Hey, by now I can tell that the B'ASS needs about a week of breaking in. I suppose this is about the amplifier chips because I recall similar from adding one chip (as the 1.0 version implied (more of 'm)). So say that in a week's time the real uncanny stuff starts to happen. Before that it's only a "ooo, nice !". But ooo nice is different from
???
By now I must be an experienced in-burner because I forgot how many times that happend, so many. But example : not necessarily in the right sequence :
Think does not work. Odd, because I can't recall that we changed something significant, and yesterday all was fine. OK, let's add some smaller capacitor value at position X and Y.
Ah, helps. Good.
Next step : let's put some load to it. Bweh, does not work. But why then ? because the situation now is the same as one week ago ? Let's put back those original capacitor values.
Does not help. Hmm ... back to the situation without load. What ? still does not work, but now all really is the same as early this morning ?
?
And so one day has passed with a result of zero.
Hey, the left channel now works all fine. Let's do the right.
Nah, right is misery. And oh, because that is misery, left is influenced too and now also does not work. Hmm ... Let again to the right channel, step by step. Remove all, re-test and ...
Left channel does not work.
? this morning is still did.
And another day had passed.
Nothing works. It must be something with the PCB. Grrr. Let's try a new one.
Does not work. But wait, this is the situation with the larger capacitors. Didn't we had one working with smaller ? let's try.
Hooray, works ! now carefully add the 2nd channel.
Hmm, seems to work, but does not look stable. Is the power supply OK ? Almost. OK, let's change that a bit (buy a few parts).
Next day, work all day on PSU. Yes, all good and as desired. But over-speced only a little.
B'ASS now works ? what ? B'ASS does not work at all. Nothing nowhere.
?
Hey if we move this wire a little, things move. Hmm ... bad solder. Re-do that.
Yeah, but now all looks odd all over.
?
Another day (possibly there's a weekend break in between) again carefully start a new board. Just finish it in one go, because this must be something in the board. Don't move it ! don't touch wires !
It works !
That consumed a day.
Now let's test balanced.
, really nothing works any more. F8ck, it must be the output wires. Something must be breaking inside of the board all the time. Let's bridge to another point.
Works.
See ?
Now balanced can be tested. Looking good ! Ha, now can listen. Let's do that tomorrow, because my crucial chips must be put over, and it is too late to experience failure and build back.
Nah, doesn't work.
?
Why ?? what ? no supply on that chip ? what ?
?
Let's bridge that too then.
Ah, works again. But not really good. Must build back my chips or otherwise no music tonight.
I didn't count, but say this were 7 days, while 7 weeks have been spent on this total nonsense. Mentioned 10 days ago I finally got one working so I could listen to it and I suppose if I don't breath on it that it will keep on working (it still does today). But at the same moment it finally worked I also decided that this seriously can't be the way to proceed. So at that same day I started to work on again a new version. I thought to change the internal topology so things were not so tight, thinking that all was too squeezed to meet (Chinese drilling etc.) machine's specifications. I also dicovered a fault in my own design or (too quick) thinking and this was about making traces thicker (2oz instead of 1oz) instead of re-designing the board with wider traces (mind you, this was 3 months or so ago, say the day before I put out the order for 3.0 to China). But I discovered that by kind of accidence.
The processing of my regular China PCB producer took longer and longer (think a month of throughput time by now). So the thing is/was : I discover something wrong and while I could adjust that in 5 minutes, it is waiting for 4 weeks again (this happened the other day to the Stealth PSU board). This is intolerable. So I thought to hand B'ASS 4.0 to a for us new manufacturer. German maybe. Belgian ? At first glance this looked fine at only ~150 euros more expensive, but I soon saw that adding other parameters also added another 300 and by then it was 3 times more expensive and not even fast as well to cut on the costs. So
Anyway ... because of ending up at the say more decent or less careless manufacturers, I learnt that I had been able to squeeze things out to China in a way which can't be. I saw that when a German manufacturer started to throw error messages at some combination I made. Like "hey boy, if you want more copper, then you must add trace witdh at the same time because else you run into machine tolerances and we can not guarantee that all works.". And I just could not get in that order. So ...
Yeah. I did exactly that with the Chines manufacturer and while nothing complained, they still could not do the thang I wanted (this is all very manufacturer/ring specific and while the one can drill up to 0.3mm the other can do 0.050mm and if I, say, for the one which can do 0.3 into 0.1, two holes may shortcut or ... and this is worse : won't because this is tested for, but don't sneeze at it !
And that is what we were doing of course with all of our changes.
In the end it is all my own fault, starting with a too small board for the lot I wanted to have on it. Now the board for 4.1 is larger (and larger again than a 4.0 which also was larger already but never went out) plus I carefully selected the factory and its capabilities, which btw are now not extraordinary again (because of the board size).
For fun I will add this :
One of the manufacturers I tested, offered a kind of 50% off per board if I in advance told that I'd take the over production. Ehh, what ? Yes, that is how it works. They produce say 100 on your desired 75 and they do this because many fail (like in my case maybe 50% failed because of too tight tolerances which they did not check in advance). So they use a kind of "lousy" machinery and just overproduce and next try to sell you all they could produce more. Not really a bad thing in itself (can work out cheaper for you) but from this you learn how these things go.
I must learn a lot, but I for sure learnt a lot these past weeks !
Meanwhile my poor man's B'ASS 3.0 measures better as anything I've seen here and it also listens the best and most normal at the same time.
Now I must try to avoid a cold.
Peter
PS: B'ASS 4.1 was put out for production (China somewhere again) first thing last Monday. Should arrive somewhere next week. And then the party starts all over ...
