Have not researched it recently since I'm trying to put off for as long as possible having to build a separate computer system to source audio. 🙂 Not clear that the quality of shelving is that great on software land, in any event.
In any event, having the pre with low impedance output drive a shelving treble boost would be the bees knees. I mentioned my hi-F fall-off to my audiophile papers who are about my age or older 59 and my desire for treble boost and after laughing (because we are now old enough to admit our faults and limitations), they all agreed it would be a good thing.
After some informal discussion, almost none of use would use bass controls. Midrange is a better bet but is more complex as the network has to do filtering at highs and lows. We all kinda agreed that a treble shelving boost (and shelving is the key word here) would be the best.
This is probably the last I will say of this in this thread as it is going OT. Please forgive me.
No worries about the OT. It's no the first time this thread has strayed away from the regularly scheduled program.
You don't necessarily need a separate computer to serve audio. I use the Mac Mini in my lab/office to serve audio through an Apple Airport Express (AX). The AX has TOSLINK out. I run that through a DAC and a pair of MOD86 amps. Remote control is via an old iPod Touch. The app on the iPod controls iTunes on the 'puter. Works quite well, actually.
I use the computer for other things as well. All my circuit boards were designed on it...
Sonos would be another option. They're coming down in price.
Tom
You don't necessarily need a separate computer to serve audio. I use the Mac Mini in my lab/office to serve audio through an Apple Airport Express (AX). The AX has TOSLINK out. I run that through a DAC and a pair of MOD86 amps. Remote control is via an old iPod Touch. The app on the iPod controls iTunes on the 'puter. Works quite well, actually.
I use the computer for other things as well. All my circuit boards were designed on it...
Sonos would be another option. They're coming down in price.
Tom
Measured leakage inductances well above 1000uH are commonplace for AC-output "wall-warts" such as the picture below. It happens in all sizes; I've measured huge inductances on both high power (~ 30VA) and low power (~ 1VA) wall-warts.
_
Attachments
Last edited:
Well that's a shame, maybe you will change your mind some day. I haven't finished my test build of the original Mod86's I bought yet, and not ready to invest in the PAR86. I do plan to either bridge/parallel multiple Mod86s (I have about 20 LM3886 chips lying around, another reason not to go PAR86), but not sure I will do it with your board or attempt my own layout (eek).
Maybe you are right that not too many folks wanted more power, but there were a few complainers at the onset of this thread about cost/power and I was just thinking except for power, the Mod86/PAR86 is kind of an ideal amp in my mind.
Boba
Umm, most Baxandall implementations I've come across use an inverting buffer and hence introduce a 180 degree phase shift. That can be OK if all signals route through it but can be a problem if selectively applied to channels in multi-way, 5.1, 7.1, or other configurations favouring a lot of power amps. Otherwise, yeah, the phase bop/wobble/wave/ripple/whatever you want to call it is often 10-30 degrees.
However, Bill and I were discussing shelving filters in DSP. Usually these are implemented in a biquad patch bay offering a variety of EQ flavours from the Bristow-Johnson cookbook. Unlike a Baxandall, this often encourages folks to select only the high or low shelf desired rather than implementing both. Even if both are selected the topology remains distinct as they're two separate filters in the chain. Sometimes the UI allows Q (slope) selection, sometimes not, so they may or may not be first order.
Good to know. Happen to have a link to distribution data? I'd be curious to get a sense of typical versus upper bound values.
No. Heck no. Even if you had perfect advance knowledge of the appropriate probability density function to fit to your measured data, you'd still need at lest 30 datapoints and ideally more than 300 datapoints, to get a reasonable estimate of the center parameter and of the spread parameter. (often called the "location" and the "width" respectively). I sure as heck have not measured and recorded leakage inductance for that many units of any given transformer part-number.
And I would suggest that it is not "obvious" which pdf to use. Leakage inductance is not directly measured during manufacturing test. So it is neither controlled nor observed. (Observability and Controllability were the cornerstones of effective manufacturing test when I took the class...) Pew. Pew pew. P.U. No thanks.
Last edited:
I have looked at Baxandall implementations as well, some interesting input impedance variations that I don't really like. Inverting design was necessary for fail safe operation when potentiometer wipers wear out or have not so ideal contact. At some point you also need to get the polarity back as well.
A good starting point for analog preamp building is Douglas Self, "Small Signal Audio Design". There's a whole chapter in there on tone control if I recall correctly. There's another chapter on volume control. It's a worthwhile read.
Tom
Tom
Tom, I am considering putting your amp into a box with 1/4" TRS inputs and outputs. What is the preferred mounting and what kind of specs and characteristics should I be looking for when choosing the sockets. Here are two I came across on Ebay as examples:
example A
example B
I've read your thoughts about parasitics so figured you would the best person to tell me what to look for in my choice.
example A
example B
I've read your thoughts about parasitics so figured you would the best person to tell me what to look for in my choice.
This type are useful in that the signal lines can be muted to ground when no plug is inserted.
Dan.
MaxHeadroom
Thanks for the input. I'm assuming you mean chassis ground.
tomchr
I've always been a fan of best bang for the buck and Neutrik connectors are only affordable in bulk. Given the high resolution capability of your amp designs, I'd hate to introduce crap on the signal that your amp is likely to amplify. Please elaborate on what makes the neutrik connector superior so that I can seek these attributes in a competing Chinese product. Of course, the alternative would be someone who happens to have a stock of Neutriks that would let me have a set at bulk prices
Thanks for the input. I'm assuming you mean chassis ground.
tomchr
I've always been a fan of best bang for the buck and Neutrik connectors are only affordable in bulk. Given the high resolution capability of your amp designs, I'd hate to introduce crap on the signal that your amp is likely to amplify. Please elaborate on what makes the neutrik connector superior so that I can seek these attributes in a competing Chinese product. Of course, the alternative would be someone who happens to have a stock of Neutriks that would let me have a set at bulk prices
A metal shell around the connector that is electrically connected to the Chassis seems to be important. It helps maintain the Chassis shield around the signal conductors all the way to the other Chassis where the same shielding effect of the metal shell is continued into that Chassis.
It's called the Pin 1 ground problem....Google it.
Pin 1 should be bonded 'directly' to chassis, and circuit local ground seperately bonded to chassis using minimal resistance/inductance connection.
In practice, with very short direct connection from xlr/6.5 socket shield connection to chassis at each connector all should be well.
The problem arises when multiple connector ground connections are soldered to a long earth track common which then connects to chassis using single jumper wire connection.
The inductance/resistance of the earth common track and jumper wire is the source of the problem.
Dan.
Pin 1 should be bonded 'directly' to chassis, and circuit local ground seperately bonded to chassis using minimal resistance/inductance connection.
In practice, with very short direct connection from xlr/6.5 socket shield connection to chassis at each connector all should be well.
The problem arises when multiple connector ground connections are soldered to a long earth track common which then connects to chassis using single jumper wire connection.
The inductance/resistance of the earth common track and jumper wire is the source of the problem.
I mean that shield/pin-1 should be bonded directly to chassis, and tip/pin-2 and ring/pin-3 short to circuit ground when plug is removed causing muting of + and - phase inputs.
Dan.
Last edited:
Wanted: experienced UK person to source, stuff, solder 2 x Modulus86
An acquaintance in the UK would like to have two modulus-86 boards purchased, stuffed, assembled, and soldered by someone knowledgeable and experienced. Anyone interested in helping, please either reply here (s/he is a DIYA member and reads this thread) or by Private Message to me, whichever is most comfortable for you, and I will put you in touch with the final end-user. S/he is a music lover not a parts sourcing expert and soldering wizard. I personally don't know whether the Modulus-86 or the Parallel-86, is wanted. You and s/he can hammer that out when you communicate. I feel confident s/he is willing to compensate you fairly for your time, effort, and expertise.
An acquaintance in the UK would like to have two modulus-86 boards purchased, stuffed, assembled, and soldered by someone knowledgeable and experienced. Anyone interested in helping, please either reply here (s/he is a DIYA member and reads this thread) or by Private Message to me, whichever is most comfortable for you, and I will put you in touch with the final end-user. S/he is a music lover not a parts sourcing expert and soldering wizard. I personally don't know whether the Modulus-86 or the Parallel-86, is wanted. You and s/he can hammer that out when you communicate. I feel confident s/he is willing to compensate you fairly for your time, effort, and expertise.
I would be interested in having a look at and measuring Tom's creation, but they don't really fit in what I am doing right now, so give me pm. Will do it at cost plus a bottle. Traveling till 7th, live in the Netherlands, so shipping nominal cost.
You said you're looking for these for both inputs and outputs. The 1/4" TSR is okay for line-level connections, but I don't like 'em for speakers (despite seeing such use on guitar amps and even PA gear).
There's the usual "universal" banana jack connectors and spring-clip connectors, it seems every stereo or "home theater" amp or speaker uses one or the other. I use 14-gauge solid wire ends on my cables that connect well to both of these. They appear to work okay, but I've heard about Speak-On as a high-current (rated 30 amps) two or multipin connector designed specifically for amplifier outputs and loudspeakers.
SpeakOns are apparently rare or unknown in hifi/home audio and I can only wonder why they're not used, other than the inertia of other things being seen as "good enough." The ultimate example of this is the original RCA phono connector (originally used underneath the turntable chassis to connect the cable that goes to the tube amplifier in those really old, pre-stereo record player/radio consoles), still in ubiquitous use for unbalanced line-level connections. I even remember cheap stereos that used RCA connectors for the speakers.
Here's a gratuitous rant about 1/4" phone connector use and abuse, vs. SpeakOn:
Speakon vs 1/4" connections - Peavey Forum
Regardless of the connector, you want the "signal" and "ground" of the speaker connector to go straight to the "speaker output" and "speaker ground" connections of Tom's board, and do NOT connect the "ground" of the speaker connector to the chassis.
There's the usual "universal" banana jack connectors and spring-clip connectors, it seems every stereo or "home theater" amp or speaker uses one or the other. I use 14-gauge solid wire ends on my cables that connect well to both of these. They appear to work okay, but I've heard about Speak-On as a high-current (rated 30 amps) two or multipin connector designed specifically for amplifier outputs and loudspeakers.
SpeakOns are apparently rare or unknown in hifi/home audio and I can only wonder why they're not used, other than the inertia of other things being seen as "good enough." The ultimate example of this is the original RCA phono connector (originally used underneath the turntable chassis to connect the cable that goes to the tube amplifier in those really old, pre-stereo record player/radio consoles), still in ubiquitous use for unbalanced line-level connections. I even remember cheap stereos that used RCA connectors for the speakers.
Here's a gratuitous rant about 1/4" phone connector use and abuse, vs. SpeakOn:
Speakon vs 1/4" connections - Peavey Forum
Regardless of the connector, you want the "signal" and "ground" of the speaker connector to go straight to the "speaker output" and "speaker ground" connections of Tom's board, and do NOT connect the "ground" of the speaker connector to the chassis.
I look at connectors the same way I look at tools. I can buy a cheap knife. It will be OK, "good bang for the buck" at first, but quickly dull. Then it slips and cuts my finger rather than the tomato I was trying to slice. I can buy a good knife. I'd have to pay a bit more, but it stays sharp, is worth getting sharpened when it does eventually dull, and will slice the tomato rather than my finger.
Same with connectors. I've had cheap connectors that didn't make a good connection. They didn't fit the published dimensions, leading to mechanical rework. In some cases, the plastic was of a type that would melt easily when wire was soldered to the connector. No good. The Neutrik NCJ6FI-S I mentioned above is $3.35 @ QTY = 1 at Mouser. That's less than what many people spend on a cup of coffee. It's also about half the cost of the LM3886...
Dude... Just buy the Neutrik. You won't regret it.
Tom
SpeakON connectors are ***AWESOME***. That is a well-engineered connector and they're actually not that pricy. About $2.50 for the chassis connector and $5 for the cable connector. They're the obvious choice if you're building a multi-channel amp for a multi-way active XO speaker, such as the Linkwitz LX521 as they're available in 2, 4, and 8-pole versions.
I suspect the reason the banana plugs and spring clip connectors are still found on audio gear is that they work well enough and can be connected to without the use of tools. I've never been particularly fond of either.
Tom
I'll be happy to elaborate on the tradeoffs between the two and am always available via email (please use the Contact Us form on my website) or PM. Email is preferred as I check that several times a day.
I'm also willing to deliver assembled and tested modules, though it sounds like there's an offer of assembly in exchange for a bottle (of wine presumably) on the table already. I can't compete with that... Not on price anyway... 🙂
Tom