4 Ohm Airbrake??

[Wilco]

I am considering building an Airbrake-type attenuator. As I understand the Airbrake can be used with any 4, 8, or 16 Ohm amp, but has an impedance closest to an 8 Ohm amp.
I would like to use the attenuator with my '63 blonde Bassman mostly, which has just one 4 Ohm speaker output (and no 8 or 16 Ohms taps on the OT). Therefore I was wondering what would be needed to build an Airbrake with a 4 Ohm impedance.
I guess the resistor values and wattages need to be tweeked, but how?
Thanks for any help.

[d2camero]

post a schematic - we can take a look from there...

[Wilco]

I thought I wasn't supposed to post schematics of the Airbrake here, as it is still in production.
A lot of info is in this thread though:
http://ampgarage.com/forum/viewtopic.ph ... y+airbrake

Edit:
A schematic might be found on page 3 of this thread:
http://ampgarage.com/forum/viewtopic.ph ... e&start=30

[peesinstew]

To make it work wrt 4ohm, like the current design works wrt 8ohm just cut each resistor value in half. The only thing wattage matters for is to make sure that the attenuator can handle the amp you plug into it, so keeping everything the same wattage for a bassman would be safe.

The trouble is that you'll probably have a hard time finding a 12.5Ohm power resistor. Mouser has 10ohm adjustable and non adjustable 100W resistors, which will get you closer to 4 ohms total load on most settings, but it will effect how much attenuation you get.

If you want to check out the actual loads seen in each case, do a search for and download the duncan speaker load calculator. I think you have to have MS Excel, but you can plug in speaker impedance values for different hookup combinations and see what the total impedance of the system is as well as the power that each component will need to be rated for. Just plug in the resistor values where they would be in relation to your cab (which you can treat as a single speaker).

[Wilco]

Peesinstew, thanks for your reply.
I know the Duncan website, and their tone stack calculator, but didn't know the speaker load calculator. I will have a look at that.

[Wilco]

I'm sorry, but I don't think this works right, using the speaker load calculator. I don't get any results even close to the impedances that are supposed to be there.
I'm wondering if I am doing this right? If I compare this to, for example, the L-pad calculator on Webers page, I am getting totally different results.

[peesinstew]

I hadn't seen that L-pad calculator before. That looks useful too, although the connection used there is different from the Airbrake. On the Duncan Load Calculator, use the "3 s+p(B)" tab to figure out the total impedance presented by the speaker and attenuator where:
Speaker A is your speaker or cab,
Speaker B is the series resistor (the adjustable one)
Speaker C is the resistor to ground.

You can flip-flop A and B just as long as you know which is which.

for example, plugging in the values on the schem I've seen:
no attenuation(don't use the calculator, this is just your cab): 4 ohm
1st step:3.4 ohm total
2nd step:7.1 ohm total
3rd step:9.4 ohm total
4th step:11.1 ohm total
5th step with no rheostat:13.4 ohm total
5th step with rheostat at max attenuation:22 ohm total

This quickly gets out of the nominal range for your 4 ohm expected load. If you instead use a 10 ohm fixed resistor and a 10 ohm adjustable resistor with steps at 2, 4, and 6 these values go to:
2.9, 3.8, 4.4, 5.0, 5.8, and with a 50 ohm rheostat turned to max attenuation: 8.6

This is still a somewhat large variance, but it's a bit more manageable. You can play around with different values too. Mouser also has 5 ohm 100W fixed and adjustable resistors.

[mhuss]

The input impedance of the stock Airbrake varies significantly, particularly at the lower setting. I modified the design a bit to keep it close to eight ohms on all settings, and close to 3 db between steps, and it's been working fine for me. Contact me off-list for the schemo if you'd like.

--mark

[Wilco]

Thanks guys for your help sofar.

@mark: That is very interesting. Have you used the (modified or unmodified) airbrake with different speaker cabs (4, 8, 16 Ohm)?

@peesinstew: I guess what I did wrong was using the wrong speaker layout in the calculator (3s+sA). Your calculations are obviously the correct ones. Thank you for that.

The imput variance on the airbrake isn't all that constant, but as I understand it, it is assumed that's not a problem because of the tolerance of tube amps to impedance mismatch. I can't find any info wether this is actually true or not. Ken Fisher seemed to think so, so I guess there must be some truth about it.
What is your experience with this?
I also read somewhere that the attenuation steps change when using a different impedance speaker cab: the steps are -3dB and an extra -1.8dB for every step thereafter, but this is when using an 8 Ohm cab. When using a 4 Ohm cab, the attenuation steps are twice those values, with 16 Ohm they are half. Could be another reason to make seperate versions for different impedances.

[Wilco]

I have one more question: How do you calculate the level of attenuation in dB from the speaker load and the series and shunt resistor values?

[Wilco]

No ideas about how to calculate the level of attenuation when the used resistor values are known?

[mhuss]

Ohm's law. Use Ohm's law to calculate the current going through each branch and use P=I^2 R to calculate the power going through the load.

The power decibels are just 10 x log(Pout/Pin). For example, if 100 watts goes in and 50 watts goes out, that's 10 x log(50/100) or ~ -3 db.

Note that this is the *power change*. The volume change will be different, and depends heavily on the speakers. The *perceived* volume change is even more variable and depends on recent exposure as well as the starting volume.

--mark

[Wilco]

Thanks again Mark. I figured it would have something to do with good old Ohm's law, but I didn't know how to convert to dB.
I want to try to use some of your ideas to adapt the Airbrake to a more constant load and convert it from a 8 Ohm to a 4 Ohm device. Before I order my power resistors I'll check and see if my attenuation steps are approximately 3 dB steps.

[Wilco]

Just a progress-report. I discovered that the Duncan load calculator that Peesinstew suggested not only calculates the total impedance, but it also calculates the power going through each speaker (or in this case one speaker and two resistors) for you.
If I then take the method Mark mentioned for calculating the attenuation-level, I can easily calculate the right resistor values for a 4 Ohm Airbrake-type attenuator.
My variation on the design will be to use two 15 Ohm wirewound Ohmite power resistors with adjustable lugs (for both resistors in this case) and a 50 Ohm rheostat. By varying the resistance on both resistors I can keep the load very close to 4 Ohms on every setting. It will be a slightly more complicated design than KF's Airbrake, but it'll probably be a lot safer.
With a 4 Ohm cab and two 15 Ohm resistors I am going to use the following values for the series and shunt resistors (using the adjustable lugs to vary):

The cab is 4 Ohms in all attenuation positions of course.
The series resistor will be 0, 2, 4, 8, 12, 15, and 65 (with 50 Ohm rheostat maxed) Ohms respectively.
The shunt resistor will be 0, 15, 8, 6, 6, 6, and 6 Ohms.
This makes for a total impedance of: 4, 4.28, 4, 4, 4, 4.56, and 5.56 Ohms.
Attenuation level will be 0, 3.22, 6, 9.54, 11.66, 12.96, and 23.37 dB.

As you see, a much closer impedance and nice 3 dB attenuation steps almost all the way. The series resistor needs 4 lugs, the shunt resistor needs 2. This makes a slightly more complicated design, but not an extremely complicated one for that matter. Even in the most extreme attenuation-setting, the impedance is 5,52 Ohms. That is a 1 : 1,38 mismatch. No problems there.
Should be quite easy to modify the 8 Ohm Airbrake in a similar way.

[Wilco]

Another update: I discovered that Mouser doesn't sell Ohmite wirewound power resistors, adjustable lug and 100W at values of 15 Ohm. They do have 10 Ohm resistors (as Peesinstew already mentioned).
It wasn't to much of an effort to make the calculations for the 10 Ohm resistors though. The total impedance varies between 3,44 and 5,06 Ohms, with attenuation steps: 3 - 5,2 - 7,2 - 9,2 - 10,9 - 23 (rheostat maxed) dB. Quite usable.
I ordered the parts this evening, so I should be on my way in a couple of days. That is, if I get a chance to solder and test the whole thing before my ski vacation.