Which Oscilloscope for a pickup maker?

[Steve Conner]

I used Solaris and Linux quite a lot in the past. Then I changed to XP both at home and at work, where I'm developing industrial test equipment based on an XScale platform running Windows CE.

For personal use, when I saw Vista, I promised myself that I'd change to either Linux or Mac as soon as Microsoft dropped support for XP. This year I was impressed enough by MacOS that I bought one of the new 13" MacBook Pros, although I'll admit that the new Ubuntu seems to be the best MacOS rip-off ever. I recently helped a colleague install it, and it does the two-fingered scrolling thing out of the box!

I still keep my old laptop around with XP, in case I need to use Sound Forge or program some PICs. Or use my old Pico PC-based scope, which requires a parallel port. Remember those?

When it comes to oscilloscopes, I think an old analog Tek takes a lot of beating. I have a military surplus R7603 that I got on Ebay for about $200. It has custom chips, but so far none of them have blown.

If I had to live with a digital scope as my only scope, it would be a Tek TDS210, TDS220, or their replacements, the TDS2012 and 2014. I managed happily at my previous job with just a TDS220 on the bench. We had a 400MHz Philips analog scope, but it only came out of the cupboard in emergencies.

The Owon DSOs are cheesy Chinese clones of the TDS2xx. GW Instek make something similar, I'm not sure if it is a rebadged Owon.

I would hate to be stuck with a PC-based scope, I find the user interface on them a complete pain in the butt.

Your mileage may vary, as I'm an EE and heavy scoper

[David Schwab]

You should try htop. It's just like top, but better.

I just ran that for an example. I don't use the terminal much anymore.

But I generally have this open all time.

[belwar]

I do ALOT of overclocking and I've been dying to get my hands on a nehalem.. just been too busy to devote time to it. I have about ten Q6600 which is the basic Quad Core 2.4 with the (266.6) 1066 fsb. Most of the computers I did the standard conductive pen trick and uped the FSB to 333 making the processor 3.0ghz x 4 cores. But on my private machine i run the FSB much higher turning the 2.4 processor into a 4.0 processor.

Why so much power you ask? BOINC.. I just seti@home on all the machines.

What's cool about the nehalem is that they somehow figured out a way to make MMC useful. The hyperthreading actually works! The processor itself is only four cores, but it will properly process 8 threaded at ones.

What really amazes me if that now we are starting to be able to use our GPU's for general computer calculations.. THAT is fast. A Q6600 core @ 3.0ghz will process a seti work unit in 20 minutes. An nvidia 260 will do it in 4 minutes.

No, they aren't.

But here's the funny part... Apple was always boasting the power of the IBM/Moto PowerPC processor, and people would say that Macs sucked for gaming.

Now all the major game consoles use either a PowerPC RISC CPU or a variation on it, and Apple switched to Intel processors! So the XBox 360 is closer to a PowerMac than a PC. Same with the Wii, but then Nintendo as using something similar to a G3 CPU since the Game Cube.

And that's OK, because the 8-core (Two 2.26GHz Quad-Core Intel Xeon "Nehalem" processors) Mac Pro is just stupid fast.

IBM wasn't interested in Moto/Freescale's Altivec vector processor, and IBM just wasn't getting the clock speed any higher.

Meanwhile I'm typing this on an eight year old G4 with upgraded CPU.

[David Schwab]

This year I was impressed enough by MacOS that I bought one of the new 13" MacBook Pros, although I'll admit that the new Ubuntu seems to be the best MacOS rip-off ever. I recently helped a colleague install it, and it does the two-fingered scrolling thing out of the box!

The latest version of OS X, 10.6, a.k.a. Snow Leopard, does multi finger-gestures, plus a lot of other stuff, and it's 64-bit. The upgrade is only $29 (it's a full installer).

Wont run on my PPC dinosaur, because it's Intel only.

Don't forget, you can run Windoze on your Mac!

I still keep my old laptop around with XP, in case I need to use Sound Forge or program some PICs. Or use my old Pico PC-based scope, which requires a parallel port. Remember those?

Not since I had a PC... which was in 1994.

But I still have a SCSI card in my Mac, which is now only attached to a never-used Iomega Zip drive. Remember SCSI?

[David Schwab]

I do ALOT of overclocking and I've been dying to get my hands on a nehalem.. just been too busy to devote time to it. I have about ten Q6600 which is the basic Quad Core 2.4 with the (266.6) 1066 fsb. Most of the computers I did the standard conductive pen trick and uped the FSB to 333 making the processor 3.0ghz x 4 cores. But on my private machine i run the FSB much higher turning the 2.4 processor into a 4.0 processor.

Hey, I'm on a machine that had a 466 MHz CPU originally! It's running 1 GHz now, with a 133 MHz bus. But I do have a 1.5GB of RAM. So it doesn't suck... yet. But I need a new machine ASAP.

Why so much power you ask? BOINC.. I just seti@home on all the machines.

I used to do seti@home. I currently think SETI is a misguided effort, but I guess it can't hurt.

What's cool about the nehalem is that they somehow figured out a way to make MMC useful. The hyperthreading actually works! The processor itself is only four cores, but it will properly process 8 threaded at ones.

Apple have been working with Intel on what they wanted from MMC processors, so thy might have contributed to that. They have something called Grand Central Dispatch for for allocating tasks across multiple cores and processors. That's why they have been getting the new chips first. Apple had a lot of input on the PowerPC CPU too, since they were part of the AIM Alliance (Apple-IBM-Motorola).

What really amazes me if that now we are starting to be able to use our GPU's for general computer calculations.. THAT is fast. A Q6600 core @ 3.0ghz will process a seti work unit in 20 minutes. An nvidia 260 will do it in 4 minutes.

Yeah, the new GPUs have gigaflops of power, sometimes just sitting there. Apple introduced OpenCL on Snow Leopard to redirect data to the GPU.

[Possum]

....

Salvarsan how are you driving the pickup coil, how do you hook up to this box? It looks interesting and price is cheap enough, I'm interested...

[Possum]

...

Yikes I just read the installation instructions for the software on Mac platform, really complicated.....

[soundmasterg]

I've got a couple Tek 2235 100 MHz dual trace scopes for amp work that work just great for what I'm doing. I have no idea if they use custom chips or not, but I'm not worried about it all that much. I got them at a very good price of $150 for the pair 5 years ago so I'm way ahead of the game. I used to have a B & K of some sort that was still dual trace, but older, and I sold that on ebay. Tek scopes are nice.....

greg

[Possum]

...

I was thinking of getting that thing but then hooked up FuzzMeasure again, truth is most humbuckers look exactly the same to a sweep, just higher or lower curves by small amounts. Can't tell a PAF from the others. Not much use I guess...

[Steve Conner]

I never understood the point of SETI@home. Seemed like I would just be increasing my electricity bill and wearing out my computer to do someone else's work for free.

I know that Intel Macs run Windows, but considering I bought a Mac to get away from Windows... Mine came with Leopard, but I really should do the Snow Leopard upgrade.

Yes I remember SCSI, I had a SCSI card in my PC to communicate with my old Yamaha sampler. It worked fine but was slow as hell, the sampler's fault as far as I knew.

If you can't tell the difference between a PAF and other pickups by means of measuring instruments, what implication do you think that has as regards telling them apart by ear in a properly conducted test? Emperor's clothes anyone?

[David Schwab]

I never understood the point of SETI@home. Seemed like I would just be increasing my electricity bill and wearing out my computer to do someone else's work for free.

I just don't think they are going to find anything using the current methods. They are assuming too much... that some other civilization is using radio, that the radio is in a band we can pick up, and that we will understand that it's a signal. It's all based on our current technology.

It's like during WWII when the German's were trying to see if England had RADAR, and didn't detect it, because it was not in the frequency range they expected!

Also by the time any signal would get here, it would be from a very long time ago. So there's a good chance that a signal might be out there, but didn't get here yet.

There is plenty of compelling evidence that non human intelligences have been interacting with us for centuries. That's what they need to be looking into. Problem is that it's outside our current understanding of reality. They keep thinking that if we are going to be contacted, it will be with nuts-and-bolts means... radio, spaceships, etc. Time works against all those things, so a really advanced civilization would find another way that doesn't involve space-time.

I know that Intel Macs run Windows, but considering I bought a Mac to get away from Windows... Mine came with Leopard, but I really should do the Snow Leopard upgrade.

Oh I know, but it's an option if needed. Leopard is nice... except for Spotlight. I liked the version in Tiger better. Leopard is the end of the line for this PPC Mac. But it was running Mac OS 9 when I got it, so I think I got my money's worth.

Yes I remember SCSI, I had a SCSI card in my PC to communicate with my old Yamaha sampler. It worked fine but was slow as hell, the sampler's fault as far as I knew.

My current Mac came from a time when Apple dropped built in SCSI in favor of USB and Firewire. But I had several SCSI devices like a UMAX scanner, a Zip and Jaz drive (aka disk eater), and an external hard drive. So I had to get a SCSI card. Now it's not really used at all.

If you can't tell the difference between a PAF and other pickups by means of measuring instruments, what implication do you think that has as regards telling them apart by ear in a properly conducted test? Emperor's clothes anyone?

Good point. But I also think we can hear things that can't be measured... yet. That kind of gets us back to the SETI thing....

[DrStrangelove]

If you can't tell the difference between a PAF and other pickups by means of measuring instruments, what implication do you think that has as regards telling them apart by ear in a properly conducted test? Emperor's clothes anyone?

Since you can't tell the difference between pickups by measuring DC resistance
or inductance, why bother with any electrical measurement?

I don't mind the infrequent Straw Man Argument, Steve, but would you
kindly stop trying to stir shit amongst presumed adults?

Before you start throwing your shoes into industrial coil winders,
let me point out that test equipment's justification is the same as
it has always been here:
to identify defective units within the same production batch,
whether by DC resistance, inductance, or (horrors!) by using
a scope and signal generator.

The impetus for a using a particular piece of test gear varies,
but raw curiosity sounds very good to me.

-The Immoderator

[salvarsan]

Salvarsan how are you driving the pickup coil, how do you hook up to this box? It looks interesting and price is cheap enough, I'm interested...

Haven't you already done this kind of thing with other soundcard signal analysis software?

Anyway, you get a scope+sig_gen for more than one purpose -- fixing guitar amps is just one other possible use.

The impedance plots needed a resistor ~1.5k and a few clip leads.
The schematic is:



The SysCompDesign site has a paper on doing this kind of network analysis at:
http://www.syscompdesign.com/na-theory.pdf

-drh

[David Schwab]

Oh... now I have to start saving up for one of those.....

[Possum]

...

No big deal on the comment about PAFs being indistinguishable from other buckers on a frequency sweep. Its why everytime I get interested in that test method I quickly abandon its usefulness. I guess in reality it is a primitive method of looking at audio, its just a line on a graph. All humbuckers share the same parts to a frequency sweep it is just showing what they all typically do, which is a slow curve following a predictable path, period. It doesn't show harmonics, and it can't tell you how anything "sounds." Kinman who makes very high tech noiseless strat pickkups etc. says the same thing, spectrum analysis and sweeps etc. aren't really all that useful in the end...