[Roger A. Modjeski]

This is a 2 part video. Part 1: http://www.youtube.com/watch?v=Thwfz6qqklc and Part 2: http://www.youtube.com/watch?v=z46PInHdmXI

And don't forget to watch in HD and subscribe to my channel.

[wushuliu]

This is a 2 part video. Part 1: http://www.youtube.com/watch?v=Thwfz6qqklc and Part 2: http://www.youtube.com/watch?v=z46PInHdmXI

And don't forget to watch in HD and subscribe to my channel.

Really appreciate the videos. My only request is to monitor the focus - the dissembling of the tube for instance is blurred, losing some of the value of the instruction, meanwhile the schematic drawing in the background is in perfect focus and even distracting. The closer you move your hands to the lens the blurrier the image. Though it may be counterintuitive we would be able to discern a lot more if you kept the movement further back or conversely refocus the camera to accommodate your hands being closer.

You can just tape down a small piece of the schematic for instance or some text, picture etc. as both a focus aid for the camera and marker for where to confine your actions. It's cool's that a tube is being taken apart, but frustrating to not really be able see the parts!

These videos are a GREAT idea btw.

[Roger A. Modjeski]

Really appreciate the videos. My only request is to monitor the focus - the dissembling of the tube for instance is blurred, losing some of the value of the instruction, meanwhile the schematic drawing in the background is in perfect focus and even distracting. The closer you move your hands to the lens the blurrier the image. Though it may be counterintuitive we would be able to discern a lot more if you kept the movement further back or conversely refocus the camera to accommodate your hands being closer.

You can just tape down a small piece of the schematic for instance or some text, picture etc. as both a focus aid for the camera and marker for where to confine your actions. It's cool's that a tube is being taken apart, but frustrating to not really be able see the parts!

These videos are a GREAT idea btw.

Thanks for your input and tips. We shot it with a webcam and I do plan to re-do it when I get some free time.

[stevenkelby]

Agreed, great idea if a little blurry, love these vids and can't wait to see more!

[Ericus Rex]

And a non-magnetic pointer would be a 'Plus'     

Good description on mu v.s gain.  Reminds me of the 0 to 60 rating on sports cars:  pro driver, straight track, perfect road, no traffic, etc.

[Ericus Rex]

May I ask another dumb, newbie question?  Too bad!  I'm going to ask anyway!   

The signal applied to the grid is merely an AC voltage, right?  And the tube has a voltage flowing through it whether there is a signal or not, right?  Then how come we are not able to hear the idle current running through the tube?  To elaborate, if the applied grid signal merely modulates the idle current to resemble the grid signal waveform (but at larger voltage) why then is that same tube current (albeit DC) inaudible without the signal?

[Roger A. Modjeski]

May I ask another dumb, newbie question?  Too bad!  I'm going to ask anyway!   

The signal applied to the grid is merely an AC voltage, right?  And the tube has a voltage flowing through it whether there is a signal or not, right?  Then how come we are not able to hear the idle current running through the tube?  To elaborate, if the applied grid signal merely modulates the idle current to resemble the grid signal waveform (but at larger voltage) why then is that same tube current (albeit DC) inaudible without the signal?

Good question which you answered yourself.. almost. I like to be particular about the language of voltage and current as it helps to visualize and communicate. Current flows through something and voltage is across something. So we talk about the voltage across a battery or the voltage from plate to cathode across a tube. Current is flowing through the tube from the power supply, through the plate and cathode resistors to form a complete path back to the power supply. Currents must have a complete path, a circle it you like.

The current flowing through the tube at idle is indeed DC which the grid modulates or causes to increase and decrease with the instantaneous signal voltage. This current increases as the grid goes positive and decreases as the grid goes negative. This amplified signal is then capacitor coupled (unless some other provision is made to bring it to zero volts DC, such as a servo) to the outside world or the next stage.

Since DC is constant it does not pass through the capacitor. However, any noise on the power supply will appear with the signal and be passed through the capacitor to the outside world. Thus, power supplies have to be very low noise.

Putting some numbers to things, the plate typically runs at 100-300 volts and the signal at the plate of the output of a preamp is typically one volt. Therefore the signal is a very small percentage of the plate voltage and plate current. This results in low distortion at small signals however when the plate swing gets to be a large percentage of the idle current the distortion rises. For a 6922 I have found that the distortion (mainly second harmonic) is about 0.1% per volt of output with 100 volts on the plate. At 5 volts it would 0.5% and so on with a maximum of about 30 volts output before clipping. Feedback will reduce these numbers, however most 6922 output tubes are low enough distortion to not need feedback.

[Ericus Rex]

Duh.  Coupling Cap.  I don't know why I didn't think of that before asking.  Sorry to have wasted your time on that dud, Roger.  Your clarification about current and voltage is brilliant!  Thank you for that one.

[mboxler]

Roger...

Thanks for the videos.  I started a thread "How Do Vacuum Tubes Work?" a month ago to try to gain a better understanding of the circuit.  Wish I had waited!

A few other questions...

1)  Is this a good illustration of the relationship between the input voltage, the grid voltage, and the voltage output the plate?  I think the 150 should really be +150.



2)  The coupling cap can be very small because the load after it is very high?  I'm used to crossovers, in which case the driver is the load, and an 8 ohm load with a 1uf cap would only send frequencies above ~ 13k Hz to the driver.  With the load on your drawing, everything above 4 or 5 hz will pass.

3)  The voltage drop measured from the "bottom" of the cathode resistor to the "top" of the cathode resistor equals the grid bias?

4)  What, if anything, happens during "tube break in"?  Seems like the only thing would be the getter finishing the gas removal and from that point forward maintaining the vacuum.

Thanks again for the "class".

Mike

[kc8apf]

Thank you for taking the time to do these videos.  Your explanation is the clearest I've heard of how tubes work and how that behavior can be applied in an amplifier.  I look forward to the rest of the series.

[Ericus Rex]

Hi Roger,

I've seen a number of newer circuit designs that use LEDs in place of cathode bias resistors.  What are your thoughts on this practice?

I look forward to video #3 with great anticipation!

[Roger A. Modjeski]

Hi Roger,

I've seen a number of newer circuit designs that use LEDs in place of cathode bias resistors.  What are your thoughts on this practice?

I look forward to video #3 with great anticipation!

Not a good idea. The cathode resistor allows the tube to "find" the proper bias over a wide range of tube parameter variations that are always present. The LED provides a fixed voltage (1.8 volts typical) much like a battery. The use of the LED will cause larger variations in plate voltage than a resistor would for the range and aging of tubes. Given the range of various 12AX7 tubes these days its a really bad idea. In some cases even cathode bias is cannot deal with them. For instance a Sovtek 12AX7 wxt is so far off standard bias as not to work in an RM-10 which is very tolerant of tube variations.

The other problem with a LED is that they pretty much come in distinct voltages, the more common being 1.8 volts. (the whites and some colors are higher). If the LED is not bypassed with a large capacitor its non-linear diode characteristic will be added in the signal path. 

[Roger A. Modjeski]

Roger...

Thanks for the videos.  I started a thread "How Do Vacuum Tubes Work?" a month ago to try to gain a better understanding of the circuit.  Wish I had waited!

A few other questions...

1)  Is this a good illustration of the relationship between the input voltage, the grid voltage, and the voltage output the plate?  I think the 150 should really be +150.



2)  The coupling cap can be very small because the load after it is very high?  I'm used to crossovers, in which case the driver is the load, and an 8 ohm load with a 1uf cap would only send frequencies above ~ 13k Hz to the driver.  With the load on your drawing, everything above 4 or 5 hz will pass.

3)  The voltage drop measured from the "bottom" of the cathode resistor to the "top" of the cathode resistor equals the grid bias?

4)  What, if anything, happens during "tube break in"?  Seems like the only thing would be the getter finishing the gas removal and from that point forward maintaining the vacuum.

Thanks again for the "class".

Mike

Hi Mike,

Good questions, here are answers.

1. Yes that  and all the DC voltages are positive. However on the other side of the coupling capacitor the sine wave is centered around zero and that voltage will have equal positive and negative values. Non-sine audio will be averaged around zero having equal area over time positive and negative.

2. Yes the cap is small because the load is high, typically 100K to 1 Meg ohm to the next stage. However if this stage goes to the outside world the cap may be several microfarads as the outside load may be 10 K ohms or less. Several preamps that John Atkinson has tested have failed in this rather badly and have no bass into some typical modern power amp inputs.

3. Yes the voltage drop across the cathode resistor becomes the grid bias as the grid is held to zero by its resistor to ground and the cathode becomes positive making the difference negative as the grid sees it looking at the cathode. Its a very clever thing.

4. Break-in of audio gear has become the biggest crazy thing I have seen in recent years. The reviewers talk about it, manufactures now rate it in their specs..(I dont) and other than a few specific things like speaker suspensions I can give little scientific support to the concept. As to tubes, there is an aging process that goes on after the tube leaves the factory. The cathode emission will increase over time and this depends on how well the tube was aged at the factory. Given that electricity has gotten rather expensive the factories are aging less than in the good old days. Some batches of tubes I get are more aged than others, it's inconsistent. Power tubes show the most change over time going up as much as 20% in cathode current from the time I tube an amp till it burns in here for 24 hours. It may go up another 10% over the next 100 hours and then stay there for the next 1000 hours. Eventually the current goes down as the cathode ages further and I get about 10,000 hours out of my power tubes through careful choice of operating parameters.

I offer another aspect to the break-in phenomenon. What is the break in time for the listener to get used to the new piece of equipment in his system?

As to the gettering, it doesn't influence the sound in any way and just keeps doing its job trapping off-gassing till it is expended. Then the tube does become gassy and is now susceptible to run-away current, red plates and melt down.

[rbwalt]

the break in of audio gear is interesting. try this. when you get a new piece of gear ( amp, cdp, pre or other). plug it in and listen for a few hrs. do not listen to it again for another 40 to 50 hrs. during this time just have the system up and running in order to break in the new piece. go back and then listen.if true then there should be no change in how things sound from the first  few hrs or so. some gear may need it some may not or to put it another way it is not going to get any better no matter what.