ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2]  All   Go Down

Author Topic: What is the max i/p level a DI is likely to see?  (Read 6922 times)

Peter Morris

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1286
Re: What is the max i/p level a DI is likely to see?
« Reply #10 on: May 06, 2013, 03:18:13 am »

It is not an off the shelf unit. It may have been built for a specific situation, I can't say.

The design is a hybrid, each channel using the two opamps in a lf353 cascaded and then feeding a transformer output stage.
Currently a variable attenuator on each channel gives from +44dbu to +4dbu of gain.
The overall gain of the unit is not an issue, I can easily make it whatever I need to. Knowing the highest level likely to be encountered allows me to setup the first opamp stage so it is unlikely to be clipped. If I use a max gain of 1 for the first stage it will handle around +22dbu before clipping at +-15v rails. I think this would be enough, unless someone decides to plug a speaker o/p into it! I guess it's impossible to totally idiot proof something!

Both active and passive DI's have their uses. FWIW though, some years ago I did some testing of transformers and I can say that a well executed active DI will outperform 99% of passive units at high level, low frequency conditions. Have a look on a cro at the o/p of some respected passive DI's with a very hot 30hz signal, it can be really ugly! Some people may like the non linearity as the core saturates, but these people probably also like to clip Midas input stages to get "that" sound. That's not for me!

Darren

Hi Darren,
If you want something special check out Ianís designs 

http://www.leonaudio.com.au/passive.htm
http://www.leonaudio.com.au/active.htm
Logged

Darren Scaresbrook

  • Full Member
  • ***
  • Offline Offline
  • Posts: 123
  • Adelaide, South Australia
    • Concert Systems Australia
Re: What is the max i/p level a DI is likely to see?
« Reply #11 on: May 06, 2013, 03:47:43 am »

Hi Darren,
If you want something special check out Ianís designs 

http://www.leonaudio.com.au/passive.htm
http://www.leonaudio.com.au/active.htm

Ian's active unit is similar in that it is using a transformer to drive the o/p.
He has allowed for a maximum +15dbu i/p signal, if I can handle +22dbu I doubt it will ever run out of swing!
I don't know if he uses a discrete fet or is referring to the fet front end of an opamp. I've regularly used a mains powered version of this that Ian built for Shaun(bass) and it always works like a charm.

Darren

Peter Morris

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1286
Re: What is the max i/p level a DI is likely to see?
« Reply #12 on: May 06, 2013, 06:36:29 am »

Ian's active unit is similar in that it is using a transformer to drive the o/p.
He has allowed for a maximum +15dbu i/p signal, if I can handle +22dbu I doubt it will ever run out of swing!
I don't know if he uses a discrete fet or is referring to the fet front end of an opamp. I've regularly used a mains powered version of this that Ian built for Shaun(bass) and it always works like a charm.

Darren

The front end is discrete and the transformers are a special proprietary design / manufacture.
Logged

TJ (Tom) Cornish

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 4284
  • St. Paul, MN
Re: What is the max i/p level a DI is likely to see?
« Reply #13 on: May 06, 2013, 08:52:46 am »


It has been a(nother) pet peeve of mine that keyboard makers don't include a low Z balanced output, it would cost them pennies at manufacture. Same goes for any product with a power supply expected to be in a back-line and feeding a snake.
A number of them are balanced - Kurzweil PC3 series, Korg Kronos, some Roland digital pianos, etc.

A curious thing I've discovered (and have commented on before) is that a number of system providers still prefer to put a DI in the signal chain as a buffer device.
Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 16486
  • Hickory, Mississippi, USA
    • Resotune
Re: What is the max i/p level a DI is likely to see?
« Reply #14 on: May 06, 2013, 11:04:57 am »

A number of them are balanced - Kurzweil PC3 series, Korg Kronos, some Roland digital pianos, etc.

A curious thing I've discovered (and have commented on before) is that a number of system providers still prefer to put a DI in the signal chain as a buffer device.

I suspect the list that don't are longer.  Yup, lots of JIC engineering (Just in case).

Getting back to the OP's question there is no spec but it's safe to ASSume that most electronic keyboards use +/-15V rails or less (often less), so unity gain with similar rails should work at their output.

Optimizing a DI input impedance for a piezo pickup, is again more like a dedicated preamp variant than a typical DI (while any cheap active DI can pretty easily deliver near 1Meg input z, I did).

Adding a DC to DC convertor so you can isolate pin 1 is clever (and quite expensive), but I prefer consoles that ignore noise current injected into pin 1, so safety grounds can stay connected. If there is hum from the connection, maybe it needs the safety ground to be in place.

JR

Logged
Don't tune your drums half-ass. Listen to what a properly "cleared" drum sounds like.   http://circularscience.com/

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6709
  • Audio Plumber
Re: What is the max i/p level a DI is likely to see?
« Reply #15 on: May 06, 2013, 12:09:09 pm »

The design is a hybrid, each channel using the two opamps in a lf353 cascaded and then feeding a transformer output stage.
Currently a variable attenuator on each channel gives from +44dbu to +4dbu of gain.

That sounds like a high impedance mic pre, not a DI. Since you have the control, I'd set the input gain low enough to handle the +22dBu and the overall gain to 1. There will be plenty of signal going down the wire to maintain goos S/N, and the final gain stage at the FOH console will fit it into the mix.

Quote
Both active and passive DI's have their uses. FWIW though, some years ago I did some testing of transformers and I can say that a well executed active DI will outperform 99% of passive units at high level, low frequency conditions. Have a look on a cro at the o/p of some respected passive DI's with a very hot 30hz signal, it can be really ugly! Some people may like the non linearity as the core saturates, but these people probably also like to clip Midas input stages to get "that" sound. That's not for me!

Darren

Which active and passive DIs in particular did you test? Did the passives have Jensen or equivalent transformers? There are many respected passive DIs with undersized transformers for high level low frequency signals. There are also active DIs with transformers after the active stage, these may or may not be better that what is in most passive DIs.

Mac
Logged

Darren Scaresbrook

  • Full Member
  • ***
  • Offline Offline
  • Posts: 123
  • Adelaide, South Australia
    • Concert Systems Australia
Re: What is the max i/p level a DI is likely to see?
« Reply #16 on: May 06, 2013, 08:29:28 pm »

Which active and passive DIs in particular did you test? Did the passives have Jensen or equivalent transformers? There are many respected passive DIs with undersized transformers for high level low frequency signals. There are also active DIs with transformers after the active stage, these may or may not be better that what is in most passive DIs.

Mac

I wish I could remember, it was around 20 years ago!
Actually I would be interested in doing this testing again and posting the results. I would need a good selection of the "usual suspects".
The results may surprise some people and dispel some myths.
 "Yes Mr. Touring engineer, this $50 DI does perform better than your esoteric, diamond encrusted unit that you can't possibly tour without"  ::)

Darren

ProSoundWeb Community

Re: What is the max i/p level a DI is likely to see?
¬ę Reply #16 on: May 06, 2013, 08:29:28 pm ¬Ľ


Pages: 1 [2]  All   Go Up
 



Page created in 0.073 seconds with 25 queries.