Loadline calculator for power stages with reactive load

This calculator allows you to compute and tune loadline and operating point of tube amplifiers’ power stages.

You can:

  • choose among most common power vacuum tubes (I am slowly adding more),
  • set the operation modality, among ultralinear, pentode, and triode (suppressor tied to cathode) mode,
  • set Single-Ended or Push-Pull configuration
  • set the Power Supply V+ voltage, the screen voltage, the screen tap, the quiescent bias current.

The calculator:

  • plots the anode characteristic graph, for the chosen operation modality,
  • plots the loadline,
  • highlights the operating point,
  • estimates the grid bias voltage,
  • estimates the expected output power at max displayed Grid Voltage, when max Grid Voltage is 0, when operating in Class A or A2, and at the selected output voltage headroom.
  • estimates the needed cathode resistor, if you want to use self-bias
Single Ended
Push-Pull

Did you buy my book? Why don’t you leave a review on amazon?

Select a Tube: Check full list of available tubes
Operating mode:
Ultralinear
Pentode
Triode
PP/SE:
PP SE
V+ (V): Grid Bias Voltage (V):
Quiescent Operating point: Iq(mA):
Vq(V):
Output Power (W):

Load (Ohm):
Screen Voltage (V): Cathode resistor (Ohm):
UltraLinear tap (%):
Out. headroom (+/-V):

If you are not familiar with loadlines, their purpose, and their use, I suggest you checking my guide: Section 3.2 (for resistive loadlines), Section 4.1.3 (for reactive loadlines), and Section 4.2.2 (for AC loadlines).

The underlying model and equations, for tracing tube characteristics curves, are those discussed by Norman Koren, in his web site.

214 thoughts on “Loadline calculator for power stages with reactive load

Add yours

    1. When you use two parallel tubes, each tube sees twice the load. So, when you use two parallel tubes in your project, you should simply use a load that is half the one that you test here (without parallel tubes). If you use three parallel tubes, each tube sees a three times larger load, so you have to use one third load.

    2. Hola me gustaría si por favor podés agregar a la lista la válvula 6dq6b /6gw6 ya que no está en la lista Muchas gracias

    1. The only value that is not referred to a single tube is the load, which in case of PP is from anode to anode. Clearly, even if all other values are per tube, they must be the same between the two tubes. And, yes, it is two tubes!
      Estimated max output power is at the speaker, with an ideal transformer.

  1. Thanks for your reply and clarification. I was wondering what arithmetic I can do on my end to extrapolate to a 4 tube push/pull amp from the 2 tube push/pull results given here?

    1. In a 4 tube push/pull the load seen by every tube is doubled, with respect to the transformer impedance. So, if the transformer is 4k, from anode to anode, with 4 tubes it is as if it were 8k. In other words, if you use 4 tubes and a 4k anode to anode transformer, 8k is the value that you have to specify as load in the calculator. Viceversa, if you want the 4 tubes to operate along a loadline corresponding to a 8k anode to anode load, you have to use a 4k anode to anode transformer.

  2. Very well done!
    I was actually making something similar myself. lol
    May I give some suggestions?

    What I always like to calculate, is what the ratio of Class-A vs Class-B power is compared to the total output power.
    This really helps with designing a certain output stage.
    In case of guitar amps, it really helps designing a certain tone of the output stage.

    Also I think it would be nice to give a notification if certain limits are being reached, like maximum power dissipation, V+ etc

    Would is be possible to add EL91 and EL95 tubes at a certain point?
    Or maybe that users can make their own list?

      1. Thank you so much for adding these models!
        If it’s not to much to ask for, is it possible to also add the ECL80?

        Maybe a feature request is to add the whole ECL8x family (ECL80, ECL82, ECL83, ECL84 and ECL85(or 805) )
        These are all great tubes to use.

        You’re work is really appreciated!

          1. Thank you so much, your work is really appreciated!
            Very interested because on paper the EL91 and the ECL80 should be very close to each other (I am talking about the power Pentode), but in these graphs they are pretty different?

          2. Did you check that you are using the same screen voltage in both cases? My impression is that using the same screen voltage (for instance 250 or 200) you get very similar graphs for the two tubes.
            I also double checked comparing my graphs with the datasheets (http://www.r-type.org/pdfs/ecl80.pdf and http://www.r-type.org/pdfs/el91.pdf) and my impression is that they overlap quite nicely.
            Can you be more specific on the differences you found?
            Thanks again for your feedbacks

          3. I double checked and you’re actually right, they match pretty nicely indeed!
            One small thing, but I think the Iq is the bias current per tube (and so the cathode resistor as well).
            This is a little confusing with a PP and I think people can easily take the wrong value maybe?

          4. That’s ok! Thanks.
            For what concerns Iq. All values are referred to a single tube. So Iq is the current in one tube, also in PP configuration. The only exception is the load, whcih in PP is referred to anode to anode load.

  3. It’s possible to have the GU81 as pentode (there is a graph of it as triode but i dont know really how close it is) or a way to plug in datasheet numbers into the formula to plot custom graphs? many thanks!

  4. Very nice calculator! This may be the single most useful thing I’ve ever found for amp design. I am very thankful. If you could, it would be very helpful to me if you can add the EF86 to this.

  5. It would really be interesting to see how a 6BQ6GTB behaves as a UL pp amp at B+ of 200V , since the DC voltage for G2 for the GTB is also 200V . Would you be so kind as to show us ?

  6. Oh ok , I missed the part that 6BQ6GTB in your pending list. But why do you show 6JQ6 limited to 140v on g2 when the maximum is +330 volts ? And could the 6JQ6 handle 350v peak on g2 is a UL configuration ?

    1. Hi,
      You can use any value for g2 in pentode mode. Just type a new value in the corresponding field.
      However, the value is locked in ultralinear mode and triode mode.
      In fact, in pentode mode, g2 is not fixed and it is automatically and dynamically computed according to the ultralinear tap percentage.
      In triode mode, g2 is equal to Va (equivalent to pentode mode with 100% ultralinear tap).

  7. Dear Guiseppe, I like this very well. I wonder if you could add some more triodes. I am using 45, 2A3, 101D, 205D types. Your 2A3 simulation is spot-on to a real amplifier I have built.

    kind regards,
    ciao, Adam

  8. Dear all,
    thanks for using this calculator and thanks for suggesting additional tubes.
    The pending list of tubes that will be added now includes:

    ECL84, ECL85
    kt120
    EL86, 6AH6, A2134, 12E1
    GM70, GU50
    6GU7 6CG7 12BH7
    45, 101D, 205D

    Hope to be able to add them all in a couple of weeks.

  9. Thank you again for adding all these wonderful new tubes!
    One small thing, the ECL85/805 (and PCL805) has a anode dissipation (Wa) of 8W instead of 7W

  10. Great tool !
    i’m looking for a tool that uses spice models and gives me Ia for a given tube, plate voltage, screen voltage, grid voltage… That would be usefull when testing tubes. I think what you did could do that easily ?

    1. My tool does exactly this … even if, it does not actually start from Spice tube models. However, my model and parameters can be used to build Spice models, provided grid current, screen current, and internal capacitance are modelled as well.

  11. would you be so kind as to add 6П3С to your list . Thanks , and thanks for adding 6JQ6 which in some ways seems to be better than the 6BK5 tubes.. As it seems the 6JQ6 can put out 9.89Wrms as a UL PP amp , which is better than the 6BK5 UL pp amp. That was quite surprising.

  12. I have a request for an oddball, the QQE03/12 (or CV2798)
    Originally this tube was created for transmitters, but I have actually seen some very interesting circuits for audio amplifiers.

    Maybe also ECL81 and ECL84 to make the list complete? 😉

    Thank you so much for the great work!!!!

  13. Added also these to my todo list. By the way, I considered ECL84 also before. However I did not find anode characteristics graph for the triode section. I just have it for the pentode section. Does anyone have it?

      1. Thank you!
        I am starting to feel a little guilty, but of all great output tubes, I would like to ask for the 5881 as well.
        Thanks for all the great work! It’s really appreciated!

  14. Thanks for all the updates so far!
    I have been really spending a lot of hours tinkering with some tubes now!
    I saw the very interesting 6N7 tubes (which can be used in a cute little amp), so I thought maybe it’s a nice idea to also add the 6SN7.
    I will be hunting down some more interesting high power double triodes 😉

  15. Hi Giuseppe, I just discovered your website and have been playing around with the load line calculator. This calculator is GREAT! Thank you for providing such a powerful tool and making plotting load lines so easy. I recently learned how to plot manually and your calculator helped me check my work. A great advantage of this calculator is being able to adjust the screen voltage, as we know the prints only provide Vg2=250v and 300v if you are lucky.

    I have a request: If possible, it seems advantageous to be able to print out calculated load lines. This could be used for build/repair documentation and comparison of various plots.

    Thanks for your hard work. I am looking forward reading your book that I will be purchasing soon!

    1. Hi,
      Many thanks for appreciating my work.
      I am glad discovering that it is becoming useful to many vacuum tube enthusiasts and professionals.
      Adding the possibility of printing the calculated graphs might be a nice feature. I’ll see how I can add it to the calculator.
      Thanks again 🙂

  16. I found another interesting one; EFL200/PFL200 (same as 6Y9, 17Y9 and Russian 6R4P)
    Double pentode with max 5W.

    Btw, would it be possible to show the power dissipation of the current bias setting?
    Of course this is easy to calculate by hand, but it helps if it’s shown right away.
    (ideally in combination with percentage/ratio compared to the total max power)

      1. Hi Giuseppe,

        Yes that’s what I mean.

        It would be nice to add the same power with input power as well, but I am not quite sure if you can calculate that from the parameters you have?

          1. Well, basically you’re just multiplying two field you already have 😉
            The thing is, normally you don’t want to go higher than about 70% of the total power dissipation of a tube. Maybe even more like 50-60%, depending on the situation.

  17. Hello Giuseppe, i’m looking to the KT88 load lines and i’m thinking if the tube max dissipation shouldn’t be 42W instead of 35W as is shown.

      1. Thank you so much!

        I think there is something strange with the 6P30b.
        Not all grid-lines show up (on the bottom), and according to the calculations, it would be able to output something like 8W.
        Which is way to much for those sub-miniature tubes.

  18. Un ottimo servizio, grazie!
    Volevo segnalarti che su Android appare il calcolo della Rk ma non l’opzione resistive/reactive… sul PC invece viceversa.
    Da cosa è dettata la scelta di mettere una Vg consigliata fissa?

  19. Hi Giuseppe,
    how about calculating THD (at least H2 and H3)? Is there a chance for such feature? You have been doing a great job so far.

  20. Dear firends,
    recentely I have been very busy so my list of tubes to add became very long. Sorry for taking so long to satisfy you requests.
    Currently my pending list contains:
    ECC808, 3A5, 4P1L, PX25, 300BXLS, T100, 5902, 6p30b, 6n16, 6n17, 6C41, 6U8A
    6528, 6sk7, 6ba6, 6j7, 6Y6, 811A, 2E22, 6dq6b, 6gw6, 6V6S

  21. Thank you for your efforts here. Very helpful calculator.
    Was hoping to get an idea on power output for these setups, but those fields are blank.
    Thanks again

  22. Dear friends,
    Sorry for not being reactive during last months.
    The lockdown, and the consequent “work from home”, absorbed me and I did not have enough time to dedicate to hobbies.
    When you are in office, the moment you lock the door and return home makes a clear distinction between your job and family/hobbies/entertainment time.
    Doing office work from home makes this distinction fuzzy and you end up dedicating less time to hobbies and entertainment.
    Hope next months to be able to dedicate more time to this

  23. Hey Giuseppe, i have a few questions, i tried some tests by loading in some known data from a Fender tube amp (Princeton AA1164) that runs x2 6V6’s with a plate voltage of 410 and a screen voltage of 400, a grid bias of -34 and an output transformer of 6600Ohms. the load lines seem to well rise over the power dissipation cruve and im wondering if i did something wrong. Is your calculator right and showing that this fender amp runs very hot and this is normal or did i misuse your calculator? im struggling with trying to understand what is happening.

    1. Hi,
      In fact the loadline stays well above the max dissipation curve. However, consider that with these parameters the tube works in class AB (almost class B, since quiescent point is close to cut-off). This means that it stays above max dissipation for half cycle, and conducts much less in the other half cycle. So it might be fine.
      Is the standard output transformer 6.6k CT for your project? Sometime I saw using 8k CT transformers.

      1. its very rare for guitar Push Pull amp transformers to be 8k. My particular project is using an identical fender amp 6.6k trasnformer for 2x 6V6’s. Almost all fenders transformers were 6.6k for the 2x 6v6 tubes, 4k for the 2x 6L6 tubes, and 2-3k for the 4x 6L6 tube amps. ive even tried looking for higher primary impedance transformers but companies like Hammond dont tend to make them. but yeah, i was confused if i was setting up your calculator wrong, because every time i plugged in a different known guitar amp, they were similarly over the power curve of the tube. which kind of scared me a bit that the load line was going so far above.

        1. With correct operating point (es. 285V 35mA) a 6.6k loadline stays just below the max anode disspiation of the 6V6. With 4k, in fact, the loadline sometimes crosses the maximum dissipation of the 6L6GC. However, it is the average disspiation that suould be taken into consideration. If the curve stays below for a reasonable amount of time, the average disspiation does not exceed max plate disspiation

  24. Giuseppe,

    Great work here! saves a LOT of time doing them by hand with pencil andpaper!

    Another request, i know you’re busy so just when you can. 6973 pentode, either from the original datasheets or Electro Harmionx’s re-issue.

    Les

  25. Hello Giuseppe, with great interest I have read your articles and looked at the calculator. I cannot say I already understand everything but hope I can ask you a question, Where do you really want the operating point to be ? Only determined by the most power output ? I’m trying to determine for 2 x KT88s in UL push pull what the best output transformer is (B+ 410 V and set for 80 ma bias / tube) , I can buys 4Ks , or 5Ks the output calculator gives the most output at 4K so that’s it ? I need a 4K transformer ? No other parameters or decision factors ? or impact on frequency response ?

    1. Hi,
      sorry for late reply.
      Difficult to answer this question 🙂
      Operation point and transformer impedance might depend on transformer availabiltity, wished output power, distortion. Frequency response depends mostly on other aspect of circuit design (high and low pass filters in the circuit), and quality of output transformer.

      1. Ciao Guiseppe!
        This is a great tool, it has been very timesaving while designing and finetuning my last amp build!
        Would it be hard to add expected max Rms output power?
        Grazie mille!

        Johannes

  26. Hi,
    Thanks a lot for this powerful tool! It is very helpful to design tube amps.
    In my opinion, the calculated output power at headroom is a bit confusing.
    The tool seems to display Va@headr.*Ia@headr., But I assumed it to be the maximal sine power at headroom, which is in fact transformed to the Output.

    My Proposal:
    dVa_peak*dIa_peak/2
    with:
    dIa_peak = Ia@headr. – Ia_bias
    dVa_peak = Va_bias – Va@headr.

    It would also be very practical to see the average anode power dissipation for a sine and probably also for signals with other amplitude distribution (don’t know if there’s a normed amplitude distribution for music signals)

    I know this would be a huge load of work, but it would in fact be very helpful when you want to go to the power limit of the tube, without blowing it. On the other hand, this can also be simulated in Spice.

    1. Hi,
      expected (approximate) output power is calculated as (Vpeak*0.707)^2/load. For instance, an headroom of 100V (around the quiescent point) and a load of 2.5k gives (100V*0.707)^2/2500=1.9999W, which correspond to an ideal sinewave output power.

      I’ll think about anode power dissipation 🙂

      1. Yes, inductace depends on number of wire turns, but I mean that tubes have its internal resistance that has impact on needed inductance in transformer. For ex triodes have low internal res. and transfomer could have low primary Inductance for ex 10-20H. Other pentodes have higher internal res and need higher primary inductance. It can be automaticly counted in your calculator.

        1. I see. You mean calculating primary inductance to have optimal frequency response in relation to plate resistance. It should be L = Z/(2*Pi*f), where Z is plate resistance parallel to primary impedance, and f is lowest frequency response (-3dB). Isn’it?
          I’ll think about it.
          Thanks.

          1. yes. that’s right. I may be wrong, so corect me, but I think when tubes work in deep class AB at low current then pri inductacne should be higher to reach low frequences. And another thing I’m thinking is that we talk about resistance, do you mean plate resistance as Ri which I can find in datasheets?

          2. Not sure inductance depends on current.
            Z is plate (or anode) resistance parallel to primary impedance. Values for plate resistance can typically be found in datasheets.

  27. Hi Giuseppe, very helpful calculator, congratulations on your idea and skills. I am writing this post to make it even better. My comments:
    * when I choose a tube, e.g. KT88, 6L6, 6V6 etc. the choice between triode and pentode does not change and these are tetrodes
    * I just finished building an amplifier on ECL86 and calculated my own operating point for a specific transformer, I put my data into your calculator and I had some differences. To confirm this, I substituted the data from the Philips catalog and there were also some differences. Maybe they are due to the fact that the calculator works for a slightly different scheme. You are using external polarity. For example, I calculated the cathode resistor for automaric polarization. Maybe you can add the option to choose the type of polarization.
    *for the pentode I give the grid voltage, but which grid? It can be noted that for 2
    * you can also add the ability to calculate the capacity of the cathode capacitor as in this calculator: https://www.fonar.com.pl/bonus/programy/bonus_fkatod1.htm
    Best regards

    1. Hi,
      Thanks for your comments.
      I go point by point:
      1) Pentode, Triode, and Ultraninear indicate the operating modality, not to the tube type. You can choose one of the three operating modality both for tetrodes and pentodes (not for triodes): (triode modality: screen connected to anode; Pentode modality: screen with fixed voltage; ultralinear modality: screen with a percentage of the anode voltage (e.g: 43%)).
      2) The choice of a cathode resistore is driven by the choice of the operating point (Vq/Iq). Once you decide the the operating point, you check what is the corresponding grid voltage (Vg), and then you chose the cathode resistor as R_k=Vg/Iq. The power stage calculator computes it for you (https://www.vtadiy.com/loadline-calculators/power-stage-calculator/). Which ECL86 section are you referring to (triode section or pentode section)? What is your configuration (V+, Vq, Iq, Vg, Screen Voltage (in case you are using the pentode section), operating modality, Reactive/Resistive load)?
      3) In pentode mode you can specify both grid voltage (g1), and screen voltage (g2), just below “Next stage AC Impedance”.

  28. Hi there Giuseppe. thanks for the great tool of yours it was very useful to me up to now. Wondering if possible to add the EIMAC 4-400A tetrode tube

    Thanks
    Take care

    Chris

  29. Hi Giuseppe Amato.
    So great to know this page.
    Thanks for good jobs.
    Could you add: 6n1p, 6n2p (Pre tube) and 6c33c (Power tobe) both of Russia.
    Thanks so much

  30. Add 6K6 output pentode?

    Also, would it be possible to add an option for fixed bias? I’ve figured out how to work around it, but it isn’t going to be spot on.

    1. Not sure I understand what do you mean? You want to directly set the grid bias?
      Typically, you do it the other way around. You set the grid bias to obtain the wanted/needed operating point. Here, you can set the operating point and see what is the corresponding grid bias (either fixed or self bias).
      By the way 6K6GT in available since some time.

  31. Giuseppe, i loved your book, but i wanted to see if there is a way to change the power line of the tubes on this load line calculator and possibly add power to the load line data points that show voltage and current when you move you mouse over them. for example, i love building amps that are less than the full power spec, such as using a 6V6 for a 10W. another reason for possibly being able to change the power line is due to different manufacturer specs. TAD makes both a 12w and 14w 6V6 tube. but JJ electronics and others only makes a 14w tube. could it be possible to change the power line? that would be much appreciated.

  32. Hello very good work
    I write since Spain
    I have one tube curve tracer and i add and ampliate experience to compare whit you app.

    i work wwhit one tube some popular in “very low impedance” (ideal for OTL amps how the EL86 include in you list)

    this is: **** 12k5 tube ****

    and is good for low voltaje since 12v at 30v aprox. of +B, in real work, (double impulse electrons) oriented at old batery car amps in others times.

    I construct some Otl whit this very playables whit speakers of 8 and 16 ohms,(and others whit 800 ohms), for Guitar player.

    ideal tube for practice, formation, DIY, and low descharge, 🙂

  33. the 12K5 tube

    is very interesanting for 12v aplications, formation, low discarge alert, batery equip for eample for guitar player portatil or not portatil, and incluive OTL GUITAR player whit speakers of 8 and 16 ohms, not only for 800 ohms speakers.

    why this tube is of low impedance of charge.

  34. Thank you for creating this. It’s an incredible resource to have. I was wondering if you could add the EL3N tube to your list? Thanks!

  35. Hi Giuseppe,

    Thank you for this very useful tools.

    I am sorry but I am confused.

    If, for example, I am going to use 2A3 in parallel connection single-end and I have 1.25k output transformer, in the simulator, I should use 2A3 with 2.5k load, am I right?

  36. Been using this load line calculator for a while, it’s great. My question is examining 6BQ5 tube in single ended operation, what exactly is meant by the power at G/max, G@0, vs power at headroom voltage, and finally in class A operation? How do all these figures compare to putting a load on an amplifier, putting it on the scope and setting the amp just before the sine wave clips and extracting the power from the voltage measurement?

    1. I calculate an approximate Output power obtained when
      1) g1 swings up the maximum displayed value for g1;
      2) g1 swings up to 0;
      3) g1 stay within the A or A2 class (that is NOT in AB class)
      3) when g1 stays between the desired headroom (which you have to set)
      Whe you use a scope on a real circuit you get the real values at clipping: In most tubes this corresponds to 1) or 2).

  37. I would like to do single end non reactive DC load driving Cap + Pri to ground – e.g. for a 6k6 with a 10K plate resistor driving a DC blocking cap (large enough tp ignore reactance…e.g. 2 uf) and 10k primary. The reactive load would then be ~5K but the DC load is 10K. App is transformer coupled driver stage for PP output stage.

  38. can your simulator work with the fixed bias configuration? I’m talking about a scheme where I drive the bias voltage on the grid with the cathode connetted to the ground. I would like t use a 300B tube. Thank you.

    1. My simulator calculates the needed grid bias voltage for your chose operating point. This is independent from the self or fixed bias modality.
      In your real build you can obtain the calculated grid bias voltage using a fixed bias circuit or a cathode resistor. It is up to your preferences.

  39. Thank you for your reply.
    I need your help because I can’t understand the relationship between the dynamic working point in the load line and the static working point.

  40. Yes, you are right because I was talking about AC and DC load lines, but for a single ended output stage based on KT88. In such final stage the output transformer impedance can be, for instance, 3 Kohm but the transformer primary resistance is 50 ohm. If I calculate the load line and working point based on the AC impedance when my final stage is in idle state (no input signal) the plate voltage, the plate current and the grid voltage are completely different compared with the AC ones. I don’t know if I’m explained clearly, but this is an issue that I can’t understand.
    Thank you for your patience.
    Flavio.

    1. Actually, I am not sure I correctly understood your question.
      However, when you use a reactive load (a transformer), the load is just seen in AC operation (when a signal is applied to the grid). With a reactive load, when you change the grid bias voltage, the operating point simply moves vertically up and down, given that the plate voltage is fixed and there is basically no load (the transformer primary resistance is basically not significant and lead to an almost vertical loadline).

      1. Let me do an example to explain my not understanding.
        When I work with a resistive load I can define the grid bias according to a right load line and I’m sure that in the idle state (no input signal) my plate current is known a stable.
        I can do the same job with a reactive load in a final stage (ie 300B). In this case I can define a load line with a center point at 250V 60mA assuming a primary load of 3Kohm. Now, in this condition, when there is no signal and having a power supply of 400 V, with the grid bias defined by the AC load line, the plate current rise to levels able to break the tube.
        I hope that this time my trouble description is clear.
        Thank you for your time and patience.
        Flavio.

        1. With a reactive load, the load “exists” only in AC operation. Therefore, at idle state, if the power supply is 400V, this is the power supply seen by the valve, given that there is no voltage drop across the load (as happens with a restistive load).
          If you want the idle operating point to be 250V 60mA, the power supply (applied to the transformer) must be 250V as well, given that there is no voltage drop across the transformer at idle state.

Leave a Reply

Proudly powered by WordPress | Theme: Baskerville 2 by Anders Noren.

Up ↑