XLR Isolation Transformer with Amp DI

Discussion in 'Amps and Cabs [BG]' started by voided3, Jan 17, 2018.

  1. voided3

    voided3

    Nov 11, 2008
    I always bring a Radial Ice Cube isolation transformer with me to gigs to patch between my amp’s DI and the mixing board, but I was doing some testing with my audio interface last night at home and noticed the isolation transformer significantly attenuated the output signal.

    My GK 1001RB-II has a level control on the DI and I noticed that to match the signal strength without the transformer and the level at noon, I had to set the level all of the way up with the Ice Cube. I also tried it with my Tone Hammer DI and had similar findings. Is this normal behavior? Being a passive device without a buffer, I suppose this is possible, and I still had a workable gain level, but I was surprised as it seems like a roughly 15 dB reduction (more or less going from line level to mic level output).
     
  2. beans-on-toast

    beans-on-toast

    Aug 7, 2008
    For a DI transformer, yes that is normal, it is called insertion loss. Some companies recommend an active DI with a passive bass for this reason. The Jensen JT-DB-E DI trasformer has around a 22dB loss, it is 14:1 winding ratio. I've included an active DI schematic using the JT-DB-E transformer.

    The isolation transformer is different.
     

    Attached Files:

    Last edited: Jan 17, 2018
    basscapes, Crater and voided3 like this.
  3. voided3

    voided3

    Nov 11, 2008
    True, though I was surprised that it attenuates that much with an active source (my amp's DI). It still gives off a "mic level" signal that any mixing board's channel strip can work with and it's worth having the piece of mind that it prevents ground loops and phantom power damage. If I ever need to run a preamp into a power amp, though, I definitely won't be using one!
     
  4. beans-on-toast

    beans-on-toast

    Aug 7, 2008
    But it still works for you with the attenuation. I use a PBass into a Radial (passive) JDI, there is attenuation but works well enough for me. With an active source, the signal would be stronger.

    When recording digitally, you don't need a red line signal to have enough dynamic range. The level can be adjusted in the DAW when mixing. You want to leave some headroom for any plugin processing that you may apply.
     
    voided3 likes this.
  5. Yup. You don't ever get something for nothing (except on MTV).
    Everything has loses. Even a piece of wire.
    It's just a matter of how much is too much.
    Sounds like it works OK for you in this case.

    If you have a ground lift switch on your D.I. that will "fix" a ground loop problem as well.
    Phantom power should not affect equipment because pins 2 and 3 on the XLR are at the same voltage potential. (see drawing)
    So you may not really need it in many cases.
    But as long as the transformer does not attenuate or cause a frequency roll off that is beyond what you want, it doesn't hurt anything.

    phpwr.png
    Lifted from: TheEC: Phantom Power
     
  6. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    The factory spec says 0dB gain/loss at a 10K load and -5dB and change at 600 ohms. Insertion loss for the line level iso transformers (i.e. nominally 1:1) I've used is typically 1-3dB, given best standard loading practices.
     
    beans-on-toast likes this.
  7. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    No, this is not correct.

    The winding ratio on these transformers is 1:1, the insertion loss should be close to zero (in practice with a typical source impedance and a standard ~3k mic input impedance it will end up being about 1dB).

    Exactly how are you using this? I suspect that you are dealing with an application error, since transformer interfacing requires some specific termination details to insure that you are doing what you think you are doing. I spent a lot of time troubleshooting transformer coupled circuits, which were common in the older pro audio days and found a lot of un-intentional mis-wiring because it's very easy to misunderstand the application notes.
     
    n1as and ObsessiveArcher like this.
  8. voided3

    voided3

    Nov 11, 2008
    I patch the Ice Cube in between my amp head/preamp pedal (GK 1001RB-II or Aguilar Tone Hammer DI) and the mixing board channel strip live. Yesterday, I was testing it out into my audio interface (Presonus Audiobox) set to analog pass-through monitoring with headphones. On my GK head, the manual states that the output impedance is 500 ohms, not sure on the Tone Hammer.
     
  9. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    The levels should be identical from the DI out into the console with and without the isolator in line.

    Where problems typically occur is when one of the sources or loads are unbalanced, specific terminations are needed. This applies ANYTIME a 1/4" connection is used, because even on a TRS plug, if the receiving or transmitting end is really unbalanced and the ring is tied to shield, there will usually be a 6dB difference and if the ring is floating it will be all over the map.

    What are the connections for the interface?
     
  10. voided3

    voided3

    Nov 11, 2008
    My signal chain was: Bass --> GK 1001RB-II ---(via balanced XLR output)---> Radial Ice Cube --> Presonus Audiobox (XLR/1/4" combo jack input on front panel labeled Mic/Line). The interface also has XLR/1/4" combo jack inputs labeled Mic/Instrument, intended for plugging a passive guitar/bass directly into, but I did not try these.
     
  11. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Ok, I'm assuming that you are using the iTwo model based on your description. Note that the XLR inputs are mic only but have a wide range of gain control. They are not line inputs (blame that on the poor labeling of functions on the front of the unit). The line inputs are in fact unbalanced so it is possible that the unit functions in unbalanced mode if the line switch is selected. Try this using all XLR connections, and with the interface set to the mic position. You may have to turn the input gain on the interface down a fair amount, but there's plenty of headroom available.

    I see this kind of thing pretty regularly, the new interfaces try to be so multifunctional that they can become less functional to normal applications.
     
    havnjero and Wasnex like this.
  12. voided3

    voided3

    Nov 11, 2008
    Sorry, I should have specified: I have an older AudioBox 44VSL four channel unit. I was plugging into channel 3, which is labeled as Mic/Line. I was using all XLR connections from amp to Ice Cube, Ice Cube to interface (other than the 1/4" cable from my bass to the amp's instrument input). The channels labeled Mic/Instrument have more gain on tap, I believe.

    Plugged into channel 3 (Mic/Line), I was able to get the input clipping light to come on at about 10 'o clock on the preamp gain without the isolation transformer, withe the amp's DI output level at noon. With the Ice Cube, I had to set the preamp gain almost to 1 'o clock to get any preamp clipping on the interface, with the amp's DI output level up all the way.
     
  13. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    This should work just fine. There should be almost no insertion loss. You might double-check the manual on the interface/DAW software just to be sure there's no routing quirks that cause the clip detect circuitry differently, though I doubt this is the case based on your description.

    I would be looking at either a bad XLR cable (open pin 3 would be my first guess) or a defective transformer... POSSIBLY a DI transformer rather than the correct 1:1 transformer. A DI transformer is wound differently with a 6:1 impedance ratio. There is also the remote possibility that there is a problem with the DI out, an effectively open pin 2 or 3 which would cause the transformer to pass only common mode signal.

    Good luck with your troubleshooting.
     
    Wasnex likes this.
  14. beans-on-toast

    beans-on-toast

    Aug 7, 2008
    My first reply about insertion loss was intended to be about a DI transfomer, such as the 14:1 ratio of the Jensen, not about a 1:1 isolation transformer. Sorry for any confusion.
     
    BassmanPaul and Passinwind like this.
  15. voided3

    voided3

    Nov 11, 2008
    I checked the cables I used on a multimeter just a moment ago and they are wired correctly, no shorts. However, what should I expect to read resistance-wise when measuring across the pins of the Ice Cube? Pin 1 input and output measured 43.4 ohms (interesting, as this is ground), but both pins 2 and 3 read open. Here is the product page with the specs for reference: IceCube™ IC-1 - Specifications
     
  16. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Across the device pin 2 to pin 2 should be open and pin 3 to pin 3 should be open. Looks like they installed some ground current management into the pin 1 circuit which is fine.

    What do you measure pin 2 to 3 on each end. This will tell you more about the winding ratios based on the DCR's being somewhat proportional to winding lengths.
     
  17. We're supposed to read the application notes? :wacky:
     
    Kijuer and agedhorse like this.
  18. Wasnex

    Wasnex

    Dec 25, 2011
    I read through the thread and didn't pick up on why you are using the isolation transformer. It should generally be unnecessary, unless something in the signal chain has a malfunction.

    I would trouble shoot in the following order: Confirm your XLR cables are good and properly wired (use a multi-meter or cable tester), try the isolation transformer on another amp, try another isolation transformer on your amp, try a different mic pre. It appears you have done most but not all of these steps.

    Beyond that, I don't find the GK DI to be particularly good sounding to begin with. Perhaps you would be better off using a quality DI.
     
  19. voided3

    voided3

    Nov 11, 2008
    Pins 2 and 3 on the input side read 239 ohms and pins 2 and 3 out the output side also read 43.4 ohms.
     
  20. I would expect a 1:1 (aka isolation) to measure the same on both sides.
    You may have an impedance matching transformer rather than an isolation.

    Can you post a pic or give any info (model # and such) that is printed on the device?