Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dan Mortensen]

A couple of days ago I was in a disagreement about how DI boxes work.

Some people at an event I was working were using a couple to serve as distance extenders for a laptop headphone jack to wireless headphones transmitter, i.e. jack --> 1/8" stereo --> 6' RCA's to 1/4" -->DI Boxes --> 20' XLR's --> L-R XLR inputs on Wireless Headset transmitter.

There was no output from the transmitter, and when I asked if the inputs to the transmitter were line or mic level and they replied line, I pointed out that the DI boxes outputted mic level. The response was that all they did was change unbalanced signals to balanced and were passive so they couldn't change levels since the pads were not engaged, to which I responded that all DI's did the same thing and the position of the pads didn't further reduce the signal but that they were now mic level going into line level inputs.

There was further discussion, and the most bull-headed and positive-in-his-opinions person prevailed (me) and the DIs were removed, but the thing still didn't work and other problems were chased down until the chain worked. Time was extremely tight and remained tight and we didn't go back to see if it would have also worked with the DI's back in the chain, but I've been wondering about it and if I was indeed right, since I am largely self-and-forums-taught and my opponent in the argument was specifically audio educated, and lots at that.

This paragraph (bolding added) from the  Whirlwind DI Box Page

"Although DI transformers are often referred to as having an intrinsic impedance, they actually reflect the impedance connected to the output back to the source and vice-versa. In a DI, the turns ratio of the transformer will usually be about 11.5:1, primary to secondary. This results in an impedance ratio of about 133:1. It's this step down characteristic of the transformer that reduces the voltage level of the line signal to mic level."

indicates that I was right, but thinking about it gives rise to more questions.

While the reduction of level from line to mic level may be a feature when running into mic level inputs, often DI's are used for direct connection to instrument pickups, which without help are definitely not line level AFAIK. So we're taking something that is already a tiny signal and reducing it further, to have to amplify it again.

Why aren't there DI's that keep things like pickups at mic level and have a switch to reduce line to mic levels?

If my opponent in the argument were right, why couldn't we then use DI's as 1:1 isolation transformer for situations requiring that?

Even if they were wrong like I think, why can't we have transformer DI boxes with 1/4" inputs and pass throughs as well as an XLR input so they can be used as iso transformers? Has there ever been such a thing? Or is it that impedance ratio thing that rules this out?

[Paul G. OBrien]

There was no output from the transmitter, and when I asked if the inputs to the transmitter were line or mic level and they replied line, I pointed out that the DI boxes outputted mic level.

Bottom line is even if there was a signal level mismatch it won't produce a "no output" condition at the transmtter it would just be a bit high or low. I wouldn't be surprised if the real problem was that they were feeding an unbalanced stereo signal into a balanced mono input somewhere along the line, this is a common mistake that can result in complete cancellation of the signal since balanced inputs are designed to cancel everything common on the hot inputs.

[Geoff Doane]

What is "mic level"?  What is "line level"?

If you've been in this business any length of time, you know there are no hard and fast answers to those questions.

1 Vrms (+2 dBu) is generally considered to be line level.  But so is anything from 300 mV to 20 V, a difference of 36 dB.

And for microphones, the output voltage depends on how loud the source at the mic is.  There could easily be a 40 or 50 dB difference between a whisper and a snare hit.

Same for guitar pickups.  It all depends on how hard the strings are plucked, and how "hot" the pickups are to start with.  I think if you actually measure the output of the pickups, they can be as high as a volt or so, into a high impedance.  So that's "line level", although the impedance part can be a bit of a problem.  As the Whirlwind page states, a transformer transforms the impedance, and along with it, the voltage.  Their example of a 11.5:1 turns ratio will drop the voltage to 1/11.5 of its original value, or a "loss" of about 21 dB.

If you have a weak signal and don't want to drop it further with a passive DI, an active box is the answer.  Most of the ones I've used (the various Radial models, perhaps others) have unity gain if there are no pads switched in.  If 1V is going in, 1V comes out the XLR.  If only 50 mV goes in, 50 mV comes out the other end.  The active circuitry takes care of supplying a low impedance to feed the balanced cable, and also presents a high impedance to the guitar pickup.  The trick is to do it without adding noticeable hiss or clipping to the signal.

A 1:1 transformer (commonly called an isolation transformer) preserves the voltage, but can't do anything about transforming impedances, which is important for guitar and bass pickups.  The capacitance in a normal mic cable (not to mention snake) will very quickly roll off the top end of a 100KΩ pickup.  If you use a DI instead, you may lose 21 dB of level, but the effective impedance driving the mic cable is now about 750Ω, much more robust for travelling through long cables.

GTD

[Matthew Knischewsky]

Why aren't there DI's that keep things like pickups at mic level and have a switch to reduce line to mic levels?

If my opponent in the argument were right, why couldn't we then use DI's as 1:1 isolation transformer for situations requiring that?

Even if they were wrong like I think, why can't we have transformer DI boxes with 1/4" inputs and pass throughs as well as an XLR input so they can be used as iso transformers? Has there ever been such a thing? Or is it that impedance ratio thing that rules this out?

DI boxes primarily are an adapter to connect an unbalanced "line" output to a balanced microphone input. There's 4 things a DI box does in this capacity:
-Impedance matching & balancing: high impedance unbalanced input to low impedance balanced output,
-Level shift: "Line" input to microphone level input,
-Isolation: through a passive transformer or active electronics, and a ground lift,
-Adapter/splitter: It has the 1/4" connections and pass throughs that are needed to connect instruments, plus an XLR output for the mixing console(s).

Historically, there was a time when the only input source to a channel would be a microphone, as mixing consoles only had microphone inputs. The DI box was created as an adapter to Direct Input or Direct Inject an unbalanced instrument signal into the console. Now most consoles to have line level inputs (either on a separate connector or on one XLR that can accommodate any signal from mic to line level) but we still use the DI box for it's convenience especially when long cable runs are needed.

Answers to your questions:
"Why aren't there DI's that keep things like pickups at mic level and have a switch to reduce line to mic levels?"

Passive DIs aren't able to do this while maintaining the correct impedance matching. Matching the impedance also means reducing the level in this application. Some active DIs have a "hot" output that need the pad engaged for any true line level source (Example BSS AR133)

"If my opponent in the argument were right, why couldn't we then use DI's as 1:1 isolation transformer for situations requiring that? "

You may use DI boxes for isolation in a pinch for you have to remember the level drop that will occur. Turning up the output of the source may help overcome some of the drop but too much isn't ideal because it's possible to overload the transformer or active DI electronics. It's better to use a true 1:1 isolation box.

"Even if they were wrong like I think, why can't we have transformer DI boxes with 1/4" inputs and pass throughs as well as an XLR input so they can be used as iso transformers? Has there ever been such a thing? Or is it that impedance ratio thing that rules this out?"

Some DI boxes have XLR inputs on them but they're usually not 1:1 if they're a true DI box.Two that come to mind are the BSS AR133 that has an unbalanced XLR in parallel with the 1/4" input, and Radial's  ProAV1 DI that has a selection of unbalanced inputs, or a balanced XLR available via a recessed switch.

It really boils down to the fact that a DI box and a 1:1 isolation box do different things and should be used accordingly. BUT that's not to say that they can't be used in other situations, just be aware of the trade offs involved so that when you just need to make something happen quickly you may make good decisions based on the equipment that's available.

[Scott Helmke]

"If my opponent in the argument were right, why couldn't we then use DI's as 1:1 isolation transformer for situations requiring that? "

You may use DI boxes for isolation in a pinch for you have to remember the level drop that will occur. Turning up the output of the source may help overcome some of the drop but too much isn't ideal because it's possible to overload the transformer or active DI electronics. It's better to use a true 1:1 isolation box.

Actually you can sort of use passive DI boxes as isolation transformers - just run the XLR output of one box into the XLR output of another (using a gender changer) and what comes back out will be the same exact signal, with no galvanic connection. Or you could go the other way, plug two DI boxes together with a 1/4" cable. There might be some compromises with respect to signal quality after going through two transformers, but that would pale in comparison with the damage caused by the very common IL-19 isolation transformer.

[Matthew Knischewsky]

Actually you can sort of use passive DI boxes as isolation transformers - just run the XLR output of one box into the XLR output of another (using a gender changer) and what comes back out will be the same exact signal, with no galvanic connection. Or you could go the other way, plug two DI boxes together with a 1/4" cable. There might be some compromises with respect to signal quality after going through two transformers, but that would pale in comparison with the damage caused by the very common IL-19 isolation transformer.

Either of these may work of course with some trade offs not necessarily just because of having 2 transformers in line. For the sake of this discussion I was keeping it simple in answering the OP's questions in regards to the signal chain as described.

Another little trick is running signal through a DI in it's reverse direction to gain some boost if needed, or if using the DI in it's normal way results in too much attenuation. A couple of weeks ago I needed isolation for a talkback mic on a festival gig- The mic is passively switched between multiple locations that I may need to talk to. I didn't have a 1:1 Iso but I did have a Radial ProAV1. When I put it inline there was too much attenuation, and the make up gain needed increased the noise floor unacceptably. So I ran signal through the box "backwards", problem solved. I'd consider this a kind of intermediate level audio hacking that requires the use of some know how and adapters that some techs may or may not have on hand.

Analog audio is great: If you know what the rules are you can learn how to break them and not make too many compromises.

[John Roberts {JR}]

We've had this discussion at length before.

In the good old days it was easier because there was only one kind of DI box, a step down transformer that "transforms" hi Z signals down to low Z, dropping level at the same time.

This is conflicted by more modern active designs that buffer, pad, whatever.

JR

[Dan Mortensen]

We've had this discussion at length before.

Hi JR,

I assumed this was true, but did two or three searches with slightly differing terms and could not find anything specifically relating to my questions, although I didn't look at more than the first couple pages of each.

Can you give me a link to the previous discussion? I didn't want to duplicate settled territory, but couldn't remember or find relevant discussion.

Thanks,
Dan

PS Thanks to all for good and interesting info! I'll look at that Radial box. I do carry one or two of their Iso boxes around.

[Lyle Williams]

It's a much better question to see than the "which brand of DI?" questions.   :^)

[Steve M Smith]

As JR says, if there is a transformer, it will usually step down the level.

If it's active, I expect it to pass the signal without amplifying or attenuating it - unless a pad switch is engaged.

The clue is usually on the printing near the pad switch (or switches) which normally attenuate by 20dB.  Without them, I assume 0dB.

I make my own active DIs.  The most basic of them is a transistor with equal resistors in the emitter and collector giving opposite polarity outputs at the same level as the base input (which is biased to about half the supply voltage.



Steve.

[ProSoundWeb Community]