Which Transformers for passive transformer box ? (Edcor, Carnhill, Lundahl )

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
ruffrecords said:
As usual the answer is it depends.

API (and Neve) output stages were designed back in the day when outputs were expected to be able to drive +26dBu into a 600 ohm load. That is 400mW of power. API output transformers, like Neve ones, are step up transformers. They are typically 200:600 or even 150:600 so if you happen to load the output with 600 ohms this will be reflected back to the input as 150 or 200 ohms. This is almost certainly not a load your interface is designed to cope with (ditto for a 5532).

But as long as I will use 1:1/600:600, etc. transformers I should be on the safe side, right?
Also, can't these API/NEVE transformers also be used as step down when wired the other way round?

The reason I bought the VTB2281 was the following blog article, where the author added additional switches to use the transformer also in a 4:1 scenario (its down at the end of the page, where he talks about it).
I added his schematic at the attachments on this post.

https://www.sonarworks.com/blog/learn/transformers-magically-colorize-your-sound/

"Below are links to download the plans, parts list, and schematics to build two awesome sounding transformer boxes. I built these with switches to change between 600-ohms and 150-ohms on the primary winding to provide for more or less saturation.
Of course, you can leave the switches out and wire them for either 600-ohm or 1:1 ratio or at 4:1 with 150-ohm primaries for about 6dB of gain and more saturation and color.
The Carnhill transformer has more taps so I wired 600/150-ohm switches for both the primary and secondary.  With four switches total and the ability to run the box at 150-ohms, you can try this box between a dynamic microphone and your mic pre-amp for extra transformer color when recording."

Would this still be safe?
 

Attachments

  • StereoCarnhillTransformerSchematic.pdf
    175.6 KB · Views: 37
I just built a stereo box for the OP's exact purpose. I used Edcor PC 10k:600 with exceptional results. 12 bucks a pop and sounds great. I'd go with Edcor.
 
Erazorblade said:
But as long as I will use 1:1/600:600, etc. transformers I should be on the safe side, right?
Would this still be safe?
The VTB2281 is a very good quality transformer. Its 600 ohm winding has an inductance of just over 12 Henries at 100Hz which means its inductive reactance at 20Hz is over 1500 ohms so it should be perfectly safe.

Cheers

Ian
 
ReRibbon said:
I just built a stereo box for the OP's exact purpose. I used Edcor PC 10k:600 with exceptional results. 12 bucks a pop and sounds great. I'd go with Edcor.

Thanks for your suggestion! But
isn't there quite a big level drop at the output of the transformer in this scenario?
 
ruffrecords said:
The VTB2281 is a very good quality transformer. Its 600 ohm winding has an inductance of just over 12 Henries at 100Hz which means its inductive reactance at 20Hz is over 1500 ohms so it should be perfectly safe.

Cheers

Ian

Thanks for your time! The transformers arrived today. I'm gonna hook them up when I got free time and report back how it worked out :)

Cheers
 
Erazorblade said:
Thanks for your time! The transformers arrived today. I'm gonna hook them up when I got free time and report back how it worked out :)

Cheers
You are very welcome. I am very interested to hear what you think of the the tonal changes they introduce.

Cheers

Ian
 
ruffrecords said:
You are very welcome. I am very interested to hear what you think of the the tonal changes they introduce.

Cheers

Ian

I could test it today for some minutes on a mono mixbus and liked what I heard, if time allows I will try to record some audio samples for an A/B comparison if anyone is interested.

Thanks and Cheers!
 
Erazorblade said:
Thanks for your suggestion! But
isn't there quite a big level drop at the output of the transformer in this scenario?

Is what the schematic called for. One of those Motown wolfboxes.
 
It would be nice to have a box with different transformers that you could change with a switch.
3 or 4 pairs of transformers.

I would definitely try OEP transformers since I always found them pretty coloured.
 
Not all interfaces are created equal and many modern ones cannot drive a 600 ohm load. Which interface do you have?

Cheers

Ian


Hi Ian, I have an AUDIENT ID22.

Preamp Specs:
INPUT IMPEDANCE (Line): >10kΩ Balanced

DAC Output Specs:
MAXIMUM OUTPUT LEVEL: +18dBu
DIGITAL REFERENCE LEVEL: 0dBFS = +18dBu
OUTPUT IMPEDANCE: <100Ω Balanced

I'm trying to figure out which output transformers I could get for some de-digitalising before my stereo buss compressor.

I like the idea of Carnhills, though am new to DIY hardware so something cheap to experiment would be preferable at this stage.

Any insight you can offer would be greatly appreciated.
Thanks
Ed
 
WOW, the Audient web site is SOOOO SLOOOOW. It maxes out both cores of my ancient AMD64 processor.

The one thing the Audient spec does not say is what load it will drive those levels into. I expect most of the time this will be 10K but to drive a 600:600 transformer you need much better than that. Yes, the quote and output impedance of 100 ohms but that gives no indication of drive capability. It does say that the topology is the same as that used in the ASP8024. I can find data for the ASP8024HE but not for the plain 8024. Anyway, the ASP8024HE is specified to drive a load as low as 2K so it might actually struggle to drive a 600:600 transformer. If you want to experiment over relatively short distances then you could try the Carnhill VTB2291 and wire it for 2K4:600. You will get a 6dB drop in level but the DAC should be able to drive it OK. If you want more 'tone' then go for a transformer with a smaller core. I use a very small TRIAD transformer to isolate the inputs to my VU meter buffer amps. The one I use is the TY-141P which costs about £5. Despite their small size, thay have a decent primary inducatance and their actual frequency response is much better than their data sheet would lead you to believe.

Cheers

Ian
 
One important point I forgot to mention. Just because you connect a 600:600 transformer to the output of your interface does not mean it is driving a 600 ohm load. As I has said many times before, transformers do what it says on the tin - they transform voltages, current and impedance according to their turns ratio. A transformer has no intrinsic impedance of its own. If you stick a 600 ohm load on the output of such a transformer it will reflect it (transform it) to look like a 600 ohm load at the input and hence only then will your interface see a 600 ohm load. If instead you connect it to a 10K input impedance line level input, the transformer will do its best to reflect that impedance to the primary and show it as a load to your interface.

A 600:600 transformer is only called 600:600 because it gives its best audio performance, particularly in terms of frequency and phase responses at the extremes of the audio range, when working at that impedance. If you drive it and load with with different impedances then its performance may not be optimal.

The bass end response is determined primarily by the inductance of the primary. This inductance appears directly in parallel with the primary.A good 600:600 transformer will have sufficient inductance such that its inductive reactance will be a lot higher than 600 ohms across the greater part of the audio band. At the bass end of the spectrum, as the inductive reactance decreases, it will eventually drop to 600 ohms and the response of the transformer will therefore be 3dB down at this frequency. Really good transformers have sufficient primary inductance that this happens well below 20Hz but to achieve this means more turns of wire, possibly a bigger core and hence more expense. Cheaper 600:600 transformers cut corners on both core size and number of turns so the primary inductance suffers. The inductive reactance will often be 600 ohms or less at 20Hz. How do they get away with it? The answer is simple. Most modern circuits have an output impedance much lower than 600 ohms - 100 ohms or less is typical so even if the transformer primary inductive reactance falls to 600 ohms at 20Hz, the response will be less than 1dB down.

The other thing to rememeber is that transformers do not care how heavily you load them; all that happens is you lose some bass but transformers do not distort when loaded heavily. They will reflect that heavy load back to your driving amplifier and it may distort but the transformer will not. The only external factor that directly affects a transformers distortion or colour is signal amplitude. High signal amplitudes cause the core to saturate and it will do this more easily at frequencies below 100Hz than at any other frequency. Inside the transformer, what determines the signal level at which saturation occurs is the core material, the core size and the number of turns. Good quality transformers have big cores made of steel with lots of turns so they can handle higher levels. The VTB2291 can easily handle +20dBu at 20Hz with very little distortion.

The more exotic core materials like radiometal and mumetal produce a lot less distortion than steel but saturation occurs at much lower signal levels (for a given core size and number of turns). This is why they are use a lot for mic input transformers where the signal levels are low. They still distort more below 100Hz though just like steel cores.

So if you want to produce transformer colour across the audio spectrum it might be an idea to use a low turns ratio mic input transformer, roll off the input below 100Hz to even out the distortion across the entire audio spectrum, then restore the bass at the output.

Cheers

Ian
You should be charging money for advice this good.
 
WOW, the Audient web site is SOOOO SLOOOOW. It maxes out both cores of my ancient AMD64 processor.

The one thing the Audient spec does not say is what load it will drive those levels into. I expect most of the time this will be 10K but to drive a 600:600 transformer you need much better than that. Yes, the quote and output impedance of 100 ohms but that gives no indication of drive capability. It does say that the topology is the same as that used in the ASP8024. I can find data for the ASP8024HE but not for the plain 8024. Anyway, the ASP8024HE is specified to drive a load as low as 2K so it might actually struggle to drive a 600:600 transformer. If you want to experiment over relatively short distances then you could try the Carnhill VTB2291 and wire it for 2K4:600. You will get a 6dB drop in level but the DAC should be able to drive it OK. If you want more 'tone' then go for a transformer with a smaller core. I use a very small TRIAD transformer to isolate the inputs to my VU meter buffer amps. The one I use is the TY-141P which costs about £5. Despite their small size, thay have a decent primary inducatance and their actual frequency response is much better than their data sheet would lead you to believe.

Cheers

Ian
Hi Ian

Thanks for your help! I'll have a crack with the carnhills and I'll pop some TY-141P's into something else.

Ed
 
It's easy enough to determine if you can drive 600 ohms without spending $$ on a transformer. Inject a tone around 400Hz and set the output to a comfortable level (not clipping). Measure the open circuit output voltage. Hang a 5K pot on the output in series with 600 ohms. If you can run that pot down from 5K to 0 and the output voltage is equal to or greater than 1/2 the open circuit voltage you have something that can drive 600 ohms. If it drops below 1/2 the OCV ,it can't drive 600 and you would be better off with a bridging transformer in the 10K 1:1 range. And if it drops significantly below 1/2 OCV you probably don't want to leave it running very long. Also, while you are at it, listen for distortion. It might drive 600 but sound like carp. Also, when you do get to 1/2 OCV the resistance of the pot plus the 600 ohms is your output impedance, or at least the resistive component of it which should predominate above 50Hz. The 600 ohms in series is to be sure you don't go below 600 because I do not want to be the one someone blames when his output driver gets fried.
I own some test equipment that can't even drive 600 ohms, some of it well-known brands like HP. I finally built buffer amps for these things with 5534 opamps. For one piece or equipment I use regularly (an old Genrad microvolter) which has an input impedance of about 275 ohms, I had to add a transistor output stage to my 5534. Sad, I suppose, but it gets me there.
@Erazorblade, regarding your question, can overloading the output stage destroy it? Ian's answer, "It depends," really means it depends on how well the output stage is designed. It is unlikely to destroy the output stage, but that doesn't mean it won't sound bad or won't give up some smoke if driven too hard too long. Opamps use feedback not only to control the gain but also to reduce the output impedance. They will do their darnedest to make that output voltage, sampled on the inverting input through a voltage divider, match what is on the noninverting input. That is their job. They don't know or care if an external load is dragging their output down. This characteristic renders the effective output impedance of an opamp in the few ohms to tens of ohms range, often less than that. So an unprotected opamp can get hot driving a low impedance load in a poorly designed circuit. Good opamps will have some sort of thermal protection, but good output drivers will be designed so that thermal protection is never needed.
 
I just built a fun color box using Lundahl LL1521’s.
Since I couldn’t find a map/datasheet for this transformer, I have named my two main outputs as 5 and 6 (upper right corner, facing the label). I love this setting on my mixes. There is quite a bit of color. - I also utilized a second pair of outputs on 6 and 8 for an extremely Lofi setting that destroys most of the top end. I love it on programmed drums and bass already. 4PDT switches between the 2. Color shifting blue to purple paint with iridescent gloss. The photos don’t really do the finish justice.
 

Attachments

  • IMG_8410.jpeg
    IMG_8410.jpeg
    1.1 MB · Views: 0
  • IMG_8409.jpeg
    IMG_8409.jpeg
    1.2 MB · Views: 0
  • IMG_8411.jpeg
    IMG_8411.jpeg
    1.1 MB · Views: 0
  • IMG_8414.jpeg
    IMG_8414.jpeg
    774.8 KB · Views: 0
A lot depends on the drive capability and output impedance of your existing buses (note there is only one S in bus and only two in total is BuSeS). Many old mixers were expected to be able to drive a 600 ohm load so you could just use 600:600 transformer and be reasonably certain of preserving the entire audio spectrum. If your bus outs cannot drive 600 ohms then you probably need to think of using a step down transformer, say something like 2K4:600. The only downside is you will lose 6dB of level.

Both Edcor and Carnhill do decent 600:600 and 2K4:600 transformers. If you want tone you are best sticking to transformers with steel lamination (like the Edcor and Carnhil). Jense, Sowter and Cinemag tend to use hybrid cores for lower distortion which I guess is not what you are looking at. I do not know about Lundahl.

Cheers

Ian

Why going for 600:600 transformer. Any reason why not eg 10K:10K type ?

Edit: I see you've addressed this and more in later and very informative posts.
Cheers.
 
Last edited:
What determines the output drive capability? My Mixer has an output impedance of around 30 Ohm. Is there more to it than just the output impedance? Or is it safe to say, that my Mixer can drive 600 ohm?

No. The output impedance does not tell you the load that can be driven. It is constrained by the maximum current capability.
 

Latest posts

Back
Top