A useful Excel spreadsheet for tube amps

Bauhausler

Rational Subjectivist
All;
I just stumbled on a spreadsheet I worked up a few years ago that I can share if anyone wants it. It computes gain and frequency response for 2-stage SE tube amps at 1/3 octave frequencies. Basically, you put in as variables the tube characteristics and values of the passive compnents and some info on the OPT and it calculates and plots the gain and frequency response at various parts of the circuit. I ginned this up because I was working on a design for a 2-stage SE that would have 12 db/oct hi-pass without extra filter networks. I found that by fiddling the coupling cap and cathode bypass caps I could get exactly the response that I wanted.
The spreadsheet does not have instructions, so right now it's probably suited to experienced excel users. I'll 'donate' this to AK and maybe somebody will want to write up some instructions. It's pretty complex and will take a long time to type up. I'd get to it eventually, but not right away.
In this example, the line 'spkr' represents the response of a passive speaker hi-pass crossover. 'tot db down' represents the response of the tube amp with the component values selected to mimic the speaker crossover responce. 'error' is the difference between them. You can get more accurate responses than this, this was just a handy example.
Here's a pic of what one of the charts looks like:
 
Last edited:
Here's the chart of another spreadsheet. I think this is a 3-stage amp: 76 - 12b4 - 2A3. I was going for 6db down at 500 hz and mimicing a first order passive speaker crossover. As you can see, the error is less than 1 db, right down to 20 hz. It's so easy to fiddle the component values and see the chart change that it's really addictive.
 
me! me!

Yes Bauhausler, I'd be very interested in a copy. I'm right in the middle of trying exactly what you are - I building a 45 SET that I want to try on the horns of my Altecs in a bi-amp configuration with my 300B amps on the woofer and I'd like nothing better than to eliminate the need for an active crossover. I wanted to try passive line-level but this sounds even more elegant.

If you don't mind, could you detail your progress on this in the DIY forum?

Thanks, Billfort
 
Re: me! me!

Originally posted by Billfort
Yes Bauhausler, I'd be very interested in a copy. I'm right in the middle of trying exactly what you are - I building a 45 SET that I want to try on the horn's of my Altecs in a bi-amp configuration with my 300B amps on the woofer and I'd like nothing better than to eliminate the need for an active crossover. I wanted to try passive line-level but this sounds even more elegant.

If you don't mind, could you detail your progress on this in the DIY forum?

Thanks, Billfort

Bill;
If you think this belongs over on DIY we can ask the moderator to move it there. I picked 'tubes' because this is the one I read.
Basically I'd like people who want this to PM with their 'offline' email address and when there's a bunch I'll do a 'mass mailing' of the zipped spreadsheet. Its about 150K right now. I don't want to burn AK bandwidth unnecessarily.
Actually, if someone wants to post this on their own site and link to it maybe that would be better. I do not have a site of my own.
This excel workbook actually has a model of a se 6sn7/45 amp with high pass characteristics. The first plot shown was taken from that. It's a piece of cake, and if you're only using the amp above 500 hz or so you need next to no inductance in your OPT, which really widens up the hardware options.
 
Very cool, I'll PM you. I'll be using a 6SN7>300B and a 417A>45 - will this work on the single stage 417?
 
Originally posted by Billfort
Very cool, I'll PM you. I'll be using a 6SN7>300B and a 417A>45 - will this work on the single stage 417?

Billfort;
You can plug in the characteristics of any triodes and it will calculate. For a single stage I think it would be easiest to treat it like a 2-stage with the first stage having gain = 1 and a very large coupling cap and plate load resistor. Otherwise, you could modify the sheet a little to exclude one stage's effects.
I'll email those sheets out tonight. Right now I'm punching up my resume for a 3:00 deadline.
 
My PM box is flushed out now and ready to receive your requests/adulation/abuse, etc.
I'm going to have to poke around a bit more. I could swear I had a version of this spreadsheet that incorporated OPT distributed capacitance and power tube Miller effect into the response calculations. It gets really complex, but it works. That might be the one that got nailed by a virus at work & died a horrible death.

Detroit Rock and Roll: Making the world a better place one garage at a time.
 
Okay, I found a workbook that includes the effects of OPT stray capacitance. That one also incorporates features for loop NFB, but I don't know how accurate they are since I don't think it takes into account phase shift of the output signal due to OPT reactance loss at LF.

There's another interesting one where I incorporated NFB and also allowed actual speaker impedance magnitude (but not phase) to be entered for each frequency. The final frequency response graph shows the effects of the woofer resonance very prominently. Another graph line shows the response with NFB being much flatter. This is useful for computing values for a SE amp to drive a speaker that was designed for a high damping factor amp. In some cases, by changing the loading of the woofer cabinet the bass can be flattened out and extended. Don't ask me how to do this. It requires too much use of T-S parameters and an intimate feel for woofer loading. There was a useful artcle on woofer loading and SE amps published in audio Amateur a few years back that takes Zout and OPT L into account.

There's another where I think I was trying to do a PP 300B thing but I'll keep that to myself because I think it's wrong.
 
Back
Top Bottom