My Take on X-BOSOZ

metalman said:


***
The negative rail can be operated at as little as -7.5V, but at those levels the CCS behaviour is less than optimal. -10V would be the minimum I would recommend for decent performance, and there are further sonic improvements up to around -15V. Beyond that the performance difference is negligible, at least to my ears.

Exactly correct!

Cheers, Terry


I've built kari's boards and designed them for -9 VDC as stated by what I thought was unaminuous consensus. Are you saying that I should go back in there and adust the DC to -15 to optimize it? It would be the 3rd time that I monkeyed with the negative voltages.
 
fcel said:
Even though the free shipping is over, are you just ordering enough parts for 20 kits or parts for a lot more kits? I'm still waiting to see what's the decision on the transformer spec and where to order the transformer from, before placing my kit order.


Hi Fcel,

We will keep getting boards and kits made as people need them. For this batch I ordered 70 sets of PCBs. The reason is that often people's time tables for buying PCBs/kits do not necessarily fit nicely into a group buy timeline. So our goal is to keep these PCBs and kits generally available so people can jump in whenever they like.

The part express transformer seems like a good choice, but it is huge (like 7X more than the amp needs) in terms of capacity. 🙂 But hey, thats cool.

Cheers!
Russ
 
Originally posted by lgreen
Are you saying that I should go back in there and adust the DC to -15 to optimize it?
The gate of the gain transistor sits at ground, so its source typically sits between -3V and -4V. The source resistor of the CCS FET drops another 2V, so that the V(DS) for the CCS FET is the value of the negative rail less 6V (i.e. for -9V rail, the CCS FET V(DS) = 3V). The CCS goes unstable if the CCS FET V(DS) drops below ~1V, so the absolute minimum value for the negative rail is -7V, but you want a extra margin to account for possible variations in parts behaviour etc. so I arbitrarily picked an extra 3V, but 2V should be fine, i.e. -9V will be fine.

The best CCS performance comes into play when the CCS FET V(DS) is about 1V larger that the V(GS) or more. The main difference in adding this extra voltage is in the CCS's ability to reject high frequency noise, so if the power supply is a good one (and yours should be very very good) then there will be little or no noticeable effect. If you live in an area with a lot of RF noise etc., then changing to the larger negative rail might worthwhile, otherwise I would say that the change isn't worth the effort.

I've said it before, when I did my circuit design I was aiming for the best I could achieve with my knowledge and experience. I tried to optimize everything to its best irregardless of whether it affected the overall performance. To give you an idea, I am running a little more than -20V on the negative rail, clear gross overkill.

Hope this gives you some perspective for your decision.

Cheers, Terry
 
Lucky you. At least you get service. I spent the morning complaining to Amazon and UPS about a Christmas present that was supposed to be delivered on the 22nd and now will be delivered on the 27th.

Did you actually compute the heat sink size needed for the power supplies? I have 1 inch and 2 inch, but no 1.5. Probably just go with the 2 inch, but wondered if you had actually done the math. It would seem we are dropping about 4.5 volts at under 100 ma (mono, 200 ma stereo) per side, so the 1 inch would probably work. Comments?
 
Yep, after the PhD in Physics, no more math. Now that I am a manager........

I will just stick with the 2" ones. They don't stick up above the caps, so there really is not a problem with them. I also found the 1.5" hard to find.
 
So, I'm about to start matching FETs. I only want to do this once 😉

Here's the jig I'm using. The 100R resistor is a 10W I have hanging around. I am measuring DC voltage across the test points. Anyone see any problems or have any recommendations?
 

Attachments

  • fetmatcher.gif
    fetmatcher.gif
    3.2 KB · Views: 672
Brian,

I hope you did not follow your drawing. TP2 is tied to S and to ground. The way it shows, your meter is shorted. I suspect it is the drawing, not the circuit.

As for my request to get some Jantzen caps as you might have extras if they are optional for the kits, Parts Express is out of stock of them. I wonder why that might be.....😀

Craig
 
Ooops, you are right. Drawing was wrong. Here is the correct one.

I am responsible for the out of stock condition. Actually, I was only able to get 75 of the 100 needed for the initial buy, but they assure me I will have the rest shortly.

The only options I have seen for volume would be either a 4-deck stepped attenuator or a multi-digipot. I guess if you could find a 4-gang pot you could do that too.

This would be a good topic for discussion actually.
 

Attachments

  • fetmatcher.gif
    fetmatcher.gif
    3.2 KB · Views: 623
Much better.

PE lists 1/8/06 as their expected date for more caps, but they usually do better than the date they list.

There are also some interesting variations out there where people have used two dual pots geared together with a set of 3 gears so that each dual pot tracks with the other. Will try to find the examples.