ribbon mics and the sound of transformers

[ioaudio]

i have been playing with a cheap thomann ribbon mic, replacing the stock transformer with a couple of variations i wound myself.
besides trying high nickel (80%) lams, the obvious choice for high impedance/low dc resistance, i also tried a core with lower Al value, but better quality (different alloy made in europe) and was really stunned by sonic differences. its not subtle at all - even with the same inductances.

the ribbon and its transformer work very close together, it seems.
i wonder if thats old news.

-max

 

[CJ]

well, tell us what the NI sounds like .

 

[ioaudio]

the high nickel sounds like the energy would be eaten up in the bass region, leaving no energy/clarity/openess in the highs. so, with some close-up vocals when the bass comes up, the highs are clouded.
the bass "tightens up", it´s less boomy with the other alloy. i rewound the high-nickel lams (the stock transformer )
and while there is some room for improvements, the core get´s "smeary" when loaded with low-freq energy.

next, of course, there´s the think with alternating the step-up ratio,balancing noise vs sound.

the transformer i (now) like best has a secondary with 132 ohms - the Al of the core is only 5000nH , while the stock core´s Al= 40000nH ...

 

[rafafredd]

interesting. Can you share what type of core you are talking about?

 

[skipwave]

[quote author="ioaudio"]the transformer i (now) like best has a secondary with 132 ohms - the Al of the core is only 5000nH , while the stock core´s Al= 40000nH ...[/quote]

What's the turns ratio on that?

I wonder if we need to account for higher output ribbon motors. Do they have higher output impedance, and should therefore have lower turns ratio transformers for optimal impedance matching? Maybe there is yet another tradeoff to consider, this one between sensitivity and saturation.

 

[ioaudio]

i´m still trying different ratio´s and winding scheme´s , but 1:42 - 50 seems to work best with this ribbon (and with my two preamp´s input impedances)

i have seven variations now, need to make a plug-system with low dcr to make comparison easier.

here´s two pics of the core, it was mady by vacuumschmelze hanau in germany. one single lam is 0.2mm(they are put together from the factory already) i´m not sure if they offer the same alloy nowadays.

 

[Rossi]

[quote author="ioaudio"]the high nickel sounds like the energy would be eaten up in the bass region, leaving no energy/clarity/openess in the highs. so, with some close-up vocals when the bass comes up, the highs are clouded.
the bass "tightens up", it´s less boomy with the other alloy. i rewound the high-nickel lams

and while there is some room for improvements, the core get´s "smeary" when loaded with low-freq energy.

next, of course, there´s the think with alternating the step-up ratio,balancing noise vs sound.

the transformer i (now) like best has a secondary with 132 ohms - the Al of the core is only 5000nH , while the stock core´s Al= 40000nH ...[/quote]

Yeah, but doesn't that mean that the new transformer will produce a low cut due to lower inductance? That was basically the problem I was facing when I rewound the stock transformer for an active ribbon (i.e. 1:100+).

Slight LF loss may actually be good depending on application. My main goal was a vocal mic, so a little LF loss was welcome because it compensated for the thick proximity effect.

Regarding winding ratio, you have to take into account the ribbon's impedance which is more or less its DC resistance. It's pretty difficult to measure such low resistances with any accuracy, though. The RM700 has a slightly higher output impedance @ about 400 ohms. That's not too bad, but this mic shouldn't be run at low impedance mic inputs (you loose gain, bass response and also some treble).

So much for today, have to get back to my books :sad:

 

[ioaudio]

yes, of course, otherwise i would have used only one primary winding :wink:
the difference i meant wasnt really the quantity but the quality of bass.

for now i gave up the big ratios, to much dc resistance on the secondary (the last one had nearly 200 ohms, and noise is unacceptable)
i will try 1:27 or so next.

 

[Rossi]

I usually used the stock core. I tried a different core once or twice (stuff I found in an old radio) I also once rewound a wall wart transformer (a search for McGyver will produce the thread) and I also reconnected a 100V loudspeaker transformer for ribbon mic operation. The radio ones had too little bass. The McGyver tranny and the reconnected 100V tranny sounded very nice but had hum problems and were a little noisy.

The stock cores produced good results. It'll be hard to get really low noise with lower Al-cores. If you want less bass, try the stock core with fewer primary windings and/or thicker wire. My best rewound active ribbon transformer had .8 mm wire for the primaries, wound pri sec pri, double bobbin on the RB500's UI-Core. The double bobbin arrangement is quite nice for experiments as you can wire the various coils in series or in parallel, thus varying the transformer ratio and other parameters.

You can also try winding the secondary first and the primary on top (sec - pri). That's what Beyer did on the transformer that CJ dissected. I tried that, too, putting as much wire on the bobbin as I possibly could, leaving just enough room for one layer of 1mm wire. That transformer is very easy to wind (remember it was meant for an article). The primary inductance was pretty low, so I had to connect the primaries in series. Still, it sounded very usable, low noise and just enough bass response.
If you want a regular passive ribbon transformer, just use thicker wire on the secondary. I used .1 mm for something 1:100ish.

 

[ioaudio]

saw your thread. i wouldnt even bother to try winding a ribbon transformer on a iron core - you will not find nickel-cores in old radios.
chapter 5 from radio designer´s has a nice table for core´s and their permeabilities and other parameters. the initial permeability is a quite important parameter in this application. (and unfavourable nonlinearities in the b-h curve might be the reason why i don´t like the stock alloy)

i do prefer to have (at least one section of) the primary close to the core, pri-sec-pri is giving the very good results (you also have some kind of shielding with the pri around the sec)
my last experiment used stripes of copper-foil for the primary, but its a pain in the ass to wind that. (and a pain in the ass to cut&insulate copper-foil )

 

[Rossi]

Yeah, I think the pri sec pri arrangement is the best, unless you have a special core and excellent shielding like Beyer. I thought about foil, but never actually tried. Twin wire like on the RM700 you dissected isn't a bad strategy, I think, for getting a flatter primary layer than you get using very thick wire.

You obviously know more about core materials than I do. I did have a look at the RDHB, but I don't have a handy source for different transformer materials. The whole affair was meant for an article anyway, so the main attraction for winding your own transformer was a) building an active ribbon transformer and b) get as much fun as possible without spending significant amounts of money. Remember you can have a very nice Lundahl ribbon tranny for about 50 Euro. Also, I was more or less limited to an active ribbon tranny, because any passive ribbon transformer would require a precise ratio and hence a winding machine with a counter. For an active ribbon tranny you just need something that rotates. Any winding ratio above say 1:80 will do, the electronics will take care of the impedance issue.

 

[ioaudio]

some links

ribbon transformer

beyerdynamic
http://www.groupdiy.com/index.php?topic=20482

discussion xlvk & prr
http://www.groupdiy.com/index.php?topic=2017&postdays=0&postorder=asc&start=15

diy ribbon trafo
http://www.groupdiy.com/index.php?topic=218986

lundahl impedances
http://www.groupdiy.com/index.php?topic=8033

 

[Rossi]

McGyver Ribbon Tranny :grin:
http://www.groupdiy.com/index.php?topic=19046

 

[CJ]

break into Mariks house, steal all his data, he is going to be a professional
piano player, not a conventional hall junkie, so he won't care. :twisted:
send any lams you find to Gus, he will re-anneal them in his DIY adapted pressure cooker that he used to stearlize mushroom bottles with.

you need the Square loop, for riboons, eh?

 

[Marik]

[quote author="CJ"]break into Mariks house, steal all his data, he is going to be a professional
piano player, not a conventional hall junkie, so he won't care. :twisted:
[/quote]

Oh yeah, sure!!!
Let me know when you come and I won't be home and leave the door open :twisted:.

Look how old junk like RCA or EV was made--HUGE core size in less than 1/4" stack, with "I"s missing.
Shorter wire, lower copper resistance, good coupling, low noise. Sounds like crap 'cause of used materials (who cares in broadcast mics?), but they knew what they are doing.

 

[CJ]

don't forget to experiment with copper ribbon, the flat tape stuff, if you can get enuff turns.