If you’ve ever worked with guitar pedals or analog audio gear, you’ve probably realized the value of a resistor decade box. They substitute for a resistor in a circuit and let you quickly flick through a few different values at the twist of a knob. You can still buy them if you know where to look, but [M Caldeira] decided to build his own.
At its core, the decade box relies on a number of 11-position rotary switches. Seven are used in this case—covering each “decade” of resistances, from 1 ohm to 10 ohm and all the way up to 1 megaohm. The 11 positions on each switch allows the selection of a given resistance. For example, position 7 on the 100 ohm switch selects 700 ohms, and adds it to the total resistance of the box.
[M Caldeira] did a good job of building the basic circuit, as well as assembling it in an attractive, easy-to-use way. It should serve him well on his future audio projects and many others besides. It’s a simple thing, but sometimes there’s nothing more satisfying than building your own tools.
We’ve seen other neat designs like this in the past, including an SMD version and this neat digital decade box. Video after the break.
Nice to see the resistors are easily-replaceable robust through-hole components. Because, inevitably, one resistor gets burnt one day, rendering the whole box markedly less useful.
agreed, and I like how the pcb is broken up into smaller repeatable patterns. I’ve done that for keyboards, and it’s both economical and facilitates repair. No need to desolder everything to replace a trace, just the pcb in question will do.
Well that’s nicer than mine. I built one that’s (externally) based on the Elenco RS400. Two rotary switches, each one selects a single resistor value, and a switch to choose between the two rotary switches. I opted for banana jacks instead of the alligator clips the Elenco uses. I didn’t see any need for a PCB with such a straight forward design.
Now I’m wishing I had built a decade box instead, especially one as polished as that one.
Meh, to big for me.
Recently I bought a few resistor decade PCB’s from aliexpress. Cost around EUR5 and 8 decades. Size of the PCB is around 30 by80mm. These have SMT resistors and you set the value with a bunch of jumpers. I had bought similar versions before, but the new batch has a 0.1 Ohm resolution. That’s quite handy if you want it for example to emulate temperatures of a PT100
How do you get a decently accurate .1ohm res? Ive never found decade boxes very usefull under 10ohms, every time ive tried to make a low ohm box the traces and switches can have up to .5ohms of resistance.
If the one they were on about is similar to one that I have it doesn’t use rotary switches, instead it uses jumpers.
To make one you would need very low tolerance resistors and low resistance connectors or jumpers and thick traces for the lower resistance.
I would still doubt the accuracy of one that cheap from AliExpress. Unless it has been tested by something accurate then I wouldn’t trust it. It is probably only good for relative values at that low of resistance increments. Once you get beyond even say 10 Ohms then the 0.1 Ohm increments will be useless because the tolerance range of the resistors will probably be greater than 0.1 Ohm depending on which tolerance they use. 1 % tolerance on 10 Ohms is +-0.1 Ohms.
You don’t get a high accuracy for EUR 5 from Ali, but you still have the resolution.
Making a resistor box with 7 orders of magnitude of accuracy as presented in this topic is above my budget.
years ago, when I was doing some analog design, I was fortunate to share an office with a really first-rate analog designer. I learned a lot from him, including
“one, two point two, and four point seven, plus a multiplier. If you need any other value except in a filter, you had better be able to defend why you needed it in the design review.” Meaning that in most instances, resistance values are just not very critical.
In my experience, if you need to fiddle around to find the proper resistance, the circuit is WAY too sensitive to be immune to all the capacitance, noise and stuff that a decade box adds to the circuit. Corollary: after you get the circuit working WITH the decade box, when you sub in a resistor of that value, the circuit won’t work any more. This is especially true if your design actually needs precision resistors (i.e. any place you’d use the lower decades of that box).
Bottom line: IMO. decade boxes are worthless.
Exactly my thoughts.
Yeah, there’s some truth to that… A power decade box—that could be of value.
Sure, you can’t use a resistor decade box in the feedback loop of a 100GHz amplifier, but that does not mean it’s “worthless”. There are plenty of (quasi DC) applications where you quickly want to set a resistor to some specific value. From the PT100 simulation I already mentioned to balancing some bridge circuit in a measurement setup.
I agree. As with so many specialized tools, they are far from useless. Over the last 20 years, I’ve used, and seen others use, decade boxes in all sorts of places from big metrology labs to small repair shops. Some of the metrology ones had active liquid cooling and temperature coefficients down the low single-digit PPM with matched and serialized test leads. Decade boxes are a simple way to dial in a series of repeatable values.
yeah in my life i can’t see the use for it either.
resistors in my life either i don’t care too much about the value, 1k and 10k alone will cover a lot of use cases. but then the other case is where i don’t care about the value but i do care about the ratio. and like you say, i know the values in my pile… 100 150 220 330 470 680 and the 10x of those, from a jimpak or radioshack assortment i bought years ago. i can usually come up with the ratio i want pretty easily from that.
the one thing in my life i have learned to avoid is tiny trimpots. so convenient to make a fixed-voltage power supply with an LM317, a couple caps, a fixed resistor and a potentiometer. but i left one of those float-charging a 12V (13.6V) lead acid battery for a decade in my humid basement and it developed a degraded spot on the film of the pot (presumably from carrying a not-quite-zero current all those years) and started drifting until it killed the battery. rebuilt it with a ratio of two resistors instead
Absolutely great, but it´s metric. Can´t we have an imperial units decade box instead ?
Well, maybe not in Freedom Units, but you could build one using standard resistor values e.g. 1.1, 1.2, 2.2, etc.
Ah an “Analog CAD Work Station” like on Jim Williams workbench photo from the cover of “Analog Circuit Design: Art, Science and Personalities”
https://books.google.com/books/about/Analog_Circuit_Design.html?id=b-fRKbJ0cC0C&printsec=frontcover#v=onepage&q&f=false
Nice. Coulda also used push-wheel switches (which also come in BCD), like I did on my cap box:
https://www.youtube.com/watch?v=3-NMrHscaXw
11-position switches?! Nigel Tufnel would be proud!