DIY builds featuring DI

[mnordbye]

I recently discovered the Tech 21 Bass driver: http://www.tech21nyc.com/products/sansamp/bassdriverp.html , and i wonder if there are any DIY projects featuring Direct Injection for use with mixers etc. I tried recording with one, and i'm surprised by how good DI recording can sound, if done properly. Any such projects out there?

If not, that's good too, because i'm intrigued to start designing a circuit which involves a DI out. I'm not very familiar with the DI part of the cicruit itself, but i reckon it's a feature which takes the signal and changes the impedance to a higher level, suitable for connecting a bass guitar (for example..) directly to a mixer. I'll be a bit bold and suggest that this is a buffer?

Anyway, anyone have any suggestions to what to include in the circuit? My first thought is:

Preamp --> Tonestack --> DI circuit

This is a roughly thought out plan, and i would like anyone who has any suggestion to post them here in this thread. Of course, this is something both bass guitarists and "regular" guitarists can benefit from, and specially those who live in apartments, where you're not able to record with those 4 monstrous stacks of yours! 

[tempus]

http://sound.westhost.com/project35.htm

Also look here for some passive ideas:

http://www.jensen-transformers.com/db.html

although you may need a buffer to use these.
And some cash..

[mnordbye]

Thanks tempus!

In the first link (which was really interesting) he says to use a transformer with a 1:1 ratio and at least a 600 ohm rating. Would this little fellow work?; http://www.musikding.de/product_info.php/info/p406_1-1-Transformer-TM018.html

Thanks

[tempus]

I'm not sure...

I'm looking into putting a ground loop interrupting circuit in my pedalboard and I'm wondering about how good the transformer needs to be myself, so I'd be glad to hear from someone who knows or who has tried different transformers. You can also look at the Hum Free A/B/Y schem at geofex.com - there's some transformer recommendations there too.

[ashcat_lt]

You know what's a great idea to add into an active DI?  An all-pass filter with variable cutoff frequency.  This would effect the balanced output only, and might be used to help alleviate phase issues when mixing the DI signal with a miked amplifier. 

[mnordbye]

You know what's a great idea to add into an active DI?  An all-pass filter with variable cutoff frequency.  This would effect the balanced output only, and might be used to help alleviate phase issues when mixing the DI signal with a miked amplifier. 

You mean like a tone control with a variable frequency? I can see how that would be a smart solution to phase things, yes. I guess you would place it before the DI circuit?

I think i'll go ahead and try that 1:1 transformer from musicding, it's such a low cost part, so i'll include it with my next order.

[lvs]

I'm wondering about how good the transformer needs to be myself, so I'd be glad to hear from someone who knows or who has tried different transformers.

I tried several small transformers :

Atech ATS-003
Etal P1200
Etal P2781
Mouser 42TM018

Some I got from PC modem cards. From all of them I got acceptable results. Acceptable meaning, I don't expect hifi response from guitar amplification.

One imo important experience I had, is that the output impedance of the stage that directly precedes the coupling transformer, influences it's low frequency response. The first prototype I made used an opamp, ensuring this impedance is small enough. Nowadays I use a Zoom 9010 which has 4 outputs, to split my guitar signal to two amplifiers and to a Firepod for recording at the same time. Right after each Zoom output (300 Ohms output impedance, the manual says) which goes to an amplifier, I tried just inserting a transformer, no opamp and other active stuff involved, and I got good frequency response, so I kept it that way.

[earthtonesaudio]

In my mind the "bare minimum" for a DI output is just balanced and low impedance.  The balancing part is easy enough, just a buffer plus an inverting op-amp, like shown here: http://sound.westhost.com/p35-f3.gif plus the op-amps are low output impedance already so that's all you need. 
A ton of that extra circuitry can be left out if you don't need/want phantom power.

[tempus]

Atech ATS-003
Etal P1200
Etal P2781
Mouser 42TM018

Some I got from PC modem cards. From all of them I got acceptable results. Acceptable meaning, I don't expect hifi response from guitar amplification.

But did the transformers degrade the sound at all? I agree, a super hi-end transformer is probably not necessary here, but I just want to make sure that it sounds the same coming out as it did going in.

In my mind the "bare minimum" for a DI output is just balanced and low impedance.  The balancing part is easy enough, just a buffer plus an inverting op-amp, like shown here: http://sound.westhost.com/p35-f3.gif plus the op-amps are low output impedance already so that's all you need.

Agreed, but the nice thing about a transformer is that it should end ground loop hum problems, since the ground will be effectively disconnected from the other grounds it interacts with.

[lvs]

Re-reading my previous post I see the slightly negative connotation in "acceptable" - actually, comparing the sound with versus without transformers using a switch was a rather positive experience. I didn't do any measuring but let my bandmates' ears help me judge. All transformers worked well distortion- and frequency-wise, and certainly well enough in our live playing situation. A GR-1 synth signal also goes through them without giving me reasons to try something else.

[bighead]

Would this be useful for connecting an acoustic guitar, with active or passive saddle pickup, directly to a mixer?

[bighead]

Someone please help me understand what the Line/Speaker switch is doing. This is from the site:

The switch selects either line or speaker level from the phone jacks, and the 1k pot allows you to set the level when using a speaker source. When using speaker input, the attenuator is variable to allow for the widely differing output levels available from amps.

Some examples when you would use one or the other?

Can I just leave out the series resistor and pot with no switch?

[ashcat_lt]

You mean like a tone control with a variable frequency? I can see how that would be a smart solution to phase things, yes. I guess you would place it before the DI circuit?

Well, sort of, except it doesn't apply boost or cut to any specific frequency.  It applies a frequency specific delay.  It's the sort of thing that's done in things like the Little Labs IBP box, which people use all the time to "correct" phase-induced comb filtering when two similar sources (like a DI and a miked cab, or a close and distant mic...) are mixed together.  This would have to go after the buffer section, on the way to the balanced out.

As for the speaker/line switch:  I personally would never use the "Speaker" option.  If you've got something like a 50W head connected to this thing, seems like you'd be trying to dissapate a lot of heat across those resistors to get it down to a voltage level that's not going to destroy your microphone input.  Maybe I'm missing something, but there's a bit more that goes into a power-soak type circuit.  I think you'd be safe to leave those resistors out altogether.  This passive DI requires either some Hi-Z active stage ahead of it, or a stepdown transformer with a relatively high ratio.  I think most passive instrument DIs are like 17:1, aren't they?

[Andre]

Maybe I'm missing something

Yes, you are   

The input signal for the DI is only tapped from the speaker cord.
The speaker stays connected to the amp, that's why there's an input and an output jack in the DI circuit.

Also the 22k together with the 1k pot are no powersoak but only a voltage divider.
I don't think there will be hardly any heat generated with this circuit.
Let's say your amp output voltage is 50V.
If we use the formula P = (U * U)/R  we get (50 * 50)/(22000 + 1000) = 2500/23000 = 0.11 W
No problem for 0.25W resistors.