THe conclusion about the output coils

[cotdt]

Wow they let me buy it a month ago but not you? Very strange. well, capacitance/inductance meters are dirt cheap these days. mine was $40 on ebay and when i got it i was surprised by the excellent quality and additional features. really good stuff.

 

[dwk123]

Re: Response to the availability of the 11uH Jantzens......

I guess I'll have to buy the 50uH and unwind...... which means I need to buy an inductance meter also.

Okay, dredging up my very stale memory of inductance, doesn't L go as the square of the turns ratio? If so, a single 50uH unwound half-way would drop to ~1/4 value which is what we're looking for. Thus, a single 50uH could be transformed into two ~12uH units with a bit of work and a second wooden core. You'd still probably want a meter unless you could count the turns, though. (although if you had an identical sized core maybe you could eyeball it...)

Right? Did I just embarass myself?

 

[theAnonymous1]

I think I'll just buy an inductance meter and wind my own coils using this calculator....

http://www.pronine.ca/multind.htm

 

[cotdt]

i think it's better to use foils than magnet wire, due to the high frequencies involved.

 

[infinia]

hi
Has any one measured the self resonant freq of these foil air coils used in your amps. I would think they would be less than 1MHz for large wide foil coils. Think about the parasitic capacitance from one winding to another with a wide foil, the capacitance per turn could be estimated from an outside turn to one from the inside and then est. for all turns, series summed to go in parallel with L total. I don't think its really a problem up to 20KHz, but at the class D switching harmonics range it's not gonna be so good here folks. Not to mention the stray fields or EMI issues I have with air coils vs powdered iron toroid designs. I guess if it gets really with bad these air foils, your output caps should start to get warmer.

 

[BWRX]

The problem with air core designs is that the field is not well contained. It can couple to nearby traces and circuitry, and in the case of small amplifier layouts (i.e. most class d designs), the input or signal circuitry; which is a bad thing. Air core inductors with flat foil windings have high interwinding capacitance but low start to end winding capacitance. The start to end winding capacitance is a large factor when it comes to coupling high frequency noise from the switching node to the output, and thus should be minimized regardless of inductor or winding type.

The SRF is a good indicator of winding capacitance, and the higher the figure the better. I would think air core foil wound inductors would have a SRF higher than 1MHz, but not as good as other types.

 

[soundcheck]

Nothing is for free. That's for sure.
In the end it's the sound impact, by looking at all parameters, that counts.

The Mundorf foil coils do generate up to 20db less distortions than
wire-wound coils due to low vibrations.
This has been measured by an "independent" DIY Hifi magazine.

This IMO has a much higher impact than the discussed capacitance
issues.

Cheers

 

[infinia]

I think we all agree that air coils including foil types are the worst performing EMI wise. In addition, depending on foil coil construction, parasitic capacitance can be an additional source of EMI due to feedthru and possibly being non inductive due to SR at higher freq i.e. harmonics of Class D sampling frequencies. The better qualities of foils include possibly lower resistance at audio frequencies and an immunity to saturation. What is traded for this, besides the higher cost of these parts, is increased EMI pollution. The wider foils being worse than narrow types.

Air core inductors with flat foil windings have high interwinding capacitance but low start to end winding capacitance. The start to end winding capacitance is a large factor when it comes to coupling high frequency noise from the switching node to the output, and thus should be minimized regardless of inductor or winding type.

Hi BWRX
I think if one uses air coils esp. foils types the start of the winding (being the inner winding) should be the noisy connection or the input, and the end (outer) should be the quiet side or output connection. But please explain or elaborate what you mean in the quoted part above, in particular "start to end winding capacitance".

.

The Mundorf foil coils do generate up to 20db less distortions than
wire-wound coils due to low vibrations.
This has been measured by an "independent" DIY Hifi magazine.

Hi Soundcheck
Not sure if I understand your intention correctly. Are you saying "low dist. due to vibration" is a first order effect in choke designs in class D amps. I think maybe your unnamed DIY Hifi magazine was measuring in regards to speaker crossover applications?

 

[cotdt]

Mundorf showed that miniscule vibrations on inductors generate a lot of distortion, and this problem is minimized with foil inductors. My listening tests tell me that air core inductors are indeed superior, and I never had a problem with EMI. I did shield my inductors, and also the Tripath chip itself.

 

[infinia]

Ihave done quite a bit of EMI testing. Mostly coming from a design perspective in both power electronics and RF communications. I have tested and reworked designs from inception through qualifications of many various standards inc EMI, RFI, TEMPEST, on-on. I have found it very difficult to shield sources of noise in general. There are not many good materials that are effective against both E&M fields. The best way to minimize EMI is the keep loop areas small i.e. return noise currents very quickly to their origin, esp. keeping alternate paths of said noise currents in regard to leakages and other parasitics under wraps.

IMO As to microphonics of inductors probably a 4th or 5th order effect here,

 

[BWRX]

I think if one uses air coils esp. foils types the start of the winding (being the inner winding) should be the noisy connection or the input, and the end (outer) should be the quiet side or output connection. But please explain or elaborate what you mean in the quoted part above, in particular "start to end winding capacitance".

Agreed that the switching node should be connected to the start (inner) winding. Parasitic inductor capacitances were discussed in another thread (can't remember which one right now), and the two major ones to take into consideration are the interwinding and start to end winding capacitances. The interwinding capacitances are just that, the capacitance between each adjacent winding. The start to end winding capacitance is the capacitance between the start winding and the end winding. In a foil inductor the start and end windings are physically separated by all of the intermediate windings, which results in a low start to end winding capacitance. In a toroid inductor wound the whole way around the core with one layer of windings the start and end windings end up being right next to each other resulting in a higher start to end winding capacitance. Interwinding capacitance is a lot lower than a foil inductor, however.

Mundorf showed that miniscule vibrations on inductors generate a lot of distortion, and this problem is minimized with foil inductors. My listening tests tell me that air core inductors are indeed superior, and I never had a problem with EMI. I did shield my inductors, and also the Tripath chip itself.

Mechanical stability can be achieved with non-foil type inductors by dipping them in varnish. If listening tests tell you that air core inductors sound superior, than the cored inductors you used previously may have been saturating. Different size cores, materials, wire gauge, and winding technique can make all the difference.

 

[cotdt]

Mechanical stability can be achieved with non-foil type inductors by dipping them in varnish. If listening tests tell you that air core inductors sound superior, than the cored inductors you used previously may have been saturating. Different size cores, materials, wire gauge, and winding technique can make all the difference.

This is definately worth trying out. About the saturation though, Tripath's own calculations showed that saturation should not occur with their recommended cored toroid, so I don't think that it saturated.

 

[PWatts]

I'd be careful with foil coils. I used a 6.8uH choke in a full-bridge application, and the standing current was very high (100mA) due to the high capacitance. Switched to a normal wire version, still air-core, and the current dropped more than tenfold. Eventually I settled for a ferrite bar though, its lower DCR made the bass much tighter and the large airgap causes high powerhandling before saturation. However, that was for an open-loop application where the output impedance was of much higher concern than the self-oscillating feedback type ones predominantly used today...