Would star quad arrangement of the cores achieve significant attenuation?
Would the resulting doubling of cable capacitance be an issue?
Not if the cable is only 0.5m long!
Quad-star... dont know how well it contains the EM fields. should be tested for EMI. Who makes sample lengths to test?
But, cable C IMO is not an issue for speaker cables.... there is plenty of current available to charge the C very fast. Ls will be a larger issue. However, with line level cables, it is just the opposite as usually the opamp drive or whatever doesnt have as much current avail and Ls is small X/Z compared to load z.
THx-RNMarsh
But, cable C IMO is not an issue for speaker cables.... there is plenty of current available to charge the C very fast. Ls will be a larger issue. However, with line level cables, it is just the opposite as usually the opamp drive or whatever doesnt have as much current avail and Ls is small X/Z compared to load z.
THx-RNMarsh
Star Quad, such as microphone and interconnect made by Belden should be excellent in terms of both inductance and capacitance. In rough terms, twisted pair, whether shielded or not, has a characteristic impedance in the range of 100 ohms, so I'm temped to guess that the star quad might have a characteristic impedance in the range of 50 ohms - but this is a real SWAG on my part.
Since the star quad that is available that I am aware of is for microphone or interconnect, I think the disadvantage might be its resistance. I'd guess the equivalent gauge would be no better than 18.
Cheers,
Bob
Many of Neslon Pass's class-A power amplifiers omit the output inductor. To name one example, his 1978 "A40" amplifier has a zobel network but no output inductor (link). Same goes for a lot of his other amps.
Why? Is he crazy to leave it out?
Is everybody else crazy to put it in?
Or is the output inductor an optional component, which (given sufficient design genius) can be designed-in, or designed-out, either way?
Why? Is he crazy to leave it out?
Is everybody else crazy to put it in?
Or is the output inductor an optional component, which (given sufficient design genius) can be designed-in, or designed-out, either way?
For actual lines made of copper (or even silver) at audio frequencies, figures like 50Ω or 100Ω are essentially meaningless: below the characteristic impedance transition frequency (mainly determined by the L/R ratio), the impedance rises, is not real anymore and tends towards ∞/-45°.
The transition happens somewhere between a few KHz and a few hundreds KHz, depending on the amount of copper.
Maintaining a consistent characteristic impedance for the whole audio spectrum (down to 20Hz f.e.) would require at least superconductive materials, and probably some additional precautions
Last edited:
Elvee,
I did not intend to suggest that impedance matching on speaker cables is desirable, or that constant impedance across the band is necessary. My mention of the estimated impedance was simply informational, and meant for comparison to typical coax. Specifically, the capacitance was a concern, and the fact that the impedance of the star quad was not so low as to suggest undesirably high capacitance.
Cheers,
Bob
In reply to Mark Johnson:
All you have to ensure is that with a worst case capacitive and inductive load, your amp still retains sufficient phase and gain margin to remain stable.
In Nelson's case his amplifiers (not familiar with A40 circuit details - is it ZGF?) they tend to be low feedback types, so the HF poles fall close to or below the 0 dB gain point. This type of amp can tolerate nasty loads without an output inductor much better than high loop gain types.
OTOH, if we talk about a classic VFA with high loop gains, the HF poles fall well above 0 dB, and you really have to think about overall phase shift and compensation design. Most designers (me included) would shoot for 60 deg phase margin in an audio amp. If you add an inductor on the output of this type of amp, you have unconditional stability - you can drive any load.
The Zobel in designs that do not have an inductor are there primarily to compensate for cable L at HF, and inevitable speaker L.
There is a lot of discussion around terminating both ends of the speaker cable. I have not tried this as of yet - something I need to look into.
So, to summarize, no genius required. Just engineering.
All you have to ensure is that with a worst case capacitive and inductive load, your amp still retains sufficient phase and gain margin to remain stable.
In Nelson's case his amplifiers (not familiar with A40 circuit details - is it ZGF?) they tend to be low feedback types, so the HF poles fall close to or below the 0 dB gain point. This type of amp can tolerate nasty loads without an output inductor much better than high loop gain types.
OTOH, if we talk about a classic VFA with high loop gains, the HF poles fall well above 0 dB, and you really have to think about overall phase shift and compensation design. Most designers (me included) would shoot for 60 deg phase margin in an audio amp. If you add an inductor on the output of this type of amp, you have unconditional stability - you can drive any load.
The Zobel in designs that do not have an inductor are there primarily to compensate for cable L at HF, and inevitable speaker L.
There is a lot of discussion around terminating both ends of the speaker cable. I have not tried this as of yet - something I need to look into.
So, to summarize, no genius required. Just engineering.
Everybody needs an output inductor except one person? Remarkable. The (link) in post #66 includes a schematic and a design analysis. Anybody who wants to, can determine the amount of global feedback in the 1978 Nelson Pass amplifier called "A40". Perhaps your own personal analysis may conclude that the A40's second stage is a CE with current source load, having enormous voltage gain. Or perhaps not, it's up to you.
My EL34 tube amp doesn't need the output inductor for stability, but to reduce any Rf coming into the feedback path via the speaker cable, I broke the negative feedback path into two R's; a 100R and a 6kohm (or so - can't remember exactly - doesn't matter here). The output goes thru the 100 ohm and then a small cap to ground, and then continues thru the "real" feedback resistor (the 6k), to the negative input of the earlier stage. My front end 6SN7 is actually outside of any negative feedback loop. It looks real good on the scope with a 10kHZ squarewave (very little over or undershoot). I'd use the same technique in a transistor circuit if I were building one.
Try holding the EMI meter in line with the driver pole pieces
Dan.
Yes, getting anywhere near a speaker cabinet increases the field intensity a Lot. I watch the meter dancing to the music beats.... Largest peak levels are with bass (highest current), of course. Behind the equipment where there are a bunch of wires... ac power, signal lines, speaker and digital is going to be interesting.
18AWG is fine for a test of quad construction comparing ratio of EMI intensity of 'standard' zip cord of same gauge.
A 50/75/100 ohm speaker cable is perfect for testing against RFI pickup etc as the test equipment is calibarted for those Z's. BTW -- calculations of Romex(R) ac power wiring used in USA buildings/homes is about 100 Ohms.
THx-RNMarsh
Decades ago, I took the coil/core assy out of a standard car ignition coil.
I built a portable amplifier box with meter, speaker, and headphone out.
The coil was mounted in a length of pvc pipe and connected via length of coax.
This made a really good tool for sniffing out radiated aether disturbances, and taking a listen to these disturbances....much better than just a meter display.
This setup was sensitive enough to pick up the driver pole pieces radiations from the seated listening position
.Interestingly the magnetic signal sounded better than the acoustic signal, and seemed to arrive earlier.
I feel that focusing the drivers directly at the listener does more than just focussing the acoustic output.
Sensitive enough to pick up electric rail pantograph supply lines variations from 80m away also
.The 'throttle' being applied by the driver was perfectly audible.
I could monitor the station arrival and departure times and compare to the published timetable.
The modern world is full of these kinds of aether disturbances.
Living as I do in the northern most parts of Perth, I have ample opportunity to spend time in radio/ac power 'quiet' places, and note the difference.
I should build an updated version....it would be re-educational and fun.
Alarming too probably, to acoustically listen to the junk magnetic disturbances throughout the modern home/workplace/test bench.
Switch mode supplies would be worth a close listen and correlate to acoustic damages in audio gear.
Cassette tape heads make a useful pickup coil....it's interesting to listen to the internal currents of an amplifier stage.
Dan.
Last edited:
Indeed, it's up to to the designer as to whether an inductor is used or not, and how much 'risk' they are prepared to tolerate wrt load.
I think the chase to eliminate the output coil is a matter of chasing 'purity', the same reason to want to go with a class A amp. But it limits topology / loopgain choices for a design. Topologies and loopgains that would otherwise result in better performance. Luckily, a coil is not an active device and doesn't generate harmonics. I was against coils also in the beginning, but realized later that they are not harmful to performance in the audio band.
I think coils offer more good than bad in the end.
I think coils offer more good than bad in the end.
I also live most of the time in remote area where it is rather quiet in all respects. No noise coming in on the ac utility lines but plenty of noise generated within the home... more so now with digital and smps in many items used.
Your idea to use a cassette tape head is a good one. I have used a VCR tape head for wider bandwidth pickup. Epoxied to a BNC allows a cable to be attached to it and other end to test equipment (spectrum analyzer). It is how and why I developed ac line filters for everything. In this case of learning about the amp/spkr interface, the 'probe' is small enough to get inside the amp without direct contact and see how effective shielding, Common-Mode and Diff-Mode coils and RC filters on the OPS effectiveness can be.
Pls go ahead and try it again. I'm in S.E.Asia for a little while goofing off. WoW! Fireworks over by the river.... great view from my 30th floor!
THx-RNMarsh
Yeah, that's what I was alluding to.
Thanks.
Good idea/implementation, what sensitivity do you get...VHS heads have only a few turns iirc.
Btw, the ignition coil pickup I described had a laminated iron core, with 300 or so primary turns, and 5000+ secondary turns.
I removed the primary and used just the secondary.
The straight rod core made the pick up very usefully directional.
Good info. These are the kinds of areas that I intend to revisit/explore.
Yes, for now I figure my D&N test set can perfectly easily give me balanced input up to 80dB+ amplification, plus metering and monitoring/measurement output ...I can try a range of pickups, including your video head idea.
Another battery portable version would be especially handy.
I got to see fireworks myself tonight....I worked a Katy Perry show....great views there too
.Lots of staging, effects, dancers etc, but she does sing very nicely....pity mega PA systems do vocals so badly.
Dan.
Last edited:
In reply to Mark Johnson:
All you have to ensure is that with a worst case capacitive and inductive load, your amp still retains sufficient phase and gain margin to remain stable.
In Nelson's case his amplifiers (not familiar with A40 circuit details - is it ZGF?) they tend to be low feedback types, so the HF poles fall close to or below the 0 dB gain point. This type of amp can tolerate nasty loads without an output inductor much better than high loop gain types.
OTOH, if we talk about a classic VFA with high loop gains, the HF poles fall well above 0 dB, and you really have to think about overall phase shift and compensation design. Most designers (me included) would shoot for 60 deg phase margin in an audio amp. If you add an inductor on the output of this type of amp, you have unconditional stability - you can drive any load.
The Zobel in designs that do not have an inductor are there primarily to compensate for cable L at HF, and inevitable speaker L.
There is a lot of discussion around terminating both ends of the speaker cable. I have not tried this as of yet - something I need to look into.
So, to summarize, no genius required. Just engineering.
It is certainly true that some amplifiers are more able to operate into a variety of loads without an inductor. There is also the matter of a judgment call as to what kind of worst case load will ever be encountered by the customer.
In addition to global feedback loop stability, some capacitive loads can cause local instability in the output stage; remember, emitter followers often do not like to see a significant capacitance across their output.
Even John Curl agrees that a 1uH inductor is inaudible, so there is not much penalty in erring on the safe side with a small inductor - as long as the inductor is properly implemented.
Zobels at the speaker end of the cable are a good thing, although few do it. They help prevent un-terminated line effects at high frequencies. The impedance seen from an un-terminated line can be nasty at some frequencies. I have a plot in my book that illustrates that. Maybe some speaker manufacturers should include a Zobel across their input terminals if they present an uncontrolled high impedance at RF.
Cheers,
Bob
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.