First time amp build - Abletec buyout power supply, Sure IRS2092 board

[scottvalentin]

First time post here, but I feel like I know lots of you just from reading many of the threads here!

For background, last year I built an LCR set of Overnight Sensations (MTM ones) that I ordered from Meniscus. I have read the PE Forums and almost every thread here at Mission Possible DIY over the past year, and so am beginning to understand a bit about the speaker side of things. However, the amplifier side is totally new to me.

Therefore, I honestly know next to nothing about electronics. So first, I want to make sure I don't zap myself, and second, that I don't fry the new equipment. I'm looking to learn a lot here and have downloaded quite a few articles from Elliot Sound to see what I can learn. However, getting to the practical side, I have attached my wiring diagram (currently incomplete) and have a few questions for the minds here:


First - The Sure manual says that the ground connection for the amp board needs to have the Positive of the -54V supply and the Negative of the +54V supply. On the Abletec manual, it only lists Pin1 and Pin2 as the -54V, and Pin5 and Pin6 as +54V. Is there a positive and negative to these, and if so, which is which?

Second - Grounding - How do I go about grounding the mains input, Abletec output, and the amp board (if I do such a thing) to the Chassis? Are there separate points for each, do they all go to one point?

Thanks all for taking a look and helping me out! I have a multi-meter, soldering iron, wire stripper, etc. I also have a few odds and ends like IEC power cable, IEC female connector, Power on switch. Anything else like do i put a fuse somewhere (and how would I do that), how do I wire the on/off switch, etc.

[JonMarsh]

IR uses this kind of terminology often- they like to think of the power supply rails as individual supplies, which are only connected together on the application board/amplifier. Whereas, most off the shelf supplies for Class D do it already at the power supply. So, for the +54V supply, the minus leg is just the supply ground connection; similarly, for the -54V supply, the plus connection is just the supply ground. The reason why this isn't the best way to do things is that return currents and voltage drops for each supply become common through the common ground connection. Use dual ground wires to the IR module, and make them as beefy as you can.

[bwaso]

You have it drawn correctly, but you need to also connect the 0V pin of the Abeltec supply to the amplifier ground.

You also need to load the LOW voltage supply pins of the Abeltec (5.6V and 7.5V) each with a resistor (about 150 ohms, 1/2Watt type each to 0V) or something else that will draw about 30mA from them or the Abeltec will cycle on and off! Because these need a minimum load to keep the power supply stable.

Chassis ground connects to the "GND" pins of the AC connector on the Abeltec. This should also connect to the "green" safety ground of the AC power cord.

Bill

[scottvalentin]

Jon thanks for the input - that makes sense and I will use heavier wire from the connector lead from the two 0v pins to the amp ground. Also get well soon! Going out to cut wood for the amp case in just a few minutes!

[scottvalentin]

Bwaso,
Thanks I read elsewhere about connecting resistors to the secondary outputs and should have some arriving any day. Also, thanks for explanation on chassis grounding. I am going to make the case out of MDF, would a large bolt/washer on the bottom of the case suffice as chassis ground? Or would a metal plate work better?

[wkhanna]

i am now subscribed to this thread.
great project!

[scottvalentin]

Did some layout for sizing, cut the wood, now just need some cable and some connectors to begin soldering.

*Edit to add photos*

[wkhanna]

nice.

btw, i have had V good results using mu-metal to shield x-formers & PS's.

i modified all my equipment.
LINK to thread

[scottvalentin]

Bought some connectors (.156" pitch) 4 and 6 pin connectors from a local shop. They only sold wire in 25' spools, so instead of buying 4 spools for $40, I bought 2m of 4 strand cable for $10 and stripped it down. Need one more 6 pin connector which I will get today.

[---k---]

These amp projects fascinate me. I enjoy reading them. Please continue to post updates.

[scottvalentin]

Not much progress other than finangling wires into the clips for the connectors.

I have a power switch, and I don't know how to wire the power switch in. Does it go between the IEC plug and the power supply input? In which case the above wiring is wrong and I need to put the switch inline before the connector.

[bwaso]

Scott, the safety ground connection is just so that if you do touch any metal on a chassis, you can't get electrocuted. If there's no metal, you don't need to connect anything to the 'chassis' (or add metal to connect to, unless you want it for emi shielding), wood is an insulator. But you should still have the power supply's AC ground connected to the wall socket's ground wire (usually green in USA) for safety. I would also connect the amp's 'pwrin' ground to that safety ground (for further safety), but it wouldn't be absolutely necessary depending on local code. If the amp's input or output metal circuits might be expose to human contact when powered on, that would be best.

(edit -- looks like the safety ground would be the white wire you have on your AC power connector, if I'm seeing it right)

Bill

[bwaso]

I have a power switch, and I don't know how to wire the power switch in. Does it go between the IEC plug and the power supply input?

Yes, the power switch goes in series with the hot "live" (not 'neutral') wire between your AC connector and the power supply board.

When testing it out, be careful, that amp and supply combo can put out a lot of power at least for a short time -- best to test first time with a speaker you don't love too much.

[scottvalentin]

Hi Bill,
Thanks for the explanation. I will switch out to a green wire for the ground. Also thanks for the explanation on the chassis ground as well - safety first so I might as well do it with no shortcuts so I learn as much as I can.

I have just the speaker in mind - I have a pair of old JL Audio 8W0 car subs that I will be using for testing and until I buy an Ultimax (or 2) in the fall.

[scottvalentin]

Ok, I have wired everything up for a test. I'd like more eyes to tell me if I will zap anything if I plug it in. Planning just to use phone 3.5mm to RCA input and hook up the two speakers.

I will add the switch, increase the gauge of the PS-Sure board ground wires, and add the secondary output resistors after I've tested...but I really want to see if it will work! Can someone tell me if it is safe to plug in and test?

Thanks all

[scottvalentin]

With some help on the PE board, it was identified that my input stage was incorrect. I had the pin-out order reversed, so I will swap the IEC inputs (flip top to bottom) so red goes to pin 1 on the bottom (not the top) as in the first photo.

Also, here are some larger images that an actually be seen:

[BobEllis]

First things first. Where is your line fuse? Anything connected to the 120V mains should have a fuse. It may not protect your PSU from a fault, but will help prevent fires. As mentioned earlier, once you get it in a case, the line ground goes straight to the case and the PSU input ground goes to the same point.

It looks like the wrong IEC connector on the power end - what I think you have should be on the cord, and a cabinet mount "PEM" (Power Entry Module) on the PSU side. Many PEMs include fuses.

Another bit of safety for testing that is usually recommended is a light bulb tester. You put a regular incandescent bulb in series with the hot lead from the line voltage. A 60W bulb will glow with the initial surge, then go dim if all is well. That's saved me a couple times from miswiring causing catastrophic damage. I'm not sure how it will work with a switching PSU, but I think it should be OK if it has decent line voltage tolerance.

As far as I can tell, the power wiring looks in accordance with your schematics.

[scottvalentin]

Bob thanks for the input. The IEC plug is above - it is the female end for an extra computer cord I had laying around. How would I wire a fuse in between what I currently have between the PSU and the plug? What should the fuse rating be? Wish I had found one with the fuse built in!

The light bulb test is a good idea, I may give that a shot. I think tonight will be test time!

[BobEllis]

OK, that's expected. I use these for my amps. http://www.mouser.com/Search/Product...y693-4304.6090

The fuse should be rated about 20% over the PSU's rated draw. You may need to go slow blow or use a soft start to avoid blowing it on inrush.

[bwaso]

The Abeltec power supply already has fuses onboard.
But if there's any doubt about your wiring ahead of the board, another one on the hot wire of the AC line wouldn't hurt.

For EMI, you might want to twist the DC power wires (and their grounds). Won't keep it from working but might make for less interference to nearby radios.

[BobEllis]

There are fuses are on the outputs - I don't see any on the input side.

Good point on twisting. Meant to mention that, but I'm at an age CRS strikes often.

[scottvalentin]

As an update, I have the amp all wired up and 5 out of 6 sides of the box glued up. Just have to cut some squares that I will glue in to corners so I can screw the top on.




Here it is hooked up in my basement. The amp is running the 2 8" JL Audio subs. Those are old car ones that have been in my garage for about 5 years. I will be starting a sub build sometime in the summer. I think I might go super cheap and get some of the GRS 12" ones (22.50/driver), as they actually model very well, similar to the Dayton Classic drivers. They have 8.5mm of xmax, and this gets me to 90+ db at 20hz in a sealed box according to Unibox. I think we listen at pretty moderate volumes, so that should be fine, plus 2 of them should get me to 95+, and I can get 4 of these for the price of one Dayton Classic 15" (my previous choice). My income tax bill is due this month and property taxes next month, so the cheaper option is looking pretty good right now.

[---k---]

Wow. That's a nice screen. Looking good. :T

IMHO. Do the subs right. They last darn near forever. And one 15" will provide as much output as 2x 12". And, you need more than you realize to get that movie theater performance your screen and setup deserve.

[scottvalentin]

That is a good point that they last forever. Here's a bit more info and you can tell me if it's still a not great idea to try the cheapies out:

1. My only point of reference for subwoofers is my old Logitech Z5500 10 inch one. Pretty sure it is a one note wonder full of mud. BUT, it still added the bottom portion that was missing. Perhaps a false frame of reference.

2. Here is the model in a 90l box, which is about my 18X20X20 box (which more or less matches the size of the Uhaul boxes) and a bit larger DCS-380 for comparison. So the DCS-380, at 80 bucks, X2 is $160 US, which is over $200 CAD for me. The GRS is $45 US for 2. So, 1/4 the price.





That is how I'm doing this comparison. It's not really a budget issue, I could do the DCS-380's, but feel like I would be 90% as satisfied with the GRS, PLUS I could brag about how cheap I am.

Maybe I'm missing something - does the peak for the GRS (2.3db at about 50hz) cause concern?

[---k---]

Well, maybe I'm a sub snob. ...

I don't think I've ever seen anyone use the DCS-380 in a sub. Just haven't. Seen people use the now-discontinued DVC drivers and lots of RS and Utilimax drivers. I've never thought of the Classic as a home theater subwoofer driver. Don't know. Box modeling is one aspect, but distortion and bass quality is another. It is probably the same driver in PE's $198 sub. http://www.parts-express.com/dayton-...oofer--300-634 so, not sure if it is worth building versus buying.

Those graphs don't get me excited. Based on your screen, I'm guessing this is for movies and not music. For movies, I like things to go boom -- really loud. And I like deep deep bass extension. I like to see good output down to less than 20 hz. F3=36 The GR F3=30 seems off to me. I think the steep roll-off between 30 and 40 will make is sound much more rolled off. You could use a bigger box and smooth out the GR response and reduce the peak (I think you need a bigger box), but I doubt that will turn it into a boom machine.

Have you considered using a big ported enclosure to increase the low end output? (The GRS w/ a Fs=32hz, might not be possible)

If you go the two driver route, are you using 2 amps or now dividing the amp between the two?

In Unibox, you can use the "Save Graph" tab to overlay the graphs. If you did one with the Datyon 15" and then a second one doubling your box volume and set it to 2 drivers, you could compare apples-to-apples on terms of output.

Me personally, a sub snob, a ported Dayton 10" RS sub with a 240w amp would probably be my entry point. CJD demoed one a while back that was real nice with his Pico something or other speakers. The amount of effort to build the box is the same for an ok driver and a great driver. But, I'm old enough that if I spend my limited time building something, I want the best I can afford.

For cheep subs, might search PE for the Digger 8 thread???

http://techtalk.parts-express.com/forum/tech-talk-forum/53195-digger8-sub-diggity-dig-it-all-the-way-to-20hz?234058-Digger8-Sub-diggity-dig-it-all-the-way-to-20Hz=&highlight=digger8

It uses a cheep $50 Infinity or JBL driver. The box has way less output, but it will have deeper extension and may be funner than the sealed boxes you proposed (assuming you're trying to keep box size down). A pair of these may be more in keeping with your cheep theme and match the size of the Overnight Sensations. The driver is also available from Amazon, so might save you on the importing to Canada. There are a couple versions in that thread and different PR options. There are other small cheep sub designs around the web.

Side note, I think the Tuba Horn guys use the classic in TubaHT sub. I built one for my brother with a non-discontinued 15" DVC and it absolutely rocked the house. More costs for wood and build complexity, and huge box size, but a fun fun fun fun sub. You probably don't want us talking you into this route....

I don't mean to sound rude here, but your Overnight Sensations are on the smaller side of things. If you enjoy them and aren't looking for bigger, you might be very well satisfied with the subs you're proposing.

[scottvalentin]

No rudeness interpreted at all - I've seen what you and CJD have cooked up and drooled over the Garnacha (amongst others) for some time. After I built the OSMTMs I have this strange addiction...

I'm very happy with the MTMs, but I want (wife would say definitely not need) some impact. Plus, I've never experienced 20-30hz as the Z5500 I have never could go that low. BUT, now that I'm typing and talking about this, if I want extension, I might want to go ported? I have looked at the TubaHT...my drywall in my house would appreciate NOT going down that road.

How much room gain should I expect? Will this improve the low end response on the sealed models? I could build a ported box sized appropriately and then test between sealed and ported by just plugging the port. You're making me overthink this and want to go back to the UM-15's.

[---k---]

I'm a sub snob.... I once had my two RL-P15s in sealed boxes. I even used a Beringer DSP to dial in more low end extension. They were nice, but I wanted to see what I could get with an oversized ported sub tuned to 15hz. ... I'm not going back to sealed boxes unless I have 4 or more with a lot of eq. It is just fun to have that ground shaking extension. ... I think my sub sounds good, but have to admit that I'm not going for audiophile accuracy, but rather room shaking boom. I think my end ruslt was a good balance of both.

If you corner load a sub, you will get some health bass extension. It is very room dependent. I think Jeff Bagby has excel sheets that will model this for you. I don't think it will be enough to satisfy me. But, at the flip side, there are plenty of people putting the RS 15" in sealed boxes and expressing blind happiness.


I don't think the Dayton classic will model well ported. But, the GRS might. Try it out in Unibox.

[scottvalentin]

Alright, after some consideration, some cardboard boxes for size, and a bunch of Unibox work, your argument that these will last for at least 10 years has swung me over to doing the RSS390HF's - the Ultimax 15's are unavailable and the RSS was able to do better in a smaller box.

Here's the models from Unibox:





So 135 litres I get a box of 25 X 19 X 22 (W,H,D). I have modelled both sealed and ported in the same size box. The goal is to build it ported (and I will likely make the box a bit deeper to get some additional volume for port displacement) so that I can see if I prefer sealed or ported. Both in terms of output and also to see if I notice the "tight" "fast transient" components that sealed sub enthusiasts always advocate.

Any holes in my approach? Also, the plan is buy both since if I start with just one, I'll never be able to argue the need for another. So we'll just start at two!

[---k---]

I'm glad we could convince you. I think you'll be very happy.

You're getting a slight rise in the low end. You're going to get room reinforcement which will raise this up even more. Some people like this. A nice "house curve" so to say. Usually it is cause by using a smaller box. Make your box bigger and I believe you'll get deeper extension. My sub has a lower port tuning (~15hz) and as big of box as I could use. At the time, Steve Callas was the champion of this style and he coined it LLT (large, low-tune), though it really wasn't anything new. You can probably dig up lots of old threads. But, the big broad summary is: 1) the low end is rolls off slowly between 20hz and 50hz, which will better match the room gain. 2) it is thought by having that low tuning, you're providing more protection to the driver against bottoming. If you watch a crazy movie that has content beneth your port tuning, so in the 18hz - 20 hz range and have it cranked, it is possible to damage the driver. Now, a lot of plate amps have high-pass filters in place to prevent this, but if you use a pro amp like me. I know that 135 liters is looking large, but.... mine is ~450L (for two woofers)

Take a look at your port size and velocity. What are you planning on for a port?

What does your box look like? Over at AVS, they have some standard cube designs called the "Marty" and "Jr. Marty" and others using slot ports. Slot ports look cool, can be used as bracing in the box, and can have less chuffing. I'm sure there are designs somewhere for RS390 to copy. Might want to take a look. DIYsoundgroup has a cub you could copy: http://www.diysoundgroup.com/ported-...1/15-cube.html

Might want to model the HO version too. IIRC, the HF is what you want in a ported box and HO for too small sealed boxes. They are very similar drivers, but the HO has a thicker cone and surround to withstand the more punishment. Winnerless wars have been fought over which is better, and the answer really is different trade offs. I was reminded this when I looked at the kit and saw it said for the HO version (still could be used with the HF, just get different response).

Have you considered a sonotube. After having to move my 450L monster, I've sworn off big heavy box subs. A sonotube makes a lot of logical sense, and the only reason I can think of not to is that the wife objects to the tube look. ...

What are you planning on for a amp? If you enjoy the sealed sound, one of the Crown or Behringer amps with the built in DSP might be the ticket to add some bass boost to the low end for movies.

If you're going to build both sealed and ported, build the ported first and then to test sealed, just plug the port with a roll of socks or what not.

[scottvalentin]

I've always wondered what LLT stood for - now I know!
I will look at a few models to tune it lower, and using the "house curve" approach to make up a 2-3 db shallow slope sounds like not a bad option.

For the port in my sims, I modelled it using 2X 3" ports that are 19cm long each, tuning the box to 21hz. Airspeed looks good according to Unibox.

My box dimensions are modelled to be a bit bigger than the UHaul boxes in the previous picture. They are that size more or less to my preference, bigger ones I think might dominate the room too much (but I might try some bigger ones). I will look at a slot port as the dual purpose for bracing makes good sense, I hadn't thought of that.

The HO version has a significantly higher Fs, and my goal was to get to really usable 20hz in room, so I discarded the HO option. It didn't look as good as either of the above sims with that goal in mind.

I have not considered a sonotube mostly because of the aesthetics. I want the box to match my OS MTMs with the black baffle and natural veneer on the sides and top.

The amp I'm planning to use is the subject of this thread - 2X250w. Eventually, I would likely upgrade this (years down the road).

Lastly, that would be my plan, to build ported and stuff the port to test the difference and see which I like better.

Thanks for all your input!

[---k---]

Ah... I had obviously forgotten where this thread started.

It sounds like your on the right track.

I know what you mean about size... But, a inch increase in height takes you to 141L. Going to 25x22x22 takes you up to 156L It might not take much more in size.

If you haven't yet, you might find BoxyCad http://audio.claub.net/software.html helpful in planing out size. It can be pretty handy.

[scottvalentin]

Ryan,
I did grab boxy cad and it is pretty cool - thanks for that!

Also, I modelled the HO version, and it works a fair bit better ported than the HF in the same (or smaller) box, but has a sealed response that is not appealling at all with the roll-off starting very high (like 80-100). So, now I'm in a dilemma because I'd really like to do the sealed vs. ported comparison, but really want the deepest usable response possible.

Here are the HO models for reference. I guess I could still test the sealed box with the HO, as the goal is to determine if there is anything to the "tight, controlled, fast, transient speed" attributes vs. ported, rather than the output extension.

[---k---]

Sorry to make decisions more difficult. Lot of tradeoffs always at work.

I like the look of your HO ported model. I think that will integrate nicely into a room and be fun.

Try not to overthink it. Room positioning and integration will dominate the sound more than sealed v. ported.

[scottvalentin]

Thanks Ryan - since there is a bit of time before I order them there is way too much opportunity to overthink it! I think I like the 20 Hz tuned HO as well. X2 of course!

Again, great input and thanks for the assistance. I will start a new thread once I get box building, probably a few more weeks before that. Just putting some final paint touches on my amp box and waiting for PE to confirm they are exchanging my IRS2092 board. Once that box is all closed up, then I can go wood shopping again and start making more sawdust!

[scottvalentin]

As an update, I haven't ordered the RSS390HO's yet. My wife is encouraging patience. But, I've done some listening with the IRS2092 on my OSMTMs and it is definitely provides a much better sound than my Pioneer Elite receiver. The imaging is more precise and there is just additional clarity in the upper mids and top end that I didn't even know I was missing!

However, since I built the amp for subwoofer duty, I am actually a bit worried. Something happens with low frequencies (under 40hz, maybe under 30) where even at moderate volumes there is this ugly static/popping type of sound. It is not the subs; as mentioned this is at moderate volumes, but seems to only happen when those low frequencies are being played. Is this the power supply or the amp board and is this related to the "bus-pumping" that was mentioned above? How can I make this work for my intended subwoofer duties?

[---k---]

I wish I knew. You need to grab Jon or Matt's attention.

[JonMarsh]

As an update, I haven't ordered the RSS390HO's yet. My wife is encouraging patience. But, I've done some listening with the IRS2092 on my OSMTMs and it is definitely provides a much better sound than my Pioneer Elite receiver. The imaging is more precise and there is just additional clarity in the upper mids and top end that I didn't even know I was missing!

However, since I built the amp for subwoofer duty, I am actually a bit worried. Something happens with low frequencies (under 40hz, maybe under 30) where even at moderate volumes there is this ugly static/popping type of sound. It is not the subs; as mentioned this is at moderate volumes, but seems to only happen when those low frequencies are being played. Is this the power supply or the amp board and is this related to the "bus-pumping" that was mentioned above? How can I make this work for my intended subwoofer duties?

That would be hard to trouble shoot without benching it. It sounds like a protection function kicking in- do you know the load impedance for certain? (Measured) Bus pumping will cause an offset in the supply voltages for an SMPS source; some wil have protection circuits for that, some you just need to add some extra output supply capacitance so it can soak up the energy. This one reason why most Hypex sub aimed products (like the DS4.0, the DS8.0) have big conventional power supplies.

Last, how efficient is your sub? you may just be clipping the amp- is the amp module bridged mono output from two channels? If not, with 55V rails, it's good for about 100W at 8 ohms. Self oscillating class D amps often sound kind of ugly when they clip- that's why most of the Hypex NCORE based designs use what amounts to an input drive buffer stage setup to pre-clip before the class D modulator does. IF you main speakers are 86 dB/watt, and your sub is 80 dB/watt, you'd need four times the power to get the same SPL. That wouldn't surprise me...

[JonMarsh]

I dug up the technical details on the IRS2092 amp board. The minimum rated/suggested power supply is 60V, max is 95V. A reasonable range to use this board would be a 75-80V +/- supply. Peak output voltage without clipping isn't going to be any different than for a class AB amplifier. It does have provision to use an inverting input on one channel and convert to a balanced bridge configuration, which doubles the voltage drive.

As I mentioned in the last post, that 55V is about right for a conservatively rated 100W/8 ohms. With a sufficiently stiff power supply, 200W at 4, actually a little more; might be around 225 to 250 bridged. With 75V rails this would be 200W+ at 8 ohms, 400W at 4. I don't know the specs on your Abletec supply, how much current it can deliver- keep in mind, a 300W rated supply is for DC value; RMS is not the same and it wouldn't be able to deliver nearly that to speakers, only between 55-65%. I would be inclined to recommend pairing this power amplifier with a Hypex SMPS1200A400 (65V rails) or SMPS1200A700 (85V nominal rails); I'd recommend staying with the A400 version, as the voltage depends on the line input voltage, and a max of 98V would be higher than I'd be comfortable with. With the A400 you'd be at around 180W at 8 Ohms, 300W at 4, and 600W in balanced bridged mono.

[scottvalentin]

Jon thanks for chiming in,
Had some discussion on this with Charlie Laub on the DIY Audio page for this, http://www.diyaudio.com/forums/class...y-20-a-11.html thread here.
He thinks it is likely bus pumping kicking the protection circuit in. I haven't tested the load impedence, but I did try hooking up the subs in series off of one channel and had the same thing happen (making the assumption that it is an 8-ohm load). It is not in bridged mode.

The abletec supply is spec'd at 4.13A for 5min, up to 16A for 10ms, and notes 445W for 5 min. I don't think I am clipping it, because the subs are spec'd at 75w power handling, and they are not being pushed to bottoming out, and I can increase the volume quite substantially for music without any of the static/popping noises.

Is there any way to add that extra capacitance easily to my set-up? Or will I have to add other things as well?

Thanks for your input.

[JonMarsh]

you could try that across each of your rail outputs- you'd be looking at something like 5,000 to 10,000 uF across each rail. With 55V rails you'd need to move up to the next rating step from 50V, say, 100V. I have some Nichicon 3300 uF 100V caps (a lot of them actually, purchased for amplifier bypass caps at board level) and I could send you 6 of them, 3 per rail. You'd need to mount them on some perf board and wire them up. That ought to help, but I'd be leery of using more; there might be stability issues with the power supply, depending on how it is designed.

BTW, if you run your amplifier as balanced bridge, using the two channels driven out of phase, then you get pumping on both sides of the bus and it cancels out. This would be much preferred to running as independent amplifier channels, from the viewpoint of bus pumping. Then you would have about 400W at 8 ohms if you supply can deliver the current.

What really counts is the current delivery rating at low frequencies-

Take a look at the data sheet for the Hypex SMPS600 power supply.

http://www.hypex.nl/docs/SMPS600_Datasheet.pdf

Read the details and fine print, and compare to how others do this. OK, I'm going on a bit about this, but Hypex is one of the few straight up companies in this business, IMO, and most of the others are just trying to price the product as low as possible, and waffle about the specs.

Now, I need to do a little bit of personal ranting- I HATE the way that most companies try to spec power supplies and cut corners for Class D supplies by claiming specs like 10ms power (that's meaning less for audio unless you listen to high power clicks). That's just the joules stored in the supply output caps. Even the acquired IR team does that, in order to try to cut corners on what needs to be built into a power supply. So we have long term continuous power ratings (such as 115 watts for a so called 1kW power supply) and short term output power (1 minute minimum, say, 580 watts, corresponds to your 445W rating, I would guess). But when a supply is spec'd for low frequency clipping, at full power, the supply must deliver not the average RMS power, but the peak power required by the waveform into the equivalent load.

Now, here's the thing... think about this a minute... If you're suppling 1kHz from an amplifier, the local bypass caps at the output absorb the AC stuff, and all that is required of the power supply is the average DC power. If you're suppling 20 Hz or 30Hz from the amplifier, the local caps built into these SMPS are small and can't supply diddly at that frequency, and the supply must track the continuous power demand, which for an X average level is 1.414 times higher at the peak of the waveform. So, to supply the same average power that will suffice for 1kHz output, the supply has to track to over 40% higher on low frequency waveforms. So for what might be 100W average output power, the power required by the SMPS is about 150W. This is understood and taken into account in Hypex power specs- they specify based on amplifier output power at 20Hz, not the output power of the supply into a DC resistive load!

Now, as regards your sub speakers, they might be rated 75Watt long term power handling, but easily handle 100 or 150 on short term transients- that depends on the overall driver motor design- long term power handling is continuous thermal dissipation limit of the voice coil, which is not the same thing. Are they DIY or commercial? If DIY, do a Unibox analysis of the driver and box design to see what the real acoustical power handling should be before exceeding Xmax. Go look at Scanspeak or other driver power specs- you'll usually see a 2 to 4 to 1 difference between acoustical power handling and long term thermal power handling. (what kind of sub only handles 75W? just curious... ) (and what's the sensitivity?)

Quickest short term fix is to configure for bridge operation and wire your sub to that, or series them together being driven in bridged mode.

If you want some extra supply caps, PM me with address and I'll UPS them to you.

[scottvalentin]

Jon that's a great explanation and thank you for the generous offer - I will PM you.

The IRS board has a switch that moves it to bridged mode, (and then you wire the + on one channel with the - on the other to the speaker terminals), but I'm not clear how you run both channels in "balanced bridge mode" and out of phase. Can you clarify how I would do that? I can run bridged mode and wire the two subs in series as well.

The subs are JL Audio 8W0, specs are here: http://mediacdn.jlaudio.com/media/mf...pdf?1317764946. Efficiency listed at 85db. They are old car audio subs I had in the garage for probably 5+ years (I'm sure it's longer) and they are in crappy, too small, sealed boxes with likely no stuffing or bracing (I bought them used). The plan is to get two RSS390HO's and use this amp for those subs (see Unibox models above for those). I need to solve this low frequency issue before I pull the trigger on the Dayton's as my main goal was to get usable 20hz output for movies!

Scott

[JonMarsh]

Always gotta read the fine print- that's continuous RMS power rating for those subs- depending on who you talk to, the actual non clipping power with music has a crest factor of 4:1 to 8:1 over RMS continuous average power, so let's say you've got a rating of 60W RMS, that means with music but not clipped a 300W amplifier is not unreasonable.

The IRS board has a switch that moves it to bridged mode, (and then you wire the + on one channel with the - on the other to the speaker terminals),

That's what we're talking about, I just use fancier words and more of them... :W

Suggest wiring these in series and driving them bridged to start-

What's the impedance of your subs? 8 ohms? 4 ohms? The amp module should be able to drive them in parallel, but the power supply probably not.

[scottvalentin]

I'm not sure if they are the 8R or 4R versions of that sub since the guy who built the boxes siliconed (or tarred) the subs in, so I tried to take them out of the box the other week without success to see for sure.

I will bridge it tonight and try it in series and see how that goes! Thanks again Jon!

[scottvalentin]

Just as an update for anyone who might be interested, I have wired the subs in series and for bridged mono operation, and this has improved things greatly. I think there is still some bus pumping that occurs at higher volumes, but it not longer occurs at my normal listening volumes. Jon has generously sent me some capacitors which I will wire in when I receive them in the mail.

Kudos to Jon for his help (and also to Charlie Laub on DIYAudio).

[scottvalentin]

SUCCESS!!
Jon I have the 6 caps wired in parallel, 3 on each rail as you instructed me and it works like a champ! Turned up the volume to far past listening and only the subs were struggling, not the PSU. No clicking at the PSU, no cutting out, no issues from the PSU/amp side at all. Now I just have to figure out how to get the subs out of the old boxes. They are glued in with black stuff that looks like shiny tar. I didn't build those, they came from someone's car years ago, but they are working impressively well in too-small, unbraced, unstuffed boxes.

Very pleased with the amp and with the performance of the PSU now. Thanks to all! I will have finished pics up in a few days if I get some time to flush trim the lid of the amp.