Simple Attenuators - Design And Testing

[BlueX]

Here’s my take on the M2/8Ohm/Out3. Circuit and specifications according to John's post #1 in this thread. I borrowed part of Gene’s layout in post #2207 (p 111), to have R3-R8 placed in the same order as the “descending” switches.

The bottom of the “box” is made from 1,5 mm thick, perforated aluminum sheet (5 mm holes). Hope that will be sufficient heat sink. Resistors mounted flat to the sheet, with thermal compound in between. The sheet is reinforced with 12x12x2 mm aluminum profiles. Coil strapped with plastic cable ties to a piece of wood. Top cover is a black aluminum mesh (for covering vent holes, etc.). Ventilation should be sufficient. Dimensions are 300x140x80 mm.

I did the following tests on the unit (test report enclosed), if anyone’s interested. I found a bad solder joint!

• No connection between chassis or any part of the circuit
• Continuity within each sub-circuit
• Resistance between different lugs on in- and output. Interesting thing: Resistance increases as more attenuation is added, until the last step when it decreases slightly.

I got problems with the “Cliff UK” speaker jacks. For two of them I get consistent readings when measuring on the solder lugs and measuring with a plug. However, for the other two I get OL (overload) for the plug, but reasonable reading at the lugs. For one of the bad jacks I can get consistent value for the plug, if I press the plug sideways. However, once I release the pressure reading goes back to OL.

I cannot trust these jacks, and will replace all four. Consequently, I have not yet played with this unit. Found the following two types instead, and will try to source them (in stereo/TRS configuration). Any comments on speaker jacks? Is it possible to use “switched” jacks?

https://www.parts-express.com/Switchcraft-Z15J-1-4-High-Power-Speaker-Jack-093-1056

https://www.parts-express.com/Neutrik-NMJ6HC-S-1-4-Stereo-Phone-Jack-Switched-090-976

I hope that @JohnH and Marshall Forum consider this fair use of names. I want to give credit. Finally, thanks to John and Gene for all work and support!

 

[Lancer X]

So I’m happily wiring away when I realize I’m maybe making a mistake. See black wires in pic 1 below. Should I avoid routing wires closely in parallel like this? Am I creating a crosstalk risk?

Trying to be neat and tidy, but now second guessing myself.




Here an example of where I’m only crossing wires perpendicularly. Better?

 

[Barnsley Boy]

Hi @JohnH ,

Been busy at work and pondering on the 4 ohm additional box, so haven't had a chance to respond back. I've drawn out what I think it should look like if I were to use an extension box to convert the 8 ohm to a 4 ohm unit. Effectively this entails adding another input jack which the box would connect to putting the components in parallel with the input from the amp. Below are diagrams showing operation as an 8 ohm unit, then a 4 ohm unit. Could you run your expert eye over these to check that I have interpreted the requirements correctly?

I like to draw these things out so I can get a good understanding of which wire goes where, and hopefully if I am correct, in some small way it may help others out .

If I were to build another 4 ohm unit, just for my Peavey (and including the -28db stage) would it be a case of putting all the components as drawn above in one box i.e. using using 2 coils? or would I be able to build another M2 with just a single coil, and change the resistor values to suit? If so, could you advise on what the component part values would be, and would the 50w version handle my 65w Peavey? One last question, I'm not sure that I understand the rationale behind the 2 output jacks. does this allow me to plug two 8 ohm speakers in? Or if the extension box is connected, two 4 ohm speakers?

The more that I delve into this massive post, the more it blows my mind!!!! Keep up the good work.

 

[Gene Ballzz]

Here’s my take on the M2/8Ohm/Out3. Circuit and specifications according to John's post #1 in this thread. I borrowed part of Gene’s layout in post #2207 (p 111), to have R3-R8 placed in the same order as the “descending” switches.

The bottom of the “box” is made from 1,5 mm thick, perforated aluminum sheet (5 mm holes). Hope that will be sufficient heat sink. Resistors mounted flat to the sheet, with thermal compound in between. The sheet is reinforced with 12x12x2 mm aluminum profiles. Coil strapped with plastic cable ties to a piece of wood. Top cover is a black aluminum mesh (for covering vent holes, etc.). Ventilation should be sufficient. Dimensions are 300x140x80 mm.

I did the following tests on the unit (test report enclosed), if anyone’s interested. I found a bad solder joint!

• No connection between chassis or any part of the circuit
• Continuity within each sub-circuit
• Resistance between different lugs on in- and output. Interesting thing: Resistance increases as more attenuation is added, until the last step when it decreases slightly.

I got problems with the “Cliff UK” speaker jacks. For two of them I get consistent readings when measuring on the solder lugs and measuring with a plug. However, for the other two I get OL (overload) for the plug, but reasonable reading at the lugs. For one of the bad jacks I can get consistent value for the plug, if I press the plug sideways. However, once I release the pressure reading goes back to OL.

I cannot trust these jacks, and will replace all four. Consequently, I have not yet played with this unit. Found the following two types instead, and will try to source them (in stereo/TRS configuration). Any comments on speaker jacks? Is it possible to use “switched” jacks?

https://www.parts-express.com/Switchcraft-Z15J-1-4-High-Power-Speaker-Jack-093-1056

https://www.parts-express.com/Neutrik-NMJ6HC-S-1-4-Stereo-Phone-Jack-Switched-090-976

I hope that @JohnH and Marshall Forum consider this fair use of names. I want to give credit. Finally, thanks to John and Gene for all work and support!

View attachment 103015View attachment 103016View attachment 103017View attachment 103018View attachment 103019View attachment 103020View attachment 103021View attachment 103022View attachment 103023

FIRST> Nice build so far!
NEXT> Avoid that linked Switchcraft jack, as it will connect the "-/negative" of the circuit to the chassis.
THEN> @JohnH did an analysis a few pages back concerning the effectiveness of that very nice type of "perforated" sheet and determined it's heat sink value to be not great. Depending on the amp wattage, you may be OK.
FINALLY> I'm quite surprised that those CLIFF jacks don't seem to be working well! Are they old/salvaged/used? Or did you overheat and melt them during soldering? These jacks are known to be fairly rock solid and rarely exhibit issues. It may be an issue with the way you are testing. The switching functions in the jacks can sometimes give unexpected results, especially in combination with the switching in/out and bypassing of components! The best way to test is to:
A> Plug your speaker out of the appropriate output jack with a proper speaker cable.
B> Plug either a speaker cable (or bare, unwired plug) into the "Amp in" jack. Do not connect to the amp!
C> Put your meter test leads across the positive and negative connections of the cable or plug that is plugged into the "Amp in" jack and note/write down the readings for each and every switch combination. If possible, I would try this first with an 8 ohm speaker (or two 16 ohm speakers) plugged out of each (both) of the "Speaker out" "1 X 8 Ohm or 2 x 16 ohm" jacks and then try the same with a 16 ohm speaker plugged out of the "1 x 16 Ohm" jack. All readings should stay in a range of 1 to 3 ohms, surrounding or near 8 ohms.

Please Keep Us Posted?
Gene

 

[mike_lawyer]

Okay, I rewired the attenuator as shown above, and both the first switch and second switch work perfectly. However, the -3.5 db switch makes no difference. I am thinking that the DPDT switches I ordered off Amazon might be very fragile with soldering. Does anyone have a source for high-quality DPDT switches that will hold up well? I would like to use mini DPDT switches for space.

 

[BlueX]

FIRST> Nice build so far!
NEXT> Avoid that linked Switchcraft jack, as it will connect the "-/negative" of the circuit to the chassis.
THEN> @JohnH did an analysis a few pages back concerning the effectiveness of that very nice type of "perforated" sheet and determined it's heat sink value to be not great. Depending on the amp wattage, you may be OK.
FINALLY> I'm quite surprised that those CLIFF jacks don't seem to be working well!

Thanks Gene, and thanks for explaining the function of that jack!
In the sheet I'm using there are two 5 mm holes in a square of 14x8 mm (112 mm2), i.e. 2xPix(5/2)¨2=39 mm2 or about 35% of the contact surface is gone. I will use this M2 with a SV20H so I guess it will be OK. I missed John's analysis (quite a lot of info in this thread).
The CLiff jacks were brand new, without any traces of use, when I got them. Overheating is always a risk, but I cannot see any damages. Maybe the contact lugs lost their spring properties? I will do some trouble-shooting and share the results. Could be of interest to others.

 

[Lancer X]

The CLiff jacks were brand new, without any traces of use, when I got them.

I had heard of some people receiving counterfeit Cliff ones. Any chance of that?

 

[JohnH]

So I’m happily wiring away when I realize I’m maybe making a mistake. See black wires in pic 1 below. Should I avoid routing wires closely in parallel like this? Am I creating a crosstalk risk?

Trying to be neat and tidy, but now second guessing myself.


View attachment 103029

Here an example of where I’m only crossing wires perpendicularly. Better?

View attachment 103030

I don't think there's any issues there. It's all low impedance output wiring, there are no sensitive input stages nearby. So a nice neat build is all it needs to be.

 

[JohnH]

Okay, I rewired the attenuator as shown above, and both the first switch and second switch work perfectly. However, the -3.5 db switch makes no difference. I am thinking that the DPDT switches I ordered off Amazon might be very fragile with soldering. Does anyone have a source for high-quality DPDT switches that will hold up well? I would like to use mini DPDT switches for space.

Hi Mike I've never had an issue with mini switches, but I'm buying locally and I don't know what brand. Can you test with an ohm meter setting directly onto the lugs, in circuit, to see if the contacts work?

 

[JohnH]

Hi @JohnH ,

Been busy at work and pondering on the 4 ohm additional box, so haven't had a chance to respond back. I've drawn out what I think it should look like if I were to use an extension box to convert the 8 ohm to a 4 ohm unit. Effectively this entails adding another input jack which the box would connect to putting the components in parallel with the input from the amp. Below are diagrams showing operation as an 8 ohm unit, then a 4 ohm unit. Could you run your expert eye over these to check that I have interpreted the requirements correctly?

I like to draw these things out so I can get a good understanding of which wire goes where, and hopefully if I am correct, in some small way it may help others out .

If I were to build another 4 ohm unit, just for my Peavey (and including the -28db stage) would it be a case of putting all the components as drawn above in one box i.e. using using 2 coils? or would I be able to build another M2 with just a single coil, and change the resistor values to suit? If so, could you advise on what the component part values would be, and would the 50w version handle my 65w Peavey? One last question, I'm not sure that I understand the rationale behind the 2 output jacks. does this allow me to plug two 8 ohm speakers in? Or if the extension box is connected, two 4 ohm speakers?

The more that I delve into this massive post, the more it blows my mind!!!! Keep up the good work.


View attachment 103031



View attachment 103032

I like the idea of the separate box, it's a convenient addition and it spreads the heat. But if you want a dedicated 4 ohm M2, basically all values halve I have an optimised set of nearest standard values if you want them.

 

[BlueX]

I had heard of some people receiving counterfeit Cliff ones. Any chance of that?

Could be, but difficult to say. They are properly branded, look well made, and were supplied by what seems like a professional company ( in business since 1972). I will have a closer look at them.

 

[Barnsley Boy]

I like the idea of the separate box, it's a convenient addition and it spreads the heat. But if you want a dedicated 4 ohm M2, basically all values halve I have an optimised set of nearest standard values if you want them.

Hi @JohnH, if you have a set of values that would be great as I'm not sure whether I should be rounding up or down!

I'm guessing that L1 would be a 0.45mH/ 18awg coil? Closest that I can get is:

0.47mH, 1.0mm Wire, DCR 0.30 Ohms

Is that close enough?

 

[mike_lawyer]

Success! I got everything to work, all switches work well. I am contemplating adding the -28 db option as well.

This unit sounds great and it was a really fun project. Thanks to Gene and all the contributors for a great design, this thing sounds awesome!

 

[JohnH]

hi @Barnsley Boy , this post has the 4 ohm values: post #2127

Here's my suggestions for 4 Ohm, to keep everything even and in proportion but based on standard values:

R1 8.2
R2A 12
R2B 10
R3 8.2
R4 5.6
R5 2.7
R6 8.2
R7 18
R8 2.7
R10 27
R11 4.7

L1 0.45mH or 0.5mH

If you are getting from a different range, try to keep the proportions within each stage close to that of the 8 Ohm set

This is the set of frequency response- attenuations that are predicted for those 4 ohm values:

View attachment 97569

 

[Barnsley Boy]

hi @Barnsley Boy , this post has the 4 ohm values: post #2127

Hi @JohnH ,
Many thanks for that. Incredible to think that this thread has 2555 posts!!! From small acorns, large oak trees grow.
So, adding my -28db stage, am I correct to just halve the values? If so, the closest I can get (according to my diagram) are:

R9A - 25w, 3.3 ohm
R9B - 5w, either 0.62 or 0.56 ohm
R9C - 1w, 8.2 or 10 ohm

Could you advise which would be the correct values to use?

I'm torn between building the 4ohm extension box, or building a new 4ohm M2 (mainly because it is so bl**dy good). I'll need to check back on the previous posts to brush up on the theory and evolution of dropping the L2 inductor in favour of a single L1. I did read about it, but I was so busy lapping up all the good stuff in the thread that it probably went in one ear and out the other!

 

[JohnH]

hi @Barnsley Boy ok, I can run numbers on that -28db stage, to find the best values for a 4ohm build.

On how to get the 4 ohm, could you describe how you'll use it some more? is it an important use that you'll want often? or just occasional? Also, what was the power of the 4ohm amp? Will you need loud but attenuated settings, like a loud jam or gig? or just home /studio use with low volume?

 

[JohnH]

hi @Barnsley Boy , I reckon the best matching set of values for a -28db stage at 4 Ohms is 3.3, 0.56 and 10 ohms

 

[Barnsley Boy]

hi @Barnsley Boy ok, I can run numbers on that -28db stage, to find the best values for a 4ohm build.

On how to get the 4 ohm, could you describe how you'll use it some more? is it an important use that you'll want often? or just occasional? Also, what was the power of the 4ohm amp? Will you need loud but attenuated settings, like a loud jam or gig? or just home /studio use with low volume?

Hi @JohnH ,
I am currently a home user with aspirations to get back into a bit of recording so low volume is a key consideration. If we don't get burnt to a crisp over the coming days/ weeks/ months, I've been contemplating getting a band together or at least looking for a couple of like minded jam buddies.

The Peavey Classic VTX is a 65 watt amp, and is incredibly loud, hence my temporary fix of using a volume control pedal in the fx loop. At the moment it's the only way to be able to use the lead channel and tweak the saturation control. The M2 with -28db stage is absolutely perfect for my JCM900 - full attenuation on gives me exactly what I want and does a fantastic job. If I can get the same results for the Peavey, then the jobs a good 'un!!

Thanks for the resistor values, I shall mull over the options and start assembling components!

 

[JohnH]

Maybe you could try the simple 3-part addition for the 8 ohm M2 to work at 4ohm. It's a lot less to build than another full one. Then, you could decide if you'd like a separate 4ohm unit?

 

[Barnsley Boy]

Maybe you could try the simple 3-part addition for the 8 ohm M2 to work at 4ohm. It's a lot less to build than another full one. Then, you could decide if you'd like a separate 4ohm unit?

I'm thinking that is the best bet. I'm just about to place my order for the components! I'll report back once it is built and tested.