Rupert Neve Designs - RNDI - DI Box inside photos

[Bo Hansen]

First of all, an active DI-box design is no big rocket science, and as you all here know, it usually consists of an impedance converter that also acts as a buffer to drive a transformer for the balanced line to the console microphone input.
(unless it is an electronic balanced DI-box with two OP-amps without transformer on the output)

So I write this with a twinkle in my eye.

A year ago at the beginning of this thread, you talked about both Rupert Neve's "RNDI" and my "1975 active DI box" and I can add in this context that both of these two designs have a very great similarity how we solved this impedance converter/buffer compared to many other active DI-boxes on the market.

In 1975 when I made this very simple design, (published here on the groupdiy forum in 2004) I chose two cascaded emitter followers, the first a NPN driven with low current for high input impedance and low noise, and the second a PNP driven with higher current for good driving of the transformer.
I chose one transistor polarity of each i.e. NPN and PNP, because already at this time I thought this gave a nicer clipping behavior and more stable bias point.

Around 2014 when Rupert's RNDI came on the market, it was the same idea, except for the first transistor where Rupert chose an N-channel JFET instead of a BJT NPN as in my case, but otherwise very similar in all other respects, except for true output, speaker mode pad and ground lift, where instead I only have low impedance monitor/amp output and fixed ground loop suppressor.

Both of us also chose about 10 times lower input impedance than active DI boxes usually have, to imitate the standard input impedance that most guitars and bass amps have, that is around 1 Mohm.
We also both chose almost the same current consumption from the phantom feed, Rupert 4.5 ma and I 3.5 ma, this is a good compromise between output swing and drive capacity.

Ps! I know how RNDI design is made because I helped RND in Texas to service their products here in Sweden for a couple of years before they got a dealer with their own service workshop.

It is with some pride to know that I had the same thoughts back in 1975 as Rupert when he made his DI-box 40 years later.

--Bo

 

[JohnRoberts]

ancients are always stealing our ideas...

JR

 

[Whoops]

Both of us also chose about 10 times lower input impedance than active DI boxes usually have, to imitate the standard input impedance that most guitars and bass amps have, that is around 1 Mohm.

Hi Bo, most Active DI boxes, I would say around 90% have 1Mohm input impedance and not 10Mega.
1 Mega is the standard in Active DI boxes.

The RNDI has not 1Meg like the Bo Hansen DI (a great circuit) it has actually 2,2Meg input impedance

 

[Bo Hansen]

Whoops,

Yes, you are right that many of the newer active DI boxes have around 1 Mohm Ri, but some of our older classic references have 10 Mohm Ri, for example the CountryMan 85, Avalon D5.
And yes, RNDI are 2.2 Mohm, and I said around 1 Mohm referring down that way.

Ps! It's good that 1 Mohm has become a DI-box standard, because passive guitars and basses sound and behave the way musicians are used to when their instrument sees that load.

--Bo

 

[Khron]

... And just how much difference in sound is there, between 1M, 2.2M, 10M?

 

[Whoops]

... And just how much difference in sound is there, between 1M, 2.2M, 10M?

More the difference from 1Meg to 10Meg, it’s quite noticeable when you connect a Piezo pickup directly into it.
You have to try it out

Not so different with passive guitar pickups or unbalanced line level sources.

 

[Bo Hansen]

Khron,

It depends a lot on the type of guitar or bass you have, but normally not much difference around a few Mohms, but when you approach 10 Mohms or the other way down towards 100 kohms you will notice the difference.

But different types of loads (resistive, capacitive, inductive) are also a useful sound in many contexts, especially on passive basses, that's why we use passive DI-boxes containing only a transformer on some instruments, and active DI-boxes with much higher input impedance on other instruments, depending on taste.

--Bo

 

[Whoops]

But different types of loads (resistive, capacitive, inductive) are also a useful sound in many contexts, especially on passive basses

For sure, I use different DI boxes with different input impedances for Bass all the time, I use it depending on the sound I need.
I love the sound of a passive DI on bass (lower input impedance) for some old school/vintage bass sound, and a 1meg input active DI for a more modern bass sound.

As for Piezo pickups without a pre-amplifier I only use DI boxes with 5meg to 10meg input impedance

 

[Khron]

Ok, a difference, but in what direction?

 

[My3gger]

More the difference from 1Meg to 10Meg, it’s quite noticeable when you connect a Piezo pickup directly into it.
You have to try it out

Not so different with passive guitar pickups or unbalanced line level sources.

Why would you use DI for line level sources into preamp when 35-40dB pad at the input does it, except for advantages of transformer isolation?
We did comparison with 2ch Bo Hansen DI having OEP A262A3E wired 6,45:1 and Cinemag CMMI-7C 7:1 into tube preamp, input was line level signal. Cinemag sounded more relaxed, maybe bigger, both trafos had distinctive "compressed sound". Pad did better job at getting the same sound except for preamp coloration, compression was gone this way.

 

[Newmarket]

Why would you use DI for line level sources into preamp when 35-40dB pad at the input does it, except for advantages of transformer isolation?
We did comparison with 2ch Bo Hansen DI having OEP A262A3E wired 6,45:1 and Cinemag CMMI-7C 7:1 into tube preamp, input was line level signal. Cinemag sounded more relaxed, maybe bigger, both trafos had distinctive "compressed sound". Pad did better job at getting the same sound except for preamp coloration, compression was gone this way.

Not clear what you mean here. But first and obvious point is that pad alone does not give a balanced output from an unbalanced source ?

 

[Whoops]

Why would you use DI for line level sources into preamp when 35-40dB pad at the input does it, except for advantages of transformer isolation?

Unbalanced Line Level In with a Jack TS connector to Balanced Mic level out in XLR

I don't know if you ever performed Live in a stage or not.
On a Stage there's just XLR female connectors that connect to the Mic Preamps of the Live console (or Live console digital stage box).
If you have a sampler, SPDS, Keyboards, most have Jack TS outputs that are unbalanced -10dBV Line Level which is 0,316 Volts, in order to connect those devices to the Mixing Console you need to convert that signal and connector to an XLR with Balanced Microphone Level. So DI boxes are used for those duties, transformer isolation and ground lifts can also be handy.

Unbalanced -10dBV Line Level (0,316 Volts), is a similar signal to Instrument Level (100mV to 1V) and uses the same connector, Jack TS. An important difference is that Line level is Low Impedance while instrument signal is High Impedance (Hi-Z)

In a Professional Recording Studio DI boxes might be used also for the same devices.
In Pro studios, there's normally 2 types of inputs, Mic inputs (go into the Mic preamp) and Balanced Pro Line level +4 dBu (1,23v) inputs in the A/D converters units.
So a DI box (being it with Mic Level out or Line Level out) is used to convert Unbalanced -10dBV Line Level in Jack TS to a Balanced Mic level or Balanced Line level +4 dBu.

I'm not talking about a studio with a semi-pro or consumer soundcard that has a DI input on the front panel.

Hope this makes it more clear.

 

[Whoops]

Ok, a difference, but in what direction?

Piezo is a capacitive source, when connected to a DI input an High pass filter is formed.
An higher input impedance will lower the bass roll of frequency.

A piezo into a DI with 1Mega input impedance can form an high pass filter that can go up into the low mids area, so both low mids and Low end can be affected (attenuated), producing a sound that some people would describe as "thin".

Better than theory is trying it yourself and check the results.

If I'm not explaining this correctly please correct me, I'm just a sound engineer that experienced this problem much more in my line of work than studied the electronics theory

 

[My3gger]

Not clear what you mean here. But first and obvious point is that pad alone does not give a balanced output from an unbalanced source ?

Comment was about what we heard during studio testing, not live situations where one can't always choose what they want.

 

[L´Andratté]

Piezo is a capacitive source, when connected to a DI input an High pass filter is formed.
An higher input impedance will lower the bass roll of frequency.

A piezo into a DI with 1Mega input impedance can form an high pass filter that can go up into the low mids area, so both low mids and Low end can be affected (attenuated), producing a sound that some people would describe as "thin".

Becauase of that I´m using a charge amplifier for piezos. No tin can sound.
But you need high input impedance on a guitar/bass di for the case of using the passive onboard tone/level pots, which raise the moderatly high output Z of the magnetic pickups substantially.

 

[mhelin]

Jensen got a DI Box schematics, too, which also have a FET, a PNP and an optional transformer, powered from P48 (not sure if this was already mentioned):
https://www.jensen-transformers.com/wp-content/uploads/2014/08/as004.pdf

 

[Whoops]

Jensen got a DI Box schematics, too, which also have a FET, a PNP and an optional transformer, powered from P48 (not sure if this was already mentioned):
https://www.jensen-transformers.com/wp-content/uploads/2014/08/as004.pdf

That circuit is for Balanced Line Level output and not Microphone Level output, it's used to send guitar signal through a very long cable (balanced and low impedance signal) and being received in the other side by a device that converts it again to instrument level/signal.

It's equivalent to this product by Radial:

 

[Newmarket]

Jensen got a DI Box schematics, too, which also have a FET, a PNP and an optional transformer, powered from P48 (not sure if this was already mentioned):
https://www.jensen-transformers.com/wp-content/uploads/2014/08/as004.pdf

Bit of a different scheme there with the transistor stage at the guitar end powered from the P48 supply at the receiving (transformer) end.
Gives you a low-Z balanced line for the long cable run. The output goes to an unbalanced amp input (rather than a balanced line-level or mic input).

I see I crossed threads with Whoops on this. But as on Mastermind "I've started so I'll finish". Or given mhelin's location , should that be "finnish"

 

[FIX]

for sure,
I personally don’t like SMD stuff, so now that I know it’s all SMD I will not buy one for myself.
It’s a feeling based on technical arguments, it’s a psychological thing when I see SMD I think it sounds bad and when I see Discreet with standard sized components I think it’s sounds good.
Psychological factors impact a lot on how you perceive the sound of gear, as the brand in the front panel also…

I will also prefer to finish my Bo Hansén DI boxes

The problem with through hole parts now is that the manufacturers are transition all their parts (especially the good ones) to SMD and just retaining a very limited selection of through hole parts. Also when designing any value of a part above 1K will never burn out, and the smaller caps will last as well, so the only part that would fail is the transistor. Removal and replacement is actually easier.

 

[mhelin]

That circuit is for Balanced Line Level output and not Microphone Level output, it's used to send guitar signal through a very long cable (balanced and low impedance signal) and being received in the other side by a device that converts it again to instrument level/signal.

Yes as it is, but if you combine the parts of the circuit and build it into external enclosure (not inside a guitar) and replace the transformer with DI tx (10:1 or so) and J4 with XLR (remove the battery and extra components) , and take the power from the XLR (that's from a console via a mic cable) it could be your proper DI box.