A real basic DI box with Pad and Ground Lift

[brazilianwonderboy]

I'm making a really basic passive DI box using this 12:1 transfomer:

http://www.rapidonline.com/sku/Electrical-Power/Transformers/Audio-Transformers/PCB-Audio-transformer/82387/88-2120

It's got a ground lift and a -10dB pad (I calculated the resistor values using a calculator at http://www.nu9n.com/tpad-calculator.html). I've drawn up a schem and was wondering if anyone could take a quick peek at it and point out any problems.

Many thanks

Dave

 

[bruno2000]

http://www.jensen-transformers.com/as/as066.pdf

 

[abbey road d enfer]

Always the same problem...the pad changes drastically the input impedance. In terms of performance, you should use an L-pad instead of a T-pad, which has the benefit of increasing the input impedance and slightly decreasing the source impedance presented to the primary of the transformer.
A 10dB pad is just about the worst thing in the world when you're mixing it with transformers and inductive sources (that's what electric guitars and basses are); the absolute worst is a 6 dB pad.
Inductive sources like to see ultra high Z loads and transformer primaries like to see super low Z sources. That's compatible with higher attenuation like 30-40dB and L-pads.
Now why do you need a pad?
Is it because you think you may overload the xfmr? In that case, the additional distortion that the pad brings in is probably higher...
Is it because you think you may overload the mic pre or console to which you're going to connect your DI? Most of the preamp can handle levels as high as -10dBu, which makes it good for +12 dBu at the input of your DI?
Is it because you want to use it at the speaker output of a guitar amp? In that case, you need more than 10dB attenuation.
Is it because you want it capable of handling high level electronic sources such as synthesizers or line outputs? In that case, I would use a pad with lower input and output impedance, like a 20 dB L-pad with 10k at the front and 1k at the foot.
The way you've done it makes it (moderately) useful only for active guitars and basses and synths.

Why the ground lift?
With the RC circuit between audio gnd and Pin 1, you don't need a ground lift.
Unless you feel like getting a jolt (and accessorily die) if your amp is not properly grounded.

 

[brazilianwonderboy]

Hi, thank you both for the replies.

bruno: on the circuit you linked to, what is the function of C1 and C2? Is this to do with solving the impedance problem caused by the pad that abbey mentions above?

Now why do you need a pad?

It's the last two reasons that you mention. I would like to be able to output from an amps speaker output, high level sources as well as guitars/basses etc. I think I might need to have more than one pad setting - maybe by using the optional variable pad idea mentioned in bruno's link.

Why the ground lift?

I based that part of the circuit on this guys schematic;

http://www.electro-tech-online.com/electronic-projects-design-ideas-reviews/90739-diy-di-box.html

To be honest, I wasn't a hundred precent sure what the function of the RC circuit was.

Do you know any good sources of information of the theoretical stuff that would be useful for me to really understand how to design an impedance bridging device like a DI?

Regards
Dave

 

[brazilianwonderboy]

I've got another question that's bothering me too. Taking a look at the JT-DB-E, I see that it says;

When loaded by the 1 kOhm input impedance typical of mic preamplifiers, this transformer has a 140 kOhm input impedance

This implies to me that the impedance that a source see's on the transformers primary is dependant on the impedance of the load attached to the transformers secondary, but on the datasheet for the transformer I was looking at, the given values are as follows;

Imp. Pri Z = 150+150
Imp. Sec Z = 3.75+3.75
Imp. Pri Series Z = 600
Imp. Sec Series Z = 15

I am confused as to how it is possible that a value can be given for the primary impedance without making reference to the impedance of the load attached to the secondary. Or am I misunderstanding what was being said in the Jensen datasheet?

Needless to say, I'm a confused in general. Can anyone ease my aching brain?

Regards
Dave

http://www.jensen-transformers.com/datashts/dbe.pdf
http://www.rapidonline.com/netalogue/specs/88-2120.pdf

 

[pyjaman]

Hi,
Those are the impedance values the inductance of the windings are optimised for. Of course, the pri/sec ratio stays dominant, and can't change. But this same transformer could be used for example as pri: 100R paralleled, 400 serie / sec: 2.5 paralleled, 10 serie. Those values are "ballpark indications". That means The transformer is best optimised for this kind of loads. It couldn't be driven by way higher source imp. without a serious low freq. loss because of a too low primary inductance. but for around 600R source imp. serie connected and below, it's fine . But I would expect a quite High LF cutoff from it if used 140K:1K IF driven By tenths  or hundred Kohms source impedance. The Jensen transformer, on the other side is best optimised for higher source impedances (higher winding inductances).

I've been thinking about making a passive DI myself... considering the price of the Jensen transformer, I finally ordered two Haufe tx from Volker and will make 2 Bo's DI instead for about the same price... and I'm sure I won't reget it... Just my 2 cents...

Laurent.

Laurent.

 

[abbey road d enfer]

Hi, thank you both for the replies.

bruno: on the circuit you linked to, what is the function of C1 and C2? Is this to do with solving the impedance problem caused by the pad that abbey mentions above?

C1 & C2 in the Jensen schemo are there to prevent hypothetical DC entering the primary. It's not justified IMO. A proper solution for a non-existing problem.

Now why do you need a pad?

It's the last two reasons that you mention. I would like to be able to output from an amps speaker output, high level sources as well as guitars/basses etc. I think I might need to have more than one pad setting - maybe by using the optional variable pad idea mentioned in bruno's link.

With a12:1 xfmr, active basses and synths are not a problem. Now, for an amp output, you need about 30dB attenuation, that I would implement with a 10k resistor on the front and a 330R resistor at the foot. Beware that tube amps want to see a load when the speaker is disconnected.

Why the ground lift?

I based that part of the circuit on this guys schematic;

http://www.electro-tech-online.com/electronic-projects-design-ideas-reviews/90739-diy-di-box.html

To be honest, I wasn't a hundred precent sure what the function of the RC circuit was.

To be honest, neither do I...seems like they're attempting to cure a potential problem and doing it the wrong way.

Do you know any good sources of information of the theoretical stuff that would be useful for me to really understand how to design an impedance bridging device like a DI?

You have to read Electrical Engineering books; all the transformer theory is there.

 

[brazilianwonderboy]

Those are the impedance values the inductance of the windings are optimised for

OK, thanks for the response. I think I get it now. So does this mean that the VTX-101-003 could be used for a DI application if the secondary is used as a primary?

VTX-101-003 Specs

Ratio = 1+1:6.45+6.45
Imp. Pri Z = 150+150
Imp.Sec Z = 6.25K+6.25K
Imp. Pri Series Z = 600
Imp. Sec Series Z = 25K

 

[pyjaman]

Yes. This one (sec as pri in serie, pri as sec in parallele) will behave about the same has the Jensen one theoricaly. But as stated by Abbey, it will not reflect as high as an amp input impedance to the instrument (important if passive pickup). There is a big mood for those passive DI, But I stay reluctant to plug it in parallele to my guitar pickup straight into my tube amp... Just personnal.

Laurent.

 

[abbey road d enfer]

There is a big mood for those passive DI, But I stay reluctant to plug it in parallele to my guitar pickup straight into my tube amp... Just personnal.

Laurent.

+1 here. Passive DI's are a quick n' dirty fix, but they're without exception sonically inferior.

 

[bdubya]

Funny you guys are talking about this.  I just recently sold my radial JDI (passive, jensen-equipped device) to fund the parts for 2 Bo Hansen DI's.  I've got one built with the lundahl.  I don't miss the JDI, but I did dig the "merge" function that it used, wherein it turns the "thru" or "amp out" jack into an additional input jack which is then SUMMED with the input jack before being going through the iron to the XLR out.  You lose the THRU functionality, but that's all right much of the time when recording.  I might try that one out.  It's useful for synths in live settings where you don't really care about maintaining the stereo spread, too or for otherwise saving inputs. 

 

[fdew]

Some general thoughts on DIs.
A Passive DI has a resistor and a transformer in the audio path.  This makes the transformer very important.  Chose a good one.

A passive DI will work well and be very nearly transparent with an active input including keyboards or any guitar with a battery in it or at the output of effects boxes.

A active DI may be a better choice for passive output guitars with no effects.

The ground lift is there to control hum caused by ground loops.  BTW  a DI box is connected to a microphone cable with a tiny shield and the other end of that shield may or maynot be grounded to the building saftey ground depending on the board manufacturer.  Therefor the DI box is in no position to be a safety ground. If a guitar amp is plugged into a AC power line then it must be grounded at that plug.  As a sound guy I will not allow any piece of gear that was made with a grounded plug to be used unless that plug is still there and working. 

The capacitor from pin 1 to ground is to provide a RF ground when the ground lift switch is open.  this is to help stop radio reception.

When you build your box you may want to consider using a plastic shell XLR connector so that if some one plugs in a connector that has Pin 1 tied to the shell your ground lift switch will still work.

The top of this page has a schematic for a basic DI box
http://www.jensen-transformers.com/datashts/dbe.pdf

the same site also has some great white papers on transformers, impedance, and grounding.
http://www.jensen-transformers.com/an/Audio%20Transformers%20Chapter.pdf
http://www.jensen-transformers.com/an/an002.pdf
http://www.jensen-transformers.com/apps_wp.html

It is personal preference but I like to have as few switches on a DI as possible as there is less likelihood of things being in the wrong position and having to run back and forth from the platform to the board to figure it out.

Frank
http://lbpinc.com/DI.html

 

[skipwave]

Funny you guys are talking about this.  I just recently sold my radial JDI (passive, jensen-equipped device) to fund the parts for 2 Bo Hansen DI's.  I've got one built with the lundahl.  I don't miss the JDI, but I did dig the "merge" function that it used, wherein it turns the "thru" or "amp out" jack into an additional input jack which is then SUMMED with the input jack before being going through the iron to the XLR out.  You lose the THRU functionality, but that's all right much of the time when recording.  I might try that one out.  It's useful for synths in live settings where you don't really care about maintaining the stereo spread, too or for otherwise saving inputs. 

Just add a switch and some summing resistors in front of your Bo Hansen and you've got a "merge" or stereo-to-mono function. Easy add-on.