A "Better" idea for an attenuator

[KGW]

I've been following the "Airbrake" thread. It looks as if the Airbrake is shunt attenuator - fixed series resistor and switchable shunt resistors (using taps of a big WW). These work fine - my hi-fi preamp has one. However the impedence changes with each switch position.

If we built a ladder attenuator, the impedence would be constant at each switch position. What do you all think? Has anybody tried this?

[d95err]

You can always get a variable Hifi L-Pad. A 100W version costs perhaps $10-20. They keep a fairly constant impedance for all settings.

I've got a 50W variable L-Pad for my 18Watt(ish) build and it works quite well.

[AdmiralB]

I see it as the opposite - fixed shunt, variable series.

[KGW]

I see it as the opposite - fixed shunt, variable series.

No, that is not how they work. Read this http://www.goldpt.com/how.html

[mhuss]

On the airbrake discussed in the other thread, except in bypass mode, there is a fixed 25 ohm shunt, and a varying series resistance, as much as 175 ohms in 'bedroom mode.'

This means the load seen by the amp varies as well, with a 30-something ohm total load at max attenuation.

A better design would be an L-Pad (or other fixed-load design) that kept a fixed load on the amp, which would reduce strain on the output stage. The problem is that this gets complicated and expensive to do at higher powers, thus the design compromises.

I use the ceramic 100-watt L-Pads for 25 watt amps and below, but find they get hots spots and tend to intermittently fail over time with 50 watt and up amps. Remember, a 50 watt amp on '10' will produces over 100 watts of heating power.

--mark h

[AdmiralB]

No, that is not how they work. Read this http://www.goldpt.com/how.html

I read it. Doesn't have anything to do with the Airbrake. If the schematic in the other thread is correct, the Airbrake is a fixed shunt, variable series.

[Dai H.]

I drew it out based on the layout drawing in the other thread, and that's how it looks to me also (fixed shunt, variable series). The shunt R comes in on "1", and stays to "Bedroom" pos., and from "2" on, an R comes in series with the speaker(s). Looking at the circuit, you could use a rheostat for the tapped R, which should give finer range of adjustment, but apparently you would need something like a 200W one min. (I think this is because the contact rating for a 20-25ohms 200W rheostat wiper/element connection seems to be about 3A and you don't want something rated less), and these seem on average a lot more expensive than the fixed R w/taps, which I guess means that these would be too expensive to make money on--losing money isn't a concern if you're DIY'ing one though, of course. Also, someone feeling experimental might try a MASS motor for the shunt to see how that works, plus consider getting rid of the BR mode portion if it turns out to not be very practically useful (one user related that the sound was grainy in BR mode).

[ampdoc1]

I used a different way a couple of years ago to regulate my amp volume. Using my VARIAC on an amp gave me the inspiration.

I installed a 25W 3 Ohm rheostat in series with the output tubes heater lines (not any of the rest of the tubes), to reduce the heater voltage from 6 to ~3 volts. It did a good job down to the point that the speaker was not getting enough voltage to reproduce sound evenly. I think at the lowest usable setting it was producing about 1/2V across an 8 ohm load (.03 Watt) with a pretty good sine wave output.
Now when I first posted this, I got some flack about stripping the heaters. But I considered this before I did the build, and ran an output tube (6L6) with a 3V heater supply 24x7 for a month. It tested the same both before and after.

You might like it or not, but it's fairly cheap and easy to try.

ampdoc

[unklmickey]

...Now when I first posted this, I got some flack about stripping the heaters. But I considered this before I did the build, and ran an output tube (6L6) with a 3V heater supply 24x7 for a month. It tested the same both before and after...

i was thinking the same, when i read this.

from other things i've read, that should destroy the surface of the cathodes.

then again, i've only read about it, and you've actually tried it.


what voltage was on the plates, and how much current?




thanks,

unk

[KGW]

No, that is not how they work. Read this http://www.goldpt.com/how.html

I read it. Doesn't have anything to do with the Airbrake. If the schematic in the other thread is correct, the Airbrake is a fixed shunt, variable series.

I'm glad that you read the Gold Point link.

Now, if you read my original post, it is not about the Airbrake.

The post about the "L-Pad" is what I was talking about. A ladder attenuator is essentially a set of L-Pads. Both the series and shunt resistor are switched at each position. With 2 100W+ tapped WW resistors, this could easily be done.

However, L-Pads are purported to NOT be the best sounding attenuators. OTOH, since each dB level is switched, you could add some filtering at each position.

[Doug H]

Here's a quick and dirty 8 ohm- 25W attenutator I built earlier this year:

http://s14.photobucket.com/albums/a320/ ... ttenuator/

I used a MASS motor, Radio Shack "In wall volume control" and a power resistor. As you can see on the schem, it does not keep a constant load on the amp. It varies from 1.5 to .5 the target load, hence the "quick & dirty" part. Not bad though, and I use it mainly in the -6db to -12db range with my 10W amp and it sounds great.

I used a 9W resistor because that's what I had lying around. It is in parallel with the 8 ohm MASS motor I had lying around to get that section down to ~4 ohms. You could do that with a 4 ohm MASS, 4 ohm resistor or an inductive load that was ~4 ohms. Name your poison - it's basically the same concept and I just used what I had.

I think the "proper" way to do this would be to use a multi-tapped autotransformer instead of the single-tapped one in the volume control and set it up like the Marshall unit. That would allow you to maintain a more constant load on the amp. I have not been able to find multiitapped autoxfmrs though, and this works pretty well in the range I care about so I'm happy. One nice thing about it is the "bypass" function is built into the rotary-switch/autoxfmr- just turn it "all the way up" for no attenuation. I'm thinking of adding a switch to the resistor so I can get full 8 ohms in the "off" position for using the MASS motor as a dummy load for a line-out.

FWIW, I think it's almost never a good idea to try use an attenuator alone to get a 100W amp down to bedroom volume. Too much fizz, loss of bass, etc, etc. I suspect a combination of techniques (PPIMV, attenuator, power-scale, etc) might be a better approach, but that's just a hunch.

[KGW]

Here are the values for a 100W/8ohm ladder attenuator (I made an Excel spreadsheet in case anybody wants a copy):

Code: Select all

P     100.00								
Z     8.00    Rin     Rg      Zin     Zout    Zin-max   Zout-load    W1       W2
dB 1   4.00   2.95    13.68   16.63   2.43    8.32      24.28        17.75    82.25
dB 2   8.00   4.82    5.29    10.11   2.52    5.05      25.21        47.64    52.36
dB 3  12.00   5.99    2.68    8.67    1.85    4.34      18.53        69.06    30.94
dB 4  16.00   6.73    1.51    8.24    1.23    4.12      12.31        81.71    18.29
dB 5  20.00   7.20    0.89    8.09    0.79    4.04      7.91         89.01    10.99

[KGW]

Doug,

You're attenuator is quite cool.

Have you heard/seen the transformer volume controls used in hi-fi? I would love to get a pair (from Bent Audio) for my hi-fi preamp. For a guitar amp attenuator, those kind of transformers would not handle the power.

The auto-former idea may work, but I'm guessing that it may not sound very good.

The "Bridged-T" attenuator (http://amps.zugster.net/articles/attenuation) looks interesting. It wouldn't be much of a stretch to build a ladder that switched R1 & R2 in the Bridged-T.

[AdmiralB]

Now, if you read my original post, it is not about the Airbrake.

My mistake. I thought this statement:

It looks as if the Airbrake is shunt attenuator - fixed series resistor and switchable shunt resistors (using taps of a big WW).

pretty much "is about the Airbrake".

[mhuss]

With 2 100W+ tapped WW resistors, this could easily be done.

Watch your individual power dissipations, though. 1/10 the length of a 100 watt resistor is only a "10 watt resistor."

--mark