Simple Attenuators - Design And Testing

[JohnH]

Hi @kaludjerko , and thanks for your interest.

The values are not super-critical, and I doubt you'd hear a difference across those values of L2.. Can you find the specified value for L1?

But, to make a choice: A higher L2, if you could hear a diffference, would theoretically be slightly brighter, and a bit more 'lively'. I'd go with the higher value of 0.56mH assuming L1 is as designed.

But its a good question. Ill run the numbers to let you know what difference it really would make.

Actually, if needed, you can reduce inductance by unwinding a few turns and re-taping the coil. Ive done that to make mine, but I have an inductance range on my meter to help do it.

And, you can also increase inductance by mounting a small steel bolt on the central axis of the coil.

 

[JohnH]

yes, going from 0.5mH to 0.56mH adds about 0.25db to the high treble. id use that but that's less difference than youll hear.

 

[Gene Ballzz]

Hi @kaludjerko

And, you can also increase inductance by mounting a small steel bolt on the central axis of the coil.

Hey @JohnH

Upon reading that, can I assume that my coil mounting method has increased the inductance? Should I be using plastic/nylon bolts here?

Thanks John,
Gene

 

[Dogs of Doom]

you put in a bad url...

I was going to bed last night when I fixed the link & typed this. For some reason, soundcloud would only let you post the link, as "in a set" (list). You can only embed a file, not a list, so, you had to remove the list reference. I see you figured that out, but thought I'd confirm it & make that clear to anyone else who might be interested...

see graphic below:



you only need to post the text & the forum will automatically format the player (same w/ youtube)...

 

[Dogs of Doom]

Hey @JohnH

Upon reading that, can I assume that my coil mounting method has increased the inductance? Should I be using plastic/nylon bolts here?

Thanks John,
Gene

how much heat does that generate?

 

[Gene Ballzz]

how much heat does that generate?

Get's only a little warm with a 20 watt DSL20 and stays generally cool with my 13 watt, 5E3, Tweed Deluxe. It gets pretty toasty (almost hot) with my JTM30, which although "officially" 30 watts, I'm betting it's every bit of 40 or more watts! I've actually rebuilt it, using lower wattage (physically smaller) resistors for all except the first, always on, -7db stage.

Just My $.02,
Gene

 

[JohnH]

@Gene Ballzz
Yes, those inductors might be acting 20 to 40% higher! I don't think that's a problem if you are liking the results though, it doesn't change the basic mid-band ohms that the amp sees.

But you could try with the bolts out to see if you notice a difference. Nylon or stainless steel bolts wouldn't have that effect.

@Dogs of Doom
This design is worked out with a factor of at least 3 on calculated power going into each of the resistors compared to their spec, based on 50W amps. The first stage does most of the work, and the other stages take a lot less or may be off. With a heavy case and all resistors bolted with thermal grease, it takes a long time to heat up and so temperature rise caused by power dissipation is averaged over about 10-15 minutes on my build. With my 50W VM, I haven't been able to get it more than warm:

AttenuatorM Inside 190217

• JohnH
• Feb 16, 2019

 

[kaludjerko]

Hi @kaludjerko , and thanks for your interest.

The values are not super-critical, and I doubt you'd hear a difference across those values of L2.. Can you find the specified value for L1?

But, to make a choice: A higher L2, if you could hear a diffference, would theoretically be slightly brighter, and a bit more 'lively'. I'd go with the higher value of 0.56mH assuming L1 is as designed.

But its a good question. Ill run the numbers to let you know what difference it really would make.

Actually, if needed, you can reduce inductance by unwinding a few turns and re-taping the coil. Ive done that to make mine, but I have an inductance range on my meter to help do it.

And, you can also increase inductance by mounting a small steel bolt on the central axis of the coil.

Thank you for prompt answer and update with calculated data.
Will be ordering from banzaimusic (hopefully mentioning them doesn't break some rules here) and they have 0.33mH (for L1).
Already have resistors and am looking forward to this build.

 

[Heph333]

How would one go about sizing the inductors for much lower wattage? Let's say I want to make a version for a 5W single-ended amp. The resistors would be small enough that I could probably build on a PCB. Could I use a small axial leaded inductor? They're rated by amps, and without knowing the OT secondary max voltage, not sure how many amps the inductor needs to be rated for. Hell.... could probably just coil my own around a small sewing thread bobbin.
I'd also like to be able to use it as a load box DI. Has anyone tested this at full attenuation with the speaker disconnected? I imagine at -31db the speaker itself isn't contributing very much to the overall reactance of the output.

 

[JohnH]

Hi @Heph333 , thanks for your interest. An amp running at 5Wamp into 8 ohms sends out 0.8Amps. but for the inductors, the current rating is only part of the spec requirements. They also need to have low resistance, particularly L1 , or else the maths get messed up. We use 18gage wire for the base design. Id say you could go down to 20gage if its only for 5W (0.8mm dia)

You can safely use it as a load box with no speaker, when at max attenuation. The amp wont know about it. You miss out on the tonal effect of bass resonance however, which is not usually a big deal. In this design, this bass effect is generated at the speaker itself.

 

[Heph333]

Hi @Heph333 , thanks for your interest. An amp running at 5Wamp into 8 ohms sends out 0.8Amps. but for the inductors, the current rating is only part of the spec requirements. They also need to have low resistance, particularly L1 , or else the maths get messed up. We use 18gage wire for the base design. Id say you could go down to 20gage if its only for 5W (0.8mm dia)

You can safely use it as a load box with no speaker, when at max attenuation. The amp wont know about it. You miss out on the tonal effect of bass resonance however, which is not usually a big deal. In this design, this bass effect is generated at the speaker itself.

Thanks for the follow-up. That's exactly what I needed. I'll be using it with Impulse Responses, so bass resonance will be emulated.

 

[matttornado]

Back again! Finally getting my parts together for my build. Just ordered all of the resistors.

So with my 100 amp, will the first -7db reactive stage, if not bypassed, knock my amp down to ~20 -25 watts? What if that's too much?

My DSL 40CR on the 20 watt setting is usually loud enough but not always.
The other night I played in a fairly large sized bar with a a wide open dance floor & with my DSL set on the 40 watt setting, my master was at about 2 oclock.
I had a 1x12 extension cab hooked up too.

I'm comparing apples to oranges at this point I know, right?

I hear the new 20 watt plexi's are loud as shit so I don't know.

 

[JohnH]

Back again! Finally getting my parts together for my build. Just ordered all of the resistors.

So with my 100 amp, will the first -7db reactive stage, if not bypassed, knock my amp down to ~20 -25 watts? What if that's too much?

My DSL 40CR on the 20 watt setting is usually loud enough but not always.
The other night I played in a fairly large sized bar with a a wide open dance floor & with my DSL set on the 40 watt setting, my master was at about 2 oclock.
I had a 1x12 extension cab hooked up too.

I'm comparing apples to oranges at this point I know, right?

I hear the new 20 watt plexi's are loud as shit so I don't know.

Yes that's right, 100W reduced by -7db is 20W. That's still pretty loud on most such amps!

But in case you need to find a setting in between, if you wire up Attenuator M as drawn, then the -3.5db stage can act on its own, which will give you about 45W. ie you switch to bypass but still have the -3.5 stage engaged. This setting doesn't involve the coils, but it still sounds fine as a resistive stage for this small reduction. I put this feature into the design for just this reason. To do this, the -3.5 stage components are speced for full amp power.

If you are fully cranking that 100W through it, then the power ratings should be x2 compared to the spec, which was for 50W - what sort of resistors are you getting?

 

[matttornado]

what sort of resistors are you getting?

I'm using wire wound power resistors like what others are using. For R1, I'm using two 60 ohm 100 watt resistors in parallel. Same with R7 but with two 150 ohm / 100 watt resistors in parallel. R2 & R8 are 100 watt ratings also. All of he others are 50 watts except R9 is 20 watts.

Thanks JohnH!

 

[JohnH]

Sounds like it will be just right! Good luck with tbe build.

 

[Musicmaniac]

Glad I stumbled on this thread.

 

[matttornado]

Sounds like it will be just right! Good luck with tbe build.

Thanks but I'm still confused trying to understand the diagram. I don't want to build the "RED BOX" part & the switching has me confused.
maybe some earlier diagrams in some earlier posts will help me.

 

[matttornado]

So If I bypass the first 3 stages & want to use just the -3.5db stage, would I connect the BLUE wire between R6 & R8 before the switch? What if the 4th stage is not active? Will the signal just go straight out to the speaker(s)?
Sorry for the stupid questions..... It looks like if the last sage is off, not going to R7, it bypasses completely to the speakers for zero attenuation.
Am I on the right the track? LOL

So this diagram below shows that switch one bypasses everything if wanted. I want switch one to either go to all of the stages or just the last
like previously discussed.
S

 

[Gene Ballzz]

So If I bypass the first 3 stages & want to use just the -3.5db stage, would I connect the BLUE wire between R6 & R8 before the switch? What if the 4th stage is not active? Will the signal just go straight out to the speaker(s)?
Sorry for the stupid questions..... It looks like if the last sage is off, not going to R7, it bypasses completely to the speakers for zero attenuation.
Am I on the right the track? LOL

So this diagram below shows that switch one bypasses everything if wanted. I want switch one to either go to all of the stages or just the last
like previously discussed.
S

Mr Tornado,
The reason you are being confused is that this particular diagram depicts my own personalized version which does not allow for -3.5db attenuation all by itself. For that matter, the one that I'm currently using all the time does not even have that first, overall bypass switch. When it is plugged in, you are already automatically, by default, at -7db reactive attenuation. I opted for this configuration for several reasons:

A) The -3.5db cut is so small/subtle by itself that I deemed it to be able to be otherwise accomplished right at the amp through several other various methods. The main intention of this -3.5db stage was to "fill in the gaps" between the -7db stages, to make the individual jumps less drastic. It was my own (not @JohnH ) judgement call that if an attenuator was truly needed, at least -7db would almost always be the minimum requirement.
B) Even though this iteration is designed to be used out of the 16 ohm tap of the amp, I often use an 8 ohm cabinet and eliminating the bypass switch was intended as a safety/failsafe, to make sure I never accidentally run my 16 ohm amplifier output straight into an 8 ohm load! My advanced years have taught me to always allow for my brain occasionally passing gas! That's partly how I've been able to make it to this age! Subsequent builds will likely have a bypass switch, but will still incorporate the minimum -7db/reactive, default feature!

1) While considering "brain farts" we should ponder the fact that with all stages individually selectable, a person could inadvertently activate just one of the lower wattage rated -7db stages. HEAT UP & GO BOOM! A really good rotary switch could help here, but the search for the correct unit is a long, deep, dark rabbit hole that provides minimal, if not expensive results. Toggles do the job, but leave the operator with the need to be smarter than the box! Factoring possible operator error into the design/implementation is also not a bad idea, IMHO!​

C) The above diagram also allows all resistors, other than the first "ALWAYS ON" stage to be both physically and electrically smaller. The other alternative could be to make every stage full wattage capable, but size could become an issue.
D) With or without the overall bypass switch, the above diagram sought to minimize the complication of wiring as well as the hands on operation of the switching. In other words, whenever the unit is active, the highest wattage portion is always active also!​

Now, I belive that somewhere in this very lengthy thread (or maybe it was a personal email to me) @JohnH showed a way to increase the wattage of that -3.5db stage, to facilitate it's individual useage. The downsides of that scenario (for my tastes at least) were:

A) It required "non-use" of the of the reactive inductors in the -3.5db only mode.
B) If I recall correctly, it took away the option of being able to use an 8Ω load, as the -3.5db by itself reflected that 8Ω load unfavorably at the intended 16Ω input to this unit!
​

Sorry for the long rant! I just wanted to clear up and apologize for any confusion I may have caused by sticking my own in here!

And really think about my considered opinion that if you are in a situation that attenuation is "TRULY NECESSARY" at least -7db will be the absolute minimum required and that the level of complication of making that -3.5db stage availble "on it's own" is simply not worth the effort, in the long run. This is especially true when thinking of the possibility of brains farting and maybe even someone else who doesn't understand, utilizing your rig!

Just My Multiple Piles Of $.02 & Likely Worth A Bunch Less!
Gene

 

[matttornado]

Thanks, Gene! yeah I think you're right that the first stage should automatically be on with -7db reduction right off the bat even though it'll take me down to 20+ watts. We all know that the difference in volume between a 100 watt & 50 watt amp isn't that much, right?

I was confused by the latest design "M" drawing with switch 1B, 4A, & 4B all clumped together - hard for me to follow the signal flow but now I don't even need to worry about it.

Should I be using ON - ON switches? I see people using on / off in some of the pictures.
Oh and can I use Cliff Marshall style "SWITCHED" jacks or do I need to use un-switched type? shouldn't matter, right?