Gain Line-up Measurements

[pfsmith0]

In order to get a feel for the signal levels suitable for the HD500, I've made a few measurements of the gains available from the different inputs to the various outputs. I've also measured the gain of all the FX. A small Excel spreadsheet containing the data can be found here (unfortunately, I can only upload a PDF version of the data, which I've done here). A brief summary is shown below:

 

All measurements were made at 1kHz with Master Volume = max, S/PDIF gain = 0dB, and Input2 = Variax (although this didn't make any difference because I had Inputs 1 & 2 panned hard Left/Right).

 

• Guitar input pad = -5.2dB compared to normal.
• 1/4" (line) output has 6.1dB more signal than 1/4" (amp).
• Aux & Guitar (normal) have the same gain structure (except Guitar has programmable input impedance).
• Guitar (normal) input to 1/4"(amp) output has 4.9dB of loss with no FX, no Amp, and Mixer = 0dB.
• Compared to 1/4" (amp) output, the XLR output is 9.4dB less while the phone output is 15.4dB more.
• Compared to the Aux/Guitar(normal) input, Mic (min gain) has 6.2dB less gain while Mic (max gain) has 38dB more gain.

 

• Aux/Guitar(normal) input clips at 0dBFS (measured at S/PDIF) with 8.3Vpp
• Mic (min gain) input clips at -1.1dBFS (measured at S/PDIF) with 14Vpp
• Mic (max gain) input clips at -1.1dBFS (measured at S/PDIF) with 91mVpp
• CD/MP3 input clips at various levels near 20Vpp, depending on which output is used (XLR, 1/4", or Phones)

By far most of the FX have 0dB of gain, but there are several exceptions (e.g., Graphic EQ = 2.9dB, Tape Echo = -5dB, and others). FX were measured with Mix=0% (to eliminate comb filtering effects) and Gain=0dB (usually). The PDF files contain the whole list.

 

Using some of this data you can see the clip level is essentially set by digital clipping, although the mic input stage only gets you to within 1.1dB of fullscale (close enough to call it fullscale in my opinion). The output stages will not get close to their clip levels (unless you use the CD/MP3 input which bypasses the digital engine). That is, it's sufficient to look at the S/PDIF signal to keep your signal below fullscale. You don't have to worry about clipping the input stages. You also don't have to worry about clipping the analog output stages.

 

I hope you find this useful in optimizing the setup of your equipment.

IO Gain.pdf

FX Gain.pdf

[silverhead]

Thanks for taking the time to do this and for sharing your results.

[pfsmith0]

Hi hurghanico, that should be very easy to do. With Father's Day coming (in the US), it may take me a few weeks. I'll post the results here, so keep an eye open for it.

[pfsmith0]

Hi Hurghanico,

 

I was able to make your measurements today and attached them here. I saw almost an exact match between left & right channels for each of the 3 configurations. All 3 configurations were basically the same. I saw nothing significantly different between them.

 

File removed - see updated file in my post below...

[pfsmith0]

Good point! You told me preset, didn't you? Let me re-check that. Standby....

[pfsmith0]

Hi Hurghanico, I went back and remeasured the HD500 and got the same results as before. The inputs were indeed set to preset. As a sanity check I took your 1&2 input configuration and panned both channels to center and the level increased as expected, although the increase was 3.5dB which wasn't what I expected but it's reasonable I suppose.

 

I was using a 0 dBV signal which isn't particularly small but it wasn't clipping. Perhaps your signal is larger and the amp is being driven into saturation at different levels for the various cases. I believe this was your interpretation as well.

 

An idea may be to insert the FX Loop in one of the channels to allow me to tap into the middle of the signal chain somewhere an read the signal there. If that idea is worth pursing, please send me new presets with the FX Loops where you'd like to see the signal.

 

Wait, wait! I just read your original request again and you wanted the amp turned on/off. Sheesh... how did I not see this before? All my measurements were with the amp on. Let me get back to you (again). Sorry...

[pfsmith0]

Hello again,

 

I re-ran the tests with the amps turned off and the results are much clearer. I saw exactly 6.02dB of difference which is what I would expect (6.02dB = x2). Using just input 1 was x2 lower than the other two - Input 1&2 and seperate path A. Apparently there's enough compression/clipping/overdrive in the amp model that even though drive level changes (and you can hear the difference in crunch) the output level does not change.

 

I've deleted the earlier file and uploaded a new one here. The first tab shows the frequency response with no amp (and no FX or anything) and you can clearly see the 6dB difference in signal levels. The second tab shows the same thing but with the amp turned on. No difference in output level (even though the input has changed by 6dB).

 

This is also a great example of why guitar amps make lowsy general purpose audio amplifiers. The frequency response sucks (literally, there's a ~40dB notch at 6 kHz).

 

In addition, I can also verify that the HD500 qualifes for "HD". The raw frequency response is quite flat (+/-.2dB) over 20-20kHz and the 24 bit noise floor easily lets me discern signals near -100dBFS.

FR.zip