Debugging my circuit: loud dc hum in signal path early on.

[guitylerham]

So I'm building a Leslie Sim from the Korg CX-3 and am tracing the signal using a probe starting at the input. All goes well until I follow through like the third capacitor in the circuit and all of a sudden there is a very loud hummmmmmmmm. It is the same pitch and volume as the positive DC terminal sounds when I touch that where the power comes into the board. One on side of this capacitor I have clean signal, on the other side, I get loud hum and quiet signal. Does that mean there is DC current leaking into my audio path downstream of this capacitor? What is a good way of fishing for where the DC leaks into the path? One of my transistors is getting +10 plus volts on the E, C, and B! That ain't right from what I'm learning. I will continue probing but thought I should ask and get some ideas rolling in at the same time. Thanks!

[petemoore]

  Assuming the schematic...whoops.
  Figuring it's probly an NPN...probably, whatever the case a transistors base must be 1 transistor diode drop above the emitter voltage, the diode is what it is for Si NPN...base above emitter by the transistor diode drop, collector could be the output, but so could the emitter...referring to the _ _ _ _ _ _ _  diagram {                    }...
  IF it's a gain stage the collector should be 'middle' of the supply voltage [about 1/2v of supply, which could be {9v supply times 1/2 = 4.5v at the collector}.  
  If the emitter [emitter follower = 'buffer' stage] is the output pin then the collector would be supply voltage, and the emitter would ride high enough on the supply line to make voltage swings +/-...so biased somewhere in the 'middle' of the power supply.
 Something is wrong with the DC bias 'periphery' wiring of this stage, or a cap is letting DC through [unlikely] or a resistor miss or miswire].

[nexekho]

DC hum?  DC can't hum, DC isn't a wave, it's static.  Do you mean AC line frequency?  60Hz in the US, 50Hz in the UK.

[petemoore]

  The same sound as when DC rail is probed...
   Put the hum-tone on the backburner until after the transistor function and bias for that function are known and achieved, the way it is now the Q can't do anything except signal block [and hum].
  Hum indicates problem. Misbias is definitely a specific problem, solve the problem then test again the transistor stage performance.

[guitylerham]

Ok, forgive my lack of preparation and expecting to give you guys anything to work with! I'll post up the schematic and layout as well as get the voltages of all the transistors and IC's tonight after work!

[guitylerham]

Ok, here's the Schem



Here's my attempt at creating an easy to see layout + PCB




Here are my voltage readings:

*all transistors are 2N5088 (lavender color) except for the 2SK170 (light blue color near the input of the circuit.
** The MN3004's that are called for in the Schem are now MN3007's (with requisite clock doubling capacitors installed).

I feel like I got in way over my head with this circuit! I thought I could just follow the layout and PCB provided by a forum member. But I noticed when trying to follow the schematic, there are many differences between the two! I guess that's what happens when someone modifies an exisiting circuit from a keyboard to be a standalone guitar pedal. I'm pulling my hair out trying to figure out what's up!

Ok, first step is to redo my transistors. I'm using 2n5088's which apparently are EBC and going in the same orientation as the original 2sc945 which are ECB! Dang it!

[guitylerham]

Here are the voltages for the various components as tested with +11.95vdc at the power supply.

My god there are a lot of parts in this circuit!

Again, all transistors are 2n5089 unless otherwise noted.

*Q0 (2SK170)
This FET transistor is not in the schematic so the original designer of this special layout must have felt it necessary to include it in the circuit...
D 11.09
G 5.01
S 5.73

Q1
E 2.89
C 11.09
B 3.42

Q2
E 2.24
C 6.11
B 2.79

*Q3 (2N3904 because my last 5089 broke!)
E 2.54
C 11.09
B 3.12

Q4
E 1.9
C 11.09
B 2.43

Q5
E 0
C 0
B 0

Q6
E 0
C 0
B 0

Q7
E 0
C 2.54
B 0.31

Q8
E 3.82
C 11.09
B 4.25

Q11
E 0
C 2.18
B 0.32

Q12                 
*there is a switch wired to the "Slow/Fast" pad down at the bottom right corner of layout. When switched, a small voltage is sent to that pad to activate the change in Leslie speed.
   "off"   "on"
E   0        0
C 0.34    0.01
B  0       0.65

IC1 (MN3007)
1 11.65
2 5.80
3 1.81
4 1.05
5 0
6 5.81
7 3.45
8 3.45

IC2 (HEF4046)
1 0
2 5.81
3 5.81
4 5.81
5 0
6 1.49
7 1.49
8 0
9 5.84
10 4.5
11 4.11
12 8.95
13 0
14 11.66
15 0.50
16 11.66

IC3 (4558)
1 5.5
2 5.5
3 5.5
4 0
5 3.76
6 3.84
7 3.84
8 11.09

IC4 (MN3007)
1 11.3
2 8.6
3 4.36
4 1.56
5 0
6 5.63
7 3.8
8 3.8

IC5 (HEF4046)
1 0
2 5.63
3 5.63
4 5.63
5 0
6 1.45
7 1.45
8 0
9 5.08
10 3.76
11 3.36
12 8.62
13 0
14 11.29
15 0.49
16 11.29

IC6 (LM324)
1 5.57
2 5.76
3 5.76
4 11.95
5 5.8
6 5.77
7 5.55
8 5.77
9 0.01
10 5.04
11 0
12 4.75
13 8.76
14 0.74

This looks like a crap ton of numbers to sift through but if someone could point me in the right direction, I'm more than eager to learn and solve this. What do you all think so far?

[guitylerham]

Bump!

So I switched all the 2n5089 transistors around: bent the base around the collector to match the pinout of the circuit layout. Much more of the circuit is now carrying signal and without much hum. There is some bad stuff happening around the MN3007's (garbley distortion). I'm going to test all the transistors to make sure they're working.

Petemoore: I will try and decipher the schematic and see which transistors do what so that I can accurately assess them based on your statements. THanks for chiming in!

[petemoore]

Re: Debugging my circuit: loud dc hum in signal path early on.
« Reply #6 on: Yesterday at 02:30:20 AM »




--------------------------------------------------------------------------------

Here are the voltages for the various components as tested with +11.95vdc at the power supply.

My god there are a lot of parts in this circuit!

Again, all transistors are 2n5089 unless otherwise noted.

*Q0 (2SK170)
This FET transistor is not in the schematic so the original designer of this special layout must have felt it necessary to include it in the circuit...
D 11.09
G 5.01
S 5.73
  Seems workable fora Jfet source follower
Q1
E 2.89
C 11.09
B 3.42
  Q1 is an emitter follower ? How's the signal sounding before/after this stage ?
Q2
E 2.24
C 6.11
B 2.79
  Seems workable for a gain stage, how's the before/after signal strength.
*Q3 (2N3904 because my last 5089 broke!)
E 2.54
C 11.09
B 3.12
These look like Emitter follower voltages.
Q4
E 1.9
C 11.09
B 2.43
  As do the readings for Q4 here.
Q5
E 0
C 0
B 0
Noting nothing going in or out of this one...
Q6
E 0
C 0
B 0
Or this one//
Q7
E 0
C 2.54
B 0.31
Doesn't look workable either...
Q8
E 3.82
C 11.09
B 4.25
  Base is above emitter by at least a diode drop, collector is at V+...thse are workable E-follower voltages.
Q11
E 0
C 2.18
B 0.32
  Base isn't adiode drop above emitter, collector barely or has no room to swing negative at a mere 2.18v from Gnd.
Q12                 
*there is a switch wired to the "Slow/Fast" pad down at the bottom right corner of layout. When switched, a small voltage is sent to that pad to activate the change in Leslie speed.
   "off"   "on"
E   0        0
C 0.34    0.01
B  0       0.65
    Not sure about this one..probably is in the LFO, there are enough 'wierd' voltages present that they may have a common cause like/maybe something in the supply or bias is cause for multiple misbiased stage.
IC1 (MN3007)
1 11.65
2 5.80
3 1.81
4 1.05
5 0
6 5.81
7 3.45
8 3.45
  Don't know much about these chips, aparently pin1 is PS rail, pin 5 is Gnd., the other pins...I can't chime in usefully on.

IC2 (HEF4046)
1 0
2 5.81
3 5.81
4 5.81
5 0
6 1.49
7 1.49
8 0
9 5.84
10 4.5
11 4.11
12 8.95
13 0
14 11.66
15 0.50
16 11.66
   I don't kno about this one either, inputs and outputs need room to swing +/- from their bias voltage, so 1/2 of supply V allows equal swing on both sides of bias...look at which pins are assigned as In/Out, see if there's room to swing from bias toward Gnd, room t swing from Bias toward V+...without hitting a rail...ie somewhere's toward 'middle bias' keeps the signal AC swings from 'hitting a power rail' and clipping the stage output.

IC3 (4558)
1 5.5
2 5.5
3 5.5
4 0
5 3.76
6 3.84
7 3.84
8 11.09
  8= V+, 4=Gnd. [0.ov], the other pins are 'center-ish' to the power rails, this chip looks to have operational bias and can swing voltage at the output + and - from bias points at least a couple volts either way before 'nearing' a power rail [the chip needs 'X'amount of 'elbow room' to swing voltage, nearing a rail causes distorton [the chip can't swing all the way to the rail...needs whatever a volt? or so of 'elbow room' [see data sheet...
IC4 (MN3007)
1 11.3
2 8.6
3 4.36
4 1.56
5 0
6 5.63
7 3.8
8 3.8

IC5 (HEF4046)
1 0
2 5.63
3 5.63
4 5.63
5 0
6 1.45
7 1.45
8 0
9 5.08
10 3.76
11 3.36
12 8.62
13 0
14 11.29
15 0.49
16 11.29

IC6 (LM324)
1 5.57
2 5.76
3 5.76
4 11.95
5 5.8
6 5.77
7 5.55
8 5.77
9 0.01
10 5.04
11 0
12 4.75
13 8.76
14 0.74
The 324 biases the same way as a single or dual opamp, the in/out pins need to be biased so there's room to swing +/- within the power limitations, the power pins get V+ and Gnd. just like the dual or single opamp, it is a package with 4 single opamps with common power rails [internally connected so you only have 1 + and 1 Gnd to connect.

[guitylerham]

Thank you so much, Petemoore for taking the time to look at my numbers! It's off to work for me but I'm going to dig into this after work tonight and this weekend. I'll post back with any new developments. Thanks again, really.

[guitylerham]

Q1
E 2.89
C 11.09
B 3.42
  Q1 is an emitter follower ? How's the signal sounding before/after this stage ?
Q2
E 2.24
C 6.11
B 2.79
  Seems workable for a gain stage, how's the before/after signal strength.

Ok When probing Q1:

Base: signal is quieter (and thinner sounding due to the capacitors, I assume) than the original input signal to the circuit. There is a "non-grounded" type hum as well. The same hum I get when I lift off the probe and just listen to the hum.

Emitter: Clear and louder than Base signal and no detectable hum.

Collector: Dead quiet with the TINIEST amount of signal.


When probing Q2:

Base: Similar in every way to Q1's Base.

Emitter: Same as the Base, if not a tiny bit louder or clearer and with no hum.

Collector: TWICE as loud as the Emitter. (Is this negative feedback?)


What are your thoughts so far?

I will continue piecing these clues together.

[guitylerham]

Update:

IC1 and IC4 are the MN3007's (mn3004's in the schem. IC1 is labled but IC4 isn't labeled so it must be the other one). There are 10k trimmers used to bias these IC's up to 6v at pin 3. I can bias IC4 up to roughly 6v but the trimmer on IC1 maxes out before I can get past 5.5v. Is this an issue? Do these need exactly 6v?

[guitylerham]

Bumplestiltskin.

[PRR]

> through like the third capacitor in the circuit  ...a very loud hummmmmmmmm.

Which cap?

> It is the same pitch and volume as the positive DC terminal

The hum at your DC +12V point should be very very clean.

[guitylerham]

PRR: Originally, I had all the transistors improperly oriented since the layout uses a different transistor than the 2N5089's that I used. Once I bent the transistor bases/collectors the right way for the layout, much of the loud hum has disappeared. However there are still issues with the circuit. Based on the voltages I found, some aren't even getting any power. I need to figure out what's up with those. I'm trying to flush these issues out but I'm shooting in the dark!

So putting the probe on the +12v supply should not make any noise? It makes smooth sounding "mmmmmmmmmmmm." I'm using a cheap radio shack multi-voltage wallwart, by the way.

[petemoore]

I'm using a cheap radio shack multi-voltage wallwart, by the way.
   Debuggin hum in Ps can be as simple as using batteries to supply 'pure DC'...can't beat it because it'll be totally ripple free and simple. Other methods include what you'llprobably need anyway: A PS regualtor/filter set up.
  Suspecting a multi-voltage RS cheep WW is putting out some ripple is 'normal', verifying it isn't too hard either [use ripple free supply and if the hum goes away...there was hum in the PS...put a regulator/filter circuit in there.
   ie probably..the RS PS lets ripple through...but there could still be a ground problem in addition.

[guitylerham]

Well, I messed up. I reversed the polarity of my power supply and now my circuit doesn't pass any sound past the first transistor. I've decided I need to start completely from scratch with a new board and all new components. I have no clue how to test if ones are still (or ever were) useable. I am too frustrated for words. i can't believe I did that. It happened so fast! I'm a dummy. Thanks for everyone's help. Maybe next time I'll do a better job first time around.