Krell KSA 50 PCB

On schedule for Monday

Hi All-

This is just over half the orders.. Ran out of envelopes. Going get some more now. Hope the post office is open at 9 PM…
 

Attachments

  • little more than half.jpg
    little more than half.jpg
    71.3 KB · Views: 1,531
similar

The 2 amps are almost identical from a circuit perspective, so at this point I'm assuming where there are differences they will be absolutely minimal. There have been reports of the 50 sounding better, but if you have very inefficient speakers, the 100 may sound cleaner at high volumes because it has 3db more output and is therefore less likely to clip...

Based on the earlier entries in the thread, it looks possible, perhaps even easy to make a 50, then later increase voltage, output devices and heatsinks to end up with a 100...

Stuart
 
are the Ft charts screwed up...

Hi,

Assuming we are looking at the same datasheets, in the 500ma-5a range, there is virtually no difference between the transistors. The charts are (quite stupidly) scaled differently. One is linear, the other is logarithmic. If you plot data from one on the other they overlay each other for all the points that I tried...my charts only go down into the 100ma region, do you mean below this?

Stuart
 
tmblack said:
Why does MJ1302& MJ3281 have large drop in Ft at low Ic and their
MJL1302 & MJL3281 is more normal?

Which amp sounds better, KSA100 or KSA50? Would like to build 1 but not both!

Tom

This goes back to the engineers vs. Listeners perceptions.

IMHO the design maybe the same, but I have experienced, less glare with lower rails and less grain with bridged designs.

The lower railed KSA-50 is on the list of many reviewers as one of the top ten amps ever made.... The KSA-100 is not.
 
the chart is incomplete...

Hi,

The chart I have for the mj* parts is so poorly scaled compared to the mjl* chart I'm not sure it is valid to compare, the error from reading the values is massive. For instance on my MJ* chart, the curve barely even extends to 0.1A. As it crosses 0.5A the thickness of the line is so great it spans the range from 10-20MHz. The earliest point that has any reliability is at 1A, where they are close to identical and they remain that way to their limits.

I think the differences you are seeing are real, but not as great as the **** charts can be read to imply. AFAIK the same die is used in both, so differences if they exist are going to be minimal.

By the way the Ft charts are derived from data collected at 1MHz, my guess is they'd be even closer at audio frequencies, where package capacitance and the like will have minimal effect.

Stuart
 
Shipped!

Hi All-

Email me if your not listed.

stgrab@yahoo.com

Did I mention I took the pile of money and bought a resort and am retiring? See pic attached (JOKING of course).




Domestic Mail Recipients

Mark G
VB 487 988 609 US

Stuart E
VB 487 988 759 US

George T
VB 487 988 541 US

Dennis C
VB 487 988 555 US

Kwok W
VB 487 988 590 US

Harry C
VB 487 988 569 US

John G
VB 487 988 586 US

William C
VB 487 988 572 US

Brian W
VB 487 988 515 US

Nicolas N
VB 487 988 524 US

Arif H
VB 487 988 507 US

Mike M
VB 487 988 498 US

David B
VB 487 988 538 US


International Mail Recipients

Jonathan C
VF 243 790 156 US Insurance
LZ 690699341 US Package

Anthony T
VF 243 790 139 US
LZ 690699284 US

Frank R
VB 487 983 120
LZ 644118279 US

P.D.
VB 487 983 133 US
LZ 644118319 US

Richard P
VF 243 790 195 US
LZ 690699338 US

Jan D
VF 243 790 173 US
LZ690699409 US

Arne K
VF 243 790 142 US
LZ 644118305 US

I Chien W
VB 487 983 080 US
LZ 644118296 US

JH H
VF 243 790 125 US
LZ 69069929US

Anthony K
VB 487 983 102 US
LZ 690699372 US

Dominique B
VF 243 790 160 US
LZ 690699390 US

Arthur B
VB 487 983 116 US
LZ 644118265 US

Ed H
VF 243 790 187 US
LZ 690699369 US

Walter G
VB 487 983 093 US
LZ 644118282 US

Aaron B
VF 243 790 108 US
LZ 690699307 US

Wim S
VB 487 983 076 US
LZ 690699386 US
 

Attachments

  • tiger5.jpg
    tiger5.jpg
    25.8 KB · Views: 1,191
Did I mention I took the pile of money and bought a resort and am retiring? See pic attached (JOKING of course).

Hello Rabstg, this resort you attached a picture of is in Dubai. If you are not a guest, it costs $75 US just to get past the front gate on the mainland. It has it's own heliport and it's suites cost many, many, many thousands a night. If you had made off with the loot, you would not even have enough to pay to use the bathroom! 🙂 🙂 🙂

Regards and thanks

Anthony
 
hmm,

Tom,

The datasheets there are a little different from mine, and you are quite right, the 1995 datasheet from motorola for the mj* transistors shows the gain plummet below a certain threshold...makes the transistors look quite different from the more recent datasheets for the on-semi mjl* parts...

As an interesting exercise you should compare the motorola datasheets from the same year for the both transistors...the 1995 datasheets for the mj & mjl parts show exactly the same characteristic and as far as I can tell are absolutely identical except for differences caused by the packaging...

As far as I can tell on-semi has updated some but not all of the datasheets. The mjl docs are clearly different, the mj ones not much at all...

On semi has either learned to make the transistors better, or they have learned to measure them better...but in either case I still believe the transistors will be virtually the same, or at least as close as any pair of transistors made from the same die.

Stuart
 
Hi All-

SO I rxed the brds and sent them out..

As I was looking at them I was scratching my head over C102, 103 and 104.

I have to ask why the back to back caps(102 & 3) when you have the bypass(104)?

And please excuse me if I missed it in the earlier portion of the thread.. If it was mentioned please let me know the post # and I will read quietly in the corner...
 
For all us aussies 2sc2240 and 2sa9** found at Wagner electronics
http://www.wagner.net.au/
That's easy because they were $0.81 for one type and $0.7* something for the other 🙂 That's cheap enough for me!! I've decided to go for the MPSA56/06 i think it was combo too for the input, see how it goes, i've got MPSA42/92 here anyway in case 🙂
Anyone know anyone who sells good caps and resistors (for the board) in australia? I can buy o/s, if i have to , trying to avoid it as much as possible... 🙂
Aaron
 
While waiting for the boards, i am collecting parts and looking to the pinnings and distances on the PCB Jan Made.

It seems to me that the pinning on the PCB board from R126 and R144 (the trim pots for bias) are not correct.

i will try to explane, but my english is not too good, sorry for that:

On Jan's schematic one endpin from a pot is connected to the wiper of that pot, the other endpin goes to Q108.

On the PCB layout the two endpins from each pot are connected and the wiper goes to Q108.

I looked at the datasheet from the cermet bourns pot, and the middle pin is the wiper.

I hope i am wrong?