Jump to content

DO HST-11 Build Thread with measurements


Electrodynamic

Recommended Posts

Ok, this is pretty embarrassing but I don't care. My mistake on install may bring a laugh. :)

 

So I rig up the OmniMic this afternoon while my son is in my bedroom watching one of his shows. I begin by turning the subs up very hot to begin taking measurements and possibly push the drivers and see what kind of SPL I am generating. I start out with Ghosts N Stuff (same track used in The Art Of Flight) and begin taking measurements at/around 115 dB. The drivers are not even breaking a sweat - they're only moving maybe 1/4". So as I'm just starting to walk over and turn up the subs again I hear what sounds like the left front in-wall speaker buzzing from not being screwed in tight enough. I don't think much of it as I listen for a few seconds and it's not the speaker so I turn it up a little more. What happened next? Watch this: https://www.youtube.com/watch?v=wp4PVsD6cfk

 

The screen fell off the wall and sheared the binding posts of the enclosure clean off. Thankfully the screen did not hit the enclosure or I would have cried. It came close but did not hit or scrape the enclosure. I believe I purchased two sets of binding posts so I may do the fix this weekend. If I do not already have the binding posts I have ordered more a few minutes ago and they will be here early next week. 

 

I put the screen back up and made sure that I pressed down to ensure the aluminum screen boarder was inside the mounting hangers. Upon firing up the projector sure enough the image is about 1/4" too high, which means the screen was never properly fit in the wall hangers. Oops. 

 

...and I was just getting started with the volume level on the subs! :D

Link to comment
Share on other sites

  • 2 weeks later...
  • Replies 106
  • Created
  • Last Reply

It's time for a road trip. We'll have the A-14k meet the little 11-inchers and show them what fer.  :D

 

We need at least a 20A home run, preferably a 30A home run.

 

Either way, I'd just like to meet up at Chez Nick for a fun day of world class low end listening and measurements.  ;)

Will look forward to the results of this meet ;)

 

 

Although with my current, sleep-deprived state, I'm having crazy ideas of a monster dual-opposed Bosso setup with 24s... :lol:

Link to comment
Share on other sites

It's time for a road trip. We'll have the A-14k meet the little 11-inchers and show them what fer.  :D

 

We need at least a 20A home run, preferably a 30A home run.

 

Either way, I'd just like to meet up at Chez Nick for a fun day of world class low end listening and measurements.  ;)

 

I'm down for that as well :) 

 

Will look forward to the results of this meet ;)

 

 

Although with my current, sleep-deprived state, I'm having crazy ideas of a monster dual-opposed Bosso setup with 24s... :lol:

 

Considering you would need 30" or larger diameter pipe to start with, that indeed would be a MONSTER. My old dual opposed tubes were 24" diameter with 18's on the top/bottom and just for a single cab, the box was close to 4 1/2' tall with the air gap on the bottom. For 24's, you would need more space than that. Thinking close to 6 foot tall cabinet to give a pair of 24's enough breathing room!!!! 

Link to comment
Share on other sites

It's time for a road trip. We'll have the A-14k meet the little 11-inchers and show them what fer.  :D

 

We need at least a 20A home run, preferably a 30A home run.

 

Either way, I'd just like to meet up at Chez Nick for a fun day of world class low end listening and measurements.  ;)

 

I'm game for it. I need to hire my neighbor to install a new run into the spare bedroom / media  room. I can have him run a 30A home run - no problem. 

 

How many other things are on the same 120V circuit as your amp?  Do you have the nuke with DSP built in? 

 

At the time that I shut down the iNuke 6k DSP,  nothing was on. The entire upstairs is on the same circuit but no one was upstairs and nothing was on except the projector, my receiver and Blu-Ray player [and the iNuke]. 

 

It's time for a road trip. We'll have the A-14k meet the little 11-inchers and show them what fer.  :D

 

We need at least a 20A home run, preferably a 30A home run.

 

Either way, I'd just like to meet up at Chez Nick for a fun day of world class low end listening and measurements.  ;)

 

Yep, bring it on. I can't wait to have my neighbor ask where the bass was coming from again. In my back yard yesterday my 35 year old rejected frat boy neighbor condescendantly asked "did you hear all that bass the other day" to which I smiled and replied with "yes, that was me." 

Link to comment
Share on other sites

The inuke 6000 can trip a powersupply protection and power cycle if you use all it has on both channels, nothing to do with what else on the line that draws power... 

On the DSP version,  setting the limiter to -2 or -3dBfs prevents it from happening to my 6000dsp.

 

Funny thing though, I recently modded my 6k for flat frequency response instead of rolled off cause I had a feeling it had more to give. 

Before I had to run it flat out to reach Xmax on my subs and it did not manage to reach Xmax below 4Hz.

Now though,  with a flat signal chain it only needs to do - 7dBfs to reach Xmax. 

So I set the limiter to - 6 to be safe, no shutdowns happening anymore and the subs use less power than I thought they needed :)

The damned coupling caps were the cause of the rolloff as I expected... 

Link to comment
Share on other sites

The inuke 6000 can trip a powersupply protection and power cycle if you use all it has on both channels, nothing to do with what else on the line that draws power... 

On the DSP version,  setting the limiter to -2 or -3dBfs prevents it from happening to my 6000dsp.

 

Funny thing though, I recently modded my 6k for flat frequency response instead of rolled off cause I had a feeling it had more to give. 

Before I had to run it flat out to reach Xmax on my subs and it did not manage to reach Xmax below 4Hz.

Now though,  with a flat signal chain it only needs to do - 7dBfs to reach Xmax. 

So I set the limiter to - 6 to be safe, no shutdowns happening anymore and the subs use less power than I thought they needed :)

The damned coupling caps were the cause of the rolloff as I expected... 

 

I am sure people other than me would love to hear more about this. Could you detail your process on measurements, what all you did you achieve your newfound performance, etc? 

Link to comment
Share on other sites

I am sure people other than me would love to hear more about this. Could you detail your process on measurements, what all you did you achieve your newfound performance, etc? 

 

Might do a thread later on about it but will try to give a  loong "short" answer, major OT and lots of ramblings on the way!!

This procedure i guess can apply to any amp if the cause of the rolloff is coupling caps...

 

First I have to say that I do not have a loadbank or something to measure the amp on, so I cannot for sure know that it indeed perform as I think.

What I do is to measure cone excursion at different frequencies with a known flat signal chain to amp, and comparing which AVR mastervolume setting I have to set to reach a certain excursion (in my case xmax)

This way I think is one way to test amplifier rolloff, if one assumes that with fixed voltage a driver will have a fixed excursion at frequencies below a sealed box resonance freq (correct?)

Any rolloff in cone excursion should then be caused by the amp if the signal to the amp is flat, if i have understood things right.

 

During one of these tests I had flattened out the signalchain to the amp but it still gave some rolloff/ drop in excursion.

I have long suspected and  hoped that the amps rolloff are caused by coupling capacitors or some other kind of highpass filter since I could to some degree "swing" it with EQ boost prior to amp, until the dsp in the inuke reached its limit ( 0dBfs).

Hoped for caps cause that could mean easy mods like bypassing or replacing caps to shape the rolloff.

 

 

 

The mod process was fairly easy yet time consuming, I have a 3000DSP as well as 6000DSP, I use only the 6000DSP for sub duty so I focused on that.

 

There are schematics on the interweb of Inuke 3000, but I have not found any schematics for 3000DSP/6000/6000DSP.

To get a feel for how things work I first studied the schematics for the 3K and compared with my 3KDSP.

I then started drawing up schematics of my own for my 3KDSP using the 3K schematics as a guide and using a multimeter with continuity tester to probe my way forward and finding which thing connects to what etc.

As I don't know shit about amplifiers I focused on the signal path prior to amplifier stage.

When I had done my schematics I compared it to the 3K and they are similar but with some differences.

 

With that knowlege I moved on to my main goal of drawing up schematics for the 6000DSP as that is what I'm using to drive the subs.

Same process with multimeter to probe and draw up schematics, following audio signal path from inputs.

Looking at the schematics I tried to figure out what things do, and succeded to some degree, did not bother with everything like protections circuits and the likes, just what could affect rolloff like capacitors in series with the signal path.

I then made some loopback measurements with REW of the amp with soundcard to amp input, and measurement points of interest inside the amp feeding the soundcard input.

First checked with multimeter to see if any DC were present at those points which could harm the soundcard.

If dc prescent I did loopback with a 1000uF cap in line blocking dc from soundcard.

With this procedure I figured out which caps did what and what rolloff it caused, finally measured rolloff at the class D amplifier chip audio signal input (pin3) which was where I decided to stop as I don't know much about amplifiers and the fact that there could be leathal DC voltages there if i get stupid and touch the wrong things, a healthy dose of caution should apply to any amplifier work though...

 

 

Anyhow, the signal rolloff inside the Inuke 6000DSP seems to be caused by highpassing/DC blocking with some capacitors.

 

First the XLR/TRS inputs are buffered in the opamp IC4 and DC protected/highpassed at C20 and C32 which are 47uF electrolytic smd capacitors.

After these caps signal goes to volume pots and back to board.

Loopback before these caps showed no rolloff.

After the cap signal was -0.6dB at 2Hz.

After the volume pot, set to no atennuation the signal was -1.2dB at 2Hz.

There were no DC on either side of the caps so I decided to bypass it, I'm feeling confident that I will not connect things to the amp input that has DC prescent, except if recorded in source and if so it will be caught by the cap in the AVR...

Also I do not use the volume pot (set wide open all the time) so i decided to bypass that too.

I soldered connectionpins to the input side of the caps and connected it with wire to the pin at the board for volume pot output.

Perhaps a bit anal with rolloff but anyway, saved 1.2dB of rolloff at 2Hz.

Loopback now showed no rolloff at this point.

 

Next there are the caps C133,C134 and C38,C39 which are 47uF electrolytic smd capacitors located at the DSP input.

These caps showed no DC voltage at the input side but 2.5V DC at the output side so these could not be bypassed but perhaps replaced with higher value caps to lower the highpass frequency...

Loopback at the input of these caps showed no rolloff so any rolloff would be caused by these caps or more things down the road.

 

After the DSP signal goes to a relay and then to the amplifier stage  through ceramic smd coupling capacitors C45 and C46.

No idea on the value of these but on the 3000 schematics they are 10uF ones.

I used these caps as loopback measurement points as they gave me a way to test measure loopbackat the input of the amplifier stage.

For DC check these caps showed around 36mV DC at the input side and around 0.1mV DC at the output side.

While it was not much DC at the input I did not think it should proceed further down the amp so I decided not to remove these but to try larger value caps instead.

I think both of the sides would be fine to connect to soundcard input but I chose the output side of the cap for loopback to be sure.

Loopback here with all DSP on bypassed showed signal to be -18.3dB at 2Hz.

I also fed the input of these caps directly from the input buffer stage, bypassing the dsp and the loopback here showed signal to be -12.5dB at 2Hz.

 

So I knew these caps would need some work if one were to lessen the rolloff.

I proceeded here with soldering wire pins to each capacitors (45,46) input and output side and connected test leads here to be able to parallell them with other capacitors of various values.

I tried the values 100,220,470 and 1000uF for these capacitors, loopback measuring between every change to know what changes it made.

Stock cap gave signal -18.3dB at 2Hz, stock+100uF in parallell gave -5.1dB, 220uF gave -4.3dB, 470uF -4.1dB, 1000uF -4.0dB.

In the end I decided to desolder the stock C45 and C46 and solder the pins to the solderpads for better support and settled for 100uF for the rest of the testing.

 

Onto the input caps of the DSP, same procedure with soldering pins and using testleads to connect various caps in parallell over the existing ones and loopback testing the changes, still using output side of C45/46 as loopback point.

These electrolytics were marked with polarity so I used the same polarity on the test caps.

Stock 47uF cap gave signal -5.5dB at 2Hz, stock+100uF gave -2.27dB at 2Hz, 220uF -1.69dB, 470uF -1.5dB, 1000uF -1.4dB at 2Hz.

 

Seeing this result I got greedy B)

Selected the highest value cap where i still saw some scaling on both DSP input caps and amp input caps C45/46.

That was 470uF for all of them and loopback showed signal to be -0.27dB at 2Hz :)

As a final loopback I chose to take measurements from the input pin of the class D chip.

On this amp that is IRS 20957s (4 of them) which has audio signal input at pin 3.

I took these measurements via a cap of 1000uF in series as i found 2.5V DC here (i think, dont seem to have written that down though)...

Loopback here showed no rolloff at all down to 2Hz, actually +0.5dB there.

I attribute that to the fact that the loopback measurements earlier showed not a dead flat response but a somewhat wobbly respons, oscillating some +-0.3dB mostly in the regions of 10Hz and downwards.

I decided to leave things be and instead of pondering about  +0.5dB signal at chip input i made some nice temporary leads, connected the test caps (470uF) and ducttaped them down to the DSP box as a temporary setup for more long term testing and put the lid on the amp back on.

 

That is basically where I am now, using the amp as usual, seeing how things turn out...

If I like the sound and result after a while I'll probably buy some caps of higher quality, open it up and do a more permanent solution with better solder job and perhaps remove the capacitors I have now bypassed or parallelled, mostly to get access to the whole solder pad which are hidden underneath the smd electrolytic caps.

 

I can also say that I did a loopback with my AVR in the signal chain, measured at C45/46 and signal was -4.4dB at 2Hz.

In Jriver dsp I added a highshelf filter of -4dB at 2.75Hz Q=0.7 which gave signal of -1.5dB at 2Hz measured ar C45/46, thought that was enough and called it quits, nearly flat signal chain...

 

I also did the excursion test I mentioned in the beginning with the amp hooked up to all the subs as usual and found that with bypassed DSP in inuke it gave the same excursion at the same AVR master volume setting at frequencies from  30Hz down to 3Hz.

2Hz needed 1dB higher MV and 1Hz needed 6dB higher.

In other words, excursion test showing -1dB at 2Hz, -6dB at 1Hz, that is flat enough for me :lol:

I did these tests at xmax so I did not test higher than 30Hz, that was loud enough that I should have used earprotection.

No reason to really as with my usual dsp settings ULF is where the power demands are, not so much in the higher frequencies.

 

Also I found out that with a flat signal chain I did not need at all that much power to reach high excursions, so I ended up setting a limiter in Inuke DSP to -6dB as that was little more than I needed to reach x-max.

My subs sadly seem to give some nasty harmonic distorsions when overdriven a bit past x-max so I see no point in pushing further than that.

Also limiting the power output on these amps are a nifty trick to avoid powercycling when the amp trips some protection cicuit when giving it the beans on both channels.

It seems to be an issue of too high current draw for the psu or something, it is written about in this test:

http://forum.speakerplans.com/behringer-inuke-nu6000-vs-kam-kxd7200-bench-tested_topic69202.html

 

 

 

 

Whew, a whole lot of rambling text but there it is pretty much, hope someone makes some sense of it :D

I'm using the amp as usual now, seeing how things goes for the time being...

If it pans out over time I'll probably re-do things for a more permanent setup and perhaps, if it is of interest write up some sort of step by step guide.

 

Until then though, if people aren't worried about warranties and are somewhat handy with a soldering iron this proceedure could perhaps straighten out the frequency response of other amps as well, use caution though, amps can be dangerous things...

 

Other gear might benefit too, I have previously bypassed the output coupling caps of my ASUS STX II 7.1 soundcard which are my audio source, so no rolloff there too.

It was not nessesary it seems as after a later inspection i found out that almost all of the rolloff I saw on loopback measurement was caused by input coupling caps at the line-in, but oh well, the deed was done and the front speakers actually sounds better in the trebel now than before with the caps in place so...

My AVR I won't bother with though, will probably replace it someday and dsp filters in Jriver do a good job in compensating its rolloff...

 

Say if one is willing to mod other gear to lessen the rolloff I see no reason to want say an Oppo 105 just based on its low rolloff figures...

Well, there is the warranty part though...

 

That is all for me now, I'm off to bed -_-

Link to comment
Share on other sites

The inuke 6000 can trip a powersupply protection and power cycle if you use all it has on both channels, nothing to do with what else on the line that draws power... 

On the DSP version,  setting the limiter to -2 or -3dBfs prevents it from happening to my 6000dsp.

 

Funny thing though, I recently modded my 6k for flat frequency response instead of rolled off cause I had a feeling it had more to give. 

Before I had to run it flat out to reach Xmax on my subs and it did not manage to reach Xmax below 4Hz.

Now though,  with a flat signal chain it only needs to do - 7dBfs to reach Xmax. 

So I set the limiter to - 6 to be safe, no shutdowns happening anymore and the subs use less power than I thought they needed :)

The damned coupling caps were the cause of the rolloff as I expected... 

 

What subs are you using? You are not using HST woofers, I can tell you that right now unless you are using 5 Hz as your reference frequency in a sealed box with a Qtc of .707 or lower. 

 

I did shut down my iNuke 6k DSP during the demo to my father with nothing else on upstairs. I'm trying my hardest to get my neighbor to install a 30 amp 110 circuit into the room but I think I might have to bait him with beer in order to get him to actually install it, haha. He's a major fan of Miller High Life. 

Link to comment
Share on other sites

What subs are you using? You are not using HST woofers, I can tell you that right now unless you are using 5 Hz as your reference frequency in a sealed box with a Qtc of .707 or lower. 

 

I did shut down my iNuke 6k DSP during the demo to my father with nothing else on upstairs. I'm trying my hardest to get my neighbor to install a 30 amp 110 circuit into the room but I think I might have to bait him with beer in order to get him to actually install it, haha. He's a major fan of Miller High Life. 

 

Oh no, i'm not using the HST drivers.

I use 8 10" drivers (SEAS L26ROY) each in a sealed 43L cab, should make for Qtc of 0.5-0.6 or something.

They are wired series-parallell in clusters of 4 subs to give a 4ohm load to each channel on the 6000DSP.

The way my housecurve EQ are set up ULF frequencies demand the most power, and rarely play louder music etc with higher frequency bass than I can keep the subs within x-max anyway.

That all makes for a max power demand of about 200W peak per subwoofer or 800w peak per inuke channel according to its limiter function set to -6dBfs...

Not really powerhungry drivers which i thought was wierd as the 6000 driven to max was not enough at low frequencies to reach x-max, well, that was the rolloff...

 

Inuke6000 according to the benchmark linked worked fine with 8ohm load and both channels driven to max.

That gives the same total power as one channel in 4 ohm driven to max which also works fine, so it is not a restriction in the amplifier modules but probably something like total powerdraw protection at the PSU is my guess as 2 channels driven to max in 4 ohm triggers a shutdown/restart.

Setting limiters to -3dBfs both channels should cut power in half, so you have 2channels 4ohm load which draws the same power as one channel 4ohm load which worked fine.

If you want more power then try -2.5dBfs, -2dBfs etc and see how high you can go before it shuts down.

It will obviously give less power to your drivers and only you can say if up to 3dB worth of power per channel is worth giving up to keep it from powercycling.

For me the choise is easy though as I don't even need that much power...

 

Too high sustained powerdraw should open the circuit breaker at the back of the amp I guess but that has never happened to me.

Think that breaker is on a timer and normal content like movies and music has enough crestfactor to not give so high sustained current draw.

Long time sinewaves at high power would probably trigger it though...

 

 

Bait him with beer!

More power drawn to the theatre room comes in handy if one likes bass :)

Link to comment
Share on other sites

Wow SB, you CLEARLY don't know anything about amps. But it appears you know everything about pretty much ALL the components that are what makes amp however.....Totally kidding. Your knowledge of how the amp works is over and above 99% of the people that use them. I think you should be in pretty good shape sir :) Excellent write-up. 

Link to comment
Share on other sites

Wow SB, you CLEARLY don't know anything about amps. But it appears you know everything about pretty much ALL the components that are what makes amp however.....Totally kidding. Your knowledge of how the amp works is over and above 99% of the people that use them. I think you should be in pretty good shape sir :) Excellent write-up. 

 

Well thank you!

I do what I can and lately I have snowed in on straightening out the rolloff in the signal chain.

Before I had done that fairly well through dsp compensation but sacrificing some 17 dB of signal headroom on the altar of rolloff compensation as I had before was no fun in regards to noisefloor :(

So I thought it were best to solve the root of the problem and lowering rolloff of components instead of compensating for it.

I did the easy first of removing coupling caps from the soundcard, that was fairly well documented on the web and went smoothly.

 

The rolloff of the inuke became somewhat of a burden when I was chasing low frequency extension, but instead of selling it and searching for other amps with very low rolloff I thought they perhaps could be coaxed into something that suited my needs.

I would perhaps learn a thing or two in the process as well :)

 

The 6000DSP though required some more thoughtwork to figure out how they worked and what the possible solutions might be.

While it was very learningful to probe my way and draw up/reverse engineer a crude schematics plan I wish that they were already floating around the internet like the 3000 non dsp were.

That would easily have cut the time in half if not more.

One trouble with more than one layer of the pcb is that connection pathways can just stop, go through a via to the underside of the pcb and pop up somewhere else.

Took some time before I found where some connections lead to...

Without the schematics for the 3000 though, I dont think I would ever have started.

As the amps are somewhat similar i the Inuke family it helped a lot to have some guidelines on how it probably was meant to be wired up.

 

The amp was fine today as well, don't anticipate any problems but it feels best to give it a month or two and see how it behaves before doing anything final.

I have another 6KDSP lying around, waiting to power more sub, and waiting for cabinets to be made for those drivers.

I'll definetly wait on modding that one until I need to use it, just in case something goes haywire with my modded one.

Best to give it some time so any eventual problems might surface...

Link to comment
Share on other sites

Nick, you should come visit someday and try to convince me to upgrade to your 24's!  I will show you what 12 of your 18's can do in an IB style layout, clone power, and a closet size sealed room!  Bass heaven.  The IB style really needs very little power as I ran EoT 10 dBs hot and used 1400 watts during that 10hz part. 

Link to comment
Share on other sites

Except they left out all the good stuff below 18hz. 

 

Yeah it kinda sucks. That movie had/has the best layout for ultra-low frequency information but they cut it all off below 18 Hz. Super huge robots stomping through town, throwing ten-story tall giant aliens around like toys, etc, perfect layout for the ultra low stuff. But at least it's a cool movie to throw in and have some fun with the volume knob. :D

Link to comment
Share on other sites

Nick, you should come visit someday and try to convince me to upgrade to your 24's!  I will show you what 12 of your 18's can do in an IB style layout, clone power, and a closet size sealed room!  Bass heaven.  The IB style really needs very little power as I ran EoT 10 dBs hot and used 1400 watts during that 10hz part. 

 

Shoot me an email. Depending on where you live I could definitely free up a few days to come listen. :) 

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...