Continuing our Mic Preamp Design Adventure… see past articles:
Part 5: Designing the Printed Circuit Board
Part 4: Amplifying mic level input signal
Part 3: Distortion & Harmonics. Solid-State vs. Vacuum Tubes
Part 2: Impedances & Loading – Adventures in building a Tube based Mic Preamp
Part 1: Adventures in building a Tube based Mic Preamp
Some photos as I build the very first prototype!
On the left, you see the tube adapter board, with the tube sockets and right-angle headers installed. This whole piece will be soldered to the main board. You can see the silkscreen layout on where it should go.
Photo of finished board, with output transformer installed and the vacuum tube adapter also installed. One capacitor on the center of the board I forgot to install and didn’t realize until after I’ve taken this photo. The potentiometer for the gain level is also not yet installed and the chips are still missing.
Above is closeup photo of how the tube adapter board is mated to the main board.
Here’s the first time we’re applying power to the board without the chips installed yet. I want to check if the LEDs and relays are working, and if the 12V regulator is working.
PROBLEM #1. Yay! We found our first problem! I noticed the LED for the tube adapter isn’t lighting up. I know I tested the tube adapter board independently and it lit up during my testing. It’s just a simple LED… how can I fail that?
FIX #1: Turns out besides the current limiting resistor on the tube adapter board, I also have another one on the main board. Surely, one of them isn’t needed. The quickest solution I chose at this moment is just place a jumper across the LED current limiting resistor on the main board.
Red LED under the tube socket is now working! That was an easy fix!
So I finish building the prototype and hooked up my scope and signal generator. And that’s where I discovered:
PROBLEM #2: No Signal Output….. NO SIGNAL OUTPUT…. I turn knobs, switches, check my scope and signal generator connections, and yep… we have NO SIGNAL OUTPUT!
This is a SERIOUS problem! It is at this point that doubt and panic creeps in. I’m scratching my head, because I know this thing should work! I breadboarded it. This PCB design is just “formality” and making it look close to the final version. But the design should work.
This breadboard mess of spaghetti wiring and various FABModule boards was the validation of our design!
So now, we proceed to become serious in troubleshooting this guy. We get our scope leads out, and start probing around.
Aha! Some sort of good news…. we have signal output and amplification working on the FIRST STAGE of our design.
This output then goes to the SECOND STAGE (which is our vacuum tube stage), and from there, the signal dies and disappears!
They say you should try the simplest solution first, so in this case the simplest solution is replacing the Vacuum Tube. Maybe it’s bad? You sometimes don’t know what you’re going to get with eBay purchases. NOPE! Still the same problem. It’s not the Vacuum Tube.
We go keep looking….
I decided to forget Stage 2 for now, and check out the 3rd Stage… so I soldered a piece of wire to the PCB underside, just right after the Stage 2 output, going into the Stage 3 input path. Hooked up a signal generator to this piece of wire, and check the output of the board and we got signal! Good news…
So that means, the 1st stage is working, and the 3rd stage is working properly. The problem is isolated on the 2nd stage only.
At this point, I have the oscilloscope set to “AC” coupling. This setting allows the scope to ignore any “DC” voltages in the signal, and just get the “AC” component, i.e. our sine wave. I then switch the scope to “DC” coupling and reset the trace to the middle of the scope screen, and began probing the same points on the PCB.
WHOOOOAH!!!! Something weird is happening. A small AC-coupling capacitor separates the 1st stage and 2nd stages. At the first stage (before the capacitor), we have almost 0Volts DC (which is good!). But at the other side of the coupling capacitor, we have a DC component riding on top of the AC waveform! About -10Volts DC. That’s seriously wrong…. Because now, with that DC component on top of our AC signal waveform, we’re shifting the whole shebang downwards! Our DC biasing is all wrong!
That AC component + the DC offset is basically our Vg (grid voltage) and so now, we’re off to some lala land and no wonder we’re not getting any output out of our 2nd tube stage.
I pulled out my breadboard setup and start investigating and comparing resistor values against the resistors on the PCB. Hmmmm….. one resistor marked 10K on the PCB should have been 100K! In Eagle CAD, sometimes I start copying components and the value of this component wasn’t changed from it’s original value of 10K to the new 100K. The silkscreen was printed wrong, and I just followed the silkscreen during the build… boom! Fail there.
But that still didn’t fix the problem… we still have no output for Stage 2. Studying the breadboard again, I see one resistor was used in the breadboard experiment that is NOT in the final PCB design. Could that one little resistor be the cause of the problem?
I had to McGuyver fix this, so the missing resistor will have to be soldered on the underside.
Power up the board, and start checking again with the scope…. YES! The DC voltage offset disappeared, and now we’re left with just the AC component signal going into the vacuum tube stage. We check the output of the 2nd stage…. and we have output! IT’S FIXED!!!!
Yes, we got output and the tube is working and giving us the 2nd harmonics we saw during the breadboard testing.
Success! The board is passing signal and it’s working as intended. At this point, I haven’t done listening tests yet, that will wait for another day.