Hi guys, I read through the links on the sticky regarding the intro to tubes and tube amplification. It was very enlightening however I have a few questions that I can't quite seem to answer with the resources provided.
Firstly, I've been looking at designs using an ecl82 with the triode as the driver and the pentode as the output tube allowing me to do a low cost 1 tube per channel amp. The Data Sheet for this tube shows the anode current as being 3.5ma and the anode voltage as being 100v. In the operating characteristics section of the manual, it lists the B+ for 200v and 170v and the anode current at .52 ma and .43 ma respectively. It suggests a anode resistor of 220k. At the anode currents listed in this section, the 220k resistor drops the voltage close to the reccomended plate voltage from the beginning of the manual. Why is there a discrepancy between the anode current at the beginning of the manual and the current listed in the operating characteristics section? Is the current at the beginning of the manual the peak output current and the operating characteristics section the quiescient (sp?) current?
Also, in many designs I've seen, there is a resistor on B+ in series with a capacitor with the plate resistor paralleled in between. Here is an example ECL82 design . This seems like it would increase the voltage drop. Is this just an extra stage of filtering?
Firstly, I've been looking at designs using an ecl82 with the triode as the driver and the pentode as the output tube allowing me to do a low cost 1 tube per channel amp. The Data Sheet for this tube shows the anode current as being 3.5ma and the anode voltage as being 100v. In the operating characteristics section of the manual, it lists the B+ for 200v and 170v and the anode current at .52 ma and .43 ma respectively. It suggests a anode resistor of 220k. At the anode currents listed in this section, the 220k resistor drops the voltage close to the reccomended plate voltage from the beginning of the manual. Why is there a discrepancy between the anode current at the beginning of the manual and the current listed in the operating characteristics section? Is the current at the beginning of the manual the peak output current and the operating characteristics section the quiescient (sp?) current?
Also, in many designs I've seen, there is a resistor on B+ in series with a capacitor with the plate resistor paralleled in between. Here is an example ECL82 design . This seems like it would increase the voltage drop. Is this just an extra stage of filtering?
Hi
The first operating conditions (with 3.5mA current) is just typical characteristics of the tube, it describes a set of values for the tube that is valid as a whole but it doesn't always describe the best operating conditions.
The other set of values is described as operating charactersitics and is a recomended set of values for using the tube as AF amplifier
There is no conflict between the 2 set of values but it indicates that it is more normal perhaps to use the tube according to the operating characteristics. The typical characteristics an be used to see if the tube is OK or not a healthy tube should measure close to these values.
You are correct in your assumption that the resistor capacitor combination is for filtering, it is very common to see this arrangement
Regards Hans
The first operating conditions (with 3.5mA current) is just typical characteristics of the tube, it describes a set of values for the tube that is valid as a whole but it doesn't always describe the best operating conditions.
The other set of values is described as operating charactersitics and is a recomended set of values for using the tube as AF amplifier
There is no conflict between the 2 set of values but it indicates that it is more normal perhaps to use the tube according to the operating characteristics. The typical characteristics an be used to see if the tube is OK or not a healthy tube should measure close to these values.
You are correct in your assumption that the resistor capacitor combination is for filtering, it is very common to see this arrangement
Regards Hans
You will notice that at the operating point you describe (ie 100V 3.5mA current), the bias point is 0V. This is not an operating point you will use in audio because it will introduce grid current when the grid swings positive.
In audio you will use a bias point of -1V at an anode voltage of 100V, for a anode current of 1mA.
Hope that helps.
Shoog
In audio you will use a bias point of -1V at an anode voltage of 100V, for a anode current of 1mA.
Hope that helps.
Shoog
Shoog said:
I noticed that the bias point of the tube, along with some other triodes was 0v, and when I calculated the voltage drop at that current based on the cathode resistor it was about 1.17v, setting the grid bias point at about -1.17v. Is this general practice for a tube that specs bias voltage at 0?
Just a few more annoying questions:
Obviously, the voltage drop across the plate resistor is dependant on the current through the tube. So when the tube manual reccomends 200v in the operating charactersitics section of the manual, is that factoring in the voltage drop across the plate resistor at or around the bias current? IE, is the tube manual reccomended resistor size based on consideration of bias current and an ultimately different desired voltage at the plate?
Along the same lines, if I wanted to use a higer impedance output transformer, could I simply increase the B+ to offset the higher voltage drop across the transformer so in the end the plate was at the same voltage at the bias point as it would be with the reccomended lower impedance transformer?
Okay, finally started drawing my own load lines on graphs from the tube manuals and things started to click! Answered many of the above questions. I was really misinterperting some things!
From what I have learned, it seems that as long as I adjust my variables so that the load line is at no point crossing (tangential seems okay) the max dissapation curve and that the load line is suitably long, I can use different loads for output transformers or plate caps.
I noticed that after doing my dissipation curve and load line on the output tube,the quiescent point is tangential or close to being tangential with the power dissapation curve. On the driver tube the quiescent point seems much much lower than the corresponding X value of the power dissapation curve. Is this simply to lower distortion? What is a good guideline for output power at the quescient point?
Thanks for putting up with my learning process. The community this forum represents is made up of some terrific people!
From what I have learned, it seems that as long as I adjust my variables so that the load line is at no point crossing (tangential seems okay) the max dissapation curve and that the load line is suitably long, I can use different loads for output transformers or plate caps.
I noticed that after doing my dissipation curve and load line on the output tube,the quiescent point is tangential or close to being tangential with the power dissapation curve. On the driver tube the quiescent point seems much much lower than the corresponding X value of the power dissapation curve. Is this simply to lower distortion? What is a good guideline for output power at the quescient point?
Thanks for putting up with my learning process. The community this forum represents is made up of some terrific people!
In order to get a bias point of 0V there is no cathode resistor, otherwise you couldn't have a bias point of 0V. If you look at the datasheet it makes no mention of a specified cathode resistor at 0V bias. This is a special case with valves and I am not certain what zero bias point valves are generally used for.
Shoog
Shoog
Plate curves are generally kinked toward the bottom of the chart. Crossing this region could cause non-linearity. High currents can lift your load line out of this region. High voltages will do this too, as well as to allow greater swings.eforer said:
I think that the manufacturer recommended operating characteristics should be studied, even if you don't take their advice.
I am currently working on using the ECL82 as an interstage transformer driver. The Pentode section is good for this. The triode section isn't so good because less than one volt of input sensitivity is a pain when most sources have outputs over 1V. Also it has far to much gain for the average situation. Finally 1mA of current limits the application of feedback.
If only there was a better paired valve than the ECL82 (the ECL86 suffers from the same triode limitations).
Shoog
If only there was a better paired valve than the ECL82 (the ECL86 suffers from the same triode limitations).
Shoog
I've drawn some load lines trying my best to interpert the spec sheet's data. I am posting a picturer of my load lines along with the data to see which (if any interpertation is correct). Any thoughts/comments/corrections are as always apreciated.
The first curve (#1) is based on the Typical operating characteristics of AF amplifier using the triode section of an ECL82 as stated in the RCA tube manual.
B+ 200v
Ra 220k
Rk 2.2k
Ia .52ma
I plotted my first point on the graph at 200v, 0ma. My thinking behind this is that at 200v plate current, the plate resistor would have dropped no voltage and thus plate current must equal zero.
I plotted my second point at 0v, .9ma as at zero volts the plate resistor must drop all 200v of B+, making the plate current .9.
I then took the Ia value (not knowing what voltage it corresponds to) and multiplied it by the cathode resistor. It yeilded a voltage of 1.144 which looked like a pretty good bias point to me.
The second line I based on the same set of figures, but I plotted the first point at 200v,.52ma. I then calculated the slope of the load line based on the 220k resistor (100/220k gives a value of 45.5ma current drop per every 100v increase). I then drew a line through the point at this slope.
The third line is using the typical characteristics data found at the beginning of the maual.
100v plate voltage
0v grid voltage
3.5ma
I plotted a point at the above listed values: 100v, 3.5 ma. This pointed intersected the 0v grid voltage curve as advertised. I used the same plate resistor and thus slope on this graph.
The first curve (#1) is based on the Typical operating characteristics of AF amplifier using the triode section of an ECL82 as stated in the RCA tube manual.
B+ 200v
Ra 220k
Rk 2.2k
Ia .52ma
I plotted my first point on the graph at 200v, 0ma. My thinking behind this is that at 200v plate current, the plate resistor would have dropped no voltage and thus plate current must equal zero.
I plotted my second point at 0v, .9ma as at zero volts the plate resistor must drop all 200v of B+, making the plate current .9.
I then took the Ia value (not knowing what voltage it corresponds to) and multiplied it by the cathode resistor. It yeilded a voltage of 1.144 which looked like a pretty good bias point to me.
The second line I based on the same set of figures, but I plotted the first point at 200v,.52ma. I then calculated the slope of the load line based on the 220k resistor (100/220k gives a value of 45.5ma current drop per every 100v increase). I then drew a line through the point at this slope.
The third line is using the typical characteristics data found at the beginning of the maual.
100v plate voltage
0v grid voltage
3.5ma
I plotted a point at the above listed values: 100v, 3.5 ma. This pointed intersected the 0v grid voltage curve as advertised. I used the same plate resistor and thus slope on this graph.
An externally hosted image should be here but it was not working when we last tested it.
For a 220k resistor and a 200V supply, there is only one load line. Any operating point must be a point on that line. To set the loadline correctly, put a point on Vp at a zero current condition (i.e., 200V). Zero current corresponds to the x-axis.
Now assume the tube swings down as far as it could even in theory- 0V. The resulting "short circuit current" is 200V/220k, or roughly 0.91mA. And that's with Vp at zero V, i.e., the y-axis. Draw a line between these points. That is your one and true load line for a resistive load.
Now assume the tube swings down as far as it could even in theory- 0V. The resulting "short circuit current" is 200V/220k, or roughly 0.91mA. And that's with Vp at zero V, i.e., the y-axis. Draw a line between these points. That is your one and true load line for a resistive load.
My first load line represents those figures. My question stems more from having a hard time understanding how the tube manual presents the data. For example on the boozhoundlabs how to section, he uses a 6v6gt. The specs for this tube seems to provide the bias point as opposed to the set of figures I obtained from the RCA manual regarding the ecl82's triode section. It seems hard to distinguish weather the manual is giving you a B+ value from the power supply with zero volts dropped across the plate load, or the plate voltage at the bias point after having dropped volts across the plate load.
It seems that there are a lot of different ways to present the specs of a tube and the reccomended usage of them.
It seems that there are a lot of different ways to present the specs of a tube and the reccomended usage of them.
eforer,
The tube data, in genaral, refers to just the tube itself. So when you see "plate voltage", that refers to the voltage measured from the plate to the cathode. Stated grid voltage, will be the voltage mesured from cathode to grid. The manufacturer has no clue what kind of circuit you will use the tube in, so they derive and refer the characteristics strictly to the tube itself. Just because these curves refer ONLY to tube does not preclude you from using load line techniques in most circumstances.
Now, there are some specialized charts that are exceptions to this, but they will be clearly labeled on the chart or in its title. For example you may find distrotion versus power charts or gain versus plate load etc... In these cases, the type of circuit used will be described or drawn and component values will be identified. Component values, in particular, may be varied over a range to define different "series" or curves on the chart. You often find these types of charts for tubes which were designed exclusively for one purpose.
Google on 6550, a beam power pentode designed solely for audio power amplification, and you will find such charts. Why the charts? These were handy as hell before people had calculators and spreadsheets... not to mention full blown CAD and simulations.
Does this help?
And yes, your Load Line #1 is correct. Your bias point will be a dot somewhere along that line. Finding your AC load line is a different matter though.
The tube data, in genaral, refers to just the tube itself. So when you see "plate voltage", that refers to the voltage measured from the plate to the cathode. Stated grid voltage, will be the voltage mesured from cathode to grid. The manufacturer has no clue what kind of circuit you will use the tube in, so they derive and refer the characteristics strictly to the tube itself. Just because these curves refer ONLY to tube does not preclude you from using load line techniques in most circumstances.
Now, there are some specialized charts that are exceptions to this, but they will be clearly labeled on the chart or in its title. For example you may find distrotion versus power charts or gain versus plate load etc... In these cases, the type of circuit used will be described or drawn and component values will be identified. Component values, in particular, may be varied over a range to define different "series" or curves on the chart. You often find these types of charts for tubes which were designed exclusively for one purpose.
Google on 6550, a beam power pentode designed solely for audio power amplification, and you will find such charts. Why the charts? These were handy as hell before people had calculators and spreadsheets... not to mention full blown CAD and simulations.
Does this help?
And yes, your Load Line #1 is correct. Your bias point will be a dot somewhere along that line. Finding your AC load line is a different matter though.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.