I'm not sure I'm quite understanding your requirement here but I'll make the following comments anyway in the hope they'll help.

1. If you want a dummy load then it should match the impedance of the output if you want to test at maximum power. On the 8 Ohm setting of your amp then you would need an 8 Ohm Load. A higher resistance will mean reduced power dissipation and sub-optimal loading of the valves in the amp. IF the output of the amp is low enough then 120W RMS in to 8 Ohms would give you just 32 Watts with a 30 Ohm load.

2. If you want a low level "line" output then you certainly do not want to add another series resistor to the 30 Ohm to create a pad as this would add to the problem in 1. At full power your amp will be generating around 31V RMS across the 8 Ohm (or higher) load. Obviously this is far too high for a line input! Therefore you will need to add two series resistors in parallel with the 8 Ohm load. To bring this down to say 1 V RMS then you need a 1:30 ratio of resistor values. I would suggest 3k and 100 Ohm values. 3k to the "hot" end and 100 Ohm to the cold or ground end. Then take your "line" out from across the 100 Ohm. This will givew you 1 V RMS when your amp is delivering 120 watts RMS into 8 Ohms.

Hope that helps.

EDIT: Half watt resistors will be fine for the 3k and 100 Ohm resistors but of course the 8 Ohm needs to be rated at over 120 Watts RMS and WILL get hot in use. I would put it in a box so you can't touch it accidentally but make sure there is plenty of ventilation and/or heatsink around the 8 Ohm!

The idea behind using a 30 ohm resistor is because your average guitar loudspeaker is a reactive load with an impedance curve variable by frequency. The value of 30 ohms is supposed to represent a middle of the ground resistive load, where I believe that an amplifier running at 8 ohms into a purely resistive load of 8 ohms will not respond or sound near the same as it would through an 8 ohm speaker. I've read that a purely resistive load equal to the transformers secondary does a pretty good job at castrating the amp tonally and causes a bit of stress to the power section and transformer where a higher value load resistor would not do this. This is the assumption behind the plan as it is.

On an unrelated but relevant note, the product called the Ultimate Attenuator uses a single 30 ohm resistor as a load. When I get home I will draw up a schematic for what I think you are saying and we will take things from there. I appreciate the advice thus far.

OK, I originally thought you were talking of a dummy load in order to test your amplifier which is a different story and you should, in that case, stick with 8 Ohm loads.

While there is no problem in using a 30 Ohm resistor it is certainly not representative of an 8 Ohm speaker. Of course a speaker coil is reactive but the speaker is designed to a "representative" resistance. I don't know where you got the figure of 30 Ohms from but I would be very wary of it. In the case of a single speaker system the DC resistance of the speaker will probably be 6 - 7 Ohms and of course the impedance will vary with frequency. You should also notice that the wirewound resistor you chose is also inductively wound although the inductance will be smaller than that of a real speaker.

What WILL be certain is that the higher impedance will cause a similar increase of impedance on the anode load of the valves (sorry "tubes") and hence the PA output stage will reach voltage saturation at very much lower levels. This is not to be confused with output transformer saturation which has a very different characteristic.

If you are looking to find a particular "sound" then trying different resistor values will be worthwhile. Also note that you don't really need a resistor rated 225 Watts for this as you will be unlikely to be able to get more than 30 watts into it with your amp. There are much cheaper 50W resistors available for heatsink mounting that will be fine for this application (but you would need 60W if you used 15 Ohms). If you bought 4 x 30 Ohm 50 watt you could adjust your load over a wide range (7.5 Ohms up to 120 Ohms) with simple series or parallel combinations.

Either way it does not affect the values for the attenuator I suggested as the output voltage will not vary enough to cause a problem for most line inputs.

I can see that further explanation is needed. My main purpose for making this unit is to record or play my amp silently when circumstances do not allow the volume needed for best tone possible. I am going to take the signal from the amp's preamp or FX Send, and run it through my Countryman Type 85 (or my Interface's line in, whichever sounds better I guess, the line in is probably more "correct) and apply a previously made cabinet "impulse" to simulate full amplification and miking.

For anyone reading at home that is not familiar with impulses, the theory behind impulses is to capture the tone of a miked guitar amp and capture what exactly it is that makes that performance unique. The way that this is done is that once a tone is miked up and sounding the way we want it, you use a white noise signal generator to send the test signal out of the interface, through a reamp box, and then into the amp's FX Return to be miked up using the same amp settings and mic position as the previous great sounding recording. Once this is done, you process the before and after files using a deconvolver and the software chews on it for a while deciding what exactly was done to the original white noise track to achieve the end result of the miked track, including power amp and transformer coloration, speaker response, and microphone EQ. From there the software pops out a finished file, which you load into a convolver and feed it a tube amp preamp signal and presto, instant tube amp miked goodness! **Assuming the original take was actually good sounding that is**

I happen to have a few of these on hand, and I hope to run the preamp of my amplifier into one of these impulses so that I can play silently at my own will without damaging my amp. I did some research a while back, and I will link to the original request and proceeding argumentation on the matter of load impedances. It is actually quite an interesting read which is why I am linking it here.

[[url=http://[/URL]="http://freestompbox…"]freestompboxes.org • View topic - Help build a load box?[/]="http://freestompbox…"]freestompboxes.org • View topic - Help build a load box?[/]

Primarily** I wish to use this box as I said before, to protect my amp from any harmful effects while taking the signal tap off of the preamp to record silently. The second idea came from a desire to put two things in one box just for the hell of it so that I could see the difference between taking the preamp signal into an impulse vs taking the tap off the actual transformer. This desire is secondary to just building a dissipative load box which will easily dissipate all of the output energy from my amplifier.

And here is your suggested circuit, if I am understanding it correctly.
(Expired Link Removed)

OK, I understand your thoughts better now and, yes your schematic is correct but it doesn't change what I've already said. Let's split this into 2 parts.

1. When using your line out (or FX send) then all you are looking to do is give some load to the PA stages. Surely if you use the FX send and do not use the FX return, then the output stage will be idle and hence you can just leave the speaker connected (You may need to insert an unwired or shorted jack in to the FX return to do this). If you are using a line out then, yes, your output would still be active so a dummy load is required and a 30 Ohm load will probably be OK although, as I said already, the PA stages will not be correctly loaded - this is why you have selectable impedance outputs on the amp. in the first place.

2. When using the dummy load with the output attenuator, your aim is presumably to obtain the colouration of the output stages too. This is when the load impedance used has the opportunity to change things. Using something that is not the designed or optimal load will alter the properties of the output stages and this is exactly why I suggested using 4 x 30 Ohm resistors which will give you the ability to vary the PA loading and hence change the output stage characteristics hence giving you a wider "sound palette" to work with.

Guitarfreak, post: 376316 wrote: I am thinking that I may get some wire between 14 and 10 guage to do the internal wiring with because my speaker cable is 14gu and I don't want any splashback voltages or current throttling. I've been told that with a resistor of that wattage, I won't really need any heat sinks or fans, but I just want to make sure because I want to do it right the first time.

This makes me, in turn, think that further explanation is due. Although in well over thirty years as an engineer I have never heard the terms "splashback voltages" or "current throttling" but I think I know what you are meaning. The current drawn by the load will depend on the total impedance of the resistor, attenuator and wiring. The impedance of the resistor will dominate and hence determine currents. The wiring only needs to cope with the currents as determined by the resistor so 14 gauge will be fine.

Also you should understand that your PA is essentially a voltage source so the actual power delivered to the load will be load dependent. For 120 W RMS into 8 Ohms, the output voltage will be approx. 31V RMS. This from P = V*2/R (V*2 is Voltage squared). For a different load use the same formula to find the power that load will dissipate, so for a 30 Ohm load you get 31*2/30 = ~30 Watts.

With regard to heatsinks and fans, whatever load you use you can calculate the actual power you will dissipate. That is always real heat and 30 watts of heat is always 30 watts of heat. If you use convection cooling the resistor will get hotter than with forced air. Just be careful you protect yourself from the heat.

BTW, I forgot to mention the use of impulse responses. Be careful as there are always limitations with these. If you've downloaded these, remember that you will not be using the same guitar as input and also that the impulse will inevitably have some characteristics of the room they were recorded in. The results, while they could give you some killer sounds, may not always give you the results you expect. It is though a good way of exploring new sounds and always try them with a few different guitars as inputs if you can.