S.M.S.L DA-9 vs AO200?

[Shuna]

go to Tokyu Hands, probably the one in Shibuya they have a pretty wide variety of silicone feet, rubber can degrade rapidly in Japan's climate and either crack or stain. They have a lot of different solutions for "earthquake proofing" your speakers, or other hardware that you might not find at other DIY stores. You know what DIY means in Japan. You assemble a pre-cut, pre-made kit. Not truly DIY. Also Tokyu Hands in Shibuya can cut lumber to within 0.1mm at 3 meters. Just give them precise measurements and they will cut the material you selected and they will also ship it to you for a reasonable price. It's much better than the local DIY store that can cut wood but rarely to exact dimensions. I've had to teach the staff at other places on how to set up a table saw to make repeatable precise cuts. Basically they don't want to do precision because they don't have the time or training.

Of course you can pick up parts from Amazon if you know exactly what you need or want.

I used to go to the Tokyu Hands in Ikebukuro every once in a while, never been to the Shibuya one though, will check it out.
Yep, definitely agree about DIY in Japan, but considering the living space, I can see how it would be difficult to do 100% DIY; not enough space to store any big equipment, probably can only store the tools to build up a pre-made kit.

 

[jokan]

Not to mention woodworking tools make so much noise and dust. Tools cost 3-4x in Japan like power screwdrivers/drills/routers but i think that's a reflection of the lack of demand for tools, it's cheaper to buy from Amazon USA and pay the import fees/tax than to buy the same tool from amazon jp. It's crazy.

 

[Shuna]

The houses being cramped together in Sumida-ku doesn't make anything better.
Can't do any spray painting without completely covering up the enclosure, which isn't pleasant during Summer in Tokyo.

 

[jokan]

The houses being cramped together in Sumida-ku doesn't make anything better.
Can't do any spray painting without completely covering up the enclosure, which isn't pleasant during Summer in Tokyo.

Surely there's a space by one of the many rivers? I've spray painted in the nearby park in the dead of summer because the sun definitely speeds up drying time. Also, painting in your apartment is highly ill advised! Not that it has stopped me from doing light paint work.

 

[arkelectron]

Hey @arkelectron , did you try different speakers?

I experienced continuous popping noise (much worse than what you show in your vid), with a Loxjie A30. It is also an infineon chip based amp. I sent it back thinking it was defective. Months after, I got the AO200 to use with a different pair of speakers (DIY bookshelf with Seas drivers) and was very happy with its performance. When I connected the previous pair I was using though (C-note), I experienced the same noise out of one of the channels.

I guess there is some kind of weird interaction with the C-note crossover (or how I assembled it) and the infineon chip amps.

I don't have any other speakers to test it with.

Here is a video of the hiss on the replacement unit. It also has a pop on the left speaker like the previous unit.

 

[jokan]

pick up some ultra inexpensive working speakers from anywhere that you can think of.
Your crossovers are very suspect, as in the wiring is probably wrong. I build speakers as a hobby, and a few of my designs have been sold in kit form in japan. Crossovers can be tricky to wire. one wrong connection and it will make cause clipping, a dead short circuit. Yes I've made the mistake myself more than once. usually when I'm working late at night.

Find some cheap used 2 way speakers, one only is really all you need. a speaker with a woofer and a tweeter. Make sure that it is guaranteed to be working. use it to test your amplifier and therefore your speakers.

This is not normal behaviour for a speaker or amplifier. The easiest method of troubleshooting is to use a working, known "good" speaker. Try and find a 2 way speaker. It's actually not important if they are 2 way or not. All you need the speaker for is to check if the problem is in your amp or as I suspect, in your crossover inside the speakers.

good luck. you probably have one wire wrong inside the crossover.

 

[arkelectron]

pick up some ultra inexpensive working speakers from anywhere that you can think of.
Your crossovers are very suspect, as in the wiring is probably wrong. I build speakers as a hobby, and a few of my designs have been sold in kit form in japan. Crossovers can be tricky to wire. one wrong connection and it will make cause clipping, a dead short circuit. Yes I've made the mistake myself more than once. usually when I'm working late at night.

Find some cheap used 2 way speakers, one only is really all you need. a speaker with a woofer and a tweeter. Make sure that it is guaranteed to be working. use it to test your amplifier and therefore your speakers.

This is not normal behaviour for a speaker or amplifier. The easiest method of troubleshooting is to use a working, known "good" speaker. Try and find a 2 way speaker. It's actually not important if they are 2 way or not. All you need the speaker for is to check if the problem is in your amp or as I suspect, in your crossover inside the speakers.

good luck. you probably have one wire wrong inside the crossover.

Yep, looks like it was the speakers. Weird because no other model of amp has had issues with them. Can't really get at the crossover without cutting the bottom out unfortunately.

 

[moonlight rainbow dream]

Yep, looks like it was the speakers. Weird because no other model of amp has had issues with them. Can't really get at the crossover without cutting the bottom out unfortunately.

Yes that was my experience too. It is weird. All other amps worked perfectly fine with the C-notes. I also did impedance sweep that correctly matched the published data. It is just infineon chip amps that had that bad reaction to my speaker. Maybe an electrical engineer can chime in with what is going on here.

 

[arkelectron]

Yes that was my experience too. It is weird. All other amps worked perfectly fine with the C-notes. I also did impedance sweep that correctly matched the published data. It is just infineon chip amps that had that bad reaction to my speaker. Maybe an electrical engineer can chime in with what is going on here.

I'm actually an electrical engineer. I spent a short amount of time studying the Infineon reference design the AO200 is based on. The outputs caps are the source of the noise for sure, but I didn't get much further into it.

 

[martinl]

I'm actually an electrical engineer. I spent a short amount of time studying the Infineon reference design the AO200 is based on. The outputs caps are the source of the noise for sure, but I didn't get much further into it.

I agree, these filtering caps look a lot like class 2 dielectric MLCCs which are known to be microphonic. They are also known to not be very linear. The selection surprised me since they'd be outside of the feedback loop and thus this distortion of LC components wont even be suppressed by feedback.

Upon further reading I saw this on page 12/86 in the datasheet
"Notably, with the multi-level modulation of the MA12070, there is no tradeoff between idle power loss and inductor cost/size, which is due to the absence of inductor ripple current under idle conditions in all configurations. Due to the high filter cutoff frequency, non-linearities of LC components have less impact on audio performance than with a conventional amplifier. Therefore, theMA12070 can operate with inexpensive iron-powder cored inductors and ceramic (X7R) filter capacitors with no significant audio performance penalty."


So they found a way of pushing those non-linearities in cheaper components further out of the audible band, which is nice, but it adds the problem of microphonics. But shouldn't those vibrations in the MLCCs be way outside of the audible band, how come you can hear them ? Feedback loop instability ? Limit cycles due to non-linearities ?

One possible, somewhat cheap, solution could be to use soft terminated MLCC with X7R characteristics. These add a conductive resin on the solder pads that is a lot softer than the ceramic materials and acts as a mechanical decoupler for vibrations. From my experience you could get something like a 10-20dB dampening of those capacitors compared to the regular ones.
The other option is to change over to either class 1 dielectric MLCC (C0G, NP0 types) or film caps. Those will probably be bigger in size though and therefore have higher series inductance.


I'm too trying to decide between a DA9 and a AO200, and it seems that they both have their respective design problems. The screws to clamp down the chip to the cooling pads in the DA9 are scaringly close to the, very sensitive, ceramic caps. I can easily see how you'd get microcracks in those capacitors when you are tightening the screws. On the other hand the AO200 cooling solution also seem a bit....meh.

Edit: On closer inspection it seems that what I thought were the filtering caps are actually flying capacitors for creating an intermediate rail with their fancy modulation scheme. Datasheet recommends 2x10µF 0805, so those are 4 of the 6 larger capacitors on the side of the amp IC. Those should be humming along at the switching frequency, which is what, 600kHz ?, so way outside audible spectra. I haven't really fully understood the modulation so I guess there might be some lower frequency "superimposed" on that though. The actual filtering caps are the film caps close to the binding posts. I'd suspect some beads on the back side then. That'd leave the decoupling capacitors as prime suspects of potential noise sources. From what it looks like they could easily be swapped out for some ultra low ESR polymer tantalum capacitors for a bit of extra umphf and quietness.

Noise from ceramic caps like these are unavoidable as they expand and contract with varying voltage, causing the whole board to participate in the music.

 

[martinl]

I think it's a bad flux solder job personally. You know, where they just heat up the board and the chip almost self solders' itself.
Common practice in line assembly on a smaller scale. They certainly are not likely to be using robots to solder these amps and parts. The amount of programming hours to set up all the robots to know where each pin goes makes it less likely especially when they make so many different amplifiers with different cases and pcb's but are basically identical amplifiers with a different form factor. I don't know the correct term for this form of soldering but I know it's super cheap and fast. All that's needed for a pop noise or hiss is one bad connection.

Im sorry, but you dont seem to know what you are talking about. That is not common practice on any assembly line, unless you are doing one-offs prototype validation stuff. What you are describing is a re-working procedure in case you are doing a repair where one of the components are faulty. For a SMT production line the solder paste is screen printed, without any additional flux, components are placed, and then it all goes through a carefully controlled multi-stage oven to follow standard IPC reflow soldering profiles. During the soldering the temperature is very tightly controlled, with a specified ramp-up and down time to not stress the components.
On this line you typically have 2 automated optical inspection machines, one to verify that the solder paste is properly applied. Then one more after soldering, to verify that components were properly soldered. Some lines also do an x-ray inspection at this point as well, to verify every single solder joint


Reflow soldering as it is called is quite cheap and fast, yes, but it's THE industry standard way to do SMT components on a production line. It's a highly repeatable and mature process. Sure, you can do wave soldering if you have some mixed through hole components and SMT, but those typically put higher stress on the components, is a more complex process and generally has a lower yield.

Using soldering robots for SMT components of this scale isn't even possible, and why would you even try when there is a better process ?


Thanks for the pictures though, very helpful

 

[Pillars]

Some pics for you.

The inside of the SMSL AO200!

The second to last picture is the mess that the OEM white thermal grease made. There is a significant gap between the aluminium heat sink blocks and thin grease doesn't cut it for cooling in my book. I used the same paste on my DA-9 which ran cooler than my acrylic rack by a considerable amount.

The last picture is the paste I used, it's almost like cake batter it's so thick.

Did you remove the aluminum heatsinks from the PCB in this process? I pulled them. They were fairly loosely mounted with not much white grease below them. I ended up swapping that out and what was between the blocks and the casing with PK3 which is a pretty thick paste. I also added front sinks to the Infineon chips and some opamps. I'm pushing this little guy fairly hard so figured go all out. Thanks for the heads up!

This is a great little amp, crazy clean.

 

[jokan]

Did you remove the aluminum heatsinks from the PCB in this process? I pulled them. They were fairly loosely mounted with not much white grease below them. I ended up swapping that out and what was between the blocks and the casing with PK3 which is a pretty thick paste. I also added front sinks to the Infineon chips and some opamps. I'm pushing this little guy fairly hard so figured go all out. Thanks for the heads up!

This is a great little amp, crazy clean.

ah yes, the crazy lack of overall quality control. the aluminium blocks for the lower heatsinks should be hand tight, not over torqued, you can bend that thin pcb with extreme ease. I used TF8 which is rated at 13.8 for its heat transfer abilities, a non-conductive thermal paste. It's cheap and with the AO200, you only need a small amount. I too decided to add 9mm x 9mm x 8mm copper heat sinks too for the chips themselves. It did lower the temps a little bit more. I drive my amp hard too, 2.7ohm load everyday all day. The amp NEVER gets turned off. I see zero reason to turn it off It never gets hot enough to warrant turning it off. It's my 3rd Infineon/Merus amplifier and the first one was only turned off a maximum of 3 times over its entire life (I started with the DA-8s) which I had since new, mine was the 3rd build of the amp. I had the DA-9 which required extensive repair and it has a different construction method where you screw down the bottom lid and the lower lid is effectively the heat sink. It had 20% coverage of thermal grease. Also the antenna lead for the BT was the exact length. On my 3rd time going inside the amplifier to root out the high-pitched whine from the display driver I finally broke the antenna lead at the socket on the PCB. I replaced the cable with one that has 10cm more length. I fixed the whine, I fixed the heat sink, and was happy until the AO200 came out which allowed for the display to be shut off completely. The AO200 was okay but read bluetoot once selected. I waited about 10 days before deciding to send it back to Aoshida-Hi-Fi Japan. They paid for return shipping, sent me the replacement amplifier once the physically received the AO200 after Aoshida/SMSL ran a full test before they shipped my replacement unit which was in good working order but I still changed the grease, and applied the heatsinks for the chips themselves. It's overkill to apply the heatsinks on the chips, but better safe than not. Mine has been problem free from day one, both AO200's have had zero issues. The thermal grease was the only thing that i replaced. And I did make sure that the aluminium blocks were hand tight with both units. My DA-9 experience showed me the lack of quality control, the DA-9 had very loosely fastened heatsinks. Probably 2-3 full turns loose.

Overall I am extremely happy with my AO200. I've not had any issues with QC with the 2nd model. I rolled the op-amps as they are "buffer" op-amps. A labour intensive job so as not to overheat the surrounding parts. I swapped out the opa1678's with opa1656's. I left the 072c's alone as I suspect they are for the subwoofer outs and I'm not running subwoofer. I can swap them, but I don't see the reward. The 072c op-amps are plenty good enough for subwoofer duty and you don't need a lot of gain from the outputs anyway as we'll be running a self amplified subwoofer, also our ears are extremely insensitive to distortion at the lowest octave.

I'm glad you're enjoying yours. Avoid going in the amplifier too much. They are definitely built to a budget. And discrete surface mount parts like the tiny capacitors off the op-amp a very short distance away will easily be damaged if you decide to roll op-amps. Generally follow the 3 second rule of staying in one spot. It's not worth possibly damaging the surface mount capacitors by staying on one leg for too long. There are a few more tricks to get a little bit more bass but I don't recommend it as the gains are minimal. The op-amp rolling is safer than the PSU mod which will gain you the same amount of bass but will not potentially over load the Infineon/Merus chips. The AO200 is running at near full capacity with tolerance in mind. You can adjust the PSU to provide a little (2vdc) more output at which point you have maxed the input voltage of the amplifier. But then again there are tolerance parts in the amplifier. Some of those power supply parts on the amplifier board are using 20% parts. Not very wise to run the amplifier at maximum voltage when the tolerance of some of the parts are 20%. It's easy to get a little too greedy with tweaking. And since there is no way of adjusting the bias on the amplifier, it's not something I would advise.

Glad you're enjoying your amplifier and keep any mods simple so you don't blow parts on the pcb. Totally not worth it. Op-Amp rolling is a definite improvement over the 1678's though. I don't know why they chose the 1678's. The opa1612 is a little bit thin sounding on the bass but has far more clarity, the 1656's give you the bass and the clarity. It's labour intensive as you do not want to blow that surface mount capacitor that is on that thin trace and very close by. So you have to do multiple passes with a wick or you can use a purpose built heat gun that has a small nozzle to remove the op amp in one shot. Just make sure you have a tool to grip the tiny op-amp so you don't lose it. I don't have the small heat gun with various nozzles yet but it's on my christmas list of things to buy. I'm doing enough op-amp swaps that it makes sense to buy one now.

 

[Pillars]

ah yes, the crazy lack of overall quality control. the aluminium blocks for the lower heatsinks should be hand tight, not over torqued, you can bend that thin pcb with extreme ease. I used TF8 which is rated at 13.8 for its heat transfer abilities, a non-conductive thermal paste. It's cheap and with the AO200, you only need a small amount. I too decided to add 9mm x 9mm x 8mm copper heat sinks too for the chips themselves. It did lower the temps a little bit more. I drive my amp hard too, 2.7ohm load everyday all day. The amp NEVER gets turned off. I see zero reason to turn it off It never gets hot enough to warrant turning it off. It's my 3rd Infineon/Merus amplifier and the first one was only turned off a maximum of 3 times over its entire life (I started with the DA-8s) which I had since new, mine was the 3rd build of the amp. I had the DA-9 which required extensive repair and it has a different construction method where you screw down the bottom lid and the lower lid is effectively the heat sink. It had 20% coverage of thermal grease. Also the antenna lead for the BT was the exact length. On my 3rd time going inside the amplifier to root out the high-pitched whine from the display driver I finally broke the antenna lead at the socket on the PCB. I replaced the cable with one that has 10cm more length. I fixed the whine, I fixed the heat sink, and was happy until the AO200 came out which allowed for the display to be shut off completely. The AO200 was okay but read bluetoot once selected. I waited about 10 days before deciding to send it back to Aoshida-Hi-Fi Japan. They paid for return shipping, sent me the replacement amplifier once the physically received the AO200 after Aoshida/SMSL ran a full test before they shipped my replacement unit which was in good working order but I still changed the grease, and applied the heatsinks for the chips themselves. It's overkill to apply the heatsinks on the chips, but better safe than not. Mine has been problem free from day one, both AO200's have had zero issues. The thermal grease was the only thing that i replaced. And I did make sure that the aluminium blocks were hand tight with both units. My DA-9 experience showed me the lack of quality control, the DA-9 had very loosely fastened heatsinks. Probably 2-3 full turns loose.

Overall I am extremely happy with my AO200. I've not had any issues with QC with the 2nd model. I rolled the op-amps as they are "buffer" op-amps. A labour intensive job so as not to overheat the surrounding parts. I swapped out the opa1678's with opa1656's. I left the 072c's alone as I suspect they are for the subwoofer outs and I'm not running subwoofer. I can swap them, but I don't see the reward. The 072c op-amps are plenty good enough for subwoofer duty and you don't need a lot of gain from the outputs anyway as we'll be running a self amplified subwoofer, also our ears are extremely insensitive to distortion at the lowest octave.

I'm glad you're enjoying yours. Avoid going in the amplifier too much. They are definitely built to a budget. And discrete surface mount parts like the tiny capacitors off the op-amp a very short distance away will easily be damaged if you decide to roll op-amps. Generally follow the 3 second rule of staying in one spot. It's not worth possibly damaging the surface mount capacitors by staying on one leg for too long. There are a few more tricks to get a little bit more bass but I don't recommend it as the gains are minimal. The op-amp rolling is safer than the PSU mod which will gain you the same amount of bass but will not potentially over load the Infineon/Merus chips. The AO200 is running at near full capacity with tolerance in mind. You can adjust the PSU to provide a little (2vdc) more output at which point you have maxed the input voltage of the amplifier. But then again there are tolerance parts in the amplifier. Some of those power supply parts on the amplifier board are using 20% parts. Not very wise to run the amplifier at maximum voltage when the tolerance of some of the parts are 20%. It's easy to get a little too greedy with tweaking. And since there is no way of adjusting the bias on the amplifier, it's not something I would advise.

Glad you're enjoying your amplifier and keep any mods simple so you don't blow parts on the pcb. Totally not worth it. Op-Amp rolling is a definite improvement over the 1678's though. I don't know why they chose the 1678's. The opa1612 is a little bit thin sounding on the bass but has far more clarity, the 1656's give you the bass and the clarity. It's labour intensive as you do not want to blow that surface mount capacitor that is on that thin trace and very close by. So you have to do multiple passes with a wick or you can use a purpose built heat gun that has a small nozzle to remove the op amp in one shot. Just make sure you have a tool to grip the tiny op-amp so you don't lose it. I don't have the small heat gun with various nozzles yet but it's on my christmas list of things to buy. I'm doing enough op-amp swaps that it makes sense to buy one now.

Whew quite the writeup! The aluminum heatsink is flat mounted so there wasn't any bend for me - I had paid attention to this to make sure I wasn't bowing. Doesn't really matter a ton with the distance anyways. I've no need to ever go inside again. No interest in rolling amps in this unit either, I prefer sockets for that plus there's not much to gain there. It's a pretty good sounding amp despite the budget nature. Sure takes a beating anyways.

 

[jokan]

Whew quite the writeup! The aluminum heatsink is flat mounted so there wasn't any bend for me - I had paid attention to this to make sure I wasn't bowing. Doesn't really matter a ton with the distance anyways. I've no need to ever go inside again. No interest in rolling amps in this unit either, I prefer sockets for that plus there's not much to gain there. It's a pretty good sounding amp despite the budget nature. Sure takes a beating anyways.

The op-amp rolling totally changes the sound signature. It's a lot of work if you, like me do not have a compact purpose built heat gun for removal of surface mount op-amps. There are fairly significant gains to be found as far as SQ (sound quality) and it will not cause any issues. There are parts that allow you (me) to use a standard socket so it's not out of the question to use various discrete op-amps by Sparkos Labs amongst others. Though with the opa1656, you get nearly discrete op-amp performance and it requires no adaptors (I don't like adaptors).

That white grease that SMSL uses is purely rubbish for thermal dissipation. It doesn't cost much more to use a high quality paste and it's much less messy, plus it will help with EMI/RFI rejection also. In fact it was the second biggest reduction in noise in my old DA-9. The biggest source of noise was shielding the LED screen power driver which produced a high pitched whine regardless of volume and could only be heard with your ear on the amplifier itself, more precisely directly by the screen, the driver is millimeters from the screen in the DA-9.

The Infineon/Merus amplifier chips are originally made by International Rectifier Company out of the USA. They were bought out by Infineon who purchased all of the IP from them, then they used that know-how to make the MERUS brand. The IRC class D amplifiers were made to sound very much like a triode amplifier. IRC is used by SPEC Corp of Japan who has their own IRC variant chips made for them and they combine the chipset with extensive analogue amplifier design (especially in the power supply) to reproduce a sound that is both natural and clean without ever sounding like the typical class D amplifier. I own several TPA amplifiers that I use on occasion and they do sound pretty good, but they are not nearly as musical sounding though the TPA amps are very impressive considering the costs. Class D will continue to improve and SMSL was smart by using the Infineon/Merus chips. The fact is that they don't measure very well but they sound very good. Another case of why SINAD measurements don't tell the full story of how an amplifier actually sounds. I've worked for SPEC corp in the past doing some technical translation for them so I had to understand exactly what they were doing minus some hidden gubbins that they keep secret but they explained enough about what is hidden underneath the metal boxes painted with material that is sold by the tin, it's actually stealth technology paint that is available to the public but not advertised to the public and ridiculously expensive. I know what's in the paint as it is disclosed but the break down of the various materials is secret. They also use a lot of parts that are more tube amplifier than class D amplifier technology to get the tone of their amplifiers "right". They make 3 different versions of their amplifiers with different voicing. One for North America, one for the EU, the last for Japan. I prefer the voicing used in their north American versions. Small differences in the parts make a huge difference in voicing.

The SMSL AO200 and DA-9 are both incredible value. It offers a sound that is very different from Bruno Putzeys amplifiers, ICEpower, Pascal Amplifiers etc. The reason is the original goal by IRC was to sound as close to analogue as possible from the very beginning. I believe that IRC-Infineon-MERUS is a viable alternative to the more popular and widely known competition. I've used ICEpower in the past and it hasn't really changed much since the first generation of their amplifiers though they've tweaked it to make it better with time. The PASCAL amps, well we only have Mytek really to compare it with as PASCAL is not very popular and Mytek is having QC problems that they appear to have slowly gotten sorted. They remain extremely expensive and I would say overpriced for the performance they offer. Nuprime/Nuforce are pretty much the same even though they swear they are different companies. They've steadily improved their amplifiers and they are in my opinion very musical but still overpriced. They are pretty though.

I completely agree with you, they are spectacular value for money. I would personally happily pay double to triple for 2x - 3x the power output all within the same sort of physical footprint. I know that it can be done if they make a more powerful chip and with a better PSU.

Glad that you're enjoying your amplifier. I am too!

Heatsinks for the Merus chips pictured!

 

[Pillars]

Hah I used the same types of heatsinks. Did you end up doing the opamp swap? Did you have any pictures of that? I'd be doing it with an iron as well. a mouser link to it wouldn't help either! What are your impressions of the swap?

 

[jokan]

you're going to want to use a "heat clip", an aluminium alligator clip and use a soldering wick leg by leg (no more than 3 seconds per leg). Repeat as necessary. Get one side done first and use the heat clip to lift the op-amp with soft even pressure the whole time if you like me don't have a mini-heat gun. The specialised heat guns really simplify the process. I'm buying one for myself over christmas.

Do one leg at a time and do the side that doesn't have the capacitor nearby. You can read the specs of the opa1678 which is the stock op-amp, and compare it to the opa1612, the opa1656. The 1656 is basically the latest and greatest of the burr-brown sound plus series. The opa1612 is now outdated. On this amp, it is used as a buffer op-amp or input stage. It is what gets amplified in other words. The gain is slightly higher than the factory op-amps but it will not cause clipping or add noise. You'll get a firmer grip over bass and crisper, more detailed mid to high end, all around a better sound, along with a much more musical presentation. Make sure that you have a soldering tip that is small, narrow and only heats or touches one leg at a time. Use a waxed wick. I use Hakko's Yellow font 1.5 x 1.5mm FR 150-83 "no clean" wick. Hakko amongst others make aluminium heat clips that work in the same manner. I use one that has angled teeth as opposed to a straight toothed clip as it seems to give better leverage.

Make certain that you don't touch other components. Make sure you completely discharge the amplifier by unplugging the amplifier from the mains and letting it sit for a good half hour before you do any soldering. You don't want residual power to be in the amplifier when you start this process for safe measure. Make sure you completely solder each leg, but not to short the legs. A loose solder joint is very easy to do even though you won't necessarily be able to see that you have a poor solder joint. It will lead to popping and or a lost channel. Do not overheat the board especially that tiny capacitor that is on one of side of the op-amp. The two opa1678's are used for both RCA and XLR inputs. The 072c's are for the subwoofer output. You can swap those out for something else also, I'd probably use an LME47990 or something similar as they seem to have better bass, or a stronger sounding bass. I'm not interested in that myself as I have a self powered subwoofer and don't want to get into what I consider to be tiny gains with more risk of damaging the board. It's easy to overheat the board, I hope your soldering iron is temperature controlled. It's worth the investment if you don't have one that has adjustable temperature control.

Take your time. The procedure should take you 30-45 minutes for both op-amps. Start with the op-amps that are in the first pic as it's easier to get to and will get you some practice. I didn't clean up after the "surgery" so there's some flux leftover. Clean it all up after you confirm that it's operational before you slide everything back into place. Don't shock yourself! Take your time with it. Do not exceed 3 seconds in one spot. Once you've got all of the solder that you can see (use your mobile's camera and zoom in to look at how much solder you've removed.) Then you can take the wick and move your soldering iron across all four legs sliding it back and forth with the soldering iron to mop up whatever is left over, one side should lift at this point, then start on the other row of four legs. It's the best way of removing all of the solder, the second row should go much easier once you get the first row out of the way. Now that the op-amp is able to be lifted on one side, it'll just come off always keep the time to no more than 3 seconds, and do take enough time so as not to damage that stupid tiny capacitor! That capacitor is important! it's tiny and cannot handle prolonged heat. The copper traces on the board are rather delicate, the black epoxy coating of the board is easy to overheat also which can be easily damaged. So be patient, take your time. Without the mini-heat gun, it's labour intensive. The mini-heat gun makes this job much easier as you'll only need a heat clip to remove the op-amp in one go with no fear of damaging other nearby components. I'm looking at one and it costs over $300 as I want a professional grade one with adjustable temp and fan speed. I figure more and more we're seeing the smaller surface mount op-amps so I need the right tools.

Yes, one of the op-amps isn't straight, but it's soldered on there correctly and isn't going to cause a problem. It's a very cramped space for a soldering iron and heat clip to solder the new op-amp in place, so a little bit crooked is something I can deal with as it's not causing any problems. Also the opa1656 has a higher heat tolerance than the opa1612. Which makes me happy! lol.

 

[Doodski]

you're going to want to use a "heat clip", an aluminium alligator clip and use a soldering wick leg by leg (no more than 3 seconds per leg). Repeat as necessary. Get one side done first and use the heat clip to lift the op-amp with soft even pressure the whole time if you like me don't have a mini-heat gun. The specialised heat guns really simplify the process. I'm buying one for myself over christmas.

Do one leg at a time and do the side that doesn't have the capacitor nearby. You can read the specs of the opa1678 which is the stock op-amp, and compare it to the opa1612, the opa1656. The 1656 is basically the latest and greatest of the burr-brown sound plus series. The opa1612 is now outdated. On this amp, it is used as a buffer op-amp or input stage. It is what gets amplified in other words. The gain is slightly higher than the factory op-amps but it will not cause clipping or add noise. You'll get a firmer grip over bass and crisper, more detailed mid to high end, all around a better sound, along with a much more musical presentation. Make sure that you have a soldering tip that is small, narrow and only heats or touches one leg at a time. Use a waxed wick. I use Hakko's Yellow font 1.5 x 1.5mm FR 150-83 "no clean" wick. Hakko amongst others make aluminium heat clips that work in the same manner. I use one that has angled teeth as opposed to a straight toothed clip as it seems to give better leverage.

Make certain that you don't touch other components. Make sure you completely discharge the amplifier by unplugging the amplifier from the mains and letting it sit for a good half hour before you do any soldering. You don't want residual power to be in the amplifier when you start this process for safe measure. Make sure you completely solder each leg, but not to short the legs. A loose solder joint is very easy to do even though you won't necessarily be able to see that you have a poor solder joint. It will lead to popping and or a lost channel. Do not overheat the board especially that tiny capacitor that is on one of side of the op-amp. The two opa1678's are used for both RCA and XLR inputs. The 072c's are for the subwoofer output. You can swap those out for something else also, I'd probably use an LME47990 or something similar as they seem to have better bass, or a stronger sounding bass. I'm not interested in that myself as I have a self powered subwoofer and don't want to get into what I consider to be tiny gains with more risk of damaging the board. It's easy to overheat the board, I hope your soldering iron is temperature controlled. It's worth the investment if you don't have one that has adjustable temperature control.

Take your time. The procedure should take you 30-45 minutes for both op-amps. Start with the op-amps that are in the first pic as it's easier to get to and will get you some practice. I didn't clean up after the "surgery" so there's some flux leftover. Clean it all up after you confirm that it's operational before you slide everything back into place. Don't shock yourself! Take your time with it. Do not exceed 3 seconds in one spot. Once you've got all of the solder that you can see (use your mobile's camera and zoom in to look at how much solder you've removed.) Then you can take the wick and move your soldering iron across all four legs sliding it back and forth with the soldering iron to mop up whatever is left over, one side should lift at this point, then start on the other row of four legs. It's the best way of removing all of the solder, the second row should go much easier once you get the first row out of the way. Now that the op-amp is able to be lifted on one side, it'll just come off always keep the time to no more than 3 seconds, and do take enough time so as not to damage that stupid tiny capacitor! That capacitor is important! it's tiny and cannot handle prolonged heat. The copper traces on the board are rather delicate, the black epoxy coating of the board is easy to overheat also which can be easily damaged. So be patient, take your time. Without the mini-heat gun, it's labour intensive. The mini-heat gun makes this job much easier as you'll only need a heat clip to remove the op-amp in one go with no fear of damaging other nearby components. I'm looking at one and it costs over $300 as I want a professional grade one with adjustable temp and fan speed. I figure more and more we're seeing the smaller surface mount op-amps so I need the right tools.

Yes, one of the op-amps isn't straight, but it's soldered on there correctly and isn't going to cause a problem. It's a very cramped space for a soldering iron and heat clip to solder the new op-amp in place, so a little bit crooked is something I can deal with as it's not causing any problems. Also the opa1656 has a higher heat tolerance than the opa1612. Which makes me happy! lol.

Are you using extra liquid flux? It doesn't look like it. Get some and use a drop on each side or if you want no clean flux get some of that in a syringe.

 

[Pillars]

It's not the prettiest but it works. I'll be ordering some and doing the swap with an air gun. It's roughly $60 for one that'll do the job just fine.

 

[jokan]

It's not the prettiest but it works. I'll be ordering some and doing the swap with an air gun. It's roughly $60 for one that'll do the job just fine.

the crooked op-amp, you'll see a tiny capacitor right above it, and the resistors nearby. do your desoldering on the other side and do each leg at a time on the side with the capacitor to be safe. it's super easy to overheat the thin traces on the pcb so use the 3 second rule with all of the legs on these tiny op-amps. the other op-amp is much easier to get to and easier to remove. I suggest you get your eye in with that op-amp. The whole job should take 30-45 minutes for both op-amps including soldering the new replacements into place. Look for the specs of the opa1678 and use that as a guide so you are actually upgrading and not downgrading! There are many suitable op-amps that you can replace them with.