So I'm going to be tackling an active speaker project

[Amphiprion]

This is just a precursor to what will be the main build thread which I'll start later, but I've decided to do an active loudspeaker build. I have several nice components that have been sitting around for literally years (Seas L15 and 27TBFC/G on the shelf for 5 years, Dayton RS315HF on the shelf for 4). I don't really intend for this to be something for someone else to attempt building, mostly because it is NOT cost effective in any way. It's just a design exercise for myself using parts I have on hand that I would otherwise have to sell at a loss and ship.

The L15 and 27TBFC/G will be going into a Dayton 0.25cf cabinet and will be a sealed design crossing over to the RS315HF in a sealed 3.0cf cabinet also from Parts Express (2.1 stereo system). I'm in the process of getting baffles cut right now, which I hosed up last weekend and had to order some new ones from PE. But if anyone is interested in seeing a totally active design done from scratch, keep your eyes peeled for this one. My day job is high-rel aerospace electronics schematic and PCB design which goes way beyond anything required for analog audio, so I should be able to generate a quality design (and I have a reasonable amount of previous passive xo design experience). Practically everything I do has an analog front end at work, so no worries about me being some digital-only guy.

Design will be done with Praxis, LspCAD 6 standard, and Altium Designer for the schema/PCB. I won't be selling boards other than the leftovers from whatever minimum amount I have to buy to get the good specials at Advanced PCB, but gerbers will be made available for anyone who wants them. Remember though, I am not designing this to be cost effective, I'm just using what I have on hand.

Feel free to ask any questions.

[Saurav]

Will your PCB design make any attempt to be modular / generic, so it can be used for other active XO projects? Basically, something like Linkwitz's ASP boards in concept?

[Amphiprion]

Nope, absolutely not. Too much hassle, that's a project in and of itself. I may make the power supply separate from the active xo board, or I might consolidate them so I only have to do one board spin. If I do them separately then the power PCB would be usable for other projects. I intend on having a fairly nifty power supply, maybe even microprocessor controlled (think a little 8 pin PIC for power supervision, etc).

Like I said, this is just a design exercise for me. I don't really intend for anyone else to build this.

[Amphiprion]

Also I am likely to make extensive use of surface mount and/or fine pitch devices since I am building this myself. Probably not going to be very user-friendly for those who can't sling solder.

[Saurav]

Nope, absolutely not. Too much hassle, that's a project in and of itself.

Thought I'd check And yes, SMD pretty much writes it off for me, I doubt I have the soldering skills (or tools) to do that by hand.

[Amphiprion]

It's really all about technique until you get to 0603 or 0402 size parts. Heck I use a 1.5mm chisel tip for everything from TSSOPs to 0805s to soldering 12ga wire to binding posts (love my OKi rework station). Smaller than 0805 - then you need a microscope. But with 0805 and larger and SOIC or TSSOP parts, it's all about technique. Very little practice needed to get it right. A decent set of tweezers, lots of liquid flux, drag soldering method for IC's, and an alcohol bath afterwards. I actually find it easier than through-hole stuff, especially when it comes to rework. If I have a spare board I'll send you one with some leftover parts if you want some practice.

Actually I have some old practice solder boards already, PM me your address and I'll send you one with some cheap spare parts I have lying around. I need another person out there singing the praises of SMD technology The boards have spots to practice standard 1/4 watt axial resistors, 1206 resistors, 0805 resistors, 8SOIC components, and 28TSSOP components. I got a bunch of 28TSSOP components that failed quality when I was working at Cirrus, if I can dig them up you can practice all day long on .65mm pitch IC's at no cost (I got them for free).

Drag soldering video (done with a wave tip here,but you can use chisel tips, hoof tips, whatever really):

http://www.youtube.com/watch?v=xP1893Nhg_0

Note initial tacking of corner pin(s) to hold IC position. That's the only hard part. Then liquid flux application, get some solder on the tip, and drag across leads. Clean with 91% isopropyl alcohol from the pharmacy section of your local grocer. Here's another:

http://www.youtube.com/watch?v=wQXhny3R7lk&NR=1

This shows how fast it can be done, way easier than doing through hole. Gotta have the extra liquid flux though. Without it you are doing to get pin bridges and all kinds of cold solder joints. Extra flux is the secret ingredient.

[Saurav]

I've seen the videos, but I learned long ago that experts can make anything look easy, and it doesn't always turn out the same way when I try it This goes for soldering, woodworking, anything. But yes, getting a through-hole IC out because I put it in backwards is a royal PITA, so I can see how SMD would be easier for rework. If you know what you're doing

[Thooms]

I'll be watching this thread with interest!

Working on a similar project myself - will make a thread once the chassis for the amps and power supplies is done.

Will you be building amps too?

[Hdale85]

SMD isn't THAT bad, just practice some If you have an SMD iron though it certainly makes things easier.

[Amphiprion]

Probably no amp building for this one. Trying not to let 'feature creep' keep me from getting this one done, as has happened with so many past projects.

If you have an SMD iron though it certainly makes things easier.

I've got an OKi MFR-2241 station with a nice iron and awesome hot tweezers. Once you use hot tweezers you will never go back to through hole. It makes depopulating components so much easier. I can get an SOIC-8 off in a manner of seconds, same with the 28TSSOP. If I had to deal with 28 pin DIP's and no tweezers I would likely be destroying components as well as pads/boards.

[---k---]

To make this really interesting, you need to build both a passive crossover and a fully active and then compare them. :P

Looking forward to seeing your progress.

[Amphiprion]

I probably will, a passive crossover is much easier design-wise than an active and I will already have the target transfer functions to try and match. Although doing comparisons between the two might not have much validity b/c of the inability to do a real double blind (or even single blind) test.

[Thooms]

A passive crossover is easier to design? I would've thought the active would be much easier as you don't have to factor in messy inductors and impedance peaks etc.? That was what attracted me to active in the first place - it's a lot easier to match the 'ideal' behaviour with small caps and opamps I would have thought?

I hear you on the feature creep. It's got so bad on my project that I'm waiting to turn some new control nobs on the lathe just to have LEDs in them...

[Amphiprion]

I've designed several passive crossovers, and for THIS speaker it should be easy to design. Heck I could probably cobble one together from Zaph's projects that have used the L15 and the similar 27TDFC cloth tweeter without doing measurements and have it sound at least listenable. Not that I would ever do that, but I've seen others work with these or similar drivers and I haven't seen people resort to drastic measures with them. L15 breakup is divinely controlled, and the 27TBFC/G is pretty easy to work with as far as tweeters go.

[cjd]

Hey Mark - this sounds like fun, and will be interesting for me to watch as I know I'll be building an active stage for at least part of some upcoming projects.

Knowing the quality of work likely to result and the current lack of good flexible-ish active stuff in the market for the DIY enthusiast (non-digital...) - what can we do to sweeten the concept of something a little more modular?

It's gonna be fully balanced, right?

whee. I can try, eh?

Where I think I have the most trouble is in getting low-noise solutions. Though I shouldn't be soldering surface-mount stuff either, to be honest. I really need to put together a straight circuit from one of the PCB's I did, rather than trying to dive straight into a more complex usage. Just to see if they're inhrerently noisy, or I broke something somewhere else with a bad job of something or other.

C

[Amphiprion]

Actually, once this project is complete and I can show off a little of what I can do, I might be open to working with others depending if our goals line up for whatever I want my next project to be. But I better not get ahead of myself Since I won't be able to cut on baffles until this weekend, I better get going on schematic. Starting with the power supply I guess.

[---k---]

I probably will, ...

I was just causing trouble. But, I think you knew that. But, I'm glad you will do it. Should make it even more interesting.

[Amphiprion]

OK well I think I am going to drop the microprocessor controlled power supply... just started getting too complex. And needlessly so. So it'll be a little more basic. Right now it's looking like basic dual secondary transformer with dual bridge rectifiers, Panasonic FM filter caps, and LM1086 preregulators to knock down the unregulated output of the bridge and filter caps to a more appropriate +17V for the TL1963A regulators (which can only handle 20V input max). I'll be using positive regulators to generate the negative rail if anyone's familiar with that trick, also helps with BOM consolidation.

[cjd]

The power supplies I have here (pretty basic regulated supplies only a little beyond what most chip-amp builders use) use a dual positive regulator. MUR1606 diodes (I think that's the right model - should look it up) and a pair of 10k caps behind the regs, target output voltage 25-30V. They seem to work quite nicely.

Haven't run them at lower voltages. Were I designing them again I'd o things a bit differently.

Simple is good. Very good.

Aside from size, is there any compelling reason to limit things to only allowing surface mount components? I know sometimes you can actually get better parts that way, not to mention other occasional tricks (I used one to put a cap between the front and back legs of one of the LM3886 IC's... and I think I killed half of them soldering 'em... needed lower temp solder I think)

C

[Thooms]

Haha - good ol' 7815 and 7915 here

[Amphiprion]

Hey, they honestly perform well enough Thooms. I am just being fancy

Re: SMT, I only intend to use it where it has an advantage, but that advantage could be anything. Surface mount tantalum and MLCC caps are by far superior to their leaded counterparts and are great for bypassing. Some regulators are easier to come by in D2PAK-5 packages than in bent and staggered lead TO-220-5 packages for example, so availability also plays a part. I like using quad opamps in SOIC packaging because I can pop them off without damaging the pads (can't say the same for quads in DIP packages). So for me, advantages can also include ease of rework. Each part is picked individually based on any number of factors. AVX makes a line of 1% C0G ceramic caps in 0603 and 0805 packages, which might be nice since the Panasonic poly radial 1% caps have a discontinuation notice (I think those Panny's are what Linkwitz uses in the Orion, and they are nice caps). But I am rambling.

Also if anyone is wondering about me posting during work, I have a little more time to post this week even though I'm at work because I'm waiting on techs to repair some blown up power supply boards that I'm in the process of debugging. The only through hole parts on those guys are the headers and connectors, everything else is SMT and several pieces have to be done with hot air due to thermal pads under devices. I won't be using anything like that, using hot air to put on a part is work for actual professional technicians. All the SMT parts I use on this project will have leads and/or tabs that you can hit an iron to.

[cbark]

Lots-o-reasons to go SMT, as Mark mentioned above, the MLCC are far superior to their leaded counterpart. Mainly due to greatly reduced lead inductance that can really hurt you in a bypass situation. Also as cjd mentioned, the size is reduced, but not only the overall end product size, but as it relates to the distance to the corresponding circuitry and the overall loop created by it. All of this equates to lower noise, and higher signal integrity. And that is just a benefit of the MLCCs, let alone all the other types of components within the circuit. Some additional benefits may not apply to this particular project though, but still worth the effort to me.

Mark, funny you mention the SMT power supply. I am working on one right now that is an older legacy design of ours that we replaced the gigantic TO-247 MosFets and heat sink assembly with all SMT Fets on a standard FR-4 circuit board minus the heatsink. The new SMT layout runs cooler and is capable of greater surge due to the newer SMT Fets and more compact layout. It is a little of a PITA to re-work though, but when compared to pulling a fan and sink assembly just to get to the Fets, maybe not so much.

...and the project sound cool. :T

[Hank]

Well, Kid, you're back! And with a "different" project. I've been wondering when you'd get around to an active design, no doubt giving up on helping me with my procrastination project (the hybrid line array). Did I tell you that it's on hold because of reviews I read about the new 1.7 Maggie? On a biz trip to St. Paul, I drove over to White Bear Lake and took my second tour of the Magneplanar plant. I plan to audition a pair here in Austin. If I'm not ecstatic about the sound, I'll go back to the line array.

PSST: hey guys, Mark LOVES digital :rofl:

[Amphiprion]

Hey cbark,

Not only lead inductance, but you can't even get the newer technology parts like polymer tantalums in leaded packages at all I don't think. It makes me laugh when audio guys can't get their 20MHz GBW opamps to be stable. I routinely work with 100MHz+ GBW opamps at work and have never had bypassing issues (always used 0402 ceramics). But beyond that even IC availability is starting to suffer in leaded packages. One regulator I like to use is the TL1963A, and TI doesn't even offer the TL1963A in a TO-220-5 package at all, you have to go back to the original non-knockoff LT1963A from Linear Tech to get it in that package, and who knows when it'll be NRND.

TO-247? My boss would kill me if I tried using a package like that, the most I can get away with is D2PAK size stuff It's crazy what they are doing with MOSFETs these days, and those new GaN FETs will only make it better. I used some PowerPack SO-8 6x5mm devices (same pinout as an SOIC-8 but only 1mm high with an overmolded plastic package and a thermal pad, Vishay makes them and some other companies make an equivalent package). Crazy performance out of a teeny little device, like a 60V N-channel that does 60 amps, in basically the same footprint as a dual SMT op amp. Pain to solder or do rework on though, I really don't like having to deal with thermal pads that you can't hit with an iron.

Hey Hank! Yeah, I was kinda wondering when that active line array project was going to get going. Guess I moved on with my own stuff And yes I do love digital, but this will be a totally analog system with the exception of maybe some control circuitry.

[cbark]

Hey cbark,

Not only lead inductance, but you can't even get the newer technology parts like polymer tantalums in leaded packages at all I don't think. It makes me laugh when audio guys can't get their 20MHz GBW opamps to be stable. I routinely work with 100MHz+ GBW opamps at work and have never had bypassing issues (always used 0402 ceramics). But beyond that even IC availability is starting to suffer in leaded packages. One regulator I like to use is the TL1963A, and TI doesn't even offer the TL1963A in a TO-220-5 package at all, you have to go back to the original non-knockoff LT1963A from Linear Tech to get it in that package, and who knows when it'll be NRND.

TO-247? My boss would kill me if I tried using a package like that, the most I can get away with is D2PAK size stuff It's crazy what they are doing with MOSFETs these days, and those new GaN FETs will only make it better. I used some PowerPack SO-8 6x5mm devices (same pinout as an SOIC-8 but only 1mm high with an overmolded plastic package and a thermal pad, Vishay makes them and some other companies make an equivalent package). Crazy performance out of a teeny little device, like a 60V N-channel that does 60 amps, in basically the same footprint as a dual SMT op amp. Pain to solder or do rework on though, I really don't like having to deal with thermal pads that you can't hit with an iron.

You are completely right, I deal with a lot of End of Life products, due to the overall life span of a product. Hence the TO-247s, we still use some big modules in some of the bigger power stuff though.

But now a days, as you say the MosFets are crazy good. We can do some pretty good power in a small area that would have been impossible years ago. I think you get into much tighter density than I work with, the leadless stuff is not prefered around here. Almost have to use them sometimes because that is all that is available.

OpAmp stability, Ha! they should try getting their circuits to work in some of the noisy environments we have to play in....

[cjd]

Nice info on SMT. This is the very edge of what I consider viable for me hobby-wise, and there's not much help for folks like me except people like you.

C

[Amphiprion]

Hey cbark,

What type of stuff do you work with? My job has me personally doing everything from power supply design, to digital layouts (FPGAs and DSPs of all sizes), to analog front end design. We're a little boutique electronics design house, so I get to get my hands in a lot of stuff. I even hack a little code from time to time (only C though, haven't learned any VHDL yet).

And yep, most everything I do is super high density. I work in internal research and development at my company, so everything is new - practically no legacy concerns unless we have to interface to something else. The stuff we are doing now is getting to the point where we won't even have our techs attempt it, we just send it out for pick and place. And our assembly house isn't liking us all that much either I would be using 0201 components if my boss hadn't explicitly forbid me from going that small on anything, says it's just not worth it if we have to rework anything in house.

Anyway, back to the schematic tonight.

[cbark]

Hey cbark,

What type of stuff do you work with?

Mainly power supplies, power conversion, and lots of smaller microprocessor stuff. Design house would be cool to see all the different types of projects. 0201 would never fly here either, I love the smaller stuff when doing layout, you can jam so much stuff right where you need, instead of compromising the board to fit the parts. Damn laws of physics.

I do get to mess with digital and analog stuff, usually mixed digital/analog plus switching supplies all on one board. We have digital power supplies, analog only, and lots of communication stuff. Usually all in a very noisy industrial environment and sometimes outdoors. We play in the 10kW+ range sometimes too, and also the several hundred amp on a circuit board world, but we use standard board construction.

It's was funny when you mentioned the op-amp thing a day or so ago. One example popped in my head, an analog sensing circuit with op-amps placed on a power supply driver card that feeds a digital control circuit. The driver card sits on top of an H-bridge that swings 600V with very high edge rates at 100Amps. I would like to see some of those guys get their audio amps to even turn-on in that environment. I have had the luxury of playing with some very unique designs.

Fun stuff, love to hear what other people are working on...

[Amphiprion]

10kW+? Holy crap, I haven't seen anything like that since my days at a fuel cell company. And even then I was just doing the control and monitoring of a fuel cell, which is kinda like a really poorly regulated Dc voltage source, and not anything at all like a switcher. I wonder if some my my stuff would even work in such a high di/dt and dv/dt environment..... that's hardcore. The only real thing I have to worry about is radar beams blasting my stuff, and we're generally well shielded. And sometimes there are rad hard requirements as well. And that's really all I can say about the actual work I do

The PE shipment of the extra baffles and the sub cabinet + other extras came in yesterday but I haven't unboxed them. I'll be routing baffles for everything this weekend and should have some pics taken. Also, instead of the microprocessor controlled power supply, I think I'm just going to use window comparators based off cheapo LM393 comparators to control regulator enables and high/low side MOSFET's to indicate via LED's whether power is good or not.

[cbark]

The only real thing I have to worry about is radar beams blasting my stuff, and we're generally well shielded. And sometimes there are rad hard requirements as well. And that's really all I can say about the actual work I do

Ha! You are one of those guys. :W I know nothing about the HiRel stuff, other than it better work for a real long time.

[5th element]

Surface mount devices always seem daunting the first time you consider using them. However after having played with them for a while you start appreciating their rather large benefits. As I make my own PCBs, one of the major benefits is the reduction in the number of holes you need to drill. And I can only echo the comments made about bypassing with surface mount caps, you save so much space!

I've recently been building some LED drivers around the LT3474 and it amazes me every time how much power they can shove through a tiny SMD package with a power pad. Of course the high efficiency helps.

Active xovers are a lot easier to design with, especially if you're using a delay circuit on the tweeter. In fact you must use a delay circuit on the tweeter, otherwise you're losing one of the large benefits an active circuit can bring.

Set the acoustic target of both the mid and the tweeter to the desired filter characteristics then alter the delay circuit until they are perfectly in phase. And also you can design for variable baffle step if you want. I'd recommend that you use trim pots in place of all the resistors in the active filter sections. This way the xovers are completely adjustable. Of course if you're going to do the majority of the design work using the digital xover emulater in LspCAD, then having an adjustable analogue active xover might not be as much of a benefit.

When Saurav asked,

Will your PCB design make any attempt to be modular / generic, so it can be used for other active XO projects?

This might not be as obvious as it could be.

If your boards end up having 2-4th order electrical filters in combination with time delay on the tweeter, baffle step compensation circuits, variable gain and a notch filter or two, that'd work pretty well with 95% of three ways.

I don't know how you're going to be handling the bass with this, but a Linkwitz transform circuit might be worth including just for kicks.

[Amphiprion]

Hey 5th,

Isn't Linear Tech awesome? They are up there with TI for my favorite IC providers. They make such awesome little switchers and LTSpice is just the icing on the cake. Makes doing small switchmode design sooo much easier for us prototypers. I haven't played with their LED drivers but I use a lot of their buck/boost/inverting/sepic type regulators. Just did a board with an IC from them in an inverting coupled dual inductor topology in fact, works really nicely.

I thought about a Linkwitz Transform, but the Q of both the 2-ways and the sub is going to be below 0.7, so it's really of no benefit to this system in particular. I might throw it in for grins though.

Also, I am back to the microprocessor controlled power supply. I realized I was being obtuse and making it way too hard. All I need with my current method is a common mode choke or two for isolation.

[Amphiprion]

OK, I am cutting baffles today! The baffles for the L15 and 27TBFC/G two-ways should be done today. I got the subwoofer cabinet from Parts Express and it seems a shame to just put a 12" sub in it. It really looks like a 15" would be better. So I might buy a 15" like an RS390HF to go into it and just sell the RS315HF.

I will have the microprocessor controlled power supply finished sometime next week and will post a schematic. Right now it consists of a toroidal power transformer with dual primaries that can be configured for either 115 or 230V operation. Dual secondaries each have a 2 amp KBPM bridge rectifier feeding two 1000uf 50V Panasonic FM filter caps. Then a special overvoltage protection circuit to make sure no more than 30V ever makes it to the LM1086 regulators (in case me or someone else ever sets the 115/230V switch wrong) or if someone forgets to take into account transformer regulation at light loads or something. If you want a preview of the circuit look for National app note AN-1533. I've used it in other projects and works well. Then LM1086 series regs in TO-220 packages with big heat sinks and with 10uF tantalum output caps with output voltage tweaked to 16V. Then each LM1086 feeds an LDO TL1963A D2PAK regulator heatsinked to the PCB. The regulator output voltages are divided down, op amp buffered and fed to a PIC 18F4620 microcontroller. The microcontroller checks the 16V output of the LM1086's before enabling the outputs on the TL1963A. A single digital supply/dual analog supply pair of DAC's supplies a supplementary voltage to the ADJ pin on the TL1963A to tweak the output voltage to 15.0 volts exactly (opposed to the standard 1% tolerance on the output).

This is way overkill and totally not necessary in any way, but I am just having fun because a) I can whip up stuff like this real fast because it's my job to do stuff like this for a living, ultra fast protoypting and b) I like to show off neat things I can do even if they're not necessary and c) someone looking at the project might find a part of it they like that they can use elsewhere (that overvoltage protection ckt from AN-1533 is really really useful and can be easily modified to provide additional undervoltage protection and current inrush limiting in other apps).

Anyway, I better get out to the garage and start routing. I wanna watch the A&M and UT games tonight so I gotta get done before then.

[Thooms]

I'll check out that circuit, it sounds interesting

Did my first bit of SMT soldering, not as scary as I had imagined! High density packages are a bit of a bugger, but the low density stuff is almost fun

Keen to see the boards for this...

[Amphiprion]

Hey Thooms, what were you soldering up?

I got the baffles routed (with backside chamfer for the woofer) and the binding posts installed in the two-way cabinets this weekend. I'll be adding foam to the walls and some polyester stuffing as well this week, and soldering up the twisted hookup wire to the binding posts and the drivers. On an interesting note, for some reason the magnetic grills don't really want to stick to the new replacement baffles I ordered. Not that I really care, speakers look better naked anyway Just a teeny wtf moment from this weekend.

The boards for this are probably months away. I have to get Praxis out of storage and get some decent measurements taken, then crossover design in LspCAD, then actual schematic and PCB layout. I'm probably going to be doing a 4-layer board on this one. Advanced Circuits has pretty cheap deals on class 2 four layer boards.

Anyway, it's work time. I have to write a TPS document this morning believe it or not. And our use of the TPS initialization predates the movie Office Space by many many years.

[Amphiprion]

Um, did someone not change the DST on the forum? My post above was at 8:54am.

[Hank]

Getting kinda picky aren't you kid?

I have to write a TPS document this morning believe it or not.

Did you get the memo about the new TPS cover color?

If you need PE baffles, I've got several on hand. Last year I was routing holes in baffles for some guys and also assembling drivers, crossovers and installing in PE cabinets. Turned out that PE had a supplier error in that there were baffle boxes marked with one size P/N, but contained another size baffle. I returned 2 or 3 different times and finally got a tech on the phone who walked into the warehouse, opened boxes and agreed with me.

Confirm that you can't do a Marchand-type board/boards design that could be component-tweaked for different drivers and alignments.

[Amphiprion]

Confirm that you can't do a Marchand-type board/boards design that could be component-tweaked for different drivers and alignments.

Oh I could, but I'm not going to That's not the point of this project. This is an exposition of what I can do, what can be done, if anything to inspire others to try the same.

[cjd]

In other words, he's going to get one of us to do the dirty work for him on that. :P But he'll help, uh, oversee things...

C

[Amphiprion]

LOL Actually after this I might try my hand at active DSP based crossovers. Something like the Analog Device's ADAU1442/ADAU1445/ADAU1446 family of chips. But those are the type of overzealous endeavors that never get finished.

[Amphiprion]

OK, speakers are fully assembled! I find build threads tedious and boring after building speakers for over a decade now, so here are the finished pics. Besides, there's not a lot to slapping together some drivers and some Parts Express cabinets anyway. The subwoofer in the pic is my old Peerless XXLS with dual passive radiators that I've never actually listened to despite building it over five years ago, and it won't be the one that goes with these two-ways (I've got a PE gloss black sub cabinet to match these mains).

Anyway, time to start actually doing some measurement and design work....

[Johnloudb]

Looks real nice Mark. I like how the binding post plate matches the driver color.

What amp is that in the picture?

[Amphiprion]

It's my old triple nickel, aka an Adcom GFA-555mkII. Bought it used for a steal from the original owner back when I was in college. I might use it for sub use as it'll bridge into a 4ohm load, but you gotta watch your volume levels - it's not technically rated to do it (it is rated to bridge into 4ohms if you got the far less popular pro model that had a fan).

[Johnloudb]

Funny, I just bought an Adcom 535mkII used for $129.00 to replace my our Counterpoint for TV duties.

Have you decided on a sub yet?

[Amphiprion]

Not quite yet. I think I've got it narrowed down to two, but I don't like the porting options I have for the enclosure. I need some elbows

Also I have been looking at your Counterpoint thread, actually have the schematic printed out here in front of me right now, but haven't had time to really dig into it yet. Troubleshooting amps isn't exactly my specialty, but I'm gonna look at it real hard