DI Box for Speaker out level

[steppenwolf]

Hi guys!

I am thinking about building a DI box that takes the signal from my guitar amp's output and delivers it as balanced line level signal to my desk.
I'm toying with impulse responses right now and wanted to do some recordings of the pure output signal of my amp and running it through cap simulators.
I'm afraid to hit regular DI boxes with the output of my amp so I wanted to find a solution for that, preferably DIY ;-).
The amp is of course either connected to a speaker or a dummy load while recording.
Levels should be pretty high at let's say 15-30 Watts into 8 Ohms. I expect some 25 dBu at around 30 Watts output.
A level like this is of course way to high especially if I add additional 6dB with the output driver.
The idea is to connect a pot, let's say 5k in parallel to the OT secondary and take the signal from the wiper into an opamp configured as unity gain buffer.
After the opamp, the signal goes into a that1646 output driver adding 6dB of gain and from there into my desk.
I thought about going directly into the 1646, but the datasheet says it should be driven by low source impedance. I think the pot will be turned down to 1k or so giving me some 4k source impedance, so I thought buffering the signal in this stage might be better...
Do you think something like this could work?
Thank you very much!
Best, Stefan

 

[PRR]

You are over-elaborating.

Pad-down with 5K fixed and a 1K pot. Take signal from pot wiper and ground to your line input (balanced or unbalanced).

 

[steppenwolf]

Oh, cool, that easy? ;D
Thank you very much!
I just thought I shouldn't run a cable with high source impedance due to noise and hf roll off, but I think I only need around 30feet of cable, so I should be fine...
10meters of cable at 100pF per meter gives me 1nF. Driving it with a source impedance of let's say 5k gives me a -3dB cut off at around 30kHz. I guess that is ok...
Best, Stefan

 

[emrr]

-3 @ 30K for guitar?  Definitely.  You'll probably end up rolling off even more to duplicate what the speaker loses in air translation. 

 

[steppenwolf]

Yeah, I think so, too. Moreover, I did the math wrong! I'd wind up with a source impedance of max 1k. That gives me -3dB at 150k, shouldn't be a problem;-)
I will try it and report back, thanks!

 

[PRR]

> cable with high source impedance due to noise and hf roll off

You got a LOT of power there.

If you had to drive a loooong line, you could run the raw speaker feed over to your destination and then pad-down so your input don't smoke. How far can you run a speaker line at 6K loading? No audio signal or buzz can force its way IN against 20 Watts at 4/8 ohms. The actual line to a speaker has losses due to wire resistance and inductance, but 30 feet of lampcord works fine for any stage-amp use. Loaded with 6K the resistance/inductance is not a problem out 1,000X further, like 8 miles! You could use a lighter gauge than lampcord. (Actually there are other problems as you approach a mile, but we'll leave that to telco engineers.)

For a mere 30 feet, any darn wire will work. If it is convenient to pad at the amp, that's fine in any room you would be recording in. If you must play next to the arc-welder, run speaker-level to a pad near the recorder.

Un-balanced... well, again you won't "take" interference from anything smaller than an arc-welder. It will radiate guitar into nearby low-level lines. If that turns out to be a problem then it is wise to pad at the amp so the line runs nearer 1V than 10V, 1/10th the radiation. Your low-level lines will typically be shielded, giving 20+dB rejection, and balanced, giving another 20-40dB. That won't be enuff if you tape the gitar-amp lead to the microphone leads, but if you can lay it even a foot apart and cross at right-angle, crosstalk is nil. (Real stages have 120V/240V power lines all over, 10+ times worse than a DeLuxe's output, and humming a different tune; this is not a big problem as long as they don't run WITH the audio cables.)

The remaining problem is ground-loop through the grounded speaker jack. If you go into a "balanced" input, and take speaker common to the "-in" pin, a good balanced input will reject 20+dB of ground crap. In smaller studios even this is not essential; I've had clean results going into unbalanced inputs.

Do be SURE your guitar amp does not have a "capacitor ground". When the USA was all 2-pin wall power, we put a 0.05u cap from chassis to one pin, and flipped it for least buzz and shock. That has not been done for a very long time, was probably never legal in your land, but some cherished vintage amps may not have a proper 3rd-pin ground conversion. Technically, if you work on a dry wood floor (NOT concrete, nor even wood on concrete) and have the amp hard-grounded to studio console, you "could" lift the amp ground off the wll-ground with a ground-cheater. However if the amp's PT gets an internal short the audio cable ground is not adequate to FOR-SURE break power before someone gets shocked.

So it could be good to have a say 600:600 or 10K:10K audio transformer on the end. But 99% of the time that isn't needed.

> -3dB at 150k

Baby mice can hear that high, but don't buy a lot of records.

Pretty much anything over 6KHz off a gitar is just hash you want to lose. Some acoustic guitar may be an exception, but you don't run these through a Fender and expect 13KHz string-zing, you aim a mike.

 

[steppenwolf]

Thanks for the insight!
In fact, ground loops were my main concern. My amp is a new model amp built in germany and it is grounded with a short heavy wire from pin3 ground directly to the chassis, so I think I have a solid ground connection at the amp.
The amp sits in the recording room, the recorder in the control room. We have separated power lines for control and recording room, so I will have to see if I'm running into problems there, but I doubt it.
Actually this solution is really cool as it doesn't involve any active circuitry, that is as neutral as it gets;-)!
Best,
Stefan

 

[pstamler]

If you do run into ground-loop issues because the amp is plugged into one room and the recording gear another, a solution that's cheaper than an isolation transformer is to run a heavy-duty extension cord from the control room to the recording room, and plug the amp into that.

Peace,
Paul

 

[CJ]

if you just want to get recording, and do not care about DIY, you could go this route>
(Jensen transformer inside, yum!)

 

[steppenwolf]

If you do run into ground-loop issues because the amp is plugged into one room and the recording gear another, a solution that's cheaper than an isolation transformer is to run a heavy-duty extension cord from the control room to the recording room, and plug the amp into that.

Ok, so I basically do that to eliminate eventual ground potential differences. by doing that I'd effectively minimize the possibility of ground loops.
Good idea, great!

if you just want to get recording, and do not care about DIY, you could go this route>
(Jensen transformer inside, yum!)

Cool option, indeed, and I won't run into problems with that one. But a pot and a resistor are so dirt cheap and I might even have those parts lying around.
I'm trying that for a start and see if it works!
Best

 

[skipwave]

And if you really want a dual purpose instrument/amp DI box, I can recommend the countryhick (or tack the resistive V divider on the front of any DI you like). The transformer isolation and balanced output can solve problems that others have already posted working solutions for, so the advantage is merely the space of one less box hanging round.

http://www.diyfactory.com/projects/countryhick/countryhick.htm