Simple Attenuators - Design And Testing

[diego_cl]

I have been drilling and tapping the holes in the boxes and then cutting the screws to length, but this is a quite tedious and often "iffy" process! Sometimes fitting a nut on the inside may be a little tough on some resistor brands, but it seems that the holes in the resistors could get redrilled to a slightly larger size, for wiggle room. Alternatively, if the holes are a consistent size, simply tap threads into the resistors, avoiding nuts altogether? I don't know why I didn't think of this sooner!

Of course, its kinda hard to avoid "nuts" like me, in this thread!

Still Screwin'
Gene

With my first build I tapped the threads on the case and used some very soft screws that were very easy to cut almost flush with a fret cutter.

With my current build I tapped the threads using a cordless drill and then fasten the resistors with 5mm SS allen screws. They where 0.4mm too long, but I didn't had to cut them, because the sanding prior to painting was enough. This method was very fast and better looking, but I kind of stripped a thread because of not doing it by hand. Luckily the other 19 threaded holes were so tight that it would be easier to break the allen wrench rather than stripping the threads.

I really like these M3 screws (M4 were too thick for the resistors)

Edit: pictures of the M3 screws on both sides.

 

[Gene Ballzz]

Its really inspiring to see all the tips, inventive suggestions and comments! Lots of really great folks! I think it would be fantastic if everyone who has built one (or more) of these great attenuators were to PayPal/send @JohnH a $10 or $20 bill for all his efforts! He certainly deserves it! This is by far, one of the best DIY threads anywhere on the innernest!
I Love This Place!
Gene

 

[bikescene]

I have a question about the order of the stages in the M2 design. Is the input/output impedance relationship not maintained if the resistive 7dB and 14dB stages (2 and 3) are swapped in order and both are engaged? I came across some posts early on in the thread that alluded to electrical reasons for the resistive stages to be -7, -14, and -3.5 dB respectively.

 

[Marcomel79]

(M4 were too thick for the resistors)

Except for my 100W 30Ohm one!

 

[diego_cl]

Does anyone know an easy and inexpensive way for labeling the enclosure?

This method looks cool, but only if you leave the enclosure unpainted:

 

[Gene Ballzz]

I have a question about the order of the stages in the M2 design. Is the input/output impedance relationship not maintained if the resistive 7dB and 14dB stages (2 and 3) are swapped in order and both are engaged? I came across some posts early on in the thread that alluded to electrical reasons for the resistive stages to be -7, -14, and -3.5 dB respectively.

I think @JohnH found some small theoretical, electrical differences, but he said he felt they were minimal enough to be of little or no consequence. Those of us who've swaped the order seem to have discovered no real world. operational anomalies! It is my "theoretic" observation that the order of stages and resistor number has evolved mostly to retain some semblance of consistancy from the earliest designs to the later versions, with their added stages, etc? I can only imagine what a P.I.T.A. it would be for John to go back and change the numbering of all his earlier posts to new numbering of components and order of stages!

I hope John will correct me, if I am mistaken?

Just Thankin,
Gene

 

[JohnH]

I have a question about the order of the stages in the M2 design. Is the input/output impedance relationship not maintained if the resistive 7dB and 14dB stages (2 and 3) are swapped in order and both are engaged? I came across some posts early on in the thread that alluded to electrical reasons for the resistive stages to be -7, -14, and -3.5 dB respectively.

Good question! have a 'brownie point' for diligent study !

Yes I did put that. One of the key design principles with this attenuator is to keep both input and output impedance as consistent as possible. Ideally this applies to each separate stage, so that it is always maintained no matter which ones are switched on. To make an attenuation stage with a particular attenuation and particular input and output impedance (different values), you need 3 resistors in either a 'Pi' or a 'T' arrangement. But with the values I used, based on measurements of my amp, at about -7db, the leading resistor becomes zero (T pattern), so there's just two. This was the basis of Stage 1 and 2, and it made sense to put them first. Following stages are never first in line, hence so long as they have the right output impedance, their input impedance can vary since only a small part of that difference reaches the amp. So they can also be just 2 resistors.

Ive got analysis set up that will take any version of these designs, and crunch the numbers to run them with any speaker and amp assumption in all possible switch combinations. Based on this i found that there were actually negligible to zero real differences between having the switched stages -14, -7 and -3.5, as compared to -7, -14, -3.5

There's no real need to mount switches physically in the same order as they are wired, but if you do its neater. So the layouts on p 111 and 112 are on that basis.

 

[Gene Ballzz]

A small burr like this can work, if you have a REALLY steady hand. A less than steady hand produces "free hand" results, like in the bottom pic! Functional, but less than elegant! I've not yet tried a punch set. There are some who have made their own water slide decals and buried them under clear lacquer, or such. Others have simply printed on either self stick paper or used card stock and glued it on.
Another optin may be to go to a shop that makes trophies and such and see how much they want to do engrave it professionally? Given the rather low price of trophies, ............? Or a silk screen shop?

Just Labelin'?
Gene



https://www.amazon.com/Dremel-107-E...ocphy=9030801&hvtargid=pla-433019193279&psc=1

 

[JohnH]

Interesting discussion on bolts etc, and rivets!

Personally I quite like the industrial/steam-punk look of vent holes and shiny bolt heads on top. And I think I prefer if the case is 'upside down' at least for Hammond types, since they are slightly tapering, and it gets the hot bits up in the air. Bolts with nuts also allow max wiggle room for getting it to fit, with two clearance holes in play.

But pop-rivets are an interesting idea! We used them when I was studying my engineering, to build test structures for the lab. They do work well if you get it right. They need exactly the right hole in the right place. You can work from the outside, and if you start with good contact, then the fixing of the rivet adds positive compression. It was a good enough joint for Spitfires, Mustangs and Messersmitts to fly around without falling apart. Add some logos to your box to match your favorite airforce? Polished aluminium would look cool, but actually be the worst for heat transfer (OK for 20W studio people) . How about some curvaceous nose-art from a USAAF B17?, Or a big red star from a Soviet Mig15?

 

[Amped eMind]

Thanks @JohnH . As you note, the precision needed for the holes and the lack of wiggle room for the rivets is a concern but if I find it to be a problem, it won't be too late to switch to bolts.

It's funny that you mention the industrial/steampunk look as it is the main vibe I am going for (even though the rivets used then were hammered and not blind aka POP rivets).

 

[Amped eMind]

@diego_cl For the labeling, I will use a cheap set of metal letter punches and paint to fill the stamping and bring out the letters.

The spacing and imperfect alignment of this method is right between industrial and DIY so it may or may not fit in with the desired look and feel of your own project. It is definitely not a professional commercial look.

 

[diego_cl]

Interesting discussion on bolts etc, and rivets!

Polished aluminium would look cool, but actually be the worst for heat transfer (OK for 20W studio people) .

Could you please elaborate a bit more about that please??

 

[sah]

Whew, I started at the beginning of this thread and have finally caught up to present-day! Thanks @JohnH for working out these designs, and for sharing so much interesting info about how and why they work.

I'm putting together some parts orders to build a couple variations and experiment with some possibilities. I've been building tube amps (just for fun), but I can't really turn up even a 5W amp in my apartment, so I only get to hear them at their best when I can take them somewhere more volume-friendly — not great for learning how different circuit variations affect them at gig volumes. With a good 50W attenuator I could maybe even experiment with some Marshall-style circuits!

Anyway, the above is just to say that half the fun for me is experimenting and learning more about what's going on in a good-sounding amp.

To start with, I'm thinking I'll build a switchable combined Resistive/M2/M3 so I can easily compare all the main variations with all other parts identical, for science. (So, one switch shorts around L2/C1, and a second switch shorts around L1.)

I also definitely want a good 4ohm build to use with my 5F1/5F2A Tweed Champ/Princeton style amp that has only a 4ohm output. I play that one a lot and I'm excited to be able to get it cooking at home. I haven't decided yet whether this should be a separate, dedicated build or just a 4ohm input on one of the 4/8/16 designs. Any thoughts on how much difference it makes? I'm worried a minimum of -10db may be a lot for a 5W amp, even in an apartment.

 

[Gene Ballzz]

Whew, I started at the beginning of this thread and have finally caught up to present-day! Thanks @JohnH for working out these designs, and for sharing so much interesting info about how and why they work.

I'm putting together some parts orders to build a couple variations and experiment with some possibilities. I've been building tube amps (just for fun), but I can't really turn up even a 5W amp in my apartment, so I only get to hear them at their best when I can take them somewhere more volume-friendly — not great for learning how different circuit variations affect them at gig volumes. With a good 50W attenuator I could maybe even experiment with some Marshall-style circuits!

Anyway, the above is just to say that half the fun for me is experimenting and learning more about what's going on in a good-sounding amp.

To start with, I'm thinking I'll build a switchable combined Resistive/M2/M3 so I can easily compare all the main variations with all other parts identical, for science. (So, one switch shorts around L2/C1, and a second switch shorts around L1.)

I also definitely want a good 4ohm build to use with my 5F1/5F2A Tweed Champ/Princeton style amp that has only a 4ohm output. I play that one a lot and I'm excited to be able to get it cooking at home. I haven't decided yet whether this should be a separate, dedicated build or just a 4ohm input on one of the 4/8/16 designs. Any thoughts on how much difference it makes? I'm worried a minimum of -10db may be a lot for a 5W amp, even in an apartment.

First> to the forum!
Next> These are cheap and easy enough to build, so why not build one for each variation you expect to encounter, or like me, simply build one to install in each amp?
Just Thinkin'
Gene

 

[JohnH]

Could you please elaborate a bit more about that please??

Sure thing! The colour and finish affects the radiative part of the heat transfer, ie, that which would occur through a vacuum with no air involved. This part, in a system with no fans and of a dark colour, I reckon to be about 40% of the total.

The key parameter is called emissivity. The best colour for maximising heat transfer is perfectly black, with an emissivity of 1.0. Most paints are in the range around 0.7. A bright white paint can be 0.4 or 0.5. But the least of all is a perfect mirror finish, which is 0.0. This does not absorb, nor emit any radiation. So the polished aluminium is not good!

This is why, if your Grandma polishes her silver teapot, her tea will stay hotter, because Grandma knows all about thermodynamics and the importance of emissivity.

 

[JohnH]

Hi @sah , welcome to our thread.

That sounds all good. Without exactly knowing what will suit you best, here are some comments:

If you attenuate a 5W amp by 10db, you have 0.5W. That's still a noisy amp to play in an apartment, and I reckon you may still use a few steps less than that at home. For myself, playing pleasantly loudly in the house, enough to spoil the TV in the next room, I think I'm using about 0.2W. Where there could be an issue though would be if you wanted just a bit less than the 5W, for a jam.

Next comment, I'm really happy to hear about your idea to switch out the reactive parts to explore the differences. That will be very interesting. The resonant part, if built, is potentially the biggest most expensive part. When you also add the extra front end to convert input impedances, if you do the one with one extra coil and two resistors, it's converting the impedance very well from high bass up, but doesn't convert the bass resonance, so you'd only have a half of that resonance in the 4ohm setting. All safe and will sound fine , but just FYI.

The resonance circuit, from testing to date, doesn't seem to make much difference when using guitar speaker, which develops its own resonance in this design. But it may make a difference if run as a load box for direct recording.

Apart from that 4 ohm build, do you have other amps in mind that only have one output tap ohms? Ie, if you built an 8, or a 16, would that potentislly deal with all your other amps? or do you need 8 and 16 inputs in some way?

 

[sah]

Aside from the 4 ohm, all of my other amps at the moment have an 8 ohm output. The one borderline case is my AC15, which normally uses a 16 ohm greenback. I could use its 8 ohm tap and convert to 16 at the attenuator output, but it would add some variables to any tests I might want to do with that one. (Or I could use an external 8 ohm speaker for testing, it just won't quite make the sound I'm used to from that amp.)

I imagine eventually I'll want to take Gene's advice and build a few different units tailored to specific purposes, but for exploration I think something based around 8 ohms with a lot of switchable options is going to be the quickest and cheapest way to go.

 

[bikescene]

Thanks @Gene Ballzz and @JohnH for answering my questions yesterday. I was able to finish mounting all the components and wiring it up. I have the switchable stages set up as -14, -7, and -3.5 dB. I added an additional fixed -7dB stage before the switchable stages. I only got to play through the attenuator briefly with a 22W amp last night, but the volume reduction was a very welcome addition. I will try my 50W Plexi clone for a bit tonight.

I just saw the post about aluminum and emissivity. I'll look into a black painted enclosure if I do another build. I only get to play my amps for about 20-30 minutes at a time, so I am hoping that I am not stressing the big resistors too bad.

 

[JohnH]

Thanks @Gene Ballzz and @JohnH for answering my questions yesterday. I was able to finish mounting all the components and wiring it up. I have the switchable stages set up as -14, -7, and -3.5 dB. I added an additional fixed -7dB stage before the switchable stages. I only got to play through the attenuator briefly with a 22W amp last night, but the volume reduction was a very welcome addition. I will try my 50W Plexi clone for a bit tonight.

I just saw the post about aluminum and emissivity. I'll look into a black painted enclosure if I do another build. I only get to play my amps for about 20-30 minutes at a time, so I am hoping that I am not stressing the big resistors too bad.

It looks great, thanks for posting. Very nice wiring! Are you finding you need that extra stage? On heat, if you can touch R1 for a couple of seconds then it's not too hot, and I wouldn't think it would have any problems with 22W anyway.

My ramblings about radiation relate to a range of colours from deep black 1.0 to fully polished mirror-like 0.0. Dull or raw metal finishes are quite high, like typical paints. Not as good as black for emitting heat, but not so far behind.

 

[Gene Ballzz]

@JohnH
I enjoy your "ramblings!" What is your guesstimate of the radiation properties of the satin black anodized, extruded aluminum cases I've used in the past? At least I think they are are extruded and anodized? I guess I'm wondering if lack of good radiation qualities might be part of the source of why they heat up so much with my 50 watt amp?

https://www.ebay.com/itm/142615103868

Just Curious,
Gene