voltages and regulators

[peterg]

I'm trying to understand voltages. I've built a few boxes using the PT2399 (more accurately the DT2399) chip – Madbean's Cave Dweller Delay, Frequency Central's Little Angel and working on Merlin's Equinox Reverb. The chip requires a 7805 regulator to get the voltage to and keep it at a constant 5V. Other chips such as OpAmps use 4.5V which is achieved by reducing (halving) the 9V supply with resistors and capacitors. Correct?

Why not make all chips run off 4.5V? Manufacturers should be catering to stompbox building, right? Seriously though isn't 9V a common power source for a lot of electronics and halving it seems to be more straight forward and economical than using a regulator and the parts that go with it? Or is it that a chip such as the PT2399 can't cope with any voltage fluctuations but OpAmps can?

[R.G.]

Good question. And good insight that clustering things for a given power source is a manufacturing advantage.

The problem is that digital chips like the 2399 and practically everything else with a clock going to it is a descendant of the 5.0V semiconductor logic family, not the 6.3V tube-radio family. And the logic family is too gigantic to care what would be most effective for biases for 9V powered things. The advantage to the semiconductor makers to have logic run on 4.5V instead of 5.0V is much smaller than the advantage of not having to change the semiconductor processes that make all that logic run on 5.0V. It would cost them money to change, and the extra they'd make is vanishingly small.

Historically, the first electronic power supplies were batteries, and these came in multiples of the 1.5V carbon-zinc cell. This give us the 6.3V power supply for tube filaments, the 12.6V supply that's twice that, and other multiplies. However, this market never used (that I know of) three-cell units to get to about 4.5V, although the Japanese-radio invasion of the 1960s did popularize the six-cell 9V battery. The echos of the carbon-zinc family are still here, in the frequent use of 6V (and 6.3V), 12V/12.6V, 24V/25.6V, and so on for lots of equipment. When the makers of transistor based logic came up with the competing logic families of the early 1960s, they wound up standardizing on 5.0V. I don't know the exact reason this happened, but I do know that different logic families used different voltages, from the 1.6 to 2.0V of Integrated Injection Logic (I-squared-L) to the 3 to 18V of CMOS. Texas Instruments popularized Transistor-Transistor-Logic (TTL) and they won. TTL used 5.0V. At a guess, it may have been because their semiconductor ovens cooked fast logic circuits that worked at 5.0V better than at other voltages. Maybe.

In any case, the history is too long and well entrenched in the semiconductor processes that run in $1B - $10B wafer fabs. Until analog electronics powered by 9V batteries and equivalents rise to be a significant fraction of the total semiconductor market, it's not going to happen. And the tide is running the other way. The share of the market that uses voltages other than logic voltages, those being 5.0, 3.3, 3.0, and going smaller is actually shrinking, not expanding. We're like an ant hitching a ride on an 18-wheeler truck. The truck driver wouldn't mind that we're there, but he/she is not going to do much to make it easy or convenient for us - if he/she even knows we're there.

I guess another way to look at it is that whatever the logic is for making 9V and 4.5V powered stuff, the market is too small to notice. I'm just glad we can still get analog stuff at all. Through hole stuff is rapidly going away, for the same reasons.

[armdnrdy]

Also,
Take notice which BBDs have been "reissued" by other manufacturers.

The "low" voltage type as dictated by the "market."

[peterg]

R.G.

Thanks for the historical and technical information! I'm an electronics novice so it will take a while for it all to sink in!

Armdnrdy

I ordered PT2399s but got DT2399s instead. I've read about issues on forum strings on these "reissues" but haven't had any problems myself. Then again I don't know what my delay and chorus would sound like with PTs in them.