Bo Hansen DI layout

[kante1603]

Just to make sure I couldn't feed phantom twice or from different sources.
I thought about this for a while and decided to go this way,same as I'm used to do with external active D.I.s. meaning to switch phantom on the pre.

Udo.

 

[CheapChip]

Hi,

First, many thanks to Bo for giving us such a great design!

But I've finished my build few days ago, and have a problems I don't manage to solve:
When my DI is powered on, LED light is on but weak.
Then there's a sort of white noise and a little hum growing up when the metal box is touched or when an instrument is plugged in.
And last but not least, the instrument's sound is distorded (not really bad sound but not really what I was looking for).

Transistors and zener diodes were changed twice, BC550 and BC560 were even sweeped with BC547 and BC547, 3 PCB were etched and resistors are verified.
Can the transformer (OEP A262A3E) be the faulty part?

Any idea Doc, I'm going to be crazy!

 

[Bo Hansen]

CheapChip,

Must ask some questions to find out more about your problem.

I assume you connected the DI-box to a true balanced microphone input with functioning 48 volt phantom powering.

Have you built up the electronics on the Group DIY original PC board ?

Have you grounded metal box in the pc-board terminal "chassis" and not in any other ground / earth point ?

Have you connected the XLR connector pins 1, 2, 3 to the PC board terminals "XLR out" in the right order ?

Have you turned all the transistors, diodes, LEDs, electrolytic right way as marked on the original PC board ?

If you have a voltmeter / DVM, you should measure over 100 uF capacitor (set to dc volts) around 23-24 volts.
Also measure from the BC560 transistor emitter to ground about 11-12 volts.
If these two measurements are correct, should the two transistors  work right.

--Bo

 

[CheapChip]

Thanks Bo for your quick reply!

So, to answer your questions:
Yes my PCB's are etched from files posted at the beginning this thread.
Yes my DI box is connected to a microphone input with a decent 48v phantom power turned on.
XLR pinout connections seem to be ok.
And PCB is grounded with "chassis" point connected to metal box by the XLR ground pin.
And every component seems to be at its place according to documents.

For now, I can't proceed to voltmeter measurments.
I'll post results as soon as possible.

Thanks again!

 

[CheapChip]

Hi everyone!

So here is what I found:
Phantom power: 47.4v
100uf cap: 10.4v
BC560 (BC557 actually) emitter: 5.1v-5.2v

It looks like there is a problem with transistors, but it is the third pair I use.
1st pair: BC550/BC560, 2nd: BC550/BC560 (same bag as the 1st pair), 3rd: BC547/BC557 from an old stock I had.  :-\

 

[Bo Hansen]

CheapChip,

Ok, you have too low internal supply voltage.
This may be due to something on the PC board drawing too much current, and then there will be large voltage drop across the two 6.8 k resistors.

But while you say that the LED is dimly lit, if something were to draw a lot of current, so would the LED brighter instead.

Are you sure it is a normal red LED you have used, and that it is connected in series with the voltage rail as described on pc-card?

Are you sure the two 6.8 K is right, so it's not 68 k instead ?

Are you sure there are 1N4004 diodes and a 100 ohm resistor which is mounted on the PC board ?

Make additional measurements.
Once you have the DI-box connected to the microphone input with 48 v phantom powering, you should measure between XLR pin 1 and pin 2, and pin 1 and pin 3.

Tell me what you get for voltage at these points.

--Bo

 

[CheapChip]

Ok Bo!

I'll check it all and give you these measurements on tomorrow morning.
It's 11:30 PM here in France, and my tools are just stocked under my daughter's bedroom...  :-[
(I think it's time to have a workshop...)

Thanks again, Bo!

 

[CheapChip]

Hello!

All I can say about the red LED: a 5mm (that seems to be) standard red LED that I picked up from an old answering machine.
It is connected with 2 wires as described on silk screen.

6.8k resistors checked: 6.7k with multimeter

1N4004 are OK and 100 ohms restitor checked at 101.7 ohms are correctly in place.

Measurements of XLR pins:
From 1 to 3: 32.8V
From 1 to 2: 0.03V (39.2mV)  ???

Thanks!

 

[Bo Hansen]

CheapChip,

The voltage around 31 volt you get between XLR pin 1 to 3 is correct, but it should be same between pin 1 and 2.

In this case it can only be a fault in your microphone cable, or the XLR connector on the preamp/mixer, or one of the 6,8 k resistor in your preamp/mixer.

Try to do a measure direct on the unplugged cable in the female end.
(ofcourse plugged in preamp/mixer end)

--Bo

 

[CheapChip]

Hi  Bo,

I said in my first post that I was going to be crazy but for now, I'm a fool...

I made new measurements between XLR pins and I saw I had about 30 volts on PIN 2/PIN 3.

I checked wires between XLR plug and PCB for the 4th or 5th time and realized that ground was in place of the negative pin and vice versa... Mistake corrected: my DI box is ... quiet and sound is ... clean.  :-[

I'm really glad it works, but ashamed you wasted your time not on a problem but a mistake.
Again I apologize for it.

Thank you again, Bo, for this wonderful project and for your help.

 

[Bo Hansen]

CheapChip,

No promblem, happy to help.
The main thing is that you get your DI box construction to function as they should.

--Bo

 

[Sandersonic]

Here's my newly completed take on this project. I've used the Lundahl 1538XL and implemented the pad-mod mentioned earlier in this thread. I built it, plugged it in and it worked silently and perfectly straight away. This is an excellent circuit design and the kit from Gustav at PCB Grinder makes it an absolute breeze to build. I also very much recommend the XL variant for the transformer. Particularly whereas electric bass often gets lined and the extra low-end headroom is good to have. Also, the pad is a very good idea since this unit (or at least my build of it) puts out a lot of signal! Especially with, for example, an electric bass with active pickups.

 

[audiophreak]

Hello,
      I've completed a dual unit of this wonderful DI and it sounds fantastic, have used it on several sources and it just excels , however I have checked some voltages as described on Bo's page :
  " Use a digital voltmeter in dc mode and connect the negative test cord to the input jack ground and the positive test cord to the junction av the two 6,8k resistors, or the long leg of the LED if this is used, or the link if the LED not used, and here you should measure +24 volts, +/- 1 volt, if everything works ok.

The next measurment is to check symetrical clipping and headroom.
Move the positive test cord to the junction of 3,9k resistor, 10uF capacitor positive end and emitter on BC560 transistor, and here you should measure around +12 volts, +/- 0,5 volt. "

    In checking the phantom power from the console unloaded its about 50vdc , both circuits without an LED I get 25.77vdc at the 6.8k resistors and 12.96vdc at the 3.9k and 10uf junction.
  With the LED in circuit I get 27.9vdc and the 6.8k resistors and 11.92vdc at the 3.9k/10uf junction.

EDIT :  Phantom voltage when unit is connected to console is 37.74vdc without LED and 38.8vdc with LED

  I am using a 3mm red LED with a 780oHm resistor in line for 24vdc operation.

  Do these voltages seem correct ??  I mean it works and all , just want it to rock solid 

  Thanks,
              Chip

ps ...  Beautiful unit Sandersonic 

 

[Bo Hansen]

sandersonic

Thanks, I'm glad you like my DI box design, hope you will enjoy using it on many recordings.


audiophreak

Your measured voltages are ok, it does not much matter if the voltage varies quite a lot in this circuit, only they do it propotionelt up and down as the specified voltage value I have stated.

What will happen if phantom powering is lower or higher is that the DI-box electronics gets more or less "head-room", and the power consumption changes up or down.

The fact that your measurements are slightly higher is because your phantom power is a few volts higher.
The voltage I have given are measured at exactly 48 volts from the mixing console/preamp. (measured when nothing is connected)

The "phantom power" on the DI-box XLR pin 2 or 3 is about 36 volts, the voltage varies a tad if you use LED or not, and if you use LEDs of other colors.
And it will increase if you connect a series resistor to the LED.

Note that shall NOT be any resistor in series with the LED, it will just be connected as it is marked on the PC card.

Use always a red LED, because it has lowest voltage drop of all colors and steal minimum voltage from the supply rail.

--Bo

 

[audiophreak]

Bo,
    Thanks so much for getting back so quickly , I was using red LEDs, but thought a series resistor was always needed for voltage applied.
    I removed series resistor and got 26.68vdc at 6.8k resistors and 12.48vdc at the 3.9k/10uf junction.
  All is well and Thanks for this great sounding project, I will be building another pair real soon, and will post pics soon too !!    ;D

 

 

[00940]

Hi,

I'm about to build a pair. They are intended for live use (guitar, keyboard, connection with local pa system, etc...). I'll use the oep transformers. But the guy in charge is worried about ground isolation (he's been spoiled with bss ar133 in the past but money is harder to come by now).

An isolated dc-dc converter (such as this one http://www.reichelt.de/DC-DC-Inverters/TMA-1212D/3/index.html?ARTICLE=120453 , using the +/-12v in serie to get 24v) would offer true isolation. I'd use 6k8 phantom resistors and a 12v zener in front of it and some lc filtering at the output. Total current draw of the DI would be around 5.3ma. The other option is to use batteries but they have the sad habit to run down right when they shouldn't.

But is it really a good idea or should I just build the DI in stock form ? Is the groundloop breaker effective enough for stage use ?

Thank you in advance for any comment 

Ben

PS: I didn't know if it was a better idea to create a new thread or post here... sorry if I picked the wrong option  :-[

 

[Harpo]

Just some reason (there are some other) why it might not be the best idea:
When you add the -from Traco datasheet- 12mA for input current no load to your 'Total current draw of the DI would be around 5.3ma', you run out of beef with your mixer only/maybe delivering 10mA phantom current. Short circuit protection 1 sec.max. might be the next bottleneck. With your additional L/C filter you already exceeded the 100uF max.capacitive load.

 

[00940]

I learned a lesson. today.. don't request other letters in the confirmation form before posting. Your long message might disappear 

Back on topic. You're definitely right, I overlooked that point in the datasheet.

One possibility might be the LME0515/LME1215 (found on Farnell and mouser). Output is only 15vdc/16ma for 5 or 12vdc input but the no-load current is 2.5ma (according to the C&D datasheet, the murata datasheet is silent on that). Total load (circuit+converter) would be just under 8ma, which is ok with smallish phantom resistors.  That could work, with reduced headroom. But well, I'm just writing this down for reference... 2x9V batteries don't seem so bad anymore. I'll build them stock for now and see what happens.

Thanks. 

Ben

 

[[silent:arts]]

You could just use an external 24V PSU 

 

[Bo Hansen]

Ben,

Start to build my DI box in original version with our PC card, and follow my instuctions carefully describing grounding and housing etc. www.hansenaudio.se/techpage.htm

I think you will be satisfied with the suppression of hum, buzz and noise.

--Bo