60 Hz Quest

[Tom Phillips]

...I put my DMM on the speaker jack and it says Zero Hertz (0 Hz). Is that possible.?...

That just means that the signal amplitude is too low for that particular meter to make a reading. That's good news and your ears already told you that the hum is much lower.

[trem]

The phase inverter is supposed to be 12AU7. A 12AX7 won't bias properly unless the circuit is modified.

The bias circuit has two zener diodes regulating the bias at -35V. Is that working? (This circuit is crying out for a mod.)

Are all four diodes in the full wave bridge working? Is the ripple on B+ 60Hz or 120Hz?

Yeah...sorry...need to get an AU.
The bias circuit should be OK. RG helped me to get 2 bias pots working. One for each pair of power tubes.
I ended up ditching the Zener's...I have -65 VDC available. All new parts in the bias circuit.
Pardon my ignorance, but.....to check the ripple, I can just put my scope on the filter cap(s) and read the frequency.?
Thank You

[trem]

Well.....I put my scope on the reservoir cap.
They are two cans. 2x 50-450 //.....and then those two in series.
Anyway...the signal jumps around a bit...kind of a saw-tooth wave. I guess that is the cap charging (pretty vertical) and then the other side of the wave is more sloped. I imagine that is the slower discharge time.
Using my God Given abilities with a scope, I get:
5-1/2 VAC
125 Hz

I did not follow that through the rest of the B+ rail.

[loudthud]

Your scope or probe might be damaged if they can't handle high voltages on the input. You should be using a 10X probe or 100X if available. You should check to see if you can find a Voltage rating for the probe. It needs to be 600V or more.

Most scopes have an three position input selector for AC, Ground or DC. The AC position inserts a capacitor in series with the input, but it may not be rated at 600V. See if you can find a rating for the scope. If you are confident the scope and probe can handle the high Voltage, set the input selector to Ground, set the Volts per Division knob to the highest setting and connect the probe to the B+ at the fuse or first filter cap. Move the input selector to the AC position. You should see a sawtooth waveform. Adjust the Volts/Div, time base and triggering to get a stable display. Measure the time between two adjacent peaks of the waveform, the frequency is the reciprocal of that number. 16.6mS is 60Hz, 8.3mS is 120Hz.

If your scope or probe can't handle the high voltage, you can construct a Voltage divider from the B+ node to lower the Voltage seen by the scope. Use something like 220K 2Watt and 22K 1/2Watt. If the scope does not have an AC input, simply connect a .022uF cap between the voltage divider and the probe tip.

[trem]

Loud -
Thank You for the tutorial.
I am using a 10x probe.
Did you see my previous post.?
Thanks

[loudthud]

Loud -
Thank You for the tutorial.
I am using a 10x probe.
Did you see my previous post.?
Thanks

Saw it after I posted. That sounds about right. Check the ripple on the bias supply.

[trem]

Saw it after I posted. That sounds about right. Check the ripple on the bias supply.

OK...here we go.
I will just give you my math as it came up.
Amplitude = 4.2 VAC
Frequency = 62.5 Hz
Hopefully my pic is decent enough to see the clipping on the leading and then trailing edge of the positive and negative wave halves. I do not know what that is indicative of.
This is with just the 10-150 bias cap...no other caps are jumped in.
Thanks Again

[Tom Phillips]

...Amplitude = 4.2 VAC...

You didn't specify the units but by "4.2 VAC" I'm assuming your reading is volts "volts peak to peak."
Anyway...no mater what the units are that's way too much AC riding on your DC bias voltage.

FYI we would calculate percentage ripple as follows:
Assuming you have 4.2 Vpp we divide by 2.828 to convert to Volts RMS. 4.2/2.828 = 1.49 Vrms
If your bias supply is producing -65 Vdc at the same node where you measured the AC then the ripple percentage calculation is (1.49/65)*100 = 2.3%. This, as we suspected, is way too much for a bias supply. Less than 0.1% is expected and far less than that is common.

You modified the bias supply. Can you supply a schematic of the actual bias supply circuit you are using? I'm thinking that, rather than just a larger single filter cap, a two stage filter section may be a good solution.

Cheers,
Tom

[trem]

Yeah...sorry.....peak to peak is 4.2V
Really all I did was to get rid of the Zeners, and add a bias pot for each pair of power tubes. The component values in the bias circuit are per the schem...
Let me dig up the schem that RG drew for me.
Hold on.......
OK...took a quick poke with my DMM...about -70 VDC at the bias cap where I measured the ripple.
Here is a pic of the bias circuit schem. Your idea of a Two Stage sounds good to me.
I have just 2 bias pots.....not 4.
AND I removed the 2 Zener Diodes.
Looks like I have about -40 VDC on the power tube grids.

And a picture of the power section.....

[52 Bill]

My first guess is that without the Zener diodes the drop across the 10K resistor is a part of the problem. Try adding the additional filter cap to the second side of the 10K resistor, the spot where the Zeners used to be.

[Tom Phillips]

I agree with 52 Bill's suggestion to add a second filter cap where the zeners were.
When you added the two adjustment pot circuit sections you significantly increased the loading on the bias supply. That caused the ripple to increase because the 8uF filter cap is too small for the higher load. I'd suggest replacing the original 8uF cap with a 25 to 50uF cap and adding a similar sized cap in the old zener location. That should reduce the ripple sufficiently to solve any of your hum problem that is due to bias supply ripple.
Cheers,
Tom

[trem]

Bill Tom Loud -
Thank You.
I will see if I have a couple of 50-150 caps. If not, I will order them.
Will get back to when I have those in. I may have to make a Mouser or CE order.
Thanks Again
REALLY appreciate it

[g1]

Sorry to go off on a little bit of a tangent, but I'm still hoping someone has an explanation why pulling the phase splitter killed the hum coming from the bias ripple.

[loudthud]

This kind of bias supply doesn't do well with big filter capacitors. They can take over a minute to charge up to final voltage. See if you can find the schematic for Vox AC50/4 Amplifier Mk III drawing OS/072. It uses two 4uF electrolytics in series in place of the .25uF cap.


Edit:

Sorry to go off on a little bit of a tangent, but I'm still hoping someone has an explanation why pulling the phase splitter killed the hum coming from the bias ripple.

My guess is that the slave side of the PI has a much lower output impedance because of the feedback which kills the ripple on the lower EL34s when the PI is installed. When the PI tube is pulled, the ripple is equal on both sides and it cancels out in the OT.

[g1]

Thanks for that explanation LT. Here's that schematic you mentioned:
http://www.drtube.com/schematics/vox/ac501965.gif