Let's get real - How good is good enough?

[StarTrucker]

I see the average GDIY'er on this forum as falling under two categories :

1. The professional engineer or hobbiest looking to DIY the tools they use to make their records. Diy enthusiasts who are interested because they want the sounds they hear on other records or in their head that they want to put on tape.

2. The electronics guru who may or may not be a musician or engineer/producer. These people really know their hardcore theory and graciously share with us as part of the community. Career engineers and techs.

Of course there is overlap and that's why we all share so much.

In terms of design, how good is good enough for the regular guy (1. above)? Someone in a thread mentioned dialing in a regulated power supply for tube preamps. He suggested considering power transformer impedance and lots of technical data-sheet level details. Are these kinds of details from the pros over the top for average builders who need reliable studio grade equipment? Where is the line between optimal performance and day to day usability for a professional record.

Ex. A 3 or 4 stage RC PSU filter section and a bleeder resistor is "regulated" and common in lots of classic gear. When does it become unnecessary to include more high tech and complicated designs? Are the old ways good enough to get by on any major record?

 

[ruffrecords]

Ex. A 3 or 4 stage RC PSU filter section and a bleeder resistor is "regulated" and common in lots of classic gear. When does it become unnecessary to include more high tech and complicated designs? Are the old ways good enough to get by on any major record?

As usual, the answer is it depends but in simple terms the answer is when it is class A. In a class A system, the total power consumption is fixed and, with adequate decoupling, this means the supply current is pretty much constant. It does not matter what the signal level is, the total power drawn is always the same. Essentially class A circuits divert current from and active device to and from a load. This means you can get away with a multiple stage RC smoothing circuit and still be confident of the operating conditions of the circuit.

As soon as you move to class AB or B (99% of all op amps and most guitar amps), the total power consumption is directly related to signal level. With no signal, the current draw can be quite small but the bigger the signal the bigger the current draw. This means the output of a multiple stage RC smoothing circuit will drop as the output signal increases. This results in the well known "sag" in class B guitar amps. For pro gear sag is not acceptable so you need a regulated power supply to ensure the operating conditions of the circuit do not alter with signal level (unless you want them to of course).

Cheers

Ian

 

[daschnoz]

I've run into this same question in a few places in life, my last job for example. The engineers were worried about the number 5 places to the right of the decimal point, when in the real world application, +/- 100 was close enough.

I think it's easy to identify where high precision is needed and where close enough is close enough. After that, it's a matter of personal preference. We're not trying to send any of this gear into orbit, so no one is going to scold you for overbuilding the thing. If you want to spend the $$ on super precision components, go for it.

I take this approach:

What is the worst tolerance component in the circuit? Anything else in the circuit can be within THAT percentage of the calculated design value and it will be fine.

Tubes... they're +/- 10% devices on a good day. IMO, using 1% resistors in a tube circuit is overkill; 5% is more than close enough.

Most of that "vintage" gear was built with 10% carbon comp resistors, which have higher temperature drift than carbon film. So with the drift, it could be off-value by even more, and yet we love our vintage gear as unmolested from the factory as possible. So if I were building that circuit today, I'd use 5% resistors and call it a day.

In a power supply, unless you use some kind of regulator circuit, you're always going to be within 5%, simply because of fluctuations in the line voltage. Target is 120Vac. I've measured as high as 127 and as low as 109 in the summer, when everyone is running an air conditioner. So even with that adjustable bleeder resistor set to give you your exact target voltage, that's only at the line voltage - now. Measure it again later today or sometime tomorrow and it will likely be a bit different.

 

[ccaudle]

Are these kinds of details from the pros over the top for average builders who need reliable studio grade equipment?

That depends on whether you are building a proven design, or trying to work up a design from scratch. If you need reliable equipment, then yes you need to understand device parameter variation, line voltage variation, effects of temperature, etc. to make sure that the circuit operates reliably under all the conditions you might encounter.

Ex. A 3 or 4 stage RC PSU filter section and a bleeder resistor is "regulated"

Decidedly unregulated, the output voltage varies in direct proportion to input voltage and to load current. It helps to understand basic principles when evaluating.

Are the old ways good enough to get by on any major record?

The old ways are good enough to get old style results. Whether that is good enough for "any" major record is context dependent. Rock and roll records have different standards than jazz and classical. Horses for courses, as they say, or in other words it is up to the craftsman to know how to choose and use the appropriate tool for the job.

 

[C.B. - Boudio]

For me personally, I work quite in a 'punk' way, in that I use whatever I can get my hands on, whilst trying to get it as decent as possible (unlike those so-called Indie artists that the internet is paved with, in their quest for the lowest of lo-fi).

For tube equipment, then, I don't get the point (either) of using those low-tolerance components, regulated supplies,... It just seems all so overkill, compared to what they made in the sixties (it is of course probable that the designers back then did take power transformer impedances and such into account, since many of those companies made everything-in house).
That being said, I have an OD3 regulator valve lying around somewhere, which seems splendid to include in a compressor - something I'll definitely do.

I think the 'old ways' ("V72 level") are usually enough; they already made more than adequate classical recordings in Germany during WWII, and I don't think that there is any possible sonical improvement to speak of for a John Coltrane record.