[newb q] my ts 808 build seems to need a large base resistor adjustment?

[gaussmarkov]

i breadboarded the tubescreamer (ts) 808 clone on ggg (schem here).  i subbed the 2n3904 for the 2sc1815 as suggested on the schem.  i had to sub a few transistor values also.  the 510k resistors on the transistor bases were two 1M resistors in parallel.  the 220 resistor on the tone pot was an 820 in parallel with a 330 for 235 ohms.  and the 51k resistor in the clipping section was a 27k in series with a 33k to get 50k.

initial voltage readings were

battery:  9.38v
IC: 1 4.63, 2 4.96, 3 4.55, 4 0, 5 4.62, 6 4.63, 7 4.63, 8 9.37
Q1: e 1.08, b 1.65, c 9.37
Q2: e 1.00, b 1.58, c 9.36

i've noticed low base and emitter voltages in other reports, but not this low.  i've also seen pin 2 of the IC a little elevated in other reports, but not this much.  the output level was quite low and full of white noise.  i checked the bias on another 2n3904 with these resistor values outside the circuit and got a similar outcome. 

so after much checking (resistors all confirmed) and fiddling (i put in 2n5089 and mpsa18 transistors as well--similarly voltages), i ended up changing the 500k base resistors (R2 and R12 in the schem) for 100k.  by this time my battery had fallen to 9.22v and the other readings were

IC: 1 4.38, 2 4.50, 3 4.32, 4 0, 5 4.38, 6 4.39, 7 4.38, 8 9.12
Q1: e 1.29, b 1.82, c 9.22
Q2: e 1.38, b 1.89, c 9.22  (maybe i switched the 2n3904s in Q1 & Q2 relative to the original readings)

this sounds pretty good: tone works, drive works, level works.   :)  but the reference voltage has fallen to quite a bit less than half the battery voltage.  is this normal?  are the base and emmitter voltages too low?  they are much lower than the values suggested on ggg. and presumably i have now lowered the input impedance enough to be losing some tone.

my guess is that i should not need this resistor adjustment and that i have done something wrong.  but i have run out of ideas for things to check.  does anyone have any suggestions for tracing my error?  thanks in advance.

[gaussmarkov]

addendum:  more, better.

when i was subbing in different trannies, i was only checking the voltages:  all low compared to the values reported on ggg.  it dawned on me this morning to actually listen to them.  some work and some don't. ???  MPSA18s and 2N5089s are great.  none of the 2N3904s work properly and i tried 3 pairs.

first conclusion:  the voltages on the trannies don't tell the whole story and they don't need to match those on ggg.  (i can hear somebody saying, "everybody knows that!!"  )  here are the readings from more experiments (510K resistors now subbed with 470Ks instead of parallel 1Ms):

Transistor	|	2N3904		|	2N3904		|	2N3904		|	MPSA18		|	2N5089		|	GGG Values
	|	Q1	Q2	|	Q1	Q2	|	Q1	Q2	|	Q1	Q2	|	Q1	Q2	|	Q1	Q2
E	|	0.85	0.91	|	0.89	0.92	|	1.11	0.94	|	0.54	0.54	|	0.47	0.47	|	3.0	3.0
B	|	1.43	1.49	|	1.46	1.49	|	1.69	1.52	|	1.02	1.02	|	0.96	0.96	|	3.5	3.5
C	|	9.19	9.19	|	9.20	9.20	|	9.19	9.19	|	9.18	9.18	|	9.19	9.19	|	8.8	8.8

i got the suggestion for the 2N3904 sub from the ggg schem. the other subs are from the amz-fx schem (sorta  ).

what should my second conclusion be? ???  that i have a batch of bad 2N3904s?  i seem to find much more variation in voltage values with those trannies than the others.  or that the 2N3904 isn't a sub?  seems like others have used it successfully. >:( 

as always, comments welcome!

[gaussmarkov]

i am trying to figure out whether my 2N3904 trannies are bad.  so i decided to run them through R.G.'s transistor tester.  all of the 2N3904s that i have used in my reports above show enormous leakage and very small gains.  as i understand it, silicon transistors like these should have negligible leakage.  my voltage readings across the 2.4K resistor (1.2K 1% resistors in series in my case, measured 2.39K on my meter) were numbers like 0.830V with 4uA base current and 0.826V without.  this is about 335uA of leakage, if i understand R.G. correctly.

to check that i am doing the test correctly, i also ran the successful MPSA18 and 2N5089 trannies through it.  without the base current, i measured zero voltage across the 2.4K resistance for all of these transistors.  so no leakage.  so now i am reasonably confident that my 2N3904s are leaking.  everything is checking out.  i guess this just happens to silicon transistors sometimes.

but ... (why me?) the gains i got for these trannies seem way too low:  the MPSA18 pair measured 0.063V and 0.064V with the base current applied, implying gains of about 6.3 and 6.4 (somewhat more if i take into account that my resistance was lower than the 2472 ohms R.G. uses).  i thought these things had gains in the hundreds?  same deal for the 2N5089 pair:  0.061V and 0.052V.  so obviously i'm doing something wrong.

well, maybe not.  because i am using a breadboard for all of this i can easily measure the collector and base currents in the test circuit myself.  i get:  0.026mA/0.004mA=6.5 and 0.027mA/0.004mA=6.75 for the MPSA18 pair and 0.025mA/0.004mA=6.25 and 0.022mA/0.004mA=5.5 for the 2N5089 pair.  these alternative gains line up pretty well.  i'm getting the 4uA base current described by R.G. i don't expect better given that my battery was 9.17V (not the 9.0V used in R.G.'s calcs) and that my meter doesn't read more than one digit accuracy in uA.

so where am i? 

• i think i have a bunch of bad 2N3809 transistors and that's why my ts808 breadboard build did not work at first.
• i have a working ts808 breadboard build when i use my MPSA18 or 2N5089 transistors instead.
• i seem to be achieving this with transistors that have next to no gain.

i'm making progress and sliding backwards at the same time.  please ... someone ...    are these gains actually normal?  am i just making a bunch of errors that magically line up like this?

[gaussmarkov]

ack.  pls ignore. i failed to edit above properly.