Citation 12 non-complimentary to drive ribbon tweeter

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am interested in building a non-complimentary version of the MOSFET Citation 12 that Nelson wrote about in AA in 1981. The intended use is ultra-high current devices such as a DIY ribbon that is very sub 1-ohm.

The twist I am interested in is to use a new low RDSon FET. This one

IXTN660N04T4 IXYS | Discrete Semiconductor Products | DigiKey

The RDSon is claimed to be 0.85 milli-ohms. With +/- 10V on the output stage and no current limiting, The amp could deliver hundreds of watts into a very low impedance load.

I have a DIY ribbon design in the works and I am loathe to use an audio transformer. A low-voltage high-current amp would be ideal.

I have no amp design skills and no board layout skills. If someone (or multiple somenones) wants to participate in a collaborative effort, I would buy boards and parts in exchange for a well layed out board and a working design.

I think that the power supply caps will need some consideration for ESR and placement close to the FETs and be connected with copper bus bars.

If this is not the correct forum for such a thread, I can ask to have it moved to the Solid State forum.

...looking for inputs....
 
Thank you for replying. Its a pleasure to correspond with you. I own two Adcom GFA555 amplifiers and I enjoy them immensely.

Fat ribbon material is 30 milli-ohms for a short tweeter. Skinny ribbon material is 400 milli-ohms. I just measured the material with a Keithley 2750 on the 1-ohm range using 4-wire and offset compensation.

I would also be using this high-current topology for driving sub 1-ohm car subwoofers.

The intent is to be able to create 300A (or more) transients. Weak magnets may require high current.
 
One more question, what is the efficiency figure of the tweeter?

Good question. I have no way to directly measure it at the moment. What I did is to use a WT2 which puts out about 3 1/2 mA at any freq between 1Hz and 20 kHz. I chose 3 kHz. I compared the tweeter to an Audax DH13B 5 1/2" midrange, a favorite from the 1980s.

The midrange is lots louder than the tweeter. The midrange is 85dB @1W@1m or so. My guess is that the tweeter is no better than 75 dB @1W@1m.

I will dig out the SPL meter and also measure the T/S parameters of the midrange. Then a better guestimate of the tweeter sensitivity will follow.

The tweeter may be less than 70 dB@1W@1m. It is not surprising. There is a large air gap and the magnets are inexpensive ceramic units. That said, the tweeter sound is very good. A picture of the tweeter is attached....
 

Attachments

  • Ribbon.jpg
    Ribbon.jpg
    190.3 KB · Views: 303
The midrange T/S parameters include sensitivity of 87 dB@1W@1m.

I found the SPL meter. The ribbon is more than 26 dB less output at 3mA at 3kHz than the Audax midrange.

The midrange yields about 86dB near field at 3mA at 3kHz. The ribbon does not deflect the SPL meter on the 70 dB range which has a -10dB mark at the far left.

So, the ribbon is no better than 60 dB @1w @1m.
 
That particular gap is intended for a wider diaphragm. A narrower gap will increase sensitivity, but not enough to negate the need for a massive amount of current to keep up with any normal cone speaker.

Opposing flat magnets cause a sharp suckout at a frequency inversly proportional to the distance between the magnets. A more narrow gap pushes the suckout frequency higher. The Magnepan IIIC tweeter has a deep suckout at 9 kHz. The ribbon is narow and the magnet gap is narrow compared to the tweeter pictured above.

Using stronger magnets makes assembly more difficult and actually dangerous. The magnets want to collapse on to each other in the gap pinching a finger quite easily.
 
The sensitivity of the midrange was calculated from the T/S parameters. It just happens that the SPL meter measured very near the dB sensitivity at 3mA with the SPL meter very near the midrange dust cap.

The tweeter is so quiet that the SPL meter does not deflect on the lowest range. I have to get an amp involved so that I can drive the tweeter enough to get a measurement.

Then, ... I will have to calculate the power into each driver and correct for the disparate powers.

I will post better info when I have it.
 
Better info is below

3kHz signal. Near field measurements on on both drivers. Current through driver is set for 85 db near field for each driver

The ribbon diaphragm is about 0.5 ohm

The midrange is about 12 ohms at 3kHz

Ribbon 85 dB at 1.3A, 850 mW

Midrange 85 dB at 32mA, 12 mW

The tweeter requires 850 / 12 = 69X more power, 18 dB less sensitive

Let me know if I made a miscalculation somewhere. I used the 2 most significant digits of the current measurements and rounded subsequent calculations to two significant digits.
 
OK, so let's say you have 86 dB at 1 watt at .5 ohms, which is 1.4A:

( 1 W = I^2 * .5 ohms )

106 dB at 100 watts, 14A

120 dB at 2,500 watts, 70 amps (100a peak)

Correct me if my math is wrong....

Your math is excellent. Two things

1. My measurements were very near field, about 1 inch from the driver. At a listening position, the dBs will be lower. The 86 dB sensitive midrange only required 12 mW to achieve 85 dB right at the cone. 10*Log (1/0.012) = 19 dB. So, as a wet thumb in the air, subtract 19 dB from your numbers.

2. The ribbon measured is 1/2 inch wide. The skinny ribbon is more than 10X lower impedance.
 
Last edited:
I would be taking more measurements depending on how good that dB meter is.
At least 3 different points.

The dB meter is a vintage Radio Shack analog hand-held unit. I performed the measurements more than once.

3kHz is in the passband of both the tweeter and the midrange.

At this point, I am satisfied that the ribbon will require much more than 100A peaks. It makes sense to have peak current capability of 300 to 400 amps.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.