freestompboxes.org

I explicitly said the amp should also be plugged into a load. To be clear, plugged into speaker cab, and an interface at the same time, the cab being the load. Just another way to skin a cat - the load the amp sees is perfect, and the DI signal captured can have an IR applied and the end result is really good since the load is really good. This isn't silent recording of course, but it does mean you can record great tracks without setting up mics.

Here's the schematic. Values for the inductors were measured with my Atlas LCR. They match close to off-the-shelf values for speaker crossover inductors.
Should be an easy DIY if you're interested.
The isolated grounding is smart and works well. Can't speak to whether it would sound better with a physically larger iso transformer or not, though I do wonder.

More Pics...

Yes, it's almost exactly this:
From: https://www.aikenamps.com/index.php/des ... d-emulator
With the values changed to reflect it being for an 8 ohm load.

Can't help but wonder what those diodes, especially the blue LED, are doing to the signal.

soulsonic wrote: ↑04 Aug 2020, 06:18 Yes, it's almost exactly this:

From: https://www.aikenamps.com/index.php/des ... d-emulator
With the values changed to reflect it being for an 8 ohm load.

Can't help but wonder what those diodes, especially the blue LED, are doing to the signal.

White on white takes a bit of reading!

deltafred wrote: ↑04 Aug 2020, 09:14

soulsonic wrote: ↑04 Aug 2020, 06:18 Yes, it's almost exactly this:

From: https://www.aikenamps.com/index.php/des ... d-emulator
With the values changed to reflect it being for an 8 ohm load.

Can't help but wonder what those diodes, especially the blue LED, are doing to the signal.

White on white takes a bit of reading!

Right click -> "Open Link in New Tab", and it will magically have a black background. Thank Mr. Aiken for drawing it this way.

soulsonic wrote: ↑04 Aug 2020, 05:58 Here's the schematic. Values for the inductors were measured with my Atlas LCR. They match close to off-the-shelf values for speaker crossover inductors.
Should be an easy DIY if you're interested.
The isolated grounding is smart and works well. Can't speak to whether it would sound better with a physically larger iso transformer or not, though I do wonder.

SuhrReactiveLoad1.JPG

Quick q - were the inductor values measured in circuit or disconnected?

Optical wrote: ↑11 Aug 2020, 22:41
Quick q - were the inductor values measured in circuit or disconnected?

DISCONNECTED!!!!!!

I figured - thanks.
I'm wondering if the enclosure might affect the inductance of those coils, the bass resonance models to quite a high frequency. Are you able to test whether the inductor values change when the enclosure is assembled?

I'm definitely trying this, possibly with toroidal inductors? What you guys think?

eduardo.medeiros wrote: ↑12 Aug 2020, 04:42 I'm definitely trying this, possibly with toroidal inductors? What you guys think?

You could try something like this: https://www.parts-express.com/jantzen-a ... l--255-624
But I think it's unnecessary to use anything this fancy. The original certainly doesn't.

This would be my pick to try for a potentially better line transformer: https://edcorusa.com/pc_series
I've used these for isolation transformers in my A/B and splitter boxes and they sound great. I have one in my parts box, and I'll see about coming up with a way to do an easy comparison.

I was just playing an old 50w Traynor head into this thing. Plugged into real speakers, the amp is clean and powerful. Plugged into this, it's almost impossible to get a clean tone. Can't barely turn the volume past 2 before it starts to distort.
Now I am very strongly suspecting the diode in the fan circuit and the LED signal indicator are actually clipping the signal when they start to conduct; like this thing has THIS built into it!

Anyone else have one of these that can verify it's almost impossible to get a clean tone?

Also, you can audibly hear the sound of your guitar coming out of it when pushing alot of power into it. I'm not sure if it's coming from the fan being modulated, but that's my best guess, considering that it's powered only by the input signal.

Interesting quirk, Eons ago I built something similar. Sort of a loadbox come attenuator / line out.
I rectified the signal to drive a fan and have never noticed the symptoms you describe.
I used a full wave bridge and cap to store some energy,, maybe that helps?
If you want I can pop the lid, Circuit schematic will be lost by now and can't be sure how I wired it all up.
I only used a resistive load,, I kept it simple as it was just to load test Amps but it runs a line out which is clean until the amp under test is close to limit. So I never noticed it introduced any distortion.
Phil.

I'm going to do a test of running a clean sine wave thru it via a clean solid state amp, and see what happens.

Did my tests. Ran a 20hz-20khz sine wave sweep through the Reactive Load. I drove the signal to the load with a large solid state PA amp (Alto APX1500). I didn't measure how much power was sent to it exactly; my criteria was to get the Load's LED glowing brightly and the fan blowing vigorously. Given the hundreds of watts the amp is capable of delivering, I'm sure it was significant. That fan was going crazy Good to say this thing can take 10 seconds worth of punishment without catching fire.

My conclusions were that it doesn't appear that the LED and fan cause any distortion from clipping. The only spot where the sine wave didn't look like a perfect sine wave was at the extreme low frequencies. The line out transformer is the most likely culprit there, and it wasn't clipping, it was the wave having a bit of a bend. This could also have potentially been caused by distortions of the coupling caps in my audio interface (it uses electrolytics for coupling). Can't discount distortions caused by the interface. I used what I feel is a good one (M-Audio Delta 1010), but this isn't precise scientific equipment, so distortions of one kind or another are to be expected. There is a kind of distortion of the wave at the extreme high frequencies too, but this isn't caused by either the load, amp or interface, it's inherent to the digital medium itself. I found the original sweep file has the exact same distortion, which appears to be caused by the sampling frequency. I generated the file at 88.2khz to avoid issues of hitting the Nyquist limit at 20khz, but it seems there is another distortion caused by samples not lining up evenly with the frequency in a way that causes the curves of the wave to become lopsided at the peaks. Never noticed this phenomenon before, but I will certainly see if there's a way to overcome it. In any case, this phenomenon was replicated exactly in the recording, so it can be said that the signal chain didn't "distort" it further. I think that, regardless of less-than-perfect setup, I've proven well that the Reactive Load doesn't cause clipping of the signal, which was my main concern.

Now, I did find that the frequency response is far from linear. I've attached the spectrum analysis here for you to see: I expected it to be nonlinear, because of the nature of the reactive load having different impedance characteristics at different frequencies, but the things that stand out to me are how sharply the high frequency response drops off, and the significant harmonic content at octave intervals (the spectrum analysis puts the peaks approximately at the note "B", the next most frequent peak is at "F#"). Probably safe to say this is what happens from the use of an LCR circuit to mimic the load, though I wonder if using different varieties of inductors and capacitors may change this behavior.
I really wonder about that high frequency rolloff. It really loses alot of energy past around 6Khz. Is this from the load or the line transformer? I guess since you'd likely be filtering out most of the high frequencies with a speaker simulator anyway it must not be a big deal, but I do wonder what causes it and if using a different line out transformer would change it in any significant way.

Got my hands on a real amp again: a big old Traynor. Been having alot of fun cranking it to 11 into the Reactive Load. Really screams.

Did you try running the sweep at a very low level? It looks like you are totally saturating out the .68mH inductor (Or whatever you are using for the top end impedance rise). When an inductor saturates, it will measure as the value of the resistive component, except that it will be going in and out of saturation and causing all sorts of intermodulation distortion and waveform distortion due to flyback voltage. A square wave test would also be quite interesting.

bmxguitarsbmx wrote: ↑28 Sep 2020, 19:11 Did you try running the sweep at a very low level? It looks like you are totally saturating out the .68mH inductor (Or whatever you are using for the top end impedance rise). When an inductor saturates, it will measure as the value of the resistive component, except that it will be going in and out of saturation and causing all sorts of intermodulation distortion and waveform distortion due to flyback voltage. A square wave test would also be quite interesting.

I didn't do a low level because I was wanting to see if it would cause clipping. I could do more sometime with lower levels, though I think what you see here is probably similar to what would be happening in normal use with a amp cranked into it.

I was thinking about a square wave too, but now I'm wondering if a square wave would even successfully make it out of my computer. Probably need to build a simple 555 oscillator to get anything like a real square wave coming in. But even then, without DC coupling, not going to be so square in this signal chain. Gonna be like a trapezoid.

My apologies. I had assumed you were using the SS amp to measure the Impedance curve of the Reactive Load.

A powerful SS amp has lots of Negative Feedback and can probably drive a non-linear load quite easily to look as if there is no clipping. A tube power amp has some output impedance (~4 ohm let's say) and some output impedance is needed to act as a voltage divider against whatever non-linearities are happening in your load. The half wave rectifier circuit may distort the output of a tube amp, while doing nothing to a SS amp because of the output impedance. You could simply insert a 4 ohm (or so) power resistor in series between the SS power amp and the reactive load and take your measurement at the same spot. It would give you a rough idea of what the impedance curve of your reactive load is, and by varying the level you could identify non-linearities whether from the half wave rectifier or from inductors saturating.