How to Avoid the Drudgery of Drawing Push-Pull 6L6 Composite Load Lines?

[J M Fahey]

OK, as the Beatles said, "Let it be".
Although we all know what happened after that record.
Try to at least regulate the screens down with a nice MosFet and keep them at not more than , say, 400V.
And .... good luck!!!

[FredB]

What do you mean happier?

I believe my RC30, which is hiding at the moment, lists the recommend P2P impedance for a pair of 6L6 tubes as 6K. That's close to your 5k5 figure.

I presume guitar amp designers lower the primary impedance to get more watts out of a lower supply voltage. At least that's what my sims suggest.

Could you post a schematic of how you'd use one resistor to determine the optimal impedance for push-pull pair of tubes.

Wind you own transformers aye? Awesome.

I've looking into doing that. What winder are you using?

[loudthud]

If you look through the RCA tube manual, you see all sorts of strange plate to plate impedances. I think what they did was mockup the amp with something close for an OT, and use a variable resistance load, being carefull to calculate the power out. Then back calculate what the primary impedance was.

[J M Fahey]

What do you mean happier?

They drive that load easier than any other

I believe my RC30, which is hiding at the moment, lists the recommend P2P impedance for a pair of 6L6 tubes as 6K. That's close to your 5k5 figure.

Cool guys

I presume guitar amp designers lower the primary impedance to get more watts out of a lower supply voltage. At least that's what my sims suggest.

Yes, problem is sims presume the original tubes, what we call NOS today, and made by the original manufacturers.
What you can buy over the counter for a reasonable price simply is not the same.

Could you post a schematic of how you'd use one resistor to determine the optimal impedance for push-pull pair of tubes.

No need for that.
Just set your adjustable resistor to , say, 8 ohms, plug it into your 8 ohm tap and measure.
Then raise by 10% steps until you get to 16 ohms and lower same way to 4 ohms.
Draw the curve of power vs. impedance and select the best.
It's patience, not rocket science .

Wind you own transformers aye? Awesome.

I've looking into doing that. What winder are you using?

An old '56 vintage one, bought from a supplier when he retired.







This is one of my commercial amps, a Marshall 1987 clone, using those transformers:



With basically the same set, I can build Fender Bassman, VOX AC50 or Ampeg B15 clones if the customer so desires, so it was a good investment.

Of course, not for a couple ones, but I did wind quite a few in a hand cranked machine at the beginning, so it can be done.
Still have it, for prototype work which does not justify setting the automatic one up.

[FredB]

Sweet winder.

I have someone who will sell me an old one that can stick wind up to ten windings at a once for $150. Although it's halfway across the country from me and it weighs several hundred pounds. Shipping would cost prohibitive.

I'm still debating getting a hand cranked one with or without a wire feeder. The one with the wire feeder is like 5x as much.

I've done that reiterative output verse impedance study with sims. I need to pick up some more power resistors of various values. I have one 10 ohm 250W one that I use for testing, and some smaller ones.

How do you pick the point at where to measure the output, where the wave form starts to flatten?

[J M Fahey]

For just one or two, buy a hand cranked simple one, use good light and glasses (even if you don't ned them, consider them binocular "loupes", I use scheap street bought +4 ones), put some nice background music and go ahead, hand guiding.
Wire sort of self guides following the last turn.

[Wombaticus]

That is *exactly* the "drudgery" that I was referring to. Cutting and pasting a new chart every time that you make a significant change can be somewhat of a PITB. I guess I fall into the category of old-dog willing to learn a new trick. Like JMF said, cutting and pasting is guaranteed to provide the right answer. I'm just worried about trusting modelling software that may or not be accurate.

The idea of Spice models is interesting. Being an old-dog, I don't know how to use Spice. Tried it once and never got it working for lack of a good tutorial. I'm not sure that it's even worth the effort of learning Spice if the models are in any way less accurate than paper, scissors and tape.

Hi Bob,

For the load line sims I posted on AX84 that Jazbo8 was referring to, I really made an effort to make sure that everything is plug-and-play, with very little understanding of LTSpice required.
If you download and install LTSpice, then put the models in the correct folders, it should be no harder than opening up the file with sim and plugging in your values --- it's clearly indicated where to punch in the voltages and bias % you wish to target. The only thing you need to really know about Spice is where to find the "run simulation" button. Basically, you press on the button with a picture of a little man running, and it's done. As others point out the results will only be as good as the model used, but I have obtained highly satisfactory results from Stephie Bench's 6L6 model in particular.

Keep in mind that this is an approximate science anyway as the datasheets correspond to a more-or-less mythical "bogie tube", and real world tubes commonly vary +/- 20% or more.

[bob p]

Hi Bob,

For the load line sims I posted on AX84 that Jazbo8 was referring to, I really made an effort to make sure that everything is plug-and-play, with very little understanding of LTSpice required.

Thanks for posting!

Is LTSpice a commercial Windows application? I understand that SPICE is open source and multi-platform, but it looks like the LTSpice program is proprietary and Windows-only. Please let me know if I'm wrong about that. I couldn't find any LTSpice downloads at linear.com that were anything but Windows .EXE files. If LTSpice is Windows-only, then I'm out of luck because I don't use Windows.

I have looked at the 6 files that you posted at AX84, using the open-source version of Spice that's included in the Fedora electronics lab, but I have to admit that I don't understand why the files you've linked to have 3 different types of file extensions, or which ones (if not all of them) are needed to perform a simulation. Loading them up using the gspiceui doesn't yield anything executable.

Can anyone offer some insights?

[Wombaticus]

Thanks for posting!

Is LTSpice a commercial Windows application? I understand that SPICE is open source and multi-platform, but it looks like the LTSpice program is proprietary and Windows-only. Please let me know if I'm wrong about that. I couldn't find any LTSpice downloads at linear.com that were anything but Windows .EXE files. If LTSpice is Windows-only, then I'm out of luck because I don't use Windows.

I have looked at the 6 files that you posted at AX84, using the open-source version of Spice that's included in the Fedora electronics lab, but I have to admit that I don't understand why the files you've linked to have 3 different types of file extensions, or which ones (if not all of them) are needed to perform a simulation. Loading them up using the gspiceui doesn't yield anything executable.

Can anyone offer some insights?

Bob,

LTSpice is indeed a Windows application... It's proprietary, but a free download. While it shares much common DNA with other SPICE applications, it's unlikely that the files I provided will work with another spice variant without some specialist massaging.

The good news, however, is that LTSpice has been developed to be compatible with WINE, which is a compatibility layer that allows many windows applications to run under linux.

As to the question of the different file extensions, here's the breakdown:

.asc files -- these are the files with the schematic/netlist for the simulation. These are what you would load from within LTSpice.

.plt files -- files with this extension tell how LTSpice to format the plotted output from the simulation. If they are in the same directory as a correspondingly named .asc file, they will be automatically loaded.

.inc files -- these are library files that contain tube models. On a windows machine they would normally go under the C:\Program Files (x86)\LTC\LTspiceIV\lib\sub hierarchy.

[Steve Conner]

As you just found out, the only open source part of Spice is the original simulation engine. Everyone and their dog developed a proprietary schematic editor, results viewer, and file format to support it.

The two versions most commonly used are PSpice student version (as every EE student gets it inflicted on them in 1st year) and LTSpice. They use incompatible file formats.

LTSpice supposedly runs fine under Wine.

[kleuck]

[bob p]

[ATTACH=CONFIG]22385[/ATTACH]

My French isn't what it used to be, but there can be some ambiguity when it comes to the use of the term "spice." In the linux world, Spice can refer to the circuit emulator, or it can refer to the Simple Protocol for Independent Computing Environments, which is a protocol for connecting thin-client users to virtualized desktops.

Red Hat | SPICE in Red Hat Enterprise Virtualization for Desktops

It looks like the popular front ends for Ubuntu and Fedora are ngspice and gspiceui.

I've got them, but that isn't getting me anywhere.

[root@8coreamd]# yum install ngspice gspiceui
Package ngspice-23-3.fc18.x86_64 already installed and latest version
Package gspiceui-0.9.98-8.fc18.x86_64 already installed and latest version
Nothing to do
[root@8coreamd]#

[kleuck]

I do not use these software myself, but the best way seems to use Geda, a software suite using ngspice for simulations.