The only way to get better at LTSpice is keep practicing and probing; these models will never get built. Sometimes, things don't work out, as in this case: I am getting 9μV output, clearly something isn't right. 😆
My brain is soggy due to being unwell, maybe missed something, can someone have a quick look over the schematics and see if missed something, please?
Here is the original schematic: http://www.r-type.org/articles/art-003f.htm.
Thanks in advance! 😀
My brain is soggy due to being unwell, maybe missed something, can someone have a quick look over the schematics and see if missed something, please?
Here is the original schematic: http://www.r-type.org/articles/art-003f.htm.
Thanks in advance! 😀
I think the main problem is the transformer model. The wrong side of R26 is grounded too. The 8 pins are called TTG-EL84PP, P1, Sg1, B, O16, O8, O4, and Com. The inductor between pins #5 and pin #6 called "O8" is zero Henries. The lower side of the primary is missing and/or mixed up with the secondary.
The schematic seems to want a 7 pin transformer that has only one secondary. Probably "TTG-EL84PP" is not supposed to be a pin and there should be pins "P2" and "Sg2"??? Perhaps O16 and O4 should be unused pins, ie 9 pins total and there should be 7 inductors and not 5. I could maybe rewrite the transformer sub, but I've done enough.
The schematic seems to want a 7 pin transformer that has only one secondary. Probably "TTG-EL84PP" is not supposed to be a pin and there should be pins "P2" and "Sg2"??? Perhaps O16 and O4 should be unused pins, ie 9 pins total and there should be 7 inductors and not 5. I could maybe rewrite the transformer sub, but I've done enough.
Thank you so very much for the help, very appreciated! 😀
Darn, is there a current production OPT that satisfies this requirement? I am not understanding what it is, how differs from a PP OPT with an UL tap.
To confirm, "P" is anode and "Sg" is signal, correct?
If so, how does "P" differ from HT since both the same voltage?
Last edited:
Of course. Thank you. 🙂
So what if no screen grid?
How is it coming looking at the layout? I did as required and labeled the components. 🙂 Worse case you are too busy and have to start a new thread so can have help from someone else (title doesn't match the topic). Want to get the parts ordered, some take a couple few months to ship.
Adriel,
Primary connections:
P normally is the Plate
S normally is the Screen Tap (when there is one)
HT probably means B+ (I guess that is the HT you are talking about).
Let's consider an example of an Ultra Linear Primary (for either a push pull, or a single ended primary, we will look at it from B+ tap to UL tap, to Plate tap.
We have identical mirror of that on a push pull primary.
The exact numbers are not what I am talking about, instead I am giving the principles.
Suppose the B+ is 300V.
If the UL tap is at 40%; and the DCR from B+ to plate tap is 100 Ohms, then from B+ to UL tap is about 40 Ohms DCR, and from UL tap to plate tap is 60 Ohms.
Suppose the plate has 35mA quiescent current, and the screen has 3mA current.
From B+ to UL tap (40 Ohms) we have 35 + 3mA = 38mA.
38mA x 40 Ohms = 1.52V drop from 300V = 298.4V at the screen.
The DCR from UL tap to plate tap is about 60 Ohms.
60 Ohms x 35mA plate current = 2.1V drop from 298.4V = 296.3V at the plate.
If the primary DCR from B+ tap to plate tap is 500 Ohms, then the voltage drops from B+ will be 5 x larger than 1.52V and 2.1V.
Primary connections:
P normally is the Plate
S normally is the Screen Tap (when there is one)
HT probably means B+ (I guess that is the HT you are talking about).
Let's consider an example of an Ultra Linear Primary (for either a push pull, or a single ended primary, we will look at it from B+ tap to UL tap, to Plate tap.
We have identical mirror of that on a push pull primary.
The exact numbers are not what I am talking about, instead I am giving the principles.
Suppose the B+ is 300V.
If the UL tap is at 40%; and the DCR from B+ to plate tap is 100 Ohms, then from B+ to UL tap is about 40 Ohms DCR, and from UL tap to plate tap is 60 Ohms.
Suppose the plate has 35mA quiescent current, and the screen has 3mA current.
From B+ to UL tap (40 Ohms) we have 35 + 3mA = 38mA.
38mA x 40 Ohms = 1.52V drop from 300V = 298.4V at the screen.
The DCR from UL tap to plate tap is about 60 Ohms.
60 Ohms x 35mA plate current = 2.1V drop from 298.4V = 296.3V at the plate.
If the primary DCR from B+ tap to plate tap is 500 Ohms, then the voltage drops from B+ will be 5 x larger than 1.52V and 2.1V.
Understood.
B means battery connection where HT means high tension, meaning high voltage. I still have a little Mid Atlantic from maternal maternal side of the family muddled with Midwest and Southern. Ah America, the melting pot.
I understand this. Also, it is most often class AB, as one valve powers down as one rams up, versus class A where the valve is running constantly. My understanding is this is why PP does better with pop and rock and roll; not sure if can do EDM. 😛
How does this solve the voltage issue and get us closer to finding an OPT?
The transformer model actually saz:
which explains why the 2nd plate circuit is missing.
It would be better to just use ~5 inductors linked by the "K1 L1 L2 L3 L4 L5 0.999" statement, so no (other) transformer model or symbol was needed.
The link you provided to the original schematic provides some details about the transformer. It saz the primary is 9K and 90 to 120 Henries, with taps at 20%. 8K at 16Hz is 80 Henries? I assume that is per plate because other sources say a EL34 plate impedance is 10K. But you show two 6BQ6 in the simulation and the schematic shows a ECL86???? So maybe the transformer needs to be redesigned for whatever you intend to build. If you set Lp=90, tap=20/100, then L4=L1=90*(1-tap)^2, L2=L3=90*(tap)^2, L5=90(8/9k).
Perhaps a transformer expert can help you better than I??
- Single Ended transformer, with Ultralinear taps at 43%
- 8000 to 8 ohms, 3db 16 to 91000 hz
- speaker taps at 8 and 4 ohms
which explains why the 2nd plate circuit is missing.
It would be better to just use ~5 inductors linked by the "K1 L1 L2 L3 L4 L5 0.999" statement, so no (other) transformer model or symbol was needed.
The link you provided to the original schematic provides some details about the transformer. It saz the primary is 9K and 90 to 120 Henries, with taps at 20%. 8K at 16Hz is 80 Henries? I assume that is per plate because other sources say a EL34 plate impedance is 10K. But you show two 6BQ6 in the simulation and the schematic shows a ECL86???? So maybe the transformer needs to be redesigned for whatever you intend to build. If you set Lp=90, tap=20/100, then L4=L1=90*(1-tap)^2, L2=L3=90*(tap)^2, L5=90(8/9k).
Perhaps a transformer expert can help you better than I??
Attachments
Last edited:
I am confused . . .
1. Which output circuit will be used?
Pentode / Beam Power mode? Then . . . No UL tap required.
Ultra Linear mode? A UL tap is required (unless you are willing to use a lot more circuitry and do more engineering too).
Using a 26%, 40%, etc. UL tap is relatively simple if you have the UL tap.
Using an active stage to drive the screen to make the output mode be UL is more complex.
2. Which output tubes will be used?
I see mention of 6BQ6, EL34, EL84, and ECL86.
1. Which output circuit will be used?
Pentode / Beam Power mode? Then . . . No UL tap required.
Ultra Linear mode? A UL tap is required (unless you are willing to use a lot more circuitry and do more engineering too).
Using a 26%, 40%, etc. UL tap is relatively simple if you have the UL tap.
Using an active stage to drive the screen to make the output mode be UL is more complex.
2. Which output tubes will be used?
I see mention of 6BQ6, EL34, EL84, and ECL86.
Last edited:
Hi Adriel,
There were some more mistakes in your LTspice drawing: missing capacitor and grid connected to ground for the lower output tube. The drawing was also very cramped and cluttered. I had to redraw it (using "drag" ) to spot the mistakes you made.
I un-commented some include files you don't use, removed the torroidy, and took a different EF86 model I had to hand.
The zip contains all the files to run the simulation.
There were some more mistakes in your LTspice drawing: missing capacitor and grid connected to ground for the lower output tube. The drawing was also very cramped and cluttered. I had to redraw it (using "drag" ) to spot the mistakes you made.
I un-commented some include files you don't use, removed the torroidy, and took a different EF86 model I had to hand.
The zip contains all the files to run the simulation.
I missed that, added it, though the only difference is now no output calculation error, the gain is still basically zero.
Sorry to confuse, that was from a previous iteration (modified an existing PP model). Turned it off to see if make any difference and did not.
I have tried and failed at this. Additionally, does not allow trying different OPTs. I am willing to try again, of course. I have some challenging radio inductors to sort out and be nifty to get them working.
I had not read the text until mentioned it, appreciate this and the other help. 🙂
I am not planning to build it, just using it to practice modeling.
What is the formula providing, please?
What the original design is, which I understand not to be UL.
I have yet to quite grasped the relationship between these and how the UL differs from the other.
The ECL86 is a ECC83 combined with a pentode similar to the EL84, though not sure how differs.
Last edited:
Adriel,
Pentode / Beam Power mode: The screen, g2 is tied to a fixed positive voltage. Screen voltage does not vary, but the plate voltage varies according to the music signal that is applied to the control grid, g1.
Ultra Linear mode: Pentode tube / Beam Power tube:
The screen voltage varies, as a Percentage of the Plate voltage swing, according to the music signal (26%, 43%, 50%, etc.).
If the UL tap is 50%, then if the plate swings down by 100V, the screen swings down by 50V; if the plate swings up by 100V, the screen swings up by 50V.
An EL84 Pentode tube has the following:
Filament, Cathode, Control Grid g1, Screen Grid g2, Suppressor Grid g3, Plate. (There is no Beam Former).
Both a 6BQ6 and a 6BQ5 Beam Power tubes have the following:
Filament, Cathode, Control Grid g1, Screen Grid g2, Beam Formers, Plate. (There is no Suppressor Grid).
The EL84 Pentode and 6BQ5 Beam Power tubes have very similar specifications. You often can find a Dual-Marked EL84 / 6BQ5.
A Dual-Marked EL84 / 6BQ5 might have a Suppressor Grid, or it might have Beam Formers.
You might not get exactly what you want if you have a Dual-Marked tube.
Your Tube May Vary
A similar situation exists between the EL34 Pentode tube and the KT77 Beam Power tube.
They have very similar specifications to each other.
But fortunately I have never seen a Dual-Marked EL34 / KT77 (I hope there are none).
When I want a Beam Power tube, I want a Beam Power tube. When I want a Pentode, I want a Pentode. No difference to me, I just like to know which I am listening to.
For the range of output tubes that I prefer, I get more design choices with Beam Power tubes, than I do with Pentode tubes.
Beam Power Tubes I use: 7591, 5881, 6L6GB, 6L6GC, KT66 (I have also tried KT77 and KT88, but am not using any now). I may try the 6V6.
Pentode Tubes: EL34 and EL84 (I am not using any now).
Your preferences may, and should vary.
Pentode / Beam Power mode: The screen, g2 is tied to a fixed positive voltage. Screen voltage does not vary, but the plate voltage varies according to the music signal that is applied to the control grid, g1.
Ultra Linear mode: Pentode tube / Beam Power tube:
The screen voltage varies, as a Percentage of the Plate voltage swing, according to the music signal (26%, 43%, 50%, etc.).
If the UL tap is 50%, then if the plate swings down by 100V, the screen swings down by 50V; if the plate swings up by 100V, the screen swings up by 50V.
An EL84 Pentode tube has the following:
Filament, Cathode, Control Grid g1, Screen Grid g2, Suppressor Grid g3, Plate. (There is no Beam Former).
Both a 6BQ6 and a 6BQ5 Beam Power tubes have the following:
Filament, Cathode, Control Grid g1, Screen Grid g2, Beam Formers, Plate. (There is no Suppressor Grid).
The EL84 Pentode and 6BQ5 Beam Power tubes have very similar specifications. You often can find a Dual-Marked EL84 / 6BQ5.
A Dual-Marked EL84 / 6BQ5 might have a Suppressor Grid, or it might have Beam Formers.
You might not get exactly what you want if you have a Dual-Marked tube.
Your Tube May Vary
A similar situation exists between the EL34 Pentode tube and the KT77 Beam Power tube.
They have very similar specifications to each other.
But fortunately I have never seen a Dual-Marked EL34 / KT77 (I hope there are none).
When I want a Beam Power tube, I want a Beam Power tube. When I want a Pentode, I want a Pentode. No difference to me, I just like to know which I am listening to.
For the range of output tubes that I prefer, I get more design choices with Beam Power tubes, than I do with Pentode tubes.
Beam Power Tubes I use: 7591, 5881, 6L6GB, 6L6GC, KT66 (I have also tried KT77 and KT88, but am not using any now). I may try the 6V6.
Pentode Tubes: EL34 and EL84 (I am not using any now).
Your preferences may, and should vary.
Last edited:
@6A3sUMMER , wow, that is a lot of information! Thank you! 😀
Any advantage of Pentode/Beam over UL?
I used to prefer European valves over American, so wasn't planning on getting any 6BQ5. However, fantastic knowing they are a different kind of tube and to watch out.
The EL84s I have are solely marked.
Any advantage of Pentode/Beam over UL?
I used to prefer European valves over American, so wasn't planning on getting any 6BQ5. However, fantastic knowing they are a different kind of tube and to watch out.
The EL84s I have are solely marked.
Adriel,
Generalizations:
Pentode / Beam Power tubes in Pentode / Beam Power mode:
The most output power and the most output stage voltage gain; but the most distortion, and the lowest damping factor (before applying negative feedback around the output stage).
So negative feedback around the output stage is most often used.
Could be global negative feedback, Schade negative feedback, cathode negative feedback, etc.
Could be tricky to make it stable, depending on the output transformer, complete circuit topology, and type of negative feedback used, etc.
Pentode / Beam Power tubes in Ultra Linear mode:
A little lower output power and lower output stage voltage gain; but medium distortion, and medium damping factor (before applying negative feedback around the output stage).
(Reminder, the UL tap to the screen is negative feedback; but other negative feedback can be used at the same time as UL).
So negative feedback around the output stage is often used, but performance may be acceptable without it.
Could be global negative feedback, Schade negative feedback, cathode negative feedback, etc.
Could be tricky to make it stable, depending on the output transformer, complete circuit topology, and type of negative feedback used, etc.
A little less tricky to make stable.
Can be fairly simple.
Pentode / Beam Power tubes in Triode Wired mode:
The least output power and the least output stage voltage gain; but the lowest distortion, and the highest damping factor (before even applying negative feedback around the output stage).
So, negative feedback around the output stage is least often used.
For those who insist on using negative feedback for Triode Wired mode: It could be global negative feedback, Schade negative feedback, cathode negative feedback, etc.
Without negative feedback, it is intrinsically stable.
Generally stable, depending on the output transformer, circuit topology, type of negative feedback used, etc., and if negative feedback is used at all,
the amount of negative feedback is usually less.
Simplest without any negative feedback.
Generalizations:
Pentode / Beam Power tubes in Pentode / Beam Power mode:
The most output power and the most output stage voltage gain; but the most distortion, and the lowest damping factor (before applying negative feedback around the output stage).
So negative feedback around the output stage is most often used.
Could be global negative feedback, Schade negative feedback, cathode negative feedback, etc.
Could be tricky to make it stable, depending on the output transformer, complete circuit topology, and type of negative feedback used, etc.
Pentode / Beam Power tubes in Ultra Linear mode:
A little lower output power and lower output stage voltage gain; but medium distortion, and medium damping factor (before applying negative feedback around the output stage).
(Reminder, the UL tap to the screen is negative feedback; but other negative feedback can be used at the same time as UL).
So negative feedback around the output stage is often used, but performance may be acceptable without it.
Could be global negative feedback, Schade negative feedback, cathode negative feedback, etc.
Could be tricky to make it stable, depending on the output transformer, complete circuit topology, and type of negative feedback used, etc.
A little less tricky to make stable.
Can be fairly simple.
Pentode / Beam Power tubes in Triode Wired mode:
The least output power and the least output stage voltage gain; but the lowest distortion, and the highest damping factor (before even applying negative feedback around the output stage).
So, negative feedback around the output stage is least often used.
For those who insist on using negative feedback for Triode Wired mode: It could be global negative feedback, Schade negative feedback, cathode negative feedback, etc.
Without negative feedback, it is intrinsically stable.
Generally stable, depending on the output transformer, circuit topology, type of negative feedback used, etc., and if negative feedback is used at all,
the amount of negative feedback is usually less.
Simplest without any negative feedback.
Understood. Thank you again for so much information! 😀
You happen to know of a way to calculate damping factor in LTSpice?
As I understand, fast paced music, such as rock and roll, likes more damping. However, know the best is to have SS handling the bass as SS has way more damping then valves.
Adreil,
I do not use LT Spice or any other simulation software.
I calculate things by longhand; and / or I do a very rough estimate, and then when I design and build the amplifier, I do a test to determine the damping factor.
I have only designed and built a couple of Pentode mode / Beam Power mode output stages; one was SE, one was PP.
I do not expect to use those modes again.
I tried the push pull amp with no feedback from the output transformer primary, and no feedback from the secondary.
Then I put a "power robbing" resistor across the plate to plate ends of the primary, again with no negative feedback.
Both push pull versions had moderately low distortion with no feedback. But the damping factor was much lower than I wanted for my speakers.
(I just wanted to hear those amplifiers on various models of my speakers).
Over the decades, I have owned commercial push pull amplifiers that had pentode or beam power output stages.
They all sounded good to me . . . they did all have global negative feedback.
The damping factor in pentode / beam power modes, largely depends on the plate impedance, rp, the primary impedance, and the amount of negative feedback that is taken from either the primary or secondary windings.
Generally, the damping factor is very very low, unless negative feedback around the output state is used. And that global negative feedback is usually the main determinant of the damping factor.
I do not use LT Spice or any other simulation software.
I calculate things by longhand; and / or I do a very rough estimate, and then when I design and build the amplifier, I do a test to determine the damping factor.
I have only designed and built a couple of Pentode mode / Beam Power mode output stages; one was SE, one was PP.
I do not expect to use those modes again.
I tried the push pull amp with no feedback from the output transformer primary, and no feedback from the secondary.
Then I put a "power robbing" resistor across the plate to plate ends of the primary, again with no negative feedback.
Both push pull versions had moderately low distortion with no feedback. But the damping factor was much lower than I wanted for my speakers.
(I just wanted to hear those amplifiers on various models of my speakers).
Over the decades, I have owned commercial push pull amplifiers that had pentode or beam power output stages.
They all sounded good to me . . . they did all have global negative feedback.
The damping factor in pentode / beam power modes, largely depends on the plate impedance, rp, the primary impedance, and the amount of negative feedback that is taken from either the primary or secondary windings.
Generally, the damping factor is very very low, unless negative feedback around the output state is used. And that global negative feedback is usually the main determinant of the damping factor.