Just for Howie the next part of the saga of a comedian and his (possibly free) power regenerators Analog Corner #308: PS Audio DirectStream Power Plant 20, Thixar amplifier stands | Stereophile.com
OK, I see your point. Unfortunately I do not know the REW algorithm to calculate EPDR. I may try to compare my measured dummy load impedance with the simulated one in MC. However the impedance looks similar with your impedance plots. So the only question is probably the EPDR algorithm.
Last edited:
Holy hell, just awhile ago I was joking here about foreign objects in the AC power "signal path," and now here's this hack doing the same thing without a whiff of irony. Guess I'm in the wrong damn business again.
I understand your scepsis, however I have repaired quite a few 23.5’s with blown outputs, and although short circuits could also be a reason for failure, this seemed never be the case.
Maybe oscillation because of the combination of inductive speaker cable + speaker or voltage peaks or whatever, but at the end the amp broke down.
But it has everything to do in this very case with an inadequate protection circuit to prevent exceeding the Soa.
Hans
Violent agreement on the protection circuit not being up to the job (and we all know high enders love to live dangerously as protection circuits sound bad). But an amplifier that can dump 1HP into a 2Ohm load per channel should on paper cope with whatever can be thrown at it 🙂.
Sorry Bill,
As long as the load is purely resistive you are right.
But as you could see in my sim it’s not too hard to exceed the Soa line with rather highly inductive speaker cables.
And I cannot find the link that quick, but in some paper a Blameless Amp was blown with a specific combination of speaker + cable, resulting in oscillations.
This could be cured by connecting a non inductive 100R resistor across the speaker.
Hans
As long as the load is purely resistive you are right.
But as you could see in my sim it’s not too hard to exceed the Soa line with rather highly inductive speaker cables.
And I cannot find the link that quick, but in some paper a Blameless Amp was blown with a specific combination of speaker + cable, resulting in oscillations.
This could be cured by connecting a non inductive 100R resistor across the speaker.
Hans
It does seem wise, in an excess-of-caution sort of way, to connect a series RC terminator at the far end of the speaker cables. It's probably easiest to do it right at the speaker terminals using banana plugs and jacks (gold plated of course).
Turns out that Nelson Pass and Matthew Polk published this idea in Audio magazine in 1980. Nelson archived a scan of the article HERE
Recommended parts values are: (6 ohms + 47nF): Matthew Polk /// (5 ohms + 100nF): Nelson Pass
Matthew Polk even applied for, and was granted, a US Patent on the idea. .pdf document attached below. Cowabunga!
_
Turns out that Nelson Pass and Matthew Polk published this idea in Audio magazine in 1980. Nelson archived a scan of the article HERE
Recommended parts values are: (6 ohms + 47nF): Matthew Polk /// (5 ohms + 100nF): Nelson Pass
Matthew Polk even applied for, and was granted, a US Patent on the idea. .pdf document attached below. Cowabunga!
_
Bill, I found the paper.
It was written by Cyril Bateman with the title:
Cables, Amplifiers and Speaker interactions.
Hans
It was written by Cyril Bateman with the title:
Cables, Amplifiers and Speaker interactions.
Hans
Thanks a lot Bill! Before saying anything, I will say I have enjoyed reading a lot of what Fremer has to share. Anyway, if indeed he is hearing a problem, remotely diagnosing it from my house is not possible. It is not AC line resistance in the transfer switch, that would have bit him in the form of overheated connections (since the transfer switch is whole-house).
I am 99.9% sure the generator controls are not inserting noise onto his AC, I have troubleshot those issues in the past (connect/disconnect controls & charger) and with Generac, Kohler and Caterpillar systems and not seen any trash. PV systems; that is a totally different issue not relevant here.
It is most likely the installer either made a different or additional ground connection (an additional one at the transfer switch or generator itself not required by NEC) which is causing noise on the ground in his listening room. Ground is not ground is not ground. The earth is a complex 3D grid of crisscrossing currents and a ground rod merely samples the voltage, and feeds current to...anywhere else of differing potential. The only way to have a chance of minimizing ground currents between equipment is to make all grounds in a system equi-potential by defining a central grounding point and independently running all technical grounds to it...like designing grounding inside an amplifier...same principles. The grounds as found in an AC distribution system were designed for safety, not for low noise.
It is difficult to guess what is happening based on the lack of information regarding any factors which would cause the problem. He is concentrating on bought-in-a-box solutions, I would suggest basic troubleshooting instead...like disconnecting or employing line isolators between his cable TV connection and TV, or between the preamp and amp (as a diagnosis). A simple isolation transformer on the AC line would give useful information.
His articles could be interpreted as setups for promoting expensive power conditioning gear. Alternatively he doesn't understand the potential (punintentional) issues, so that is what he is grasping at.
Diagnosing noise issues is boring and relatively low-tech. We achieve extremely black noise floors in studios in commercial power districts with terrible AC power quality by following basic good practice and isolated power systems, filtering if necessary and tidy grounding, not by installing >$10k power regenerators. The studios we install in residential grid areas do not require as much attention...until the customer adds a 10 KW PV inverter and claims weird noises...or his housekeeper (whomever that may be) runs a vacuum cleaner overhead and his single-coil guitar hears it nicely...fixing that is more fun.
Just my 2¢ or less.
Howie
p.s. I am sure many will think that Fremer has such good hearing he is hearing something which the major studio producers and engineers I work with cannot. Maybe so, but that would be another story...
Keantoken, it seems to be obvious that the abrupt changes in EPDR impedance plot from REW are exactly at the points where the phase response crosses zero and changes sign. Don’t you think it is absolutely logical?
It is logical for some mathematical functions to do that. I am aware of it. I don't take that to mean I should make an assumption when I have the capability of actually finding out.
I think what I created is actually more like EADR, or average power, I missed that EPDR is based on peak instantaneous power. I can adjust the simulation to measure peak power instead.
I think what I created is actually more like EADR, or average power, I missed that EPDR is based on peak instantaneous power. I can adjust the simulation to measure peak power instead.
@Howie: I do note that he has his own 'technical ground'
Would be funny if that was the culprit...
Would be funny if that was the culprit...
It would be interesting to measure the currents flowing between the two ground rods. Also to check which is lower resistance to ground.
Mark, thanks for the link to the patent.
Cyril also uses a R/C but in parr. to 100R. The R C values depend largely on the cable/speaker combination, where the 100R is used in all cases.
With only 2Watt he has blown 3 Blameless and 1 Audio Research amp.
Having installed the termination network on the speaker cured all self destruction problems.
http://www.waynekirkwood.com/images/pdf/Cyril_Bateman/Bateman_Speaker_Amp_Interaction.pdf
Hans
I did try the R+C termination with some measurable improvements in MHz EMI coupled voltages, however got rid off it after some time as unnecessary under home conditions.
I do not think that the oscillations issue is tied with the topology, it is rather a question of frequency compensation and PCB design. I had to increase one of the compensation capacitors from simulated 56pF to real world 330pF to stay safe. This reduced the SR, but the stability is above all gold.
I do not think that the oscillations issue is tied with the topology, it is rather a question of frequency compensation and PCB design. I had to increase one of the compensation capacitors from simulated 56pF to real world 330pF to stay safe. This reduced the SR, but the stability is above all gold.
Interesting. I’ve terminated both pairs of speakers (at the spkr end) with 10 Ohm and 0.1 uF) but admit it wasn’t to do with taming any potentially damaging load, just to reduce ringing. I did not measure it - I still need to do that.
While we're in this neighborhood, do you fellas have any recommendations for multiple power amps that don't like having their outputs combined together into long multiconductor cables? In my experience with certain commercial brands this can cause major oscillation, overheating etc., while others seem not to mind it at all. Is there a passive termination that would be effective in squelching this sort of thing?