Alu VS laminated ribbon

OK, here are my findings about differences among pure aluminum versus laminated ribbons. The ribbon was built for testing purposes, 1 meter tall, 20 mm wide, with a gap of about 0.5mm on each side, the magnets were N40 and the measurement distance exactly 50 cm.

The alu ribbon was 8 microns thick and the laminated ribbon had 12μ of mylar foil plus three alu traces of 8μ, about 6mm wide.
The weight of the 12μ thick mylar foil is less than that of the 8μ thick alu foil.

The impedance of the laminated ribbon was 3.3 oHms, and the alu ribbon had 3.4 oHms of resistance wire in front.

It seems both have roughly the same efficiency?

URL]


URL]


URL]
 
Some more observations, pure ribbon has Fs below 5Hz, laminated has two: the fundamental one at about 9Hz and one at about 150hz, the latter is probably the result of my sloppy laminating technique. I have used extremely thin layer of spray glue.

On the plus side, the laminated ribbon is less sensitive to the slot anomaly between 5 and 6 kHz, has slightly more extended upper regions, but at the expense of somewhat more emphasized 15-22kHz region and less sensitivity below 1kHz (the baffle was 60cm wide, with ribbon at the side).

I have a Libra that shows marginally higher weight of the laminated ribbon but the difference is too small to be measurable.

Main advantage of the laminated ribbon over aluminum one is probably somewhat greater mechanical robustness and the corrugations can be shallower.

Any opinions?
 

Marveloudio

Member
2004-06-21 10:19 am
X
spl depends also on the source resistance , which should be taken into account , your levels/impedances do not match physical relationships .....

one impedance curve has a wiggle is it the laminated ribbon?

please try to estimate absolute splevel by comparing with a (flat response) driver with known efficiency/sensitivity at 2.83Volts input , on a baffle with same size/dimensions , and at the same distance! (1m distance would be better for comparison)

which glue did you used for lamination? did you measure weight of the al traces and finished lamination to determine how much mass was added?

would be nice to get the full picture without open questions ...

otherwise - nice work , well done!

bye
 

Marveloudio

Member
2004-06-21 10:19 am
X
looks like the laminated has a resonance around 17kHz that should be filtered , younger ears will be annoyed by such a large peak ...

the minor peak around 4kHz should be equalized too , our ears are quite sensitive in this range ...

the Z-bump of the lam-ribbon comes most probably from the lessend stiffness across the width by weakening the structure through the mylar section between the traces ...

with the alpureribb version take care of hidden contact resistances , left/right levels may change by inproper mounting of the terminals ...

it is probably a good idea to clamp the alpureribb softly near the middle position with small foam pads on both sides , make it less supceptible to any air puffs etc ...

greetz
 
Some more observations, pure ribbon has Fs below 5Hz, laminated has two: the fundamental one at about 9Hz and one at about 150hz, the latter is probably the result of my sloppy laminating technique. I have used extremely thin layer of spray glue.

On the plus side, the laminated ribbon is less sensitive to the slot anomaly between 5 and 6 kHz, has slightly more extended upper regions, but at the expense of somewhat more emphasized 15-22kHz region and less sensitivity below 1kHz (the baffle was 60cm wide, with ribbon at the side).

I have a Libra that shows marginally higher weight of the laminated ribbon but the difference is too small to be measurable.

Main advantage of the laminated ribbon over aluminum one is probably somewhat greater mechanical robustness and the corrugations can be shallower.

Any opinions?

The pure alloy has better sound but will require a thin layer of mylar backing for good sonics, also better to use without corrugation ....

Could you measure with 2.83 volts @ 1or 2M , it would give us a better idea of your efficiency...

Very interesting.

I have one question:
are the three laminated alu traces connected in series or in parallel?


Thanks,

Bruno

Series for it to be 3 ohms ....
 
Last edited:
The source resistance remains unknown to me, but my estimation is a few miliohms (Yamaha AX 396 has high damping factor=low output resistance).

About the impedance wiggle, read my additional observations.

At the moment I don`t have any drivers for comparison, but I was more interested in general tendencies of pure aluminum ribbon with the added resistance wire of the same impedance as the laminated one and to my knowledge on this forum nobody tried to measure this before.

I have used the "Tesa Sprühkleber" and as I said before the difference in weight was too small for my Libra to show up, my estimation for pure alu ribbon is 0.6 grams, laminate 0.8-0.9 grams.
The peak is at 22kHz not 17.

The laminated ribbon has three alu traces connected in series which gives 3.3oHms.

Unfortunately I don`t have voltmeter so everything is just relative, but I guess both lover curves would translate to somewhere between 83-85dB, no more. I am saying this because long time ago I have compared such ribbon to a dome tweeter of 90dB 1W/1m.

Almost forgot, someone from this forum reported four or eight times higher sensitivity of laminated ribbon VS pure aluminum one, whatever that means...
 
Last edited:
No!
ERROR!!!

Do humans make mistakes or only computers? :eek:
The middle trace was incorrectly wired, the new measurement shows much higher sensitivity, some 7-8dB higher than pure alu ribbon of roughly the same impedance!
Did no one ever tried to measure this? There must be clear benefits with such vastly improved sensitivity of the laminate, not only drawbacks?
I apologize for misleading you...

URL]


URL]
 
Last edited:
No!
ERROR!!!

Do humans make mistakes or only computers? :eek:
The middle trace was incorrectly wired, the new measurement shows much higher sensitivity, some 7-8dB higher than pure alu ribbon of roughly the same impedance!

Nice to read that the problem is solved.

About the top frequency response of pure Alu ribbon, did you use an impedance transformer?

Bruno
 
No!
ERROR!!!

Do humans make mistakes or only computers? :eek:
The middle trace was incorrectly wired, the new measurement shows much higher sensitivity, some 7-8dB higher than pure alu ribbon of roughly the same impedance!
Did no one ever tried to measure this? There must be clear benefits with such vastly improved sensitivity of the laminate, not only drawbacks?
I apologize for misleading you...

URL]


URL]

Could you provide us with a pic of your laminated diaphragm .....
 
Well, it`s mylar foil (without adhesive), I spray glue the aluminum foils then attach them to the mylar foil - I guess I have used the wrong terminology? Perhaps I should say multi trace ribbon versus the single trace?
This setup is close to 8dB more efficient than..., please reread everything I have already stated. It is similar to this setup who indeed used the self adhesive tape, unlike me.
I have spread the vertical graph resolution drastically to make it easier for you guys to observe the differences in impedance and the wrinkles below 170Hz (!) are in the order of 0.3 oHms, is that a lot? I could easily use the 50dB span instead of the 10dB and then you wouldn`t notice any wrinkles.
Pure aluminum foil is very efficient only with transformers, if you use ANY kind of resistors - efficiency drops hugely.

By the way, do you have any experience with multitrace ribbons (mylar/kapton & alu) VS the single trace (alu only), have you built any and observed the differences in SQ?
 
I have only built full alloy , not traced , but have compared the sound to other traced units from others , i prefer the sound of the straight alloy vs laminated multi-traced ones. Efficiency and load is ....:xeye: ..sonics...:) only good for SS, if running tubes of course traced is the way to go .

Those wrinkles in the electrical graph will show itself sonically and for sure when measuring magnitude and phase..


regards,
 
I love the sonics of pure aluminum ribbons, but they are somewhat more sensitive to the slot diffraction effect which needs to be tamed with an RCL and laminated ribbons are much less sensitive to this phenomena, perhaps because they don`t need as deep corrugations, if at all. If you don`t use the RCL trap the sound can be very delicate (in the presence region) but also tonally incorrect and somewhat metallic in character.

I don`t know, I should do more testing before making any final conclusions but at the end I might choose the laminated multitrace ribbon out of practical reasons, maybe I will trade some sonic details for greater mechanical and sonic robustness, we`ll see...

By the way, I always thought pure alu ribbons are prime candidates for all amps that need output transformers, tubes of course, but also SS amps like Susan Parker`s Zeus amp:

http://www.diyaudio.com/forums/solid-state/42259-zero-feedback-impedance-amplifiers.html

And from 6moons this news about Raal tweeters:

"Recently, we finished development of Amorphous C-cores application in our ribbons. Some years ago, we were the first to introduce large gap approach in building ribbon transformers and now we are the first to use AMCC's in ribbons for absolutely uncompressed high power transient transfer and, at the same time, even further improving already the best low-level performance in the world today, by decreasing the width of hysteresis compared to HF ferrite cores. For direct coupled tube SE-ribbon, we succesfuly developed 175:1 ratio, -3dB at 120kHz Amorphous C-core transformers that can carry DC idle current of SET. Impedance conversion is about 30000:1, bringing the 0.1 Ohm ribbon up to 3kOhm, good for plate loading of most triodes in HF."

regards,
miro
 
Last edited:
Those wrinkles in the electrical graph will show itself sonically and for sure when measuring magnitude and phase...

Almost forgot, with the expected XO around 500Hz, it will hardly show anything,
I mean these wrinkles could cause deviations in the order of less than half dB.
I`m sure that the slot anomaly in the presence range if not cured, plays a much bigger role here...

regards,
miro