Simple Attenuators - Design And Testing

[tometome]

Hey! Pleased to announce a successful M2 build that works brilliantly!! Thank you so much for sharing this design.. I ended up using a combination of 18 and 20 gusge wire for now however will require it w 18 gusge once it’s built into the speaker box I’m planning on making.

 

[JohnH]

With all due respect, a rheostat is simply a variable resistor. I see your concern with regards to the notion of just having a 'zero to full' attenuation range on the rheostat, but I've encorporated them for the 'fine control' in designs without issues. Will there be a tonal response difference across the dial? Sure, to a degree, depending on the rest of the circuit.

Or am I missing something here?

Our designs use switched stages, for a few reasons:

Passive attenuator designs based on pots or Lpads may or may not sound good enough or acceptable, but as you note, there typically is indeed some tonal difference across the dial. In the basic examples it sucks! But M2 and its related variations go for zero tonal difference across the full attenuation range. As a criterion for this, a close-miced cab at full volume or at any attenuated level, recorded and the recordings normalised, sounds and measures as identical. and can be demonstrated by listening, measuring or circuit analysis and all these have been confirmed by others. The results typically achieve this within about a dB or less. This doesn't set out to account for cone distortion or hearing FM effects, but these get blamed too often for what is really sub-optimal design.

So, having worked out what the target is, testing by circuit analysis leads to values for what the ideal sets of resistors need to be to maintain not only tbe input impedance but also the output impedance (which is key!) of the attenuator. There is no simple pot or Lpad (or pot with other resistors) that can do this across a wide sweep. So we end up with sets of values each specifically optimised for one setting.

Hence switches, and then given the conclusion that only switched settings will nail the desired results at every setting, we use a series of toggles because they are simple, robust and economical. Three toggles, each controlling stages of attenuation x2 the previous, gives 8 closely-spaced settings.

Set the amp volume for best tone, then step it down with the attenuator to get the right volume, and then forget the attenustor and play!

 

[colchar]

Sorry Sir,
But those are the crappy attenuators that don't pay attention to the critical balance of series and parallel resistance in the attenuator and have very little in common with the @JohnH design and generally sound like DOO-DOO and destroy the natural feel and response of the amp! And while yes, the JohnH principal could certainly be developed into a unit with a "continuous knob" control, the technical logistics would be very complicated and cost prohibitive. I swear to all that is holy in rock/guitar/music, the small steps of this attenuator design present no realistic issue! The standard M2 design provides ten small steps between full amp volume and babby sleeping in the next room volume! If retaining all tone, sound, dynamics, feel and response, with truly continuously variable volume control, one needs to step up to a rea-amping unit, like a Fryette, UA Ox, etc, along with the price and operating nuances! I'll take the JohnH all day, every day!

And @colchar , you really should give me a call! Real, human to human conversations can cover a lot more ground than typed messages on a discussin board and in a shorter time frame!

Also Just Sippin'
Gene

Yes, I do need to give you a call. I've just been busier than expected with work so have been posting here when I've taken a moment or two off from grading papers (and yes, I was drinking last night while grading - if you read the papers that I had to read you would be drinking too!).

But I am about to more or less say fuck work because it is a long weekend. I'll get some work done so that I am not even more behind next week (I am already playing catch-up) but the majority of the weekend is mine so I will call you.

 

[Gene Ballzz]

So here is my latest iteration of layout and packaging. It is ready to wire up, test and be done! Labeling is still a thorn in my side and I'm still looking for a more elegant means that I'm capable of pulling off. The P-Touch labels are fine and hold up well, but they still look kinda janky, to me! I'll be talking to a screen printer today, just for poops & giggles! Note that the big honkin' vent holes in the bottom are sized and mounting holes drilled to accept a standard 50mm fan or even four of them!

Also the last box I posted, ended up as a sacrificial lamb, as I learned the hard lesson that regular, clear spray paint does not adhere well to powder coat!

Still Buildin'
Gene

 

[C-Grin]

Looks amazing two thumbs up

 

[JohnH]

Hi @Gene Ballzz , that looks really great! (I think I'd keep wanting to turn it over to admire the grills!)

And I see it's getting the neat 8-16 tweak on the output with the 39ohm. I'll get a schematic diagram to show that.

 

[Gene Ballzz]

Another M2/16Ω done! Still waiting for some labels for the side, that will say: The JohnH on one side and M2/16ohms on the other side! The fan is not mounted/included in this build, it is merely there to demonstrate the fit! I think in the future, I will drill an extra hole for a wall wart input and simply put a plastic/rubber plug in it, if no fan installed! Any comments criticism of the cosmetics and/or layout?

Thanks Folks & Especially @JohnH .
Gene

 

[JohnH]

Looks like a great bit of product design and production engineering!

What's your coil-mounting arrangement? Any space to offset it a few mm up off the base?

Also, since you've got a credible looking fan, shall we try to take the next step with developing an amp-driven self-powered arrangement? To do this, I'd be interested to know what's its dc ohms is, and what dc voltage does it start to spin, and what voltage is needed to spin it moderately but not at max speed. Could rig it up to some AA batteries or wall warts or whatever you have to try different volt levels. When it's spinning nicely at whatever voltage that is, can you measure the dc current through it?

Given all that info, I can work out a first-pass at how to wire it up with a rectifier and resistor/ cap arrangement, then see if the load that it puts on the amp is enough to change our tones.

 

[Gene Ballzz]

Looks like a great bit of product design and production engineering!

What's your coil-mounting arrangement? Any space to offset it a few mm up off the base?

Also, since you've got a credible looking fan, shall we try to take the next step with developing an amp-driven self-powered arrangement? To do this, I'd be interested to know what's its dc ohms is, and what dc voltage does it start to spin, and what voltage is needed to spin it moderately but not at max speed. Could rig it up to some AA batteries or wall warts or whatever you have to try different volt levels. When it's spinning nicely at whatever voltage that is, can you measure the dc current through it?

Given all that info, I can work out a first-pass at how to wire it up with a rectifier and resistor/ cap arrangement, then see if the load that it puts on the amp is enough to change our tones.

I tried it with a fresh 9 volt battery. While it moves a little air and is "whisper" quiet, I don't think we'd want it spinning any slower. I had some used batteries and anything below 7.8 vots, the fan stops! The current rating listed on the fan is 0.08 amps at 12 volts. If you like, I can hook up that fresh battery, leave it running and see how long it lasts? That might give us some kind of ballpark guesstimate?

Don't know if that helps?

As for standing off the coli, the thickness of the end plates of the bobbin is 2.25mm. I've not been able to find any flat head aluminum or nylon screws longer than 1.5 inches. On top of that, all the stainless stuf I find is either listed as "mildly magnetic" and any I put my hands on are certainly magnetic! That's in 18-8, 316 and black oxide stainless! I've currently used 1.5 inch nylon with aluminum nut and not even enough room for a washer! And to get that 1.5 inch, I had to bite the bullet and go with flat blade screw head! Still on the hunt! FWIW, the base is die cast aluminum and the gutter grating is also aluminum. The screws holding the grate are plain jane steel! I didn't expect that to be an issue? I can run out and acquire some stainless screws, nuts and washers, but besides being kinda pricey, they'll be "mildly magnetic" also!

Thanks Again John,
Gene

 

[JohnH]

ok sounds fine. No issue with steel for the screws for other than the coil and you have that covered.

Can you check resistance of the fan as measured by a meter?

 

[yogi.1026]

Hi Gene, For the labelling google water slide decals for pedal box building they will give you some ideas. I think I did my design in Gimp, print onto water slide sticker, I used basic cannon desktop. spray with couple coats of clear to seal the text then wet and apply. Finish with some clear , jobs done!

 

[Gene Ballzz]

Hi Gene, For the labelling google water slide decals for pedal box building they will give you some ideas. I think I did my design in Gimp, print onto water slide sticker, I used basic cannon desktop. spray with couple coats of clear to seal the text then wet and apply. Finish with some clear , jobs done!

I've not had great luck with making my own waterslide decals, especially if they can't get buried under clear, after application. The downside here is that I've not yet found a clear that will adhere to powder coat. I've yet to try lacquer, but I'm not very confident. I'll do a test piece and maybe I'll need to reinvent my wheel! I really hate to go away from the powder coat, as it is so rock solid and durable! I do have a couple references for screen printers who may be able to help me!

Yours looks so nice that I specifically recall it! How did you accurately cut the square holes for the switches

Thanks,
Gene

 

[Gene Ballzz]

With all due respect, a rheostat is simply a variable resistor. I see your concern with regards to the notion of just having a 'zero to full' attenuation range on the rheostat, but I've encorporated them for the 'fine control' in designs without issues. Will there be a tonal response difference across the dial? Sure, to a degree, depending on the rest of the circuit.

Or am I missing something here?

To expound on what @JohnH said in post #2,902, there IS something you seem to be missing here! If the main and only goal is to continuously/variably control the volume of a cranked amp, many of the currenttly marketed units will certainly accomplish that! If however, you want to control that volume, "while still retaining the natural nuances of tone, feel and response while doing so" the JohnH is one of the very few answers to the equation, as that volume reduction is done in a different manner! Yes, there are are then "re-amping" units like the Power Station, Unleash, etc, but those are totally diffferent again in their approach, although most of them are truly stellar and work quite well, AT A HEFTY PRICE!

You're obviously a smart guy, know your stuff, enjoyable to banter with and I don't want to pick nits with you. Understand that I've tried nearly every passive attenuator design since they first came on the market, in the early '70s and every one was a total disappointment, until the JohnH, although some were worse or better than others!

The real downside (and altruistic upside) is that the JohnH design has been developed, tested and perfected in the public domain so that it connot really be patented and or sold to a maufacturing compny! There is no secure and/or ongoing "profit" to be made on this design, other than allowing others to share the enjoyment of the design. A few industrious individuals can certainly build some of these units and collect a fair and reasonable fee for their time, labor and investment of using their own money, up front, to collect, in bulk, the componenets needed! Because of this, I don't see these ever being available commercially, Unless some bean counters in a large company, with sharp pencils, can see a real profit!

For a home garage, under the radar built unit, I see a fair price for a simple, M2 attenuator being in the $300 range. To offer them commercially would bump that price to $500, or more, once you factor in insurance, licensing, taxes, advertising, book keeping and all the other sundries that go along with operating a real business! And then to make them available "Internationally" we get into being RoHS compliant, import/export fees, inspections, and on and on!

Yeah, manufacturing can be a messy business at best (when following all the rules), which is why most manufacturing has been moved to less scrupulous parts of the world with societies who could care less about rules, human rights, child labor, etc, ad nauseam!

Or Well?
Gene

 

[PelliX]

You're obviously a smart guy, know your stuff, enjoyable to banter with and I don't want to pick nits with you.

Likewise, and thank you! A little picking of the nits can be great fun at times. Perhaps I should reduce the picked nits in this particular thread, though.

Fully agree on the economics here, I'm also not looking into manufacturing. Did offer another forum member one the other day but at the price of the parts plus two tenners for the beers/effort.

If however, you want to control that volume, "while still retaining the natural nuances of tone, feel and response while doing so" the JohnH is one of the very few answers to the equation, as that volume reduction is done in a different manner!

True, though I insist that the Weber Mass 'concept' is a very good one, too. A coil is a coil and all that. I think where we might have got our wires crossed [see what I did there?] is that typically the off-the-shelf units offer a 0-100% or 50-100% attenuation level by means of the rheostat. Without some rather complex circuit design and bumping the cost of the BOM way up, that's never going to work very well, as you rightly point out. I was in fact more advocating the use of a rheostat as a fine adjustment, complementing the switches present in the M2 design, for example - not necessarily negating them. Naturally, one might sacrifice the balance of the circuit by using this feature, but turn the knob all the way and the rheostat merely becomes a resistor in the (multi-level switched) circuit. I've since then also considered that a fine grained control is hardly necessary as one's amp already features this.
I respect your practical expertise in the matter and appreciate you taking the time to clarify it further!

EDIT: I'd also be interested in your technical opinion of the Weber's I mentioned. Are there parts of their design that you inherently disagree with? Have you tried any, and if so, how did they line up with other attenuators in your opinion?

 

[JohnH]

hi @PelliX , I can pick up the question about the Webers.

With a good circuit model of a full speaker, we can model most passive attenuators in Spice. Here's my Weber Mini-mass analysis:



On the schematic at the bottom, a full speaker is modelled, and at the top the mini-mass driving a speaker.

The upper plot shows the signal at the full speaker in red, and the other upper lines are what the amp sees driving the attenuator.

The lower lines show the attenuated output at two settings of -10dB and -25dB

The red curve is the ideal tonal reference without attenuator and being logarithmic dB plots, ideally all curves should be parallel to red.

See how on the top chart, the attenuated tones are losing its bass peak and its treble rise? This is the classic issue with attenuators that work with pots and lpads. As you turn down, the low-impedance output of a turned down pot is damping tbe response, leading to dull tone.

The Weber is doing a reasonable attempt at showing the amp a reactive load, but tbe right response is not feeding through to the speaker.

Webers typically have a treble compensation switch. The lower chart shows that, and it does lift the treble back up towards where it should be. But it doesn't replace the bass peak, and it throws the response seen by the amp way off and so the amp can't respond as usual.

Our M2 is a lot closer to being consistent in all these areas. It doesn't need a treble work-around, and it let's the real speaker develop its own response by showing it a consistent and amp-like output impedance.

 

[Barnsley Boy]

Another M2/16Ω done! Still waiting for some labels for the side, that will say: The JohnH on one side and M2/16ohms on the other side! The fan is not mounted/included in this build, it is merely there to demonstrate the fit! I think in the future, I will drill an extra hole for a wall wart input and simply put a plastic/rubber plug in it, if no fan installed! Any comments criticism of the cosmetics and/or layout?

Thanks Folks & Especially @JohnH .
Gene View attachment 112078 View attachment 112079

@Gene Ballzz , Haven't swung by the forum for a while, but just had to say that is a cracking build. I would have been inclined to flip it upside down to show off those grilles - love that look. Is that plasterers expanded metal mesh? And how did you form the large holes? I have a hole coring bit, but I think it's more for wood so I'd be worried about totally mullering the metal.

 

[Gene Ballzz]

@Gene Ballzz , Haven't swung by the forum for a while, but just had to say that is a cracking build. I would have been inclined to flip it upside down to show off those grilles - love that look. Is that plasterers expanded metal mesh? And how did you form the large holes? I have a hole coring bit, but I think it's more for wood so I'd be worried about totally mullering the metal.

Thanks for the kind compliment! Coming from you, with your stellar builds it is truly appreciated!

The mesh is aluminum and I believe intended as rain gutter guard. The plaster and/or stucco stuff is steel and the edges of the diamonds are more sharp, as well as the diamonds being smaller.

The holes were made with one of these (any size you choose, but mine is 1 3/4 ") and the hole is perfect for a standard 50mm format fan:

And one of these is your best friend, when drilling/cutting any holes larger than 1/8" in aluminum:

https://www.amazon.com/gp/product/B01L2XR4P2/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

I'd like to know how you neatly cut the square holes in your build?

Again, Thankin'
Gene

 

[Barnsley Boy]

I got the square/ rectangular hole using a dremel to cut out roughly, then some painstaking filing. It helps if the switch/ digital thermometer has a small lip around it to hide any rough edges. I did something similar recently on this to enlarge the bridge pickup hole to convert from open to covered pickups. Now that was a proper pain to get right. Wish I'd had the de-burring tool.

 

[Gene Ballzz]

I got the square/ rectangular hole using a dremel to cut out roughly, then some painstaking filing. It helps if the switch/ digital thermometer has a small lip around it to hide any rough edges. I did something similar recently on this to enlarge the bridge pickup hole to convert from open to covered pickups. Now that was a proper pain to get right. Wish I'd had the de-burring tool.

Yeah, that de-burring tool is the schnit! I've used one bit forever on aluminum and plastic. Just a little bit of steel chewed up two bits, in no time flat! Still worth it though, for what it does. It even gets right into the corners of square holes! One of the handiest tools I own, along with this:

https://www.woodpeck.com/ppr-2019.html

I've only got a 6 inch in inches and yeah, it's a bit pricey, but also priceless in its function!

Enjoy,
Gene

 

[Barnsley Boy]

Oohh I like that a lot! One day, when I have my dream workshop maybe I could justify adding it to my fantasy list of tools!
...... but until then!