Simple Attenuators - Design And Testing

[JohnH]

ok, for a test, maybe you could plug in a 16 ohm cab, and the 16 Ohm M2 both in parallel into the amp? Then you have an 8 Ohm load and about -3dB attenuation on the speaker. For this use, the M2 can be set to max attenuation with no speaker and is then acting as a 16 Ohm load.

 

[pike14130]

Just found this lengthy thread and had a couple of questions for the designers/contributors. 1) How were the inductor, capacitor and resistor values derived ie: calculated or substitution? 2) What values to use for a 2 and 4 ohm attenuator for 50W '59 Bassman clone build. I am using a Mercury Magnetics FO50BM-2MM OT with 438VDC at OT CT (Plate Supply) . Using 2 - 8 ohm Weber 10A150Ts in parrallel, Z= 4.455 ohm at 1V 1Khz, connected to the secondary 4 ohm tap and on the primary side I show Z=4.737K at 1V 1Khz.
THX

 

[JohnH]

hi @pike14130 and welcome to our thread. Yes its a long one but the first post has all the key info which I keep updated, with links to a few other key posts. There are values for 2 and 4 Ohm amps there too, in a table about halfway down. Several have been built for the old 2 Ohm Fenders and they do a great job based on feedback from users.

For these low-ohm units, I suggest to build the simplest circuit, the basic M2.

The whole thing is pretty simple but there was a huge amount of wirk in optimising the values. The basic idea is that the amp sees a load which looks like a speaker and the speaker sees a consistent and moderately high output impedance from the attenuator, typical of a tube amp. I based it around measurements from my VM combo, but there's some magic in the maths that turned out that it adapts very well to any amp.

The most important parameter is the rise in impedance with frequency, and the designs follow measurements made from real Celestion speakers. There is also a bass peak in real cabs, but the M2 doesn't include that explicitly, instead it let's the real cab develop it for itself . Hence each cab can resonate naturally. There is also a design M4 that does model the bass peak, with a big cap and inductor. But it's not usually needed except if the main purpose is for use as a load box rather than with a real speaker I recommend not adding this for the low ohm open-back Fenders.

Design is by iterative optimisation using spreadsheets. Each value is tweaked to select from available values, and give a close balance across different amps and different speakers. I also use Spice models to try new configurations. Both ways start with a seperate model to represent a speaker, based on work by Randall Aiken. Happy to discuss further

 

[Gene Ballzz]

FWIW and whoever may be interested, I have a couple 16Ω M2s listed in our classifieds!

JohnH Attenuator, M2/16ohm​

JohnH standard M2/16ohm attenuator. $300 (via PayPal F&F) plus the ride, currently $16 through USPS in the lower 48. We need private contact to consummate the transaction! And if not comfortable with F&F, I need an extra $20 for fees, etc. I actually have two of these that are absolutely identical!

www.marshallforum.com

Gene

 

[ByrdmanFL]

Good Day! I have a SC800H on the way, but it will take a while to get here. In the meantime, I have been reading the info in this thread and figured I would probably like to build an Attenuator while I wait. If the SC20H isnt what I want when it gets here, I will probably trade it back in for an SV20H instead.

I have been watching YT videos on how various attenuators are put together, and was wondering what is the advantages/disadvantages of the John H vs something like this:

 

[Gene Ballzz]

Not any really fair comparison here. Yes, that "ELEVENATOR" will certainly shave volume off of your amp/speaker combination, but will NOT even come close to retaining the tone, feel, response, dynamics and character that the @JohnH design will! The difference is dramatic and not subtle at all! Well worth the time and effort to build for $100-$150 or purchase from a builder for around $300. And FWIW, the Marshall Studio Series (SC20, SV20, Mini Jubilee 2525 or ST20) and the simple JohnH M2 are truly a match made in heaven! In a few short days, I'm gonna find out how well that M2 works with my new JTM45 Clone, built by one of the most well respected Marshall restoration techs on the planet!
Just Sayin'
Gene

 

[JohnH]

hi @ByrdmanFL

Yes what Gene said!

This type of resistive L-pad-based design is very common but tends to muffle the sound and squash the dynamics as you turn down low. Its just the nature of how it works and the reasons are down to electrical engineering, Its predictable and measurable, to do with how real speakers have an impedance that varies with frequency, and the interactions of impedances seen by the amp and seen by the speaker. I think you can hear it in the video, and some may find it to be OK, but the full volume tone is significantly punchier and clearer IMO. Im not sure how low they are really turning it down in the video, my meters are only showing a small reduction but they may have normalised the recording.

Our designs solve this in a simple reactive circuit, backed up by extensive analysis and testing across pretty much every type of amp. The feedback from many others has been a huge advantage of this open-sourced project. The basic M2 design has a few more wires than the Elevenator, but with very simple robust parts that are easily sourced and not expensive. You can turn it down as low as you want with zero change in tone.

 

[ByrdmanFL]

Not any really fair comparison here. Yes, that "ELEVENATOR" will certainly shave volume off of your amp/speaker combination, but will NOT even come close to retaining the tone, feel, response, dynamics and character that the @JohnH design will! The difference is dramatic and not subtle at all! Well worth the time and effort to build for $100-$150 or purchase from a builder for around $300. And FWIW, the Marshall Studio Series (SC20, SV20, Mini Jubilee 2525 or ST20) and the simple JohnH M2 are truly a match made in heaven! In a few short days, I'm gonna find out how well that M2 works with my new JTM45 Clone, built by one of the most well respected Marshall restoration techs on the planet!
Just Sayin'
Gene

hi @ByrdmanFL

Yes what Gene said!

This type of resistive L-pad-based design is very common but tends to muffle the sound and squash the dynamics as you turn down low. Its just the nature of how it works and the reasons are down to electrical engineering, Its predictable and measurable, to do with how real speakers have an impedance that varies with frequency, and the interactions of impedances seen by the amp and seen by the speaker. I think you can hear it in the video, and some may find it to be OK, but the full volume tone is significantly punchier and clearer IMO. Im not sure how low they are really turning it down in the video, my meters are only showing a small reduction but they may have normalised the recording.

Our designs solve this in a simple reactive circuit, backed up by extensive analysis and testing across pretty much every type of amp. The feedback from many others has been a huge advantage of this open-sourced project. The basic M2 design has a few more wires than the Elevenator, but with very simple robust parts that are easily sourced and not expensive. You can turn it down as low as you want with zero change in tone.

Thanks for the very detailed explanation! I want to make one myself and the M2 looks like what I will need. I have been reading but is there somewhere I can buy this in "kit" form as I like to do all my own soldering? Or at least a break down of every part I will need somewhere?

 

[JohnH]

The best source of parts depends where you are. So, is FL for Florida in the US? There have been a few parts lists on the thread.

Key decisions are, are you using 8 or 16 Ohms at the amp and speaker?
Do you want to save $ and order resistors from China (as Gene and I have both done), or more expensive branded ones?
Where to get the coil depending on your location.

 

[ByrdmanFL]

The best source of parts depends where you are. So, is FL for Florida in the US? There have been a few parts lists on the thread.

Key decisions are, are you using 8 or 16 Ohms at the amp and speaker?
Do you want to save $ and order resistors from China (as Gene and I have both done), or more expensive branded ones?
Where to get the coil depending on your location.

Yes, I am in Florida and I will look for the parts list. Probably 8 ohm only since both my cabs are 8 ohm. I don't mind ordering from China, It really doesn't matter.

Where is the best place to get the coil?

Thanks again!

 

[DirtySteve]

I recently purchased a M2. The only comparison I have is a MiniMass. The MM didn't do squat for me, when turned down as low as it could go it was still too loud for me to turn my amp up any and it sounded like crap, it really screwed with my tone. The M2 as is allows me to turn my amp up to 3 or 4 on the master volume which sounds way better than without it and my MV at .5. It doesn't color or effect my tone at all like the MM does. I'm currently considering adding another stage to it so I can crank my amp up even more and still attenuate back down to bedroom volumes.

 

[NickKUK]

I'm just putting together a BOM for an order to finish off the 2W JCM800 build I have. It's connected to a 16R G12M-65.

So, rather than 100W I was simply thinking of scaling down the hardware, 100W->10W, 50W->5W and 25W->3W. It should be fine with the 1.8mH still was 18AWG, and the entire thing should be relatively small to include into the back of the combo.

 

[jfromel]

@NickKUK For a small build I would half the wattage of the 50watt build. Price and size will be similar to what you are describing and will handle anything up to a Deluxe Reverb/AC15.

 

[jfromel]

@NickKUK For a small build I would half the wattage of the 50watt build. Price and size will be similar to what you are describing and will handle anything up to a Deluxe Reverb/AC15.

 

[jfromel]

V3 is a hit. The hottest DC supply I have is 19v @ 3 amps. That's 44 watts at 8.3 ohms. I ran it for an hour with the cover and core temp of hottest resistors is 90c (40% derating) chassis temp is stable at 35c. I am pretty confident it can handle a 60 watt anything even in extreme playing conditions. I might dismantle one of my 3d printers that has a 24v supply which would equal 70 watts +/- and see if I can blow it up.

 

[NickKUK]

@NickKUK For a small build I would half the wattage of the 50watt build. Price and size will be similar to what you are describing and will handle anything up to a Deluxe Reverb/AC15.

That makes sense with the heat derating too. I can chassis mount so that would be straight forward.

 

[JohnH]

I'm just putting together a BOM for an order to finish off the 2W JCM800 build I have. It's connected to a 16R G12M-65.

So, rather than 100W I was simply thinking of scaling down the hardware, 100W->10W, 50W->5W and 25W->3W. It should be fine with the 1.8mH still was 18AWG, and the entire thing should be relatively small to include into the back of the combo.

Sounds ok to me if its a basic M2 circuit. If its dedicated to such low power, as an alternative to case mounted resistors, it'd be within range of using ceramic block resistors wired point to point (10W or a pair of 5W for R1 and the rest maybe all the same at 5W for consistency?), with air around each one, either using a tag strip, or maybe in most cases, direct to switch lugs. In that scenario, keeping with full-sized switches would give better space and access, even though mini-switches have plenty of rating.

 

[diego_cl]

After a lot of hard work, I finished my M2 attenuator for 100 watts.

It required much more effort than the two M2 50W attenuators I built before, mainly because of the case construction, the selection of components for active cooling, and the limited space for installing and soldering parts inside.

I added a thermocouple so the fans activate when R1 goes above 40°C and a fan control board with dip switches that allow for dynamic speed management.

I also included a power switch and an AC outlet so the amplifier can be daisy-chained through the attenuator, ensuring that the amplifier won't be used without the fans being powered.

I tested it with a Deluxe Reverb and, with the volume at 10, it didn't reach 21°C (fans off). I still need to try it with the 100-watt Marshall head.

I tested it with a maxed-out Marshall JVM401H (100 watts) and realized the cooling was way overkill. According to the thermometer attached to R1, the temperature couldn't go above 34°C. The thermostat, also attached to R1 and rated at 40°C, started the fans when the thermometer showed 33°C. It took at least 15 minutes to reach that temperature while the room was around 15°C.

Arctic P12 fans are great and I couldn't even heard them.

 

[NickKUK]

There seems some good 50/25/16W chassis mount wirewound from TE (HSA series) that will easily cope.
I was wondering if a rotary switch could be used. Will have to have a look.

 

[JohnH]

hi @NickKUK

All fine. A rotary is nice, they're hard to find with high current rating but you don't need much for a 2W amp. But the other issue is, there are three stages so three poles needed, and across the stages there are 8 attenuation steps, so its a 3 pole 8 position switch, which is not common. Whereas the three toggles are very simple and available.