Simple Attenuators - Design And Testing

[BlueX]

I'm having issues finding the inductor, plans call for a .9 mH 18 awg wire.

Can someone share a link or part number?

I've used this one, but availability can depend on your location:

Jantzen Audio, 0,900 mH, 17 AWG (D 1,2 mm). Two different spool sizes: 000-1413 (OD 45, ID 25, H 30 mm), or 000-1566 (OD 52, ID 20, H20 mm).

 

[dinkyguitar]

I've used this one, but availability can depend on your location:

Jantzen Audio, 0,900 mH, 17 AWG (D 1,2 mm). Two different spool sizes: 000-1413 (OD 45, ID 25, H 30 mm), or 000-1566 (OD 52, ID 20, H20 mm).

Found this one...for 8 ohm version. Should be a good fit.

Jantzen Audio 0.90mH 18 AWG Air Core Inductor Crossover Coil​

Jantzen Audio 0.90mH 18 AWG Air Core Inductor Crossover Coil

Jantzen Audio 0.90mH 18 AWG Air Core Inductor Crossover CoilThe Jantzen Audio 18 gauge air core inductors are precision wound with high-purity copper wire to a ± 3% tolerance. A special Jantzen high-temperature baking process bonds the wires throughout the inductor, eliminating the possibility...

www.parts-express.com

 

[JohnH]

One thing I'd advise, in a 100W 8 Ohm build, is a thicker coil wire, Like 16 awg.

 

[dinkyguitar]

Thanks,

Plans call for a .9 mH, but in the event I can't find one, can I use a 1.0 mH?

 

[dinkyguitar]

Nevermind, the answer was looking me straight in the face on the "alternate values" section

 

[JohnH]

Only Fans

This post is only about fans, What did you think it was going to be about?

As you feed an attenuator with more amp power, that power has to be dissipated as heat, hence we might use a big thick chassis, heatsinks, cooling vents etc. Such passive cooling can work very well but depending on the amp power, how it is run and the attenuator build, somewhere from about 40 or 50W up, some kind of an active cooling system is needed.

These days there are very good efficient and compact fans readily available, intended for cooling tech equipment. Typically they use brushless dc motors, at 5V, 12V or 24V. They can be various sizes, Ive seen them from 30mm to 120mm diameter or more. Bearings can be of different types such as roller, sleeve, hydrodynamic and mag lev.

There are two main ways to power these fans, either from a separate supply fed to the fan, or by rectifying some of the spare amp power to supply them without extra wires.

Supply from an external source is very simple, if such power is available, and a 12V fan fed from a 9V supply typically will run very well and quietly too. Also, it will keep running even when you stop playing, continuing to remove any extra heat that has built up.

But a fan supplied from the amp power is neat and self contained, and its simple enough to do with the right circuit and values. John Fromel has a 40mm one in the Lotus M2 attenuators which are being very well liked.

So I got a larger fan to test, such as might go into a point-to-point build. The key parameters are the voltage and current draw, the diameter and the noise. I got an 80mm one at 12V, with a current spec of 80mA. It has what is described as a hydrodynamic bearing, and is very low noise, just 18dB at its rated setting.

I ran some voltage vs current tests, also noting what it needed to start spinning. At very low voltages, virtually no current passes, so you cant measure its resistance with a meter, due to the transistors that control such a brushless motor.

At just under 4V, it starts to spin. Voltage vs current at that point implies an effective resistance of about 150 ohm in my example. At 9V, its spinning very nicely and pretty much silently, and this seems like a good target for setting it up in an amp set loud at its tonal sweet-spot. This fan is also very quiet at 12V. By 9 to 12V, its effective resistance has risen a bit to about 165 Ohm.

Reassuringly, if you work out an Ohms based just on the 12V and 80mA spec, it also results in 150 Ohm, ie close to as measured.

It seems very likely that many other fans with similar electrical specs will draw about the same and so may work similarly in the circuit described below.

There are a few options and decisions, such as whether to rectify with one diode, or with a 4-diode bridge rectifier. Also what capacitor is to be used for smoothing the dc. And most importantly, what resistor to feed the supply through so it runs nicely, and where the resistor should go, to suit the specific unit with the intended amps.

So I did some tests and made some numbers to work out what might be needed for different power ratings and output Ohms. The idea was to give the 12V fan about 9V of supply at about 70% to 80% of the amps power. That is a somewhat arbitrary choice but it seems to set it up to work well. I tested with my VM and DSL combos, then worked out a spread of values for other amps based on maths and Spice sims.

I chose to use a full bridge rectifier, since I could buy a small combined one for a dollar, or you can use four seperate diodes. The cap I used was 470uF, which smooths the dc well enough, but the fan stops fairly quickly if you stop playing. It does a good job but you can use a bigger cap if wanted.

Here are the parts that went into my test rig:



And here's the circuit, with suggested values and ratings to suit a range of amp power and Ohm ratings :



A zener diode is used to keep the fan voltage within the 12V spec when overdriven, and it may start to operate if the amps with RMS power as noted, are fully over-driven into saturation. Resistor values are shown in Ohms, and the colours indicate suggested power rating, with a x1.5 margin assuming the fully overdriven amps. With testing I'd expect that lower ratings may be ok.