Simple Attenuators - Design And Testing

JohnH

Well-Known Member
Joined
May 31, 2008
Messages
4,959
Reaction score
3,159
Location
Wilton NSW
Hi @brad Messier , thanks for posting. An important test is to test resistance as seen by the amp.

Plug speaker into attenuator
Plug a cord into the attenuator input, but without the amp
Measure resistance across the jack plug that would go to the amp. For 4Ohm version, we expect 3.5 to 5 Ohms in all settings. (amp is not connected in this test)
 

Gene Ballzz

Well-Known Member
Joined
Feb 19, 2013
Messages
3,456
Reaction score
3,831
Location
Las Vegas, NV
Hey @JohnH ,
I think I'm gonna start a new thread, simply for completed builds. I'll ask for only those who have finished a build, with pics to post, and direct those with questions to this thread. It might be kinda fun to actually know how many completed builds there are out there?
Watcha Think?
Gene
 

AtomicRob

Well-Known Member
Joined
Feb 4, 2009
Messages
500
Reaction score
432
Location
California
The question is, can I get away with just using one half of the DPDT? They are rated at 5A @ 120V AC, or 2A @ 250V AC.
The switches for the stages won't exceed 5A (or 120VAC) so using SPDT or half of the DPDT would be fine. Using ohms law, I = sqrt(P/R), so for a nominally 50W amp which puts out like 90W RMS, if you're building the 8 ohm version that's 3.4A, the 16 ohm version sees 2.3 amps. For 100W amps, at max 170W RMS output, you get 4.6A at 8 ohms, 3.3A at 16 ohms. That 4.6A is uncomfortably close to the limit. BUT the stages are all after the initial -7db fixed stage so the power is actually reduced to 1/5 of that and the current is worst case like 2A.

The bypass switch sees the full signal so I'd recommend a 10A rated switch for that one if you're building an 8 ohm version or especially a 100W version. I'm building a 100W right now and I'm using Carling 2M1 series for the stages and G series for the bypass.

BTW you can't really double the rating of a DPDT by ganging the poles - a 5A rated DPDT can't switch 10A on a single circuit. The reason is that mechanically the two poles don't switch at the exact same time, so one pole or the other will always see the full current initially when switching. You might get away with it if you never switched it under load - and that may be true for an attenuator - but still it's not a safe way to build.
 

brad Messier

New Member
Joined
Dec 25, 2021
Messages
8
Reaction score
1
Hi @brad Messier , thanks for posting. An important test is to test resistance as seen by the amp.

Plug speaker into attenuator
Plug a cord into the attenuator input, but without the amp
Measure resistance across the jack plug that would go to the amp. For 4Ohm version, we expect 3.5 to 5 Ohms in all settings. (amp is not connected in this test)

Well something is clearly off, my resistance reading was 14, 19 and 21 depending on the switch positions! Please forgive my crude drawing, but this is the layout I ended up with. IMG_3133.jpg
 

JohnH

Well-Known Member
Joined
May 31, 2008
Messages
4,959
Reaction score
3,159
Location
Wilton NSW
Hi @brad Messier . The switches look a bit odd too where you connect to each side of the two-pole switches. The basic design only uses one switch pole per switch, so when we use two poles, they are ganged together with links across each pair of lugs, as in Gene's or my diagrams.

Definitely don't test with the amp until all confirmed by resistance tests!
 

JohnH

Well-Known Member
Joined
May 31, 2008
Messages
4,959
Reaction score
3,159
Location
Wilton NSW
Hey @JohnH ,
I think I'm gonna start a new thread, simply for completed builds. I'll ask for only those who have finished a build, with pics to post, and direct those with questions to this thread. It might be kinda fun to actually know how many completed builds there are out there?
Watcha Think?
Gene
'

Sure! it would be nice to have a more condensed gallery. I bet there'll be random questions on it too! But happy if you'd like to manage that.

I lost count years ago on how many of these we have seen. I love seeing all the builds. I'm guessing about one new build every 3 pages or so?
 

Gene Ballzz

Well-Known Member
Joined
Feb 19, 2013
Messages
3,456
Reaction score
3,831
Location
Las Vegas, NV
Yeah @JohnH ,
Off the top of my head I was guessing somewhere roughly between 35 & 50. And then there were a few less vocal folks who likely built and never showed off their work! And then likely some that came, saw, built, never posted and are happy users! Ya just gotta bet that some manufacturer will glom on, take the design, patent it, package it and then try to sue you for infringing on it! :nutkick: :p This design is just too good to be kept a secret for much longer!
Just Teasin'
Gene
 

JohnH

Well-Known Member
Joined
May 31, 2008
Messages
4,959
Reaction score
3,159
Location
Wilton NSW
Yes indeed! Not worried about anyone else patenting it, they can't. It's well published right here as 'prior art' with dates etc.
 

junk notes

Well-Known Member
Joined
Dec 10, 2018
Messages
3,787
Reaction score
5,009
Hi all, Built my first John H. It's a M-Lite.
:cool:
:h5:
cliffcolours.jpg
 

Barnsley Boy

Member
Joined
Sep 2, 2017
Messages
81
Reaction score
89
Location
Suffolk, UK
BTW you can't really double the rating of a DPDT by ganging the poles - a 5A rated DPDT can't switch 10A on a single circuit. The reason is that mechanically the two poles don't switch at the exact same time, so one pole or the other will always see the full current initially when switching. You might get away with it if you never switched it under load - and that may be true for an attenuator - but still it's not a safe way to build.

Understood!

Picture1.jpg


Next time, I'll probably use SPDT chunkier toggle switches. These mini ones are ok for guitar pickup switching options, but I think maybe something more industrial could be on the cards for the next build!

After faffing around with soldering bits of jumper wire between terminals and then bending the wires into something half decent to solder to, I came to the conclusion that I am better off using spade connectors after all. Rather than crimping the wires onto the connector, I have soldered them in place. The connectors can then be pushed onto the switch terminals. Connection between adjacent poles is achieved using a small piece of wire bent into a u-shape, laid into the trough of the connector. Flowing some solder into the trough then connects it all together. I've used heat shrink to tidy it all up. The only disadvantage of this is that the whole assembly becomes quite bulky, but as the case I'm using is fairly spacious, It isn't an issue.

Picture2 (2).jpg

@JohnH & @Gene Ballzz - good idea on the completed build thread. During my planning stage I started to look back on some of the previous builds to see how other people had tackled things. 118 pages (and counting) is a lot to get through. I even toyed with ploughing through and compiling them myself into a single document.

I think this thread is possibly one of the most important on the whole forum for users of valve amps. Is there any way that this could be stickied (I believe that is the correct terminology), so that it is easily accessible?
 

brad Messier

New Member
Joined
Dec 25, 2021
Messages
8
Reaction score
1
@brad Messier ,
Look at R1, and the input negative! That should be connecting to the negative of the speaker out and by proxy to R7 & R3. I think the rest is correct.
Let us know?
Gene

OK, I think that has taken care of the issues. I tied the - end of R1 in to R7 and R3 via a jumper from the input ground over to the output ground. I was looking at this from a guitar/pedal perspective and figured a ground is a ground... did not understand those negative ends needed to be tied together, I separated them to keep the wiring clean.

Switch wise, I did jump the poles of the switches together, for clarity I omitted this from my illustration.

New resistance readings, 5.3 6.6 and 8.8 with various switches engaged. Clear to plug it back in?
 

JohnH

Well-Known Member
Joined
May 31, 2008
Messages
4,959
Reaction score
3,159
Location
Wilton NSW
OK, I think that has taken care of the issues. I tied the - end of R1 in to R7 and R3 via a jumper from the input ground over to the output ground. I was looking at this from a guitar/pedal perspective and figured a ground is a ground... did not understand those negative ends needed to be tied together, I separated them to keep the wiring clean.

Switch wise, I did jump the poles of the switches together, for clarity I omitted this from my illustration.

New resistance readings, 5.3 6.6 and 8.8 with various switches engaged. Clear to plug it back in?

Still maybe some gremlin I think! A 4ohm M2 shouldn't go as high as 8.8. But, the measurements should be with the speaker plugged into the output. So maybe that accounts for it?
 

Gene Ballzz

Well-Known Member
Joined
Feb 19, 2013
Messages
3,456
Reaction score
3,831
Location
Las Vegas, NV
@brad Messier ,
Where are you deriving the "-/negative" for the line out? I see from your pics that you are using isolated Cliff style jacks for all else, but the line out appears to have it's "-/negative" tied to the chassis, but nothing else is! Please note that I am avoiding the term "ground" as nothing here is actually ground, simply +plus and -minus, and all should be isolated from the chassis!

Most of the one or few issues that have come up over the course of this thread seem to have centered around allowing the case/chassis to become any part of this circuit! While it kinda doesn't make any sense that the chassis/case can't be used, most folks' issues have been cured by make sure all circuitry electrically "floats" inside the chassis!

Again, Please Let Us Know?
Gene
 

brad Messier

New Member
Joined
Dec 25, 2021
Messages
8
Reaction score
1
@brad Messier ,
Where are you deriving the "-/negative" for the line out? I see from your pics that you are using isolated Cliff style jacks for all else, but the line out appears to have it's "-/negative" tied to the chassis, but nothing else is! Please note that I am avoiding the term "ground" as nothing here is actually ground, simply +plus and -minus, and all should be isolated from the chassis!

Most of the one or few issues that have come up over the course of this thread seem to have centered around allowing the case/chassis to become any part of this circuit! While it kinda doesn't make any sense that the chassis/case can't be used, most folks' issues have been cured by make sure all circuitry electrically "floats" inside the chassis!

Again, Please Let Us Know?
Gene

Ok, seems to be in order now. I had a Switchcraft input jack in there due to space issues. Swapped that out for a slightly modified Cliff jack and got the negative side tied together and isolated from the case. Added jumpers to the jacks themselves, I have not worked with this style jack before not exactly Resistance now checks out, I am seeing 3-5 Ohms resistance to the output tip with the speaker plugged in. Reading a steady 5 Ohms on the - side of the circuit. I will have some time tomorrow to give it a full test. Not the neatest build of all time, but things are really squeezed in there tight! IMG_3146.jpg
 

JohnH

Well-Known Member
Joined
May 31, 2008
Messages
4,959
Reaction score
3,159
Location
Wilton NSW
M2 Gig

This is a possible new 'Variant of Interest'. Not sure of it will go 'viral'!

Its intended to address a issue that has come up a few times, where the attenuator is intended only for small volume reductions for gigging or loud jamming, using low or medium powered amps. It focusses on providing well-balanced tones with reductions of 3.5 or 7 db, taking say a 20W amp down to 9W or 4W. It's basically Stage 1 from the M2 design, with a DPDT switch to provide the extra -3.5db setting. (The base M2 starts at -7db)


In the drawing, the twi parts of the switch are shown at the -3.5db setting.

Values are as follows, for use at 8 ohms or 16 Ohms. Power rating for resistors are 50W min for 50W amps.

R1A 27 47
R1B 33 82
R2A 22 47
R2B 18 39
R2C 10 15

L2 0.9mH 1.8mH

Its also possible to build this as Stage 1 into a full M2 design. Can post if interested, there are a couple of other tweaks needed in that case.
 
Last edited:

Emiel

Member
Joined
Jul 22, 2008
Messages
57
Reaction score
58
M2 Gig

This is a possible new 'Variant of Interest'. Not sure of it will go 'viral'!

Its intended to address a issue that has come up a few times, where the attenuator is intended only for small volume reductions for gigging or loud jamming, using low or medium powered amps. It focusses on providing well-balanced tones with reductions of 3.5 or 7 db, taking say a 20W amp down to 9W or 4W. It's basically Stage 1 from the M2 design, with a DPDT switch to provide the extra -3.5db setting. (The base M2 starts at -7db)


In the drawing, the switches are shown at the -3.5db setting.

Values are as follows, for use at 8 ohms or 16 Ohms. Power rating for resistors are 50W min for 50W amps.

R1A 27 47
R1B 33 82
R2A 22 47
R2B 18 39
R2C 10 15

L2 0.9mH 1.8mH

Its also possible to build this as Stage 1 into a full M2 design. Can post if interested, there are a couple of other tweaks needed in that case.

Using this as Stage 1 into a Full M2 design would have my vote! :applause:
 


Top