some doubt about transistor marks and hfe range

[Dimitree]

hi everyone
I was looking at some '80s jap service manuals and found this:



I checked the 2SC945 datasheet and found that it has 4 hFe2 ranges.
The schematic indicates that "K" range is needed.
But what's the "L" for?
and furthermore, if "K" is specified, why should we care about the "color" of Q3 and Q4 in order to remove a resistor? maybe the "color" of Q3/Q4 specifies a tigher hFe range inside the "K" range? the datasheet doesn't talk about "color" (other datasheets do).

[armdnrdy]

"L" shows up as "lead free" here:

http://pdf1.alldatasheet.com/datasheet-pdf/view/413186/UTC/2SC945L-X-T92-K.html

But....it could be a factory notation of some sort as well.

What exactly is the schematic you posted? Seeing the whole circuit would probably help.

[PRR]

> "L" shows up as "lead free" here:

Unlikely in "... '80s jap service manuals".

> Seeing the whole circuit would probably help

Agree. Can't even see a value for R45.

But it *appears* to be a signal switch. (Q4 is used "inverted".) The amount of Base drive current may be critical, and would depend on Beta, which is what those colors usually meant. Insufficient drive would not keep the transistor ON. Excess drive causes large offset voltage (click).

http://www.radiomods.co.nz/transistors/2sc945.pdf

[Dimitree]

yes of course this is the big schematic
http://machines.hyperreal.org/manufacturers/Korg/MS-synths/schematics/KLM-307.GIF

it's the famous korg ms20 synth filter, but the part I'm talking about is the VCA..
by the way my couriosity was about the transistor and not directly related to the circuit.. If the datasheet doesn't talk about colours and "L", how can we undestand the correct hfe needed? Or maybe the transistors were hand marked?

[Dimitree]

so I just searched for a pic of that board, and yes, the transistors are actually hand marked!
check this:
http://nwheaton.homemail.com.au/images/Korg_KLM-307A_PCB_top.jpg

[duck_arse]

http://pdf.datasheetcatalog.com/datasheets/228/239948_DS.pdf
http://www.semicon.toshiba.co.jp/info/docget.jsp?pid=2SK30ATM&lang=en&type=datasheet
http://pdf1.alldatasheet.com/datasheet-pdf/view/253549/ETC2/2SC1000.html

seems diff manufacturers use colours for hFE ranges, and Idss for jfets.

[Dimitree]

yes I know, but in that case the colours is marked with a letter, like, in the 2SK30A, GR (green) or Y (yellow).
here instead, the letter K means the same thing, but it's not coded with colours names.. So I can't see any relationship between the green or blue color and the "K".
Furthermore as you can see on the pcb photo, the transistors are actually coloured on the top of their packages, and having seen other photos of that board, I can tell that sometimes they are green, sometimes they are blue, but it's always the "K" version of the transistor.
So I guess the colour encodes a "sub-range" inside the K range. I guess this sub-range can be determined looking at that small resistor table in the schematic.

[PRR]

So I was half-right. It is a "switch" but the important part is how it acts at "part-ON". There is a narrow range of small values where (with opamp) it acts as a variable gain.

> how can we undestand the correct hfe needed? Or maybe the transistors were hand marked?

That's my guess. KORG had piles of transistors, a test rig, and low-paid labor.

The selection may not only be about hFE. Vbe and stray resistance may matter also.

For your uses: use HIGH gain parts, then fiddle R44-R45 so that nominal voltages at Initial Gain or from the EF give expected gain. Try 250K pot plus 220K fixed in series.

Still do not know what (L) is. This *may* indicate that you take these parts from bin L, which is where the sorted and painted transistors went after testing.

[Dimitree]

many thanks, so R44-R45 can just be removed and I could put a (if it would exist) 330k trimmer in there and adjust it.
but how can I understand what is the expected gain?

[digi2t]

I get the feeling that the color code was the on the face of the factory test indicator. If the needle went into the orange or yellow, then cut R44, if it was in the green or blue, then cut R45.

Since the gain range for K was between 300 and 600, and the two had to be matched, this might explain the color coding.

I found the same info on the «L», meaning lead free plating.

[armdnrdy]

So I guess the colour encodes a "sub-range" inside the K range. I guess this sub-range can be determined looking at that small resistor table in the schematic.

Probably correct.

What I would do...is match the HFE for two 945s and socket the resistor. Try the two resistor values to see what difference there is.

Unfortunately, when working out a build from an old schematic...there is almost always missing information.

Isn't there a thread running right now, "how long does it take to build a pedal?" The R&D can overtake much of the actual building time! 

[Dimitree]

actually, on the Korg MS10 and early MS20, the circuit is the same but there is only 1 transistor, but the requirements for it (K range and color/resistors) are the same, so it's not only about matching them but actually tune them (or tune it) to the circuit, as PRR pointed out