John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Radio Frequency Characteristic Impedance
I'm half a day late, but this Jim Brown paper covers it well.

I thought we covered this before. Taking the exact RF characteristic impedance equation and substituting the values per unit length and taking the limit predicts the correct behavior for any frequency. In the limit the reflection steps blur into exactly the lump L/C solution (at audio lengths R and G are normally secondary). Look back I hate to think it was all a waste of time.
 
I thought we covered this before. Taking the exact RF characteristic impedance equation and substituting the values per unit length and taking the limit predicts the correct behavior for any frequency. In the limit the reflection steps blur into exactly the lump L/C solution (at audio lengths R and G are normally secondary). Look back I hate to think it was all a waste of time.

It's always forgotten.

My next task is to get steppers to run at high speed. Now, they are using 100 feet of 16 awg twisted pair, and for "some reason", the motor can't go fast enough. Seems a 2 ohm load at the end of 100 feet of 150 ohm cable has a problem.

Turns out the mismatch is preventing the motor torque, as the current can't rise fast enough. The settling time exceeds the pulse rate. And, that's just simple steps, never mind micro stepping with pwm.

I'm having the guys replace one cable with a 6 pair 20 awg cable, with three pairs paralleled per motor coil. Granted, the pairs are not orthogonal, but I suspect they are spaced sufficiently such that the characteristic impedance is down in the 50's.

If that isn't sufficient, I'll have them use all 6 pair for one coil, run another cable, and be done, 25 ohms hopefully will suffice.

Next worry is the sheer quantity of cable. They've purchased a very very large quantity (did I mention very?) of the standard twisted pair, I hope that the number of drives needing speed is small. I'd hate to purchase another thousand kilometers of cable(assuming 50 ohms fits the bill).

We discussed using simple 50 ohm coax, but the drives run +/- 170 volts in bridge mode, the cable jacket is not rated for significant voltage to ground, even though we use the core at 5 kv. At least the 6 pair is 500 volts.

Luckily, my compatriates here no longer try to explain to me why it cannot work, apparently seeing it do so tends to quiet them.

Of course, audio is far more complex...

J
 
Ok - simple subjective opinion then. That's fine - everyone is allowed an opinion... :)
Not simple. This is not some "opinion" I pull from a back looking cavity. You want to trivialize what I say because you probably are happy with FLAC and redbook. I was in your camp up until the last decade. It took me 35 years to figure this stuff out. I don't treat audio circuits like a pill or an equation. I look at your pedrigree on this forum and find nothing but empty comments. You don't list where you work and what you do outside of this forum. So your posts are
simple subjective opinion
unless you can provide some references.
 
The issues with line impedance for short loudspeaker cables are how they affect the power amplifier's feedback. If you have a discontinuity this can drive many feedback amplifiers to apply correction to the signal at the amplifier end of the cable which is not exactly what is really going on at the loudspeaker end. As even modest amplifiers still have gain at frequencies where the line impedance comes into play this can be an issue. Note that many high end loudspeakers have zobel (constant impedance) networks on the tweeter. That also assists in keeping the behavior from getting nasty.

Now it is certainly possible to design amplifiers that do not suffer from those issues and some are deigned that way.

The problem is to correctly identify what issues are really in play and then address those issues.

I read about a guy using Hypex Ncore based amps who had nasty, nasty audible noise whenever he used speaker cables shorter than 11ft, with his Magico speakers that use an "elliptical" crososver, which according to Alon Wolf just means extremely high db rolloff if I understood him right (music was playing when he told me).
 
Not simple. This is not some "opinion" I pull from a back looking cavity. You want to trivialize what I say because you probably are happy with FLAC and redbook. I was in your camp up until the last decade. It took me 35 years to figure this stuff out. I don't treat audio circuits like a pill or an equation. I look at your pedrigree on this forum and find nothing but empty comments. You don't list where you work and what you do outside of this forum. So your posts are unless you can provide some references.

See that's just it, it's exactly an opinion. You're asking for references and not providing your own. I don't really know what else to tell you here. Nor do I trust your generic statements and attempts to "reason by experience" (nor anyones).

If you have meat, please provide. If not, perhaps it's best to say, "hey, this is what I like and make my business."
 
JN
I would send you some of my magic cable. But I only had a sample run made for now. In quantity it will run about $3,500.00 per 1,000' roll. Oh yeah also a bit more than 200 pounds per spool. So I suspect your solution is a bit more cost effective.

I do trust you won't explain in public the techniques used in the construction.

ES

BTY zobels?
 
Last edited:
See that's just it, it's exactly an opinion. You're asking for references and not providing your own. I don't really know what else to tell you here. Nor do I trust your generic statements and attempts to "reason by experience" (nor anyones).

If you have meat, please provide. If not, perhaps it's best to say, "hey, this is what I like and make my business."
Copy the three words in my signature and go the search engine that stole it's name from Milton Sirotta. Start reading.
 
In short, "Hey, this is what I like and make my business."

As I said.

Please don't try to tell us you have some sort of magical "truth" and improvements. You're selling something because you like it and believe in your product. No shame in that.
No magic! Hard work, lots of listening, constant playing with circuits. I also bugged lots of EE's ((like you) See, I didn't make you serve me your creds. I looked at them myself ;^) ) at my past places of employment. Brad Plunkett, for example, was a big help in the birthing my product.

But the finer points of my design are all mine! Not even top EE's were able to help. And I did ask!
 
JN
I would send you some of my magic cable. But I only had a sample run made for now. In quantity it will run about $3,500.00 per 1,000' roll. Oh yeah also a bit more than 200 pounds per spool. So I suspect your solution is a bit more cost effective.

I do trust you won't explain in public the techniques used in the construction.

ES

BTY zobels?

I require low smoke, zero halogen, radiation resistant, tray rated, UL listed, and at least 300 volt capability. So far, only helucabel has product that comes close to the bill, the flame testing approval is done in house. Their drag chain cable tray stuff is pretty good as well, we are using their mega flex stuff.
Early on, we would call the companies with our requests and specs, they would make custom cables for us. First, they would caution us with either a kilofoot minimum, or say that the nre was going to be large. Pfffft.

We were the first large scale machine to require LSZH, I suspect NEC will use our experience towards code mods in the future. The vendors learned a lot building all our cables.

I may yet have to design a low impedance double braid coax at about 12 awg and 15 to 20 ohms as our high speed magnets are about 100 feet from the rack. We can't put the electronics in the tunnel, so the cables are the only viable solution. The new Brazil machine and the new one in California can use it as well, so someone like Belden could see it as a new product. Each machine requires about 30 miles of it. Oh, this is a fast feedback system, designed to keep the beam from bouncing about. When it bounces at the quarter micron level, they tend to call us in to fix it. Brazil wants 10 kHz bandwidth, so they will be calling us..

As to zobels, I don't know yet how the stepping motor amps handle what they have, if the researchers will stick to full steps, or if they have to use pwm micro stepping. It may require transient damping, so a combo r/c may be needed.

J
 
Last edited:
JN

Is that all? So what you need is obviously simpler than specialty audiophile cable. :)

Mine just has to be tray rated for 300 volts and sunlight resistant. As to getting custom cables, it requires a lot of hand holding to get the important funny stuff right. I think I mentioned they had more concerns about colors than the parts I though were important. There was also one issue they just couldn't get right. But now that I have something close we can modify it to show them a sample of the changes needed.

So how does your cable sound? ;)
 
The source output voltage is 10 mV it is delivered to the load of 1,000,000 ohms by the cable under test. The voltage drop across the cable under test is what is measured. It is around 2nV. As there is no connection between the source or load the only connection is across the wire under test. So any common mode voltage is due to leakage. The source is of course powered by batteries.
What kind of resistor, have you tried different types ?.

Dan.
 
AX tech editor
Joined 2002
Paid Member
Riight. That could be read at least 3 ways. I am going to be charitable and read it as marketing story.

To his defense I must say his prices and the tech data, although limited, are a fresh wind in hi-end audio circles.

Robert, your website is irritating, impossible to look at anything seriously as the pictures are constantly changing. Unless it is designed to prevent prospective customers to get a good look? :cool:

I just watched your YouTube from T.H.E. show, admirable effort to explain in layman's terms the tech workings, even I could follow the gist of it.
But the customer was ready to part with his money way before you were finished with your pitch ;-)

Jan
 
Last edited:
Status
Not open for further replies.