DIY Ribbon transformer Part II - The McGyver Way

[Rossi]

P.S. Does anybody have an answer to my initial questions which somehow got lost in this threat:

- is a big transformer more susceptable to hum pick up

- is there a way to reduce hum pick up (wiring techniques?) without using a mu metal can?

 

[Guest]

[quote author="Rossi"]P.S. Does anybody have an answer to my initial questions which somehow got lost in this threat:

- is a big transformer more susceptable to hum pick up
[/quote]

Like any bigger antenna on such long-long waves.

- is there a way to reduce hum pick up (wiring techniques?) without using a mu metal can?

No. Use copper foil in parallel with winding additionally to a mu metal can in perpendicular to them.

 

[Guest]

[quote author="Rossi"]I'm interested in optimizing the overall performace, the combined performance of ribbon, transformer and preamp in a real world recording setup. [/quote]

Me too. :grin:

I don't have any experience with ribbons, but I suspect that low Z load will work similar to acoustic damping. Also, on very high frequencies and very low Z load inductivity of the ribbon may cause roll-off on highs.

 

[Marik]

[quote author="Rossi"]P.S. Does anybody have an answer to my initial questions which somehow got lost in this threat:

- is a big transformer more susceptable to hum pick up

- is there a way to reduce hum pick up (wiring techniques?) without using a mu metal can?[/quote]

As I wrote in another thread, if you want to make a good ribbon transformer, find a small hi-Ni UI-core, wind it with thickest wire you can fit, for both, pri and sec, for minimum signal loading and copper noise. The UI with dual bobbin is less susceptable to the hum pickup than single bobbin EI.

Assuming the ribbon is properly acoustically damped for flat response, load it with at least 10 times higher impedance, for minimum frequency response anomalies and maximum S/N ratio.

 

[mhelin]

[quote author="Rossi"]
- is there a way to reduce hum pick up (wiring techniques?) without using a mu metal can?[/quote]

I think if you wind another similar transformer and connect the two "anti-parallel" (I mean parallel but in anti-phase, secondaries can also be connected in series) the EMI generated voltage (hum) will be cancelled out. At least this works on humbucking guitar PU's, and I think mic input transformers like Lundahl's are internally built and connected this way too (correct me if I'm wrong).

http://www.lundahl.se/pdfs/datash/7904.pdf

"When designing the LL7904, we have used our wellestablished
two-coil structure to create a transformer with a high degree of symmetry. The transformer is built up from
two primary windings (which should be used in parallel) and two secondary windings. Each secondary winding is built
up from two sections, one from each coil and is surrounded by it own electrostatic shields. The symmetric structure
results in an internal cancellation of noise signals caused by external magnetic field (humbucking). It also increases
immunity to ground noise between secondary systems and reduces the effects of input common mode signals."

Maybe the Edcors are also made this way?

 

[Rossi]

Yeah, that makes sense! I don't think the humbucking thing would work with the arrangement I have right now. My McGyver'd power tranny has two bobbins next to each other. I suppose the two bobbins are too close to each other to be used anti-parallel. Maybe I'll just try. The problem is that re-wiring the thing is a ***** and that the secondaries aren't equal.

A dual bobbin job on a UI core as Marik suggested seems to make more sense. There is a transformer producer not too far away from here. I don't think they ever made audio transformers, but maybe they can sell me suitable core materials.

Too bad I can't find any supplier of thin aluminum sheets around here. I'd love to DIY a ribbon mic from scratch. :grin:

 

[burdij]

I think you can address the hum problem with a humbucking transformer design. Here is a diagram of the operation:

Obviously, an E/I core will not do. You could break the legs of the Es to make Cs ("mmm . . . break legs"). Or you could get U/I cores. I think this company could supply cores small enough to fit in a mike:

http://www.magmet.com

Of course, that wouldn't be what McGyver would do.

 

[burdij]

Also, I forgot this link:

http://microphonium.blogspot.com/

Its Robert Crowley's blog about what is going on at Crowley and Tripp Microphones. There is some interesting information they have learned from taking apart an RCA 77DX (and putting it back together!) and some other RCA ribbons. There is also some relevant information regarding transformers and ribbon matching.

 

[Rossi]

Thanks for the diagram, Burdij! Looks like UI or CI is the way to go. I think I'll stop by this small transformer manufacturer around here and see what they have. Not sure ordering a core form overseas is worth the cause.

Robert Crowley posted in our "Ribbon dissection" thread. I really like his no-nonsense assessment of the RCA 77.

 

[Marik]

[quote author="Rossi"] Looks like UI or CI is the way to go.[/quote]

If you can't get UI you can also try hi-Ni D or L cores.

 

[Rossi]

Thanks Marik! Makes sense.

 

[CJ]

Wind the pri first so the wire length is short=low dcr.

 

[Rossi]

Thanks, that makes sense as well :grin:

So far I always wound the pri first. I figured that it would probably be good to have the input signal close to the core for best transfer. Not sure that's correct, just it seemed to make sense to me.

 

[Marik]

Don't underestimate importance of connecting wires. Some 0.08 ohm wire resistance could worsen noise figure by almost 2db.
Use some solid 1mm copper and make a humbucking loop.

 

[Rossi]

Yeah, that's true. Unfortunately, the connecting wires on some of the cheapo ribbons aren't too good. The newer models seem to use thinner wire, and you do hear an increase in noise. Plus, I don't think they use the best wire to begin with. Doesn't even look like copper.

Unfortunately it's hard if not impossible to re-wire the ribbon element without damaging the ribbon. I know, because I tried. You have to unmount the ribbon before you can put a soldering iron to the clamp.

 

[tony dB]

Rossi,

I'm trying to pimp a t-bone ribbon but have difficulties in finding out the wires, as the original trannie is not marked, only a few wires sticking and twisted together 2 times... Do you happen to have a trannie with labelled wires? or a little schemo?

 

[Rossi]

If you've taken out the tranny already, just measure the DC resistence with a multimeter. The primary is very low, a fraction of an ohm and the secondary is over 100 ohms.

Don't measure while the ribbon is still connected, you might damage the ribbon!

Actually, you don't even have to know the tranny wires. Just connect the ribbon wires to the Edcor's primary and the output plug to the secondary. The shield goes to ground/chassis/Pin 1. If the mic happens to be out of phase, just swap the wires on the secondary. Make sure the primary is well connected to the ribbon. As Marik pointed out, any additonal resistance will increase noise.

 

[tony dB]

cool, thanks!

 

[tony dB]

From my first tbone ribbon the original transformer measures:
primary 10 ohm
secondary 32 ohm

Just checked my 2 tbone ribbons (speech only so far).
After replacement of the stock transformer with the Edcor RMX-1 it's safe to say that this became a very useful mike immediatly.
The stock version sounds hi passed and rather thin, compared to the Edcor pimped one. The difference is huge. No golden ears needed to hear the upgrade :twisted:

 

[Rossi]

Thanks Tony. You probably have the older version of the RB500. Your multimeter was probably wrong on the primary DCR (the mic would have been unusably noisy at 10 ohms primary) but the secondary (which is easier to measure) is about what I have on the older tranny. It is a dual bobbin UI tranny, right?

Yes, the Edcor sounds nicer, no question.