Back in 2017 I got a used pair of those speakers and have been using them since as my main speakers for mixing and the occasional mastering. Some engineers I really looked up to used them (and still use them I think) so I was very curious how they sound like. Fortunately I fell in love with how they sound right away. Their punchy lower midrange and upper bass really amazed me, so I kept them and started getting used to work on them.

Two years later in 2019 the first speaker started some weird crackling and the mid-woofer was popping, moving erratically and in general it was super noisy, which led me to open it up and try to fix it. I found the problem to be in the active crossover input board and after reflowing most of the components I got it back to working condition.

Shortly after fixing the first one, the second speaker started having the exact same problem. And in the following years it reappeared over and over again…

After checking those input boards over and over again I realized that some of the copper tracks on the PCB had hair cracks. In hindsight I think that was actually the main cause of the whole crackling problem. This same input board was used by Adam for their other speakers too and to make it fit to the different speaker configurations they hand-soldered some SMD parts to change the filters because having the boards assembled professionally for each speaker version probably would have been too expensive – at least that’s my guess. So my theory is that this hand-soldering which is a really rough process kind of reduced the longevity of the boards. Also the problem only appears in the midrange crossover(from what I read also with other peoples speakers) which again is the part of the PCB that is hand-soldered.

In my opinion those PCBs don’t look like they’re professional products but see for yourself:

I especially like the botch capacitor soldered to some SMD pads.

At this point it kind of was clear that the problems won’t be going away and that I’d have to design my own PCB to get the speakers back to a reliable working condition. Luckily Adam provided a service manual so making my own PCB wasn’t that big of a deal. So in Jan. 2024, I finally got to it:

I tried to keep most of the component positions like they were in the original but omitted the parts for the 4-way versions (S4A I think). Luckily everything worked out great and it drops right in:

The next step was to calibrate the crossover to factory spec. I measured the old crossovers with REW (rough signal flow: input -> crossover -> power amp -> 4 Ohm load -> audio interface -> REW) and the results where all over the place. This is what Adam suggests in the service manual:

My curves looked very different. The left speaker looked ok, but the right one seemed quite off (it’s an overlay of both speakers):

Speaker levels didn’t match and the tweeter curve wasn’t as flat as in the service manual. Really strange… I’m really not sure if they tuned each finished speaker to sound the same at the factory or if those strange curves were just accidents (I suspect the latter).

So I used the left speaker as a reference and calibrated my replacement crossovers to those curves – except for the tweeter curves. I wanted to know how they would sound like with a flatter voicing as the service manual suggests. The result looked like that:

These are the curves for both speakers with the new crossovers and I left in the old tweeter curve for comparison. It looks like a lot of difference but it is just +-1dB (and sorry for the small pics, I’m hoping you still can get an impression of what’s happening).

I hooked everything back up, turned the tweeters 90° because I wanted to use them vertically and for the first time listened to the speakers with their new crossovers. It was really strange, something was off. That 1db change in the tweeter curve suddenly seemed like A LOT, the speakers felt too bright and the transition from mids to tweeters sounded strange. After a week of listening tests I knew that I won’t be getting used to that voicing and I had to do something about that. I wasn’t sure if the problem was setting the speakers up vertically for the first time and introducing another new variable because of it (classic beginners mistake…) or if that little change in the curve was responsible for my impression. At the same time I read people made the same experience with placing them vertically, saying they seem brighter and less punchy than in horizontal placement…damn!

Still the first thing I wanted to try is opening them up one last (hopefully) time and changing the tweeter curve to how they were before. So I did that:

As I said, the difference isn’t a lot, it’s just around 1dB, but after putting the speakers back together and listening to them once again I was relieved that it made such a huge difference. Finally the voicing was back to how I remembered it from before. Midrange and tweeter integration sounded just right again, or at least how I was used to it. Vertical placement also seems fine for now, I’m using them with the midwoofers on bottom, because my stands are quite high and I want the mids almost on earlevel. But maybe I change them back after some more listening tests, we will see…

So in conclusion, if you have problems with your ADAM Audio S3A speakers crackling, being noisy, or having the membrane of the midrange speaker move like crazy, the problem is most likely coming from a faulty crossover PCB. The other problem I encountered is that the tweeter volume is fading away, sometimes coming back etc. This is a classic case of a faulty relay, which I think happens a lot in power amps (at least I already changed quite some relays in some different power amps). Sometimes you can get it to work if you turn up the volume or if you turn your speakers off and on repeatedly, but to really fix it you’ll have to change the damaged relay – I used Zettler AZ764-1C-24D.

That’s it for now, let me know what you think or in case you need help with your S3As.

PS: I did some very dirty acoustic measurements before and after tweaking the high end curves. The red graph was with the flat curve as suggested in the service manual, the blue one is the final setup. As you can see 1k-6k is a little more balanced and above that the sparkle is a little less aggressive. This makes a lot of difference how the speaker sound. I though about it again and I think it is best described that with the red curve the midrange and treble sounded like “this is coming from the woofer, and that is coming from the tweeter” and with the blue curve it sounds like one speaker, midrange and tweeter just work better as a unit -> better mid+tweeter integration.