 12db variable 12db fixed (40Hz) so it could be tuned up to my Mordaunt Short MS40's(70Hz ports closed), but I've found that I need to filter down low or Bass starts to buzz/boom and is really crude, even now a bass guitar is booming a little. I have nothing to reference it with, so I usually listen to music that has a bass that starts high in frequency then drops and then the speaker has to be able to follow it. Chemical Brothers have a good track for that, can't remember the name of it. I'll dig it out tommorow.
Cheers Mike :) |
|
|
| richie00boy |
Mike it would be nice to find out the parameters 4fun is telling you, but you can manage without.
As Pinkmouse has said the PSU is preventing you from real damage.
When you get a man enough amp the sound will be tighter and more controlled, as well as hitting deeper. If you are definitely going to put a new amp in then you can leave the LT as it is and see how the speaker copes with the new amp. |
|
|
| mikee55 |
Hi all, just stuck meter on rails, 9.8 volts average also the LT runs off this supply as well at appx +/- 9.7 volts appx whilst playing Pink Floyd Dark Side of The Moon.( oops). Wasn't playing low stuff either. If the power supply is drooping that much, does it mean its not regulated, just out of curiosity?
Cheers Mike :D |
|
|
| richie00boy |
It could be unregulated, or it could just be that it's simply not strong enough and collapsing.
Time to build an amp and PSU :)
Just so happens I've been prototyping an n-channel MOSFET amp that is reasonably easy to build and will do about 150w depending on output device choice :) |
|
|
| 4fun |
Hi,
You mean constant low voltage?
The supply may take feedback from one of the other voltages, so 12V may not be directly regulated.
If no load on the rail with feedback the 12V may suffer if loaded alone.
Newer PC supplies usually have feedback on 12V, as all load is there today.
But as pointed out earlier, your supply is probably to weak. |
|
|
| mikee55 |
Hi Rich,
Sounds interesting, would 150W be enough?
I just wired in some heavey electrical wire used for wiring sockets, replacing the thin wires that would of fed harddrives and CD roms. Has made a difference, its louder and no squeaking yet.
150w, ooooh! :bigeyes:
Cheers Mike ;) |
|
|
| FullThrottleRic |
With LT's you have to resist the temptation to go for loads of extension if you have limited power and Vd. Remember that every time you add 3db of boost you double your power requirements, if you try to get mega low output it gets out of hand quickly :bigeyes:
Looking at the plot you posted, assuming you are playing everything above ~50Hz at 20W...
By 40Hz you have 3db's of boost, 40w amp required,
36Hz 6db, 80w required,
31Hz 9db, 160w required,
26Hz 12db, 320w required...
When you get to 20Hz you have 15db of boost, requiring 640w to play 20Hz at the same level as 50Hz with just 25w, not to mention your cone will be slapping you in the face from the other side of the room :D
What size is your room? It's good practice to design a sub based around your room gain, but to do this you need some measurement gear. Do a Google search for 'RoomEQ Wizard', it's a freeware software tool that uses your soundcard and a mic to take measurements.
A quick and dirty method I use to do an initial estimate of room gain is to do a nearfield measurement of a speaker (nearfield is at low level with the mic very close the the cone to prevent room/baffle effects), it doesn't matter how it measures so long as you have some 20Hz output.
Next put the mic in the listening position and the speaker where it's likely to be placed and take a measurement, keeping the levels of the upper frequencies similar in level.
Now you can compare the two plots, you should find that the second measurement appears to go lower. Compare relative output of the two measurements, for example at 20Hz, 30Hz, 40Hz and 50Hz ect. From there you can work out how your sub needs to roll off to match the room.
Using this method I get flat output to 10Hz in my room, despite having a sub with a 30Hz f3 and 0.5Q rolloff slope :devilr: |
|
|
| richie00boy |
| Whilst that's a good point illustrating the climbing power requirements, in reality he's unlikely to be listening to anything more than about 5w on his main speakers. |
|
|
| 4fun |
Mikee55,
From my post #29| quote: | | if you can see the coil through the vents in your pic #2 |
I had a pic from another thread in mind:smash:
However you may have vents below spider. Otherwise it sometimes possible to put a narrow beam light close to spider and when finding the right viewing/light angle it may be possible to se coil trough spider. |
|
|
| mikee55 |
FullThrottleRic, hello.
I usually listen at low volumes (except when driving) approx 1 watt would be more realistic, but from the sub point of view, I was after just being able to reach down as low as I can. Sometimes I'll have a mad session with the volume, but then I think hifi goes outta the window and I'm listening for the fun of it. I recently had the amp out of a pair of PC speakers and drove a pair of 12" 3 way Goodmans loudspeakers. I reckon that to have been 3 to 5 watts, and they filled my 3m x 4m lounge. And yet, I thought you'd need loads of power to get what I got with this little amp. I suprised a few people with it.
4fun,hello.
If I unscrew my driver anymore, I gonna need bigger screws. I must buy some T-nuts and bolts, when I do I'll have a look.
richie000boy, hello sir.:)
Hows the amp design coming along, sir? I'm saving some pennies now. What size Transformer are we thinking, please?
Cheers Guys :D |
|
|
| richie00boy |
This is the prototype n-channel MOSFET amp. With IRFP250 devices it spits out 150w. The final design will feature a slight rejigging of a couple of parts to space them better, but look basically like this.
And the PSU looks like this with 10000uF 63V caps.
All my projects are designed to be easy to build, i.e. decent size tracks and pads and sensible component spacings.
For a 150w amp I would recommend a 300VA transformer. But really anything from 180VA to 300VA would do. |
|
|
| mikee55 |
Wow, that looks like somthing I could I get my teeth into. Its called an 'n-channel MOSFET', does this imply that no 'P type ' transistor is involved? One thing I'd like to know is, if the quoted output of these car amplifiers is no where near what they state what would be the wattage of my car amplifier, compared to this design?
Cheers Mike:confused: :) |
|
|
| richie00boy |
There are n-channel MOSFETs and p-channel MOSFETs just like there are NPN transistors and PNP transistors. The selection of n-channel MOSFETs is much greater than p-channel MOSFETs because they are used in switch-mode power supplies. The 'n-channel MOSFET' bit is really just about the output stage. There are both NPN and PNP smaller transistors used in the design.
Roughly speaking car amps are rated peak-to-peak output, which is four times the RMS power. Also with the fact that your PSU is strangling the amp, a proper setup should see a good difference. |
|
|
|
|
