Load Line, Q-Point, and Frequency Response in Excel

[martin manning]

You only have to provide the anode characteristics data table (plate curves), by digitizing the plot from the data sheet for the tube you want to add.

You will want to put this data into one of the blank “cust” (custom) areas on the device library sheet so that it will be accessible by clicking the appropriate button on the input data sheet.

You can load the data on the device library sheet, but I suggest that you initially enter the data on the input data sheet, beginning at row 48. That way you can look at the graphs as you go, checking for errors and looking for reasonably smooth lines on the ra and mu plots. When you are satisfied, copy the yellow/gray shaded area on the input data sheet and paste it over one of the custom areas on the device library sheet. The buttons on the input data sheet will pick up the identifier (type number) from the table on the device library sheet, so include that.

MPM

[martin manning]

I've updated this tool with a couple of fixes and some improvements. Recent posts seeking help with plotting load lines for DC coupled stages and reverb drivers, plus some discussion about stages with input attenuators and the effect on frequency response inspired me to add some new features..

A reactive plate load and the grid leak resistor Rg2 have been added, and there was an issue with the interpolation used in locating the quiescent point that has now been fixed. In the previous version, there was some error with large Rk values.

While the focus is on common-cathode gain stages, it will correctly do load lines and gain calculations for cathode followers. I've found this thing to be useful as a quick check and graphical viewer, and the frequency response results seem to compare quite well with a full-up simulation.

Edit: Attachment deleted. See first post in thread for latest version.

[tubeswell]

Wow! Cool! thanks Martin

[martin manning]

My apologies... I uploaded the wrong version, the one without the CF gain calcs. Here is the correct one, reversioned 8.01.

Edit: See first post in thread for the latest version.

[Hellhammer]

Wow, this is very useful indeed... nice work! So has anyone added more tubes already? I just starting plotting the anode characteristics for the Philips ECC88, do I send them to you Martin for future updates or post here for everyone to use?

One question regarding DC coupled stages. I can't get it to draw the AC load line right. If my HT is 300V and Vg is 50V then the AC load line should start from 250V or have I missed something? I probably have...

[martin manning]

Wow, this is very useful indeed... nice work! So has anyone added more tubes already? I just starting plotting the anode characteristics for the Philips ECC88, do I send them to you Martin for future updates or post here for everyone to use?

One question regarding DC coupled stages. I can't get it to draw the AC load line right. If my HT is 300V and Vg is 50V then the AC load line should start from 250V or have I missed something? I probably have...

I don't know if anyone has digitized more tube types but I, and I'm sure others, would be happy to have them made available here. They could be posted as separate Excel files, and picked up by others as desired. I guess I'll move the main file to the first post so it is always easy to find.

Re the AC load line, it goes through the DC operating point, with it's slope determined by the AC load, which has Rl (because Co is considered to be shorted) and the reactive part of Ra combined with the DC load of Ra and Rk. The voltage at zero current falls out from that and Vg, since Vg is part of determining the DC op point.

[Hellhammer]

Here are curves for the Philips ECC88, from a 1955 datasheet. If someone wants.

Edit: Tweaked values for smoother curves.

[Hellhammer]

Re the AC load line, it goes through the DC operating point, with it's slope determined by the AC load, which has Rl (because Co is considered to be shorted) and the reactive part of Ra combined with the DC load of Ra and Rk. The voltage at zero current falls out from that and Vg, since Vg is part of determining the DC op point.

I might be mixing apples and oranges here but I'm looking at Merlins example for a DC-coupled 12AU7 stage in his pre-amp book on pages 58-59. Vg is 145V and HT 350 wich means that the maximum available anode-to-cathode voltage is 205. He draws the AC load line from this point. When trying to replicate this stage using your eminent load line plotter it doesn't seem to take max Va-k into consideration? Or am I doing/reading it wrong?

[martin manning]

Ok, good work on the ECC88. The only thing I'd maybe do is tweak the table to smooth out the ra curves a little more. You could also tag it with your name in the blank cells at the end of the first line after the Cg-k etc.

Re the example, good catch... I see what's happening; Rk needs to come out of the AC load if it is bypassed by Ck. Let me have a look at this and see if there is a way to work that in. I'm trying to handlle several circuits with the logical statements. I appreciate your effort wringing this out, that's how it will improve.

[martin manning]

Try version 8.02. It replicates Merlin's Fig. 2.28 very well if you put a very large value in for Rl. The fix is just a test on Ck. Thanks again for trying it out on this circuit!

PS I'll have a look through the gain calc for this too.

[Hellhammer]

Try version 8.02. It replicates Merlin's Fig. 2.28 very well if you put a very large value in for Rl. The fix is just a test on Ck. Thanks again for trying it out on this circuit!

PS I'll have a look through the gain calc for this too.

I was just going to ask about this! Maybe the increase in cathode current feedback is the villain here? Looks great otherwise though! This tool will come handy.

Btw I tweaked and updated the values for the ECC88.

[Hellhammer]

Pentodes then?

But seriously... how hard would it be to add pentode curves? For example EF86 curves for set G2 voltages. Could they be added as is just for load line plotting?

[martin manning]

I have another spreadsheet for that ;^) Not quite ready for prime time, but it works, and has scaling for changing screen grid voltage.

[Hellhammer]

Wow! You the man! I'd love to try it out if you need a beta-tester.

[Tonegeek]

Thanks so much for sharing your work.