Formulas used to calculate distortions for all triodes (including 845 and 6S33S) are the same. You can analyze the high value for 6S33S by comparision tube curves with the applied formulas.
Here is the link to the description of simulator fundamentals:
Program wspomagaj?cy dobór punktu pracy triody
The description is in polish, and it is not finished yet, but you can use google translator and compare formulas for distortions to the included figure which is self explanatory. I'll prepare the english version of manuam soon.
Here is the link to the description of simulator fundamentals:
Program wspomagaj?cy dobór punktu pracy triody
The description is in polish, and it is not finished yet, but you can use google translator and compare formulas for distortions to the included figure which is self explanatory. I'll prepare the english version of manuam soon.
Thanks, gsmok.
I didn't mean to imply that there was anything wrong with your program, I was just surprised at how badly the 6S33S performed in terms of H3, and was wondering if someone knew a trick for how to set it up that wasn't evident to me.
Your simulator is extremely helpful, because it quickly distinguishes bad tubes from good tubes and why - I knew the 6S33S was non-linear, but I had always assumed ignorantly that it simply made a lot of H2 - not a very bad thing - but now I know it makes a ton of H3 - which to me is a very bad thing indeed.
Question: how do you determine what amplitude of input signal to use for calculating distortion and power figures?
Very Best, Synchro27
I didn't mean to imply that there was anything wrong with your program, I was just surprised at how badly the 6S33S performed in terms of H3, and was wondering if someone knew a trick for how to set it up that wasn't evident to me.
Your simulator is extremely helpful, because it quickly distinguishes bad tubes from good tubes and why - I knew the 6S33S was non-linear, but I had always assumed ignorantly that it simply made a lot of H2 - not a very bad thing - but now I know it makes a ton of H3 - which to me is a very bad thing indeed.
Question: how do you determine what amplitude of input signal to use for calculating distortion and power figures?
Very Best, Synchro27
In any case, I do not avoid responsibility for any errors in the software. On the contrary, any comments to the computed results, are very helpful to me in finding and correcting errors 🙂.
Distortion are calculated for the input voltage, which is set using the yellow marker (value of Uinp selected by a yellow marker is displayed below the graph). You can decide what input voltage is taken into account during simulation od distortions.
Distortion are calculated for the input voltage, which is set using the yellow marker (value of Uinp selected by a yellow marker is displayed below the graph). You can decide what input voltage is taken into account during simulation od distortions.
gsmok,
just to drive you completely crazy LOL.....
.....would it be possible to add the capability of making the loadline horizontal (or nearly horizontal) and adjust it up and down for current?
This would be simulating loading the plate of the tube with a CCS; the better the CCs, the more perfectly horizontal the load line.
I find that with many tubes, such as the 2A3, the shallower the slope of the loadline (the higher Ra is), the lower the distortion, especially H3, and it is possible to almost completely eliminate H3 and dramatically lower H2 (and, by the way, the closer you can get to obtaining the highest mu possible from the tube)....but the way the program is right now, I cannot set a shallow (or horizontal) loadline and adjust it up for the optimal Ia.
Still love your program and am profoundly grateful...but us DIY'ers always want more, more more LOL.....
Very Best, Synchro27
just to drive you completely crazy LOL.....
.....would it be possible to add the capability of making the loadline horizontal (or nearly horizontal) and adjust it up and down for current?
This would be simulating loading the plate of the tube with a CCS; the better the CCs, the more perfectly horizontal the load line.
I find that with many tubes, such as the 2A3, the shallower the slope of the loadline (the higher Ra is), the lower the distortion, especially H3, and it is possible to almost completely eliminate H3 and dramatically lower H2 (and, by the way, the closer you can get to obtaining the highest mu possible from the tube)....but the way the program is right now, I cannot set a shallow (or horizontal) loadline and adjust it up for the optimal Ia.
Still love your program and am profoundly grateful...but us DIY'ers always want more, more more LOL.....
Very Best, Synchro27
Thank you for all remarks. I'll analyze them and try to implement in the simulator.
Meantime I have added the beautifull looking russian triode GM-100. I have two of them 🙂. Koren's model had some problems with such high positive grid voltages, but the model is sufficient to play with this tube. Mayby someone try to build amplifier based on this tube 😉.
Meantime I have added the beautifull looking russian triode GM-100. I have two of them 🙂. Koren's model had some problems with such high positive grid voltages, but the model is sufficient to play with this tube. Mayby someone try to build amplifier based on this tube 😉.
The GM100 is a very cool, huge triode...great potential for an insane SET amp.
However, did you realize you only have the grid-conduction curves on there? That is, the curves for grid biasing from 0 through positive values?
God knows I could be wrong, but if I were contemplating using the GM100, I'd wish to try to figure out a negative bias point and loadline that extended from negative grid territory into positive grid territory for transients; I'd be looking to have most of the operation of the amp, at low power, in negative grid territory. This would require a drive circuit, of course, that can handle grid conduction well, and of course that immediately brings up tubelab.com's power drive circuit.
Having said all this and having asked for the curves to be added that arise from negative grid biasing, it does certainly occur to me that perhaps a GM100 amp should be run completely in grid conduction mode....aside from tradition, and the inadequacy of most drive circuits to handle grid conduction, I am hard pressed at the moment to explain why a SET amp couldn't be designed to run completely in grid conduction mode. Enlightenment welcome.
Very Best, Synchro27
However, did you realize you only have the grid-conduction curves on there? That is, the curves for grid biasing from 0 through positive values?
God knows I could be wrong, but if I were contemplating using the GM100, I'd wish to try to figure out a negative bias point and loadline that extended from negative grid territory into positive grid territory for transients; I'd be looking to have most of the operation of the amp, at low power, in negative grid territory. This would require a drive circuit, of course, that can handle grid conduction well, and of course that immediately brings up tubelab.com's power drive circuit.
Having said all this and having asked for the curves to be added that arise from negative grid biasing, it does certainly occur to me that perhaps a GM100 amp should be run completely in grid conduction mode....aside from tradition, and the inadequacy of most drive circuits to handle grid conduction, I am hard pressed at the moment to explain why a SET amp couldn't be designed to run completely in grid conduction mode. Enlightenment welcome.
Very Best, Synchro27
gsmok, a little more info on the GM100.
This fellow, at Leviathan: GM100, reports finding a good bias point in negative grid territory:
"31-05-2001
I may have found a good bias point for this tube after all. 2500V/200mA and -70V using a 10K load (Rp=2700). The grid can be driven to 135V (=270Vpp) before the tube clips on it's B+ supply (Ugrid = B+ / mu). This should give upward of 140W. For those not willing to go into grid current, it will still deliver 40W with 70V of swing. Run for three hours and you'll have generated enough heat to last the entire winter. Still, very cool tube, would love to build something with it."
Attached is the graph he has on the page.
Very Best, Synchro27
This fellow, at Leviathan: GM100, reports finding a good bias point in negative grid territory:
"31-05-2001
I may have found a good bias point for this tube after all. 2500V/200mA and -70V using a 10K load (Rp=2700). The grid can be driven to 135V (=270Vpp) before the tube clips on it's B+ supply (Ugrid = B+ / mu). This should give upward of 140W. For those not willing to go into grid current, it will still deliver 40W with 70V of swing. Run for three hours and you'll have generated enough heat to last the entire winter. Still, very cool tube, would love to build something with it."
Attached is the graph he has on the page.
Very Best, Synchro27
Last edited:
Very nice tool. Similar results (with EC86 an 845) to traditional loadline on paper, with more précision and lot of informations.
Somme triodes suggestions : 6BX7, 6SL7, 6H30, 417A/5842, 6C45P, GM70.
Thanks.
Somme triodes suggestions : 6BX7, 6SL7, 6H30, 417A/5842, 6C45P, GM70.
Thanks.
I added GM-70. I couldn't find good resolution manufacturer's anode characteristics. Mayby someone has such data?
Hi gsmok,
Hate to be a pest, this is just a friendly reminder about the 6SL7 in case you've forgotten.😀
Thanks a bunch!
Hate to be a pest, this is just a friendly reminder about the 6SL7 in case you've forgotten.😀
Thanks a bunch!
Another suggestion if I may, is the possible addition of the 5th harmonic component in the distortion calculations. I know it might be very small to non-existent at low power levels but it does rise with level.
The reason I'm suggesting this, is because it's the odd order components that are more objectionable and this will make an excellent tool in designing for operating points that show favorable distribution of harmonics.
The reason I'm suggesting this, is because it's the odd order components that are more objectionable and this will make an excellent tool in designing for operating points that show favorable distribution of harmonics.
I found this data a while ago.
I don't know who produced the data, but the curves are more comprehensive than the manufacturer ones that I have seen.
I mentioned a while back that the drag doesn't work on my tablet or phone. Would it be possible to add text inputs that sync with Ua and Ia, so you could drag or type the values? That would at least make it functional on the tablet and wouldn't require any changes to the drag code.
Regarding above suggestions for the extensions of the program and corrections of errors - I appended them to the list. Please take into account that I introduce them very slowly due to the fact that I'm doing it in my free time, which unfortunately I do not have too much 😱.
I added 6SL7/6SL7GT.
Thank you for the data. Unfortunately characteristics are only for negative grid voltages. Mayby it will be a good idea to divide simulated characteristics for negative biases only and positive/negative biases. I'll try to do such simulations.
I added 6SL7/6SL7GT.
Thank you for the data. Unfortunately characteristics are only for negative grid voltages. Mayby it will be a good idea to divide simulated characteristics for negative biases only and positive/negative biases. I'll try to do such simulations.
Last edited:
Just wanna say thank you again for this marvelous tool, and I appreciate all the work you're doing...in your spare time!! Wish we could help.