Carnhill VTB 9072 DI Transformer with line level audio?

[Sam0311]

Hi all,

I’m pretty new to the word of DIY and looking to get my hands dirty by making a few different passive DI boxes.

One of these will be used for balancing a mono synth. I want the DI to be as small as possible so I can tuck it neatly away behind the synth on my desk.

I was wondering if anyone had any experiences the Carnhill VTB 9072 DI Transformer and knows how well it deals with line level audio?

I was originally looking at using the VTB 9046 (10k:600) since it says it’s designed for line level input, will attenuate less and the impedances would probably be fine for what I’m using it for. I’ve now started to lean towards the VTB 9072 (144k:250) instead since I think it will make the box more versatile - the higher input impedance means I can use it for guitar or bass if I choose and the lower output impedance would be closer to what a mic preamp would need.

Link to the Audio Maintenance site for the transformers specs:

www.audiomaintenance.com/downloads/carnhill_design_guide.pdf

Anyway just wanted to have someone to run this by, would appreciate any input.

Cheers

 

[Lee_M]

I've not used either of them, but the 9072 is listed as a "High Level Audio Signal DI Transformer", which gives me the impression that it's intended for line level signals. Colin at Audio Maintenance is a helpful guy, shoot him an email or give him a call, I'm sure he'll be able to give you some hard info on how it handles line level signals.

The secondary impedance is nice and low (250Ω with parallel secondaries, around what you'd expect for a microphone output), and it gives you up to 27dB of signal reduction (or less, depending on how you wire it up), so it'll work perfectly in front of a mic preamp.

The primary impedance (144kΩ) will be good for using with synths, electronic keyboards, drum machines, etc. but it's a bit lower than you'd normally use for guitar or bass, passive instruments typically want to see an input impedance of 1meg (1000kΩ) or more.

The 9046 doesn't seem like a good choice, I'm pretty sure that's intended for use in the 1073 line input, and possibly somewhere in the 2254 as well. The impedance ratio isn't drastic enough for use as a DI, you'd want a much higher primary impedance and a slightly lower secondary impedance.

9072 is definitely the better choice, it should be ideal for synths and electronic instruments. YMMV with passive instruments (guitar, bass, etc.), I find that active DI boxes are a much better choice for those, but the passive one might be fine. The best option is to build one (or more) of each imho, the Bo Hansen active DI box sounds amazing and is really simple to build.

If you do use the 9072, I'd suggest adding a switch for changing primary impedance, the 36kΩ option (parallel primaries) should be fine for a lot of gear and it won't steal so much of your signal level.

 

[Newmarket]

Just seeing this from a "Similar threads" link.
But given the discussion of transformer impedances I feel I should clarify that a transformer doesn't have any real Input / Output Impedances itself.
The effective Input / Output impedances primarily depend on the Impedance values of the equipment / instrument. These are 'transformed' according to the square of the transformer turns ratio. There's more to it in practice but others here are better at that than myself.
The impedance values - 10k / 600R etc indicate suitable / suggested ranges for the source / destination impedances wrt frequency response / distortion etc.

 

[abbey road d enfer]

The impedance values - 10k / 600R etc indicate suitable / suggested ranges for the source / destination impedances wrt frequency response / distortion etc.

Typically, it means that the impedance with the secondary unloaded should never be lower than 10k, at the lowest frequency indicated. That's what an honest mfgr would indicate.
So if the lowest frequency is 20Hz, that means an inductance of 80H, but if the lowest frequency is 40Hz, the inductance would be only 40H.

 

[CJ]

Actually transformers do have an impedance. They have inductance and resistance.
So they have inductive reactance and resistance..
And capacitance if you wanna take it a step further. This all adds up to an impedance which will of course be dependant on f.

Add up all the variables and you have a non linear resistance to the flow of ac current.

If we apply 20 volt/ 20 hz signal to a transformer primary and 2 ma of current flows, then that transformer primary has an impedance of 10,000 ohms at 20 hz.
This takes into account dcr, stray c and hysteresis and Eddy currents.
Spell check capitalized Eddy. Not me. Eddy currents we're not discovered by Duane Eddy.

Now that same transformer will undoubtedly have another winding, otherwise it would not be a transformer, it would be a choke.

Now we have a pri z and sec z. Depending on if we are transmitting power or audio signals, their will be an impedance of both pri and sec that will work the best for each situation.

For power, the most efficient ratio of the transformer impedance to the circuit impedance, front or back, will be 1 to 1.

For what we do around here. We usually want the transformer primary to have an impedance at least. 10 times that of our mic or instrument.

Ok .now the part you read about in the Jensen white pages which states that a transformer has no impedance. They are trying to point out that the impedance hooked up to the primary will be reflected to what ever is hooked to the secondary by a formula that involves the square of the turns ratio.

So for example if you have a 150 ohm mic on the primary of a 1:4 input XFMR, then 16 times that will be 2400 ohms reflected, or "transformed" to the secondary load, if there is one.

So if our transformer has 10 times the z of the mi c like it should, then we have a 150 ohm mic into a 1500 ohm primary, which will be seen by the load on the secondary as a 2400 ohm source. This 2400 ohm source will be sitting on a 24,000 ohm secondary minus losses.

Now what Jensen is saying is what if we hook a 600 ohm mic into that 1:4 transformer? Does the impedance seen by the load change? Of course it does. It has to because of the turns ratio formula. A 600 ohm mic into a 1:4 input XFMR now equals 9600 ohms on the sec.

So in order to write off this ambiguity, Jensen says you can not assign a definite impedance to a transformer because the Z seen by the load will depend on the Z of the source. But the transformer itself has to have a pri and sec impedance. Otherwise source and load would see a dead short and the music will hashish to exist.

Do I make my self clear young man?