I started with ECAP, and it was good! However, the slide rule proved it wrong one day. That was enough to teach me caution.
I have, just right after you began your "anti-spice campaign", which doesn't affect me in any way, btw. I know the pro's and con's of either way, as well as the sheer necessity of reality checks -- including reality checks of sims with other similar but somewhat different sims to see wether it's the specific sim or the general concept that fails...GRollins said:
Ahm, would you pls point us to the specific offending posts in the Aleph-X thread? I scanned over about the first 250 post and don't find anyone saying "it must be BS because my spice says so".
Simple-minded dogmatism is not worth discussing IMHO, regardless of the camp it comes from.
- Klaus
Pointless , amusing, insulting, ironic, or offtopic posts here by ANY participant will be deleted.
Bin time will also be distributed freely
as will moderation
Starting NOW!
Maybe we could turn back to discussing the input stage JFet configuration? The simulation issues have been dealt with in detail earlier, either in this or another thread. It really surprised me to see some going down that road again.
I used the horizontal H in the amp I'm currently fiddling with because it was what I was used to and also because I was trying to be reasonably faithful to the original JC-3. Some changes were necessary given that I'm using MOSFETs in addition to JFETs, but the front end bias wasn't one of them.
Once I get the mechanical stuff done--meaning heatsinks and other sundry metalwork--I'll sit down and decide whether to pursue the phono stage I have in mind or to try more variations of the front end of the JC-3.
It quite possible that I may use a vertical H as part of one or both circuits. Or something else entirely...who knows?
Grey
Once I get the mechanical stuff done--meaning heatsinks and other sundry metalwork--I'll sit down and decide whether to pursue the phono stage I have in mind or to try more variations of the front end of the JC-3.
It quite possible that I may use a vertical H as part of one or both circuits. Or something else entirely...who knows?
Grey
However, the JC-3 is NOT the CTC Blowtorch. They do different things, and only share the same devices and only similar topology.
john curl said:
John,
At some point--less important/interesting to me than phono stages and amps at the moment--I'll turn around and do another preamp. Do you have any commentary on the relative application of one biasing scheme or another regarding placement in the system? Obviously, lower noise becomes more important as you move towards the source end of the system and voltage swing becomes more important as you move towards the speaker end...etc., but are there any other characteristics that would prompt you to choose one biasing scheme over another for a given application?
Scott,
What I wrote above is in no way to be read as criticism of you or your work. Proprietary software just isn't a practical solution for most of us unless we write our own. Yes, I've coded in numerous languages over the years, but the investment of time required to write a state of the art simulation program would exceed the number of man-hours that I have available before I punch my final time card and go to the Great Workbench In The Sky. It's just more sensible for me to work with real stuff. And from my perspective, buying parts with the money it would cost to buy a really comprehensive program is a far better investment.
Grey
GRollins said:
This may be off the immediate topic, but there was an extensive discussion of the blowtorch aluminum chassis 'way back in this forum. I'd like to mention my (non-audio) experience with nickel-plated aluminum to mitigate magnetic coupling.
We had designed a highly reliable all-analog 28 vdc to 400 Hz static inverter for one of the next-gen (1970's) airliners. We used Westinghouse NPN alloy transistors that they made for their TV set horizontal oscillators, five parallel per side driving the output transformer. They were matched by Westinghouse into 5 hfe grades and 3 Vbe grades (15 dash numbers). When Westinghouse sold their semiconductor division in 1972 the alloys started to disappear, and we went to faster epi devices. Later, vendor after vendor, we were finally forced to use the double diffused transistors and added base resistors to allow better sharing of the output currents. hfe matching was no longer necessary since we had designed for the low gain of the original alloy devices, and we only needed two Vbe grades.
These faster transistors were now causing coupling into the front end of the analog oscillator circuit. I tried an aluminum shield plate between the output stage/transformer and the sensitive front end with no luck. Then I tried a steel plate, which interrupted the H-field coupling but made the unit overweight. Our metallurgist suggested that we plate the aluminum shield with nickel (I didn't even know that was possible at the time). This resulted in an effective shield for both E- and H-fields. However, due to the low Bmax of nickel, the position of the plate between the output and front end sections was critical. It had to be far enough away from the transformer to prevent saturation of the thin nickel plating.
I wonder if a selective area of nickel plating inside an aluminum chassis might help reduce hum pickup from the power transformer in a one-box preamp? It involves an alkaline zincate process to prep the aluminum for plating. I don't know the cost but it was more expensive than anodizing.
Best Regards, Chuck Hansen
There are many ready to use aluminum zincate salt nowdays, don't have to mix following the handbook formula (like the old days).
Go to car wheel plating shop, after the copper layer (electroless+thick layer) we can plate anything on that aluminum surface.
Go to car wheel plating shop, after the copper layer (electroless+thick layer) we can plate anything on that aluminum surface.
People who are of a mind to mock high end products frequently point to the thick, expensive metal used in building the chassis. Setting aside shielding of the circuit, I've always regarded thick metal as a safeguard against coupling eddy currents (induced by stray magnetic fields from the transformer) into the circuit. Transformers, unless well shielded, will have magnetic fields that pulse for some distance beyond the body of the transformer. This is a trivial observation. Most people stop consideration at that point. However, there's more to consider. An unshielded transformer--or a shielded transformer whose bottom is unshielded--will then induce eddy currents in the chassis. Those currents will then induce secondary magnetic fields in the chassis. Those magnetic fields will then impinge on the circuit. What's the definition of a generator? When a conductor and a magnetic field are in relative motion. If the sound pressure in the room is sufficient, a thinner piece of metal will vibrate; a thicker one will not, or at least not as much. Hence thicker metal will induce less of these parasitic currents into the circuit.
Note that this will not show up in bench testing because it's rare to see speakers used as a test load. If the test load is a resistor, there's no sound in the lab, and the chassis doesn't vibrate. The eddy-induced secondary currents are still there, but at a greatly reduced level, and only at (comparatively low) line frequencies. At higher sound pressure levels, the music itself will begin to modulate the circuit and blur the signal. In mid-fi, it's too much to expect that they would use heavy metal, but if there's anything to my idea it is one of those attention-to-detail things that make high end a separate beast.
Grey
Note that this will not show up in bench testing because it's rare to see speakers used as a test load. If the test load is a resistor, there's no sound in the lab, and the chassis doesn't vibrate. The eddy-induced secondary currents are still there, but at a greatly reduced level, and only at (comparatively low) line frequencies. At higher sound pressure levels, the music itself will begin to modulate the circuit and blur the signal. In mid-fi, it's too much to expect that they would use heavy metal, but if there's anything to my idea it is one of those attention-to-detail things that make high end a separate beast.
Grey
Good points, Grey. It is also important to realize that RFI from 10KHZ to many hundred Meg Hz is in the air, around most of us. Thick aluminum shielding is perfect for this sort of thing.
Yeah, but that's the obvious thing...well, hopefully obvious, anyway. (With some people you never know.) But the chassis can play more than one role. Sometimes I sit and think about the DC resistance represented by the chassis, given that it's used as a ground plane by some folks. No, I'm not saying that's a good idea, but that doesn't stop people from doing so. Again, thick metal is good, although surely something like 1/8" is more than sufficient from a resistance point of view. You're going to have more problems getting the current into and out of the chassis--through the anodizing or surface oxidation of the aluminum--than you'll see from the metal itself.
Another factor is sheer mechanical rigidity. Given that components (particularly capacitors and tubes) can be microphonic, any and all support is going to help the sound quality.
Thick slabs of metal also distribute heat more quickly and evenly if you're talking solid state.
All in all, thick is good for everything except the price...and possibly your back.
Grey
Another factor is sheer mechanical rigidity. Given that components (particularly capacitors and tubes) can be microphonic, any and all support is going to help the sound quality.
Thick slabs of metal also distribute heat more quickly and evenly if you're talking solid state.
All in all, thick is good for everything except the price...and possibly your back.
Grey
From the audiophile perspective, aluminum is still cheap. In the
U.S. it costs as much more to get it anodized.
U.S. it costs as much more to get it anodized.
john curl said:
I talked about this with John yesterday. We agree that if any current-carrying conductor gets within proximity of nickel plating (or any other magnetic material) it could result in a non-linear current flow in the conductor vs frequency or current magnitude. That brings up the recent proliferation of iron or steel leads on metal film resistors.
When a current flows through a wire a circular magnetic field is generated around the wire. If the excitation is alternating current the the crowding of the current to the surface is more pronounced if the material is iron/steel instead of copper.
Also, the end-connections of surface mount parts are nickel.
Best Regards, Chuck Hansen
Hi Chuck,
Just an observation here.
We can chase small effects until we are paralyzed into inaction. Engineering is all about assigning a weight to various problems so that the project does not get sidetracked chasing non-issues. It's also called cost control, and that is not a bad thing.
So, the real question becomes more of "will this have any meaningful effect on performance", instead of "here is a problem and things must be perfect". I have seen designs that have been compromised due to the avoidance of a problem that was not nearly as important as reliability issues or other sonic problems.
Therefore, the effect in question should be studied and quantified. This can occur over the life cycle of a product line. The very high end segment of the industry has the reputation of worried a lot about little.
-Chris
Just an observation here.
We can chase small effects until we are paralyzed into inaction. Engineering is all about assigning a weight to various problems so that the project does not get sidetracked chasing non-issues. It's also called cost control, and that is not a bad thing.
So, the real question becomes more of "will this have any meaningful effect on performance", instead of "here is a problem and things must be perfect". I have seen designs that have been compromised due to the avoidance of a problem that was not nearly as important as reliability issues or other sonic problems.
Therefore, the effect in question should be studied and quantified. This can occur over the life cycle of a product line. The very high end segment of the industry has the reputation of worried a lot about little.
-Chris
Anatech, this is the Blowtorch thread. It started because some people needed to criticize the exotic approach used by CTC to make an 'all out' product.
If it was called the Parasound JC-2 thread, then we would talk about quality vs cost effectiveness, and presume that the power transformer can be mounted internally with minimal shielding and that a thin sheetmetal chassis might do.
The Blowtorch was like an exotic sports car. Like the Porsche 959. Limited production, minimal throughpath, but high chassis stability. This is where the money is in the production of the design.
The Parasound JC-2 is more like a Mazda Miata. Good bang for the buck!
If it was called the Parasound JC-2 thread, then we would talk about quality vs cost effectiveness, and presume that the power transformer can be mounted internally with minimal shielding and that a thin sheetmetal chassis might do.
The Blowtorch was like an exotic sports car. Like the Porsche 959. Limited production, minimal throughpath, but high chassis stability. This is where the money is in the production of the design.
The Parasound JC-2 is more like a Mazda Miata. Good bang for the buck!
It would be nice to hear and measure that miracle. I heard another miracles and they failed against sophisticated engineering designs.
- Status
- Not open for further replies.