Slides Rules and Other Analog Computers
published 2025-10-02
By Christopher Howard
I'm still trying to work out the issue with the tariffs on my THAT electronic analog computer, and am filing a duty dispute claim through UPS. The 300% fee was, apparently, a mistake by person or persons unknown, but they are still trying to hit me with an approximately $300 fee which I think is not correct. It seems that if your product is classified as a "copper product" then you get a 50% tariff fee, but if it is an "electronics" product, that is only 2.6%, and somehow my shipment got hit with both. I think maybe that is because a small bag of copper signal cables is included. But the signal cables are only worth, at most, around $30, so it wouldn't be fair to get a 50% tariff for the value of the entire computer.
I do want to say that I am a fan of tariffs in general, and how Trump is using them. But of course I don't want to pay more of them then I have to pay, because of some bureaucratic mistakes. And I don't want to pay tariffs that I didn't know about when I placed the order, when I could have instead spent the money on some American-made product.
I've been pondering, if this order falls through, the possibility of making a 2.0 version of my homebrew analog computer. I was thinking that it would be really cool to have a 100 volt computer, so as to minimize the noise effects dramatically. But this might not be feasible, as a good 100 VDC linear supply is quite expensive, and the affordable op amps only operate in the double-digit range for supply and input voltages.
An interesting PDF survey of early (i.e., pre-1970s) computer technology is available here:
It covers some interesting topics like the slide rule as well as the electronic slide rule. The document is referenced by this also very interesting guide on how to use a slide rule:
I have a Pickett Model N-16-ES, a nice slide rule, so I have been working through the guide. For me, it is most important that I understand what is happening with the manipulation of logarithms. The basic idea of a slide rule is that, according to the logarithm product rule
log10(x * y) = log10(x) + log10(y)
and therefore you can add together the two logarithmic scales to get the product. Things get a little more complicated when you do wrap-around multiplication.
log10(x) + log10(y) = log10(xy)
log10(x) - log10(10) + log10(y) = log10(xy) - log10(10)
log10(x) - log10(10) + log10(y) = log10(xy/10)
The folded-scale multiplication is a similar idea, if I understand it correctly:
log10(x) + log10(y) = log10(xy)
log10(x) + log10(y) - log10(π) = log10(xy) - log10(π)
log10(x) + log10(y/π) = log10(xy/π)
I find that, using the slide rule for basic multiplication, my answers always have an error of less than 0.2% (two-tenths of a percent) against the output of Emacs Calc, which I am assuming to be exactly correct. This is not too bad for most applications.
Copyright
This work © 2025 by Christopher Howard is licensed under Attribution-ShareAlike 4.0 International.