DIY Ribbon transformer Part II - The McGyver Way

[Rossi]

Here's an update on my experiments re DIY ribbon transformer winding. (Thanks again CJ for the great tips)

For the time being I ditched the idea of re-winding the stock transformer in my cheapo ribbon mic. Instead I cracked an old 9V wallwart and took a look at the transformer inside. Just for laughs, I decided to just rearrange the parts the tranny is made of. This tranny consisted of E I lams, and two bobbins next to each other. One with thicker wire (.5mm) and one with finer wire (.1 mm I'd say). I unwound the wire and wound both bobbins more or less identically: 1 layer of the thick wire (8 turns I think) and many layers of thin wire, then another layer of the tick wire and more of the thin wire. pri-sec-pri-sec. I didn't have a counter on my improvised winding machine, so the secondaries were just played by ear. As it turned out I ran out of the thin wire too soon on the last secondary. DC of the secondaries, if I remember correctly was

1: 164 ohms
2: 126 ohms
3: 174 ohms
4: 51 ohms

To make up a bit for the unequal secondaries I wired 2 and 4 in series. This gives me three secondaries of about 170 ohms. The primaries are wired in parallel, DCR is very very low, about .1 ohms.

Wiring the three secondaries in parallel gives a pretty normal ribbon step up ratio. About 1:30 I'd guess. Wiring all secondaries in series gives a very high step up ratio for use with an active impedance converter.

Surprisingly, this crappily improvised tranny doen't sound bad. I made a comparative measurement and found that the treble is really good, acutally. Bass is a bit lower than on a commercial tranny, but not bad. It actually cuts just enough bass to make it sound good at "LD condenser distance"

Here's a sound clip I recorded standing about 20-25 cm in front of the mic:

As you can hear, there is a slight hum problem. My McGyver Tranny really likes to pick up hum from all sources, almost like a single coil guitar. Does anybody know a way to reduce this hum without using a MU-metal housing. I'm afraid the tranny is a bit too big to find a suitable can. Does the size of the tranny make it more susceptable to hum? Forgot to measure the tranny's size, but I'd guess about 5 x 5 x 6.5 cm. About 3 times larger than a normal ribbon tranny.

There's also some hiss, but it seems that some of it may disappear once the contacts to the ribbon become shorter and more solid.

Improvised mic:

McGyver Tranny:

 

[Jonkan]

very cool,

you rock!

/J

 

[Keith]

Rossi, I'm watching you! Actually just last week I wound a few power supply type ribbon trannies as an experiment and they all worked pretty good,but they did'nt sound at all like the Edcore nickel core which had much better bass and treble. The power supply cores gave good overall frequency and I was impressed until I compared the Edcore. Volume was similar with both types. I'm unfamiliar with staggered winding of the coils. Is there an advantage? I find that transformer humm disappears after I put them in the grounded mic case. Good Post! Keith#2

 

[Rossi]

Hey Keith, I think results vary pretty much with the core material used, especially bass response. My knowledge concerning core materials is fairly limited, but the core material of this wallwart transformer looked a bit like the material the original ribbon tranny was made of. Which CJ said was good.

For all I know, interleaved windings reduce losses. What is good energy transfer and low DCR. I personally wanted the option to change the transformer ratio. Having several primaries allows you to wire them in parallel, thus reducing DCR. Having several secondaries allwos you to change the step up ratio by wiring them either in parallel for a normal passive ribbon mic. Again, paralleled windings reduce DCR. Wiring the secondaries in series gives a very high step up ratio for use in an active ribbon mic.

That said, this simple tranny sounded fine using just one of the secondaries.

It's hard to say anything definite about the overall performance without wiring the thing correctly and putting it into a case.

I made other experiments as well. I tried a simple 100V tranny made for speakers. The 4 ohms winding was about .6 ohms DCR. It sounded quite nice, but hiss was a bit high for real use. That tranny had hum problems as well, and they didn't disappear when I put it into the mic housing (it was a very tight fit). Maybe a smaller 100V tranny would be okay.

I have a feeling hum pick up is size related. I hope CJ will chime in and tell us what he knows.

 

[Rossi]

Apart from the DIY aspect, the Edcor tranny is pretty nice and inexpensive. :thumb:

 

[Keith]

Rossi said; Having several primaries allows you to wire them in parallel

Are you matching the DCR of the ribbon and primary? K2

 

[Rossi]

No, my take is that DCR of the tranny windings should be as low as possible, i.e. lower than the ribbon's DCR if possible.

DCR means noise. The ribbon produces signal + noise, the connecting leads and the transformer winding DCR produce nothing but noise. The tranny amplifies both signal and the DCR related noise floor. Since the signal is so low, the noise floor matters a lot. The secondaries' DCR produces some noise as well.

On the other hand, you need some windings to get enough inductance (which you need for good LF response), so you can't just use one or two primary windings and 30-60 windings secondary. That's why a good core is important: you get enough inductance with a minimum of windings.

Don't confuse DCR with impedance. As far as the ribbon itself is concerned, DCR and impedance are about the same. But for the tranny, DCR and impedance are very different. The ribbon should see very low DCR but high impedance. The impedance on the ribbon tranny primary is the preamp impedance stepped down via the tranny ratio. It should be at least 10 times the ribbon impedance. If your preamp's input impedance is 2k and your tranny has a 1:30 ratio, the ribbon will see an impedance 2000 ohms / 30^2 = 2,22 ohms. That would already be slightly on the lowish side for a .3 ohms ribbon. In other words, it's the tranny's ratio that must be in some relation to the ribbon impedance.

 

[Keith]

Rossi, I see what your saying. I'm new to this. Last week in my post about ribbon volume Merik taught me about matching the ribbon and primary DCR which after months of frustration brought a real break through for me. Before my primaries were all bass or all treble. Matching put them both together. I ran the math you gave me and it makes sense, I need time to gel it through my head and see the practical application.
I don't really understand the diff between DCR and impedance very well.
I made a McGayver that had a 65 ohm secondary DCR and it sounded pretty good. I also made a McGayver from a 9 volt with a zillion turn secondary and initially it sounded good but I have'nt put it in a case yet which is when I can really give it a good listen. I did'nt even know they had McGayver in Germany. Shows what I know! :grin: K2

 

[skipwave]

[quote author="Keith"]McGayver[/quote]

Hmmmm, McGayver. Was that MacGyver's queer cousin?

All kidding aside, this ribbon transformer experimenting is really cool. Thanks for sharing.

 

[Rossi]

Well, as I said, I'm not a transformer god by any stretch of the imagination. Marik on the other hand knows a lot about ribbons. In fact, these are about the first transformers I ever wound (I made some related experiments I did not write about). Still, from what I can tell, an ideal ribbon tranny should have as low DCR as possible both on the primary and secondary side.

I'd like to hear what PRR or CJ have to say.

About impedance: As long as you're going for a regular passive ribbon, you want an output impdedance of about 200 ohms. 600 ohms would still be kind of okay; quite a numer of moving coil dynamics are actually 600 ohms. But that's more or less the max mic output impedance that makes any sense.
Usual mic preamps have input impedances around 2000 ohms, some are lower. A higher step up ratio on the ribbon mic tranny will give higher output impedance and higher output. But at the same time you lose some gain because the preamp impedance loads down the mic output. Hence 600 ohms is about the max. If you wind a transformer with higher than usual step up ratio and end up with a mic output impedance of more than 1000 ohms, you lose about as much gain as you get from the extra step up. Sound suffers too, especially treble response. Hence 600 ohms actual output impedance should be regarded as the max. Most 200 ohms ribbons are acutally about 300 ohms, probably due to DCR components.

 

[Marik]

[quote author="Keith"]...in my post about ribbon volume Merik taught me about matching the ribbon and primary DCR... [/quote]

Hi Keith,

I thought I was clear, saying that we need lowest DCR on Pri, hence the thickest gage you can fit. Sorry if it created some confusion.

Specifically for ribbons, I wound quite a few transformers with different ratios, using different cores, laminations, and winding techniques... and still learning.
And yes, I also would love to hear PRR's, CJ's, or anyone else's comments.

 

[Marik]

Oh yeah,

Loading, although helps to get rid of transformer ringing on HF, mostly affects LF response, where the impedance curve of the ribbon goes way up.

 

[Rossi]

Heavy loading will also decrease HF response significantly.

Transformer ringing doesn't seem to be a problem with ribbons as the ribbon element itself has a natural roll off. I wonder who came up with the idea that ribbons should be used with low impedance preamps?

 

[Marik]

[quote author="Rossi"]Heavy loading will also decrease HF response significantly.
[/quote]

That's actually is an interesting observation, as I never encountered that (and never saw in literature).

Edit: Just made a little simple test--loaded the mic with a 300 ohm (wanted to go to extreme) resistor and got exactly the same result as always--significant decrease on LF and only slight on HF (of course, besides significant signal loading).

Best, M

 

[Guest]

[quote author="Rossi"] I wonder who came up with the idea that ribbons should be used with low impedance preamps?[/quote]

Probably, from somebody who wanted less noisy preamps?

Theoretically, I may expect roll-off on low frequencies because of acoustic short cut, and roll-off on high frequencies because of length of the ribbon.

Do you have curves of frequency response with different loads of the naked ribbon without any tranny? It would be intresting to see them...

 

[CJ]

I really don't have anything to add, I have only taken apart one SM 57 tranny, thats it for mic iron.

I think Marik is the guy.
Just PM the hell out of him.
Sooner or later, you will catch him after a fine Caberet or a couple of Stoli's.
Thats when you get your info! :razz:

Heck, I don't even own a ribbon mic.
But you do not need to if you know the physics.
Which I don't. :grin:

But I do know that the pri dcr on that 57 iron was way low.
Just a few turns there.

But this has me interested mic transformers, an area which has been swamped out by the mic/line in and line out stuff.

Oh, Gus too, but he has to like you, or else!

 

[Rossi]

Hmm, since I know that Marik is developing something for someone, I don't feel like pressing him for more information than he is willing to share without being drunk :grin: Especially since I can't even disclose all my own findings. I'm researching for a magazine article, right now. Although I'm not sure I can even use my own research for more than a few side remarks, to be honest. Basically I just wanna know, dammit :green:

I guess I'll just make some more experiments. But it seems logical to me that any DCR that is not related to producing a signal is just dead resistance, i.e. noise.

As for loading: I made comparitive measurements with the whole ribbon mic and different loads. Light loading seems to affect LF response mostly, but heavy loading affects HF response quite significantly. By heavy loading I mean that the output impedance is actually greater than the pre's input impedance. But some slight HF loss can also occur with ligher loads. It depends on the mic's inpedance curve. If you have a transformer coupled pre, the actual input impedance may vary as well.


@Wavebourn: it would be very hard to measure the naked ribbon and I'm not sure it would help. There's no way to use a ribbon mic without a transformer. Brad once talked about designing transformerless ribbon electronics, but I think it was meant more as a kind of stunt. The ribbon impedance is just too low.

 

[Guest]

[quote author="Rossi"]
@Wavebourn: it would be very hard to measure the naked ribbon and I'm not sure it would help. There's no way to use a ribbon mic without a transformer. Brad once talked about designing transformerless ribbon electronics, but I think it was meant more as a kind of stunt. The ribbon impedance is just too low.[/quote]

I used 4 transistors in parallel for dynamic mic preamp. You may use more.

Having no measurement results you can argue forever without any success, right?

 

[Marik]

[quote author="Rossi"]
As for loading: I made comparitive measurements with the whole ribbon mic and different loads. Light loading seems to affect LF response mostly, but heavy loading affects HF response quite significantly. By heavy loading I mean that the output impedance is actually greater than the pre's input impedance. But some slight HF loss can also occur with ligher loads. It depends on the mic's inpedance curve. If you have a transformer coupled pre, the actual input impedance may vary as well.
[/quote]

On a second thought, that actually might be a case with thicker or stiffer ribbons (is it what you have?), where motional impedance curve has very pronounced peak @ around 10kHz as a result of insufficient air damping.

[quote author="Wavebourn"]
Theoretically, I may expect roll-off on low frequencies because of acoustic short cut, and roll-off on high frequencies because of length of the ribbon.
[/quote]

Hi Wavebourn,

I am not sure it is clear what you mean here. Could you elaborate, please?

 

[Rossi]

[quote author="Wavebourn"]
I used 4 transistors in parallel for dynamic mic preamp. You may use more.

Having no measurement results you can argue forever without any success, right?[/quote]

But that's only a preamp for measuring, not for actual recording use, right? I'm interested in optimizing the overall performace, the combined performance of ribbon, transformer and preamp in a real world recording setup. I'm not really arguing, I'm just telling you what I heard and measured in my experiments.

@ Marik. Yes, the ribbon on these cheapo mics may be a bit thicker than on top notch designs.