ECL: Why'd they use negative voltages?

that's

Yep. ...Jim Thompson

in

myself

Thus was born CML. ...Jim Thompson

Pathetic :)))))

Here is 10kW audio amplifier. Yes, 10kW true RMS sustained sinusoidal power. I designed the DSP and control loops for it. It indeed burns 2kW rated woofers while developing SPL of somewhat 170dB.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

One has to wonder what sort of speaker systems are used with such an amplifier (10kW) if 1.5kW is purportedly enough to kill any individual speaker -- must be quite the array!

Ah, only a 200Hz frequency limit though:

Still, a very cool accomplishment...

---Joel

You should have seen one of the first generation NTSC effects generator called the 'SqueezeZoom' It had a linear 5 V 1000 A linear power supply with a 208 three phase input.

V for

that

ular

ECL uses NPN emitter follower transistors to buffer the output signal. If you bias ECL between 0V and 5V (PECL)and short an output to ground, you can draw quite a lot current from the small and rather fragile output transistors, with a serious risk of blowing them up.

Bias ECL between 0V and -5.2V (or whatever - 100K ECL used -4.5V) and shorting the output to ground merely reveres biases the base-emitter junction, and nothing blows up.

Hewlett-Packard used ECL biased between )v and 5V in their laser- interferometer positioning system, and when I complained about the risk to one of the HP reps, he did admit that most of their repair work on duff laser interferometers was replacing ECL chips with dead outputs.

Admittedly, they had made life riskier by putting a differential ECL signal pair on the inner and outer of a singel regular coaxial connector, which left a biggish chunk of metal that service engineers could brush with the ground lead on their scope probe.

Trust Jim-out-of-touch-with=3Dreality-Thompson to be unaware of this.

-- Bill Sloman, Nijmegen

My "home speakers" could suck down 600W (each). They were 4 ft wide, 3.5 deep and 6 ft tall. They weighed 1200 pounds. The neighbors would visit and ask, "And how are the *children* today?"

Yikes! I have no idea what the entire "tester" consumed (my responsibility was primarily the CPU). We had a dedicated service installed *just* for it, though. Power cord took two of us to plug in (and twist-lock)... some idiot mounted the outlet at eye level (not very smart when you're lifting all that copper/rubber and trying to mate the things).

(sigh) As I said, disappointing to see such brute force used. Even the UUT connection was by way of a third horsepower motor driven "connector engagement system". Though I guess it would be hard to do it otherwise (today you could probably integrate much of the electronics and get some huge wins from just reducing the size of everything)

The recomended system is 4 x 18' woofers.

I have some doubts. The actual impedance of the speakers is usually much higher then rated resistance. A good big loudspeaker can dissipate few hundred W of sustained active power. The mechanical damage of the speakers (due to the material fatigue) is big problem at those power levels also.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Each cabinet had 4x15's, 9x5's and 9x2's. Cabinet walls were two inches thick (plywood + particle board laminated together). Screws every 3" (grid). I think there were 60 pounds of screws in each cabinet -- or, perhaps that was per "half cabinet"? (my first speakers taught me that doors aren't as wide as you would like them to be! :< So I cut the cabinets in half to make it possible to get them in and out of buildings! One of my first cabinets still sits in my parents' basement as it wouldn't fit through the upstairs doorway)

Somewhere I have photos of me *inside* one of the half-cabinets. And sitting *atop* one of my "first" cabinets (they were much smaller -- 4 ft tall, 3 ft wide/deep)

See....

that's

Ok, that makes sense.

Inverters don't, at least if you want a valid output level. Never had any trouble with non-inverting stages.

Vbb is generated in some chips as the reference switch level for differential inputs, usually 1.35 volts below Vcc. The on-chip Vbb supplies are usually wimpy, too weak and at the wrong voltage for terminations.

The "standard" ECL termination is usually 50 ohms to -2 volts. Really short runs can be unterminated, but still need a pulldown; a couple hundred ohms to -5 will usually do.

You can do cool delay and one-shot things with ECL by adding caps and weak pulldowns to outputs.

John

Nothing elegant, but someone's got to do it. The clocking was fun.

How about a UHF transmitter tube with a pair of 1.5 V 1000 A filaments? They had to be balanced to 1/100 of a volt, so there was a pair of 'resistors' made out of 3" * 1" buss bar that was about 18" long. There was a large steel stud in each end. You used a wrench to tighten the nuts on one 'resistor' to adjust it, to balance the voltages.

On the other end was a high current, water cooled 7 kV plate supply. Lots of places in old broadcast gear to cause an explosion, or to electrocute someone. :(

I once designed a VME module with a lot of PECL with signals terminated into +3V, and which also had some circuitry running between +3V and GND. The +3V net was shared. Since the combined current of all the PECL terminators largely exceeded the consumption of the stuff between +3V and GND, I used a negative regulator with its input connected to GND to make the +3V.

That raised some eyebrows, but it worked fine.

Jeroen Belleman

On Jun 7, 7:16=A0pm, Joerg wrote: [...]

Buzzer cartridge?? Are you allergic to the term VIBRATOR?? LOL

On a sunny day (Tue, 8 Jun 2010 01:40:53 -0700 (PDT)) it happened Greegor wrote in :

You also designing PIC programmable vibrator?