That signal across the cap is whatever the signal is being fed to the input of the amp.
Not sure if I'm doing this right, but based on the voltages and the 3k B+ resistor I calculated the first two stages as drawing 9.667mA and based on that I calculated the first two stages to be presenting a load of 31.7k and if I plug that and the 66uF cap value into a high pass filter calculator I get this.
Also if I plug the values into a parallel RC calculator I get this.
Even a 33uF filter cap seems to be enough.
Not sure if I'm doing this right, but based on the voltages and the 3k B+ resistor I calculated the first two stages as drawing 9.667mA and based on that I calculated the first two stages to be presenting a load of 31.7k and if I plug that and the 66uF cap value into a high pass filter calculator I get this.
Also if I plug the values into a parallel RC calculator I get this.
Even a 33uF filter cap seems to be enough.
Since you have a clean sine at the speaker terminals it says that at 20Hz. all is well. How about all the other frequencies? With speakers attatched? You won't know if you have a stability issue until you have one. Best practice is to decouple the stages in the original design plan. You are sort of trial and error designing. But that's DIY. New day, new toy. It's coming along, though.
Last edited:
So I do actually have a problem.
I get this oscillation. 1uS/DIV 2V/DIV 1X probe across 4 ohm load.
I can stop the oscillation by doing either of three things.
1. Plugging in the audio cable with or without it being connected to the source.
2. Shorting the input jack to ground.
3. Connecting a clip lead to the 47pF cap and 1 meg junction and holding it in my hand.
Here's the layout of the first two stages and the input and output jacks.
I can move the top capacitor some and the waveform changes.
Some things I can think of.
1. The 47k resistor should be across the terminals of the input jack.
2. The input jack should be moved elsewheres.
3. I need shielded cable from the input jack to the .68uF cap
4. The jack should be grounded right at the chassis.
5. Need shielded cable on the wiring going to the 6V6 control grids.
6. RF shorting cap on the input connector.
7. Something else.
I get this oscillation. 1uS/DIV 2V/DIV 1X probe across 4 ohm load.
I can stop the oscillation by doing either of three things.
1. Plugging in the audio cable with or without it being connected to the source.
2. Shorting the input jack to ground.
3. Connecting a clip lead to the 47pF cap and 1 meg junction and holding it in my hand.
Here's the layout of the first two stages and the input and output jacks.
I can move the top capacitor some and the waveform changes.
Some things I can think of.
1. The 47k resistor should be across the terminals of the input jack.
2. The input jack should be moved elsewheres.
3. I need shielded cable from the input jack to the .68uF cap
4. The jack should be grounded right at the chassis.
5. Need shielded cable on the wiring going to the 6V6 control grids.
6. RF shorting cap on the input connector.
7. Something else.
Put the 47K where the 1M is. Toss the 1M.
Ground the input to the chassis.
Do you have a SMPS (computer) or a flouescent bench light nearby. It looks like stray atmospheric radiation pickup.
Ground the input to the chassis.
Do you have a SMPS (computer) or a flouescent bench light nearby. It looks like stray atmospheric radiation pickup.
The .68uF input cap isn't needed unless you are afraid of DC coming in from a source device. Or is it a Van Alstine input filter? If so, remove it for now anyway. Is your scope probe getting a ground to the chassis? I usually see that kind of noise just by touching the probe with my finger.
Found the problem. The input jack is right next to the terminal the cap goes to that feeds the top pair of 6V6 tubes. I disconnected the connection to the jack and moved the wire away and no oscillation.
I guess it never was a problem before due to the gain not being as high from the input and phase splitter stage.
I guess it never was a problem before due to the gain not being as high from the input and phase splitter stage.
Got the jack moved and all is well.
The Schiit Saga S has no DC on its output, but I prefer using an input coupling cap so that no matter what I plug into it I don't have to worry about DC being on the output.
The Schiit Saga S has no DC on its output, but I prefer using an input coupling cap so that no matter what I plug into it I don't have to worry about DC being on the output.
So what's the takeaway? What would create that kind of signal in an amp with no signal being input? Avoiding the pothole doesn't fix the pothole. This kind of noise intrusion on open inputs is normal but for its amplitude.
Last edited:
Best I can figure is the stray capacitance between that terminal the 1uF cap connected to and the terminal on the input jack was just enough to start it oscillating and the cable capacitance was enough to stop it from oscillating.
So I tried an 83pF cap with 100k series resistor across the feedback resistor and I was able to eliminate the 47pF cap and 500k pot.
10kHz waveform. 1V/DIV 4 ohm load 1X probe.
At 20kHz the output waveform is very close in amplitude to the input waveform with both scope channels set to where the waveforms are equal amplitude at 1kHz.
There's a slight phase shift and if I had a phase angle voltmeter I could tell you exactly how much phase shift.
So I tried an 83pF cap with 100k series resistor across the feedback resistor and I was able to eliminate the 47pF cap and 500k pot.
10kHz waveform. 1V/DIV 4 ohm load 1X probe.
At 20kHz the output waveform is very close in amplitude to the input waveform with both scope channels set to where the waveforms are equal amplitude at 1kHz.
There's a slight phase shift and if I had a phase angle voltmeter I could tell you exactly how much phase shift.
Oh ok so with it set like I have it each division is 90 degrees and there's four divisions in 360 degrees. So one just takes 90 and divide by 5 to get 18 degrees.
It would be an interesting comparison at 20KHz. to look at the raw square wave from your sig-gen to see how the leading edge rise shape compares to the amp output. You see a slight rounding that is very common @ 20KHz. Much greater detail displayed.
With the rise time of a square wave being so fast it's not a sin for the amp not to keep up.
I'm well pleased with the sound of the amp.
Now I just gotta do some tweaking to the placement of the open back speaker so that the lower bass sounds better.
Now I just gotta do some tweaking to the placement of the open back speaker so that the lower bass sounds better.
Last edited:
Thinking about it I didn't have this issue until I went with the split load phase inverter.
The reason was that there was only one phase inversion with the long tailed pair so the signal on the terminal next to the input jack was inverted.
With the split load phase inverter the signal now has two inversions meaning the signal on that terminal would be in phase with the input thusfore allowing the stray capacitance to turn the amp into an oscillator.
Should I keep the two milliammeters or just use the one?
So while wanting to tone down a blue LED temperature display in the same size housing as the ammeters I found a piece of a thin green scope graticule and cut it to fit. Made the blue green and toned the brightness down some. Decided to try it with the blue ammeters and I got this. The picture with the room lights on shows the shade of green it is.
Well I discovered a problem.
I used a box with a pair of 100 ohm resistors inside to make the signal from the preamp mono. It sounded great on most things so I thought.
However playing a reel to reel tape of songs that tend to exploit the then newer stereo technology I noticed an issue in that the sound wasn't quite right, but I could unplug one cable and all was good again.
So I'm thinking that even though the preamp has a 75 ohm output impedance the 100 ohm resistors are two low.
I can handle it two ways.
1. Try a pair of 1k resistors and drop 11mV of signal.
2. Use a second 47k resistor in the amp since I have a stereo input jack and place two resistors after the input jack to combine the stereo to mono. Thinking maybe 10k would work. That way I have no drop in signal level.
The 1k resistors is very easy to do as I have the necessary resistors.
The stereo input mod I'd need to run a second audio cable to the amp and use a stereo RCA to stereo 1/4" adapter.
That may also explain why the bass seemingly doesn't sound as good.
I used a box with a pair of 100 ohm resistors inside to make the signal from the preamp mono. It sounded great on most things so I thought.
However playing a reel to reel tape of songs that tend to exploit the then newer stereo technology I noticed an issue in that the sound wasn't quite right, but I could unplug one cable and all was good again.
So I'm thinking that even though the preamp has a 75 ohm output impedance the 100 ohm resistors are two low.
I can handle it two ways.
1. Try a pair of 1k resistors and drop 11mV of signal.
2. Use a second 47k resistor in the amp since I have a stereo input jack and place two resistors after the input jack to combine the stereo to mono. Thinking maybe 10k would work. That way I have no drop in signal level.
The 1k resistors is very easy to do as I have the necessary resistors.
The stereo input mod I'd need to run a second audio cable to the amp and use a stereo RCA to stereo 1/4" adapter.
That may also explain why the bass seemingly doesn't sound as good.