Simple Attenuators - Design And Testing

[donwagar]

Thanks JohnH. I checked my local shop, they carry Neutrik NYS219s in stock. They look right, so I'll replace all the jacks when I finish the footpedal. Easy fix. I'm sure I have room on the footswitch jack, if the amp and speaker jacks are tight I can look into isolating the existing jacks.

 

[JohnH]

Someone that actually knows what they are talking about will probably chime in. But I don't think so, I do that when I'm setting my volume, also I'm building a -3.5dB footswitch to cut volume on the fly at gigs.

Hi @donwagar and @MarkS2 , yes I agree, the always-on first stage, together with the continuous non-switched connections through R4, R6 and R8 through to the speaker, plus the fact that you are only flicking one switch out of three at any given instant of time, protects the amp from seeing an open-circuit transient even though the toggle switches are likely to be non-shorting.

This is very different to the full bypass switch, which for an instant, leaves the amp fully open with no load which may make high transients unless the amp is off or on standby.

 

[JohnH]

Thanks John, so if I only need an 8ohm out, which is the simplest diagram/version to get to grips with, on the plus side I do have a decent iron/solder & a multi meter & I have a lot of experience building hi-fi amps including PCB stuffing, so maybe I should start to understand in a small way how what i'm doing

That sounds like you could get your head around this! if you sketch out the parts to scale, then you can draw a wiring diagram for yourself and check it three times to make sure things are connected where they should be.

 

[nzingler24]

Simple Attenuators - Design And Testing

Thanks @JohnH all parts have been ordered now

 

[MarkS2]

Hi @donwagar and @MarkS2 , yes I agree, the always-on first stage, together with the continuous non-switched connections through R4, R6 and R8 through to the speaker, plus the fact that you are only flicking one switch out of three at any given instant of time, protects the amp from seeing an open-circuit transient even though the toggle switches are likely to be non-shorting.

This is very different to the full bypass switch, which for an instant, leaves the amp fully open with no load which may make high transients unless the amp is off or on standby.

Thanks for the confirmation. That makes a lot of sense.

I finished putting mine together tonight, and will hook it up, take some measurements and test it out tomorrow!

 

[JohnH]

hi Mark, that's a nice open build, and very neat graphics on it.

 

[Graham G]

That build of Marks attenuator looks exactly the spec. I want/need for my ORI50 with 8ohm Speaker & I can easily follow what he's done to build one myself, regardless of this excellent work Mark.

 

[MarkS2]

hi Mark, that's a nice open build, and very neat graphics on it.

Thanks! I tested it out this morning and it works perfectly! What a difference for getting better sounds out of this amp.

By the way it's got a cover, it just wasn't in that photo. On that note, the sheet metal is pretty thin so I don't know how it'd behave thermally for higher wattages (I'm running 15W), but I left the R1 on the back face with room for a heat sink which could be a later upgrade.

Edit: also by the way, those graphics were applied simply with a piece of cardstock run through my laser printer, then attached with some spray adhesive. Quick and dirty but highly effective!

That build of Marks attenuator looks exactly the spec. I want/need for my ORI50 with 8ohm Speaker & I can easily follow what he's done to build one myself, regardless of this excellent work Mark.

If you're looking at that pic, 2 things to be aware of: I omitted the extra jacks, and I have a bypass switch wired like JohnH sketched up in this post #1866

 

[dbishopbliss]

A little background... I have an amp with two speaker outputs - one for 8 ohms and the other for 4 ohms. It is wired so that if you plug in two 8 ohm speakers it will parallel the two cabinets for a 4 ohm load.

Right now, I have an M2 and a M3 for 8 ohm loads. I have used both plugged into the amp. It seems to work. Any reason you can think that I should not do this?

If not, I think I'm going to build a dual M2. That is, use DPDT switches with connections for M2 on one side and the other M2 on the other side.

Either that, or is there a 4 ohm design (I didn't see one but the thread is long)? Maybe I could build it so that I could plug both speakers into the attenuator and they would be paralleled and then one cable to the amp. That saves me the cost of the resistors and an inductor.

On other hand, the attenuator would be less useful. I could always use just one side for 8 ohm loads.

 

[JohnH]

A little background... I have an amp with two speaker outputs - one for 8 ohms and the other for 4 ohms. It is wired so that if you plug in two 8 ohm speakers it will parallel the two cabinets for a 4 ohm load.

Right now, I have an M2 and a M3 for 8 ohm loads. I have used both plugged into the amp. It seems to work. Any reason you can think that I should not do this?

If not, I think I'm going to build a dual M2. That is, use DPDT switches with connections for M2 on one side and the other M2 on the other side.

Either that, or is there a 4 ohm design (I didn't see one but the thread is long)? Maybe I could build it so that I could plug both speakers into the attenuator and they would be paralleled and then one cable to the amp. That saves me the cost of the resistors and an inductor.

On other hand, the attenuator would be less useful. I could always use just one side for 8 ohm loads.

I reckon all of that would work!

You can plug your two attenuators into the amp for a combined 4 ohm load, (and set the levels of each speaker differently if you want to)

You can use just one M2 or M3 into an 8 Ohm amp output, then use it with one cab, or two 8 ohm cabs for a 4 Ohm load out of the attenuator but still using the 8 ohm amp tap

You can build a 4 Ohm version. All resistors and inductors are half their value as compared to an 8 Ohm build. But if you were wanting an M3, you would also double the capacitor, probably with several in parallel. Power ratings of resistors stay the same

 

[JohnH]

Thanks! I tested it out this morning and it works perfectly! What a difference for getting better sounds out of this amp.

By the way it's got a cover, it just wasn't in that photo. On that note, the sheet metal is pretty thin so I don't know how it'd behave thermally for higher wattages (I'm running 15W), but I left the R1 on the back face with room for a heat sink which could be a later upgrade.

Edit: also by the way, those graphics were applied simply with a piece of cardstock run through my laser printer, then attached with some spray adhesive. Quick and dirty but highly effective!



If you're looking at that pic, 2 things to be aware of: I omitted the extra jacks, and I have a bypass switch wired like JohnH sketched up in this post #1866

View attachment 94066

Looks great and great news that its working fine. With all the space in the box and a 15W amp, Im guessing it warms up just mildly?

Since Graham G and others may be referring to it, I've got a couple of general points:

In my schematic, I show the order of stages as -7db reactive, -7db, -14db, -3.5db then to the output. There are small electrical reasons for that choice of order. The -7 stages are the best balance of impedances so i put them both at the front. It actually works pretty fine with any order, but I slightly prefer it that way. It also allows, with different switch wiring, to use -3.5db on its own, which I show on design M (post 1) and in my own build, but I took it off M2 and M3 for simplification, since its rarely needed.

But anyway getting to the point: You can wire it per the diagram, but its probably more logical to physically place the switches with -7 in the middle, eg -14, -7.-3.5. Then its like binary numbers, where each digit is 2x the previous.

On choice of cases, steel vs aluminium: Steel is much stiffer and stronger, but cant be cast easily for cases, so its made by folding. Hence steel cases are much thinner than aluminum ones. At the same time, aluminium as a material is about 5x as thermally conductive as steel (stainless steel pots often have an aluminium base). So an AL case (or heatsink), being thicker and more conductive, is usually better when heat dissipation is needed for a high power.

Thermal grease under the resistors can help to spread heat due to high power.

There's a couple of watch-it's with mounting the inductor. Some air-coil inductors are on a plastic bobbin with a bolt hole. Using a normal steel bolt through this hole can add 40-50% to its inductance, so better to use a plastic or stainless steel bolt, or another method like zip ties or silicone, as on this build.

Also, having the coil close to the chassis can increase inductance if its steel, or decrease it if aluminium due to two competing effects of magnetic permeability and eddy currents, Something non metallic to space the coil a few mm (say 5mm) off the case avoids this. That being said, Ive only understood those points aver time, and my build works just fine without accounting for them!

Jacks - should be plastic (as you have here) so insulated from the case, eg Cliff type. I like to show at least two output jacks in parallel and I use them both for one of my rigs which is a combo on a cab. But the main reason is that when I'm plugging it all together, I know that the single jack is the input and the group of jacks is the output, even if i get my right and left mixed up!

 

[dbishopbliss]

You can use just one M2 or M3 into an 8 Ohm amp output, then use it with one cab, or two 8 ohm cabs for a 4 Ohm load out of the attenuator but still using the 8 ohm amp tap

Am I understanding this correctly? I can plug an 8 ohm M2 into the 8 ohm output of my amp and then plug a 4 ohm load (two 8 ohm speakers in parallel) into the M2??? I thought the speaker load was part of the load the amplifier sees.

It would certainly make things easier if all I had to do was add another speaker jack in parallel to the M2.

 

[JohnH]

Am I understanding this correctly? I can plug an 8 ohm M2 into the 8 ohm output of my amp and then plug a 4 ohm load (two 8 ohm speakers in parallel) into the M2??? I thought the speaker load was part of the load the amplifier sees.

It would certainly make things easier if all I had to do was add another speaker jack in parallel to the M2.

Yes indeed. The attenuator between the speaker and the amp dilutes the effect of the speaker impedance from reaching the amp. The greatest difference is if all the switched stages are switched off. Then with an 8 Ohm speaker, the impedance seen by the amp (at 440 hz) is about 8.6 ohm. If you change to a 4 Ohm speaker pair this only drops to 7.6 Ohms, a negligible change.

If you switch on more stages, then the difference is even less. At max attenuation, you can pull the speaker out and use it as a load box. M2 and M3 are safe with any speaker ohms, if you match the attenuator to the amp. There's a very slight tonal change in theory, with about 1db more presence and resonance, probably not an issue and easily adjustable at the amp if it is.

Conversely, you can run a 16 Ohm cab from an 8 Ohm attenuator and its safe. The tonal change is now a reduction in presence and resonance, and its a bit more, nearer 2db which I can hear. So for this use, i have the added optional output 3 which corrects this tone change, and I use this feature myself sometimes.

if you try them with the 4 ohm cab combo, and you notice and want to adjust for that very slight tonal change within the build rather than elsewhere , its an easy addition with a resistor and a switch or a switched jack.

 

[Graham G]

I had decided that because I don't have the technical knowledge required to follow Johns schematics etc., ( I do have the building ability to build one), i'd have to buy an off the shelf box, but it would have to be at the budget end of the market, at 76 I'm past buying expensive equipment, I may get the chance to only use it once , so after reading some reasonable reviews(& some negative ones) I bought a Bugera PS1 & used it at Saturdays gig, it's going back Today no more comment needed.
So i'll have to try to get to grips with the schematics & measuring etc & build one, wish me luck, i'll need it .

 

[JohnH]

Hi @Graham G , way to go! keep in touch, I'm happy to discuss each step. What type of amp and speaker is it for? After that, the next step is to identify the parts to buy, and then lay them out to plan the wiring.

 

[Graham G]

Hi John, Origin 50C(with HF ORI20 Mod,which I did myself) Celestion Redback 8ohm, only used for gigs(no home use).
I've started to draw out an M2 circuit, I only need 1 8ohm output, so I'm trying to draw it out with just an input & 1 output(8ohm),I also only need Band level reduction.

 

[Graham G]

Hi, attached is my drawing of the schematic, "if" this is correct I could follow this to start my build, please don't laugh too much, there's a first time for everything .
If the drawing makes sense i'll post my interpretation of what I think is going on in the schematic.

https://i.imgur.com/37hN2Dn.jpg

 

[JohnH]

Hi @Graham G , neat drawing and everything is right, except at switch 2, the left lug has ended up touching the 90 degree corner of the red wire so it looks connected to it. It should be like switch 1 or 3 which I expect was the intention.

As sketched, the switches have the poles connected to the left lug and all the attenuation stages are working. Now imagine them all to the right. In that setting, R3, R5 and R7 only have their lower ends connected, the top ends are floating, so they aren't active. R2, R4 and R6 are all bypassed by their switches, so they are also inactive. So like that, only the first fixed stage is working, for minimum attenuation of -7db.

Next job, order the parts and make a sketch to scale, to work out where you will physically place the parts. This is easier when you have the parts to hand. For an easy life, Id suggest to get a box that is bigger than you might really need. Mine on post 1, which has a few more parts, was a squeeze! Though it would be an easier fit with what you are building.

There can be a mental leap involved in going from the schematic to a more pictorial wiring diagram. While you can lay it out very close to how the schematic looks, you will probably shuffle the parts to make them fit better, and run the wires to give the shortest runs and neatest wiring. Wires will go between the lugs of the components rather than joining each other in mid-wire. So long as the correct lugs are connected to each other, it doesn't really matter the wiring route between the parts. Some circuits are more critical for this (eg within an amp), but not this one, so there is much flexibility.

 

[Graham G]

Hi @Graham G , neat drawing and everything is right, except at switch 2, the left lug has ended up touching the 90 degree corner of the red wire so it looks connected to it. It should be like switch 1 or 3 which I expect was the intention.

As sketched, the switches have the poles connected to the left lug and all the attenuation stages are working. Now imagine them all to the right. In that setting, R3, R5 and R7 only have their lower ends connected, the top ends are floating, so they aren't active. R2, R4 and R6 are all bypassed by their switches, so they are also inactive. So like that, only the first fixed stage is working, for minimum attenuation of -7db.

Next job, order the parts and make a sketch to scale, to work out where you will physically place the parts. This is easier when you have the parts to hand. For an easy life, Id suggest to get a box that is bigger than you might really need. Mine on post 1, which has a few more parts, was a squeeze! Though it would be an easier fit with what you are building.

There can be a mental leap involved in going from the schematic to a more pictorial wiring diagram. While you can lay it out very close to how the schematic looks, you will probably shuffle the parts to make them fit better, and run the wires to give the shortest runs and neatest wiring. Wires will go between the lugs of the components rather than joining each other in mid-wire. So long as the correct lugs are connected to each other, it doesn't really matter the wiring route between the parts. Some circuits are more critical for this (eg within an amp), but not this one, so there is much flexibility.

Thanks John, Yes I drew it out by hand & the connection you mentioned is meant to be seperated,
So there is no actual positive negative as such ? & what wire/cable would you advise, as in AWG stranded, solid core etc ?.
Am I reading it correct in that the Reactive Stage(not sure what that means) is always connected(so -7db) & the other 3 stages are switched in or out? as needed.
By the way i'm in the UK.
Thanks for your help.
Cheers, Graham.

 

[JohnH]

Hi Graham
Id suggest 18awg wire. Solid is good if all your connections are between items fixed to the chassis, or stranded if some are on the lid, as in my build. Stranded for all is ok too.

Yes there's not really a ground, although the lower part of the diagram looks ground-ish, it may not be, it's all really just connected to two taps coming off the amp output tranny. So best to use the plastic Cliff jacks so the case is not part of the circuit.

Yes stage 1 is always on, and sets everything up. It's reactive because it has the coil, and it's impedance changes with frequency. Another way to say it is resistors have resistance and inductors and capacitors have reactance, changing with frequency and shifted in phase.