ECL: Why'd they use negative voltages?
in [Design]
|
From: Jim Thompson on 7 Jun 2010 20:34 On Mon, 07 Jun 2010 17:27:00 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon(a)On-My-Web-Site.com> wrote: >On Mon, 07 Jun 2010 17:16:06 -0700, Joerg <invalid(a)invalid.invalid> >wrote: > >>Joel Koltner wrote: >>> I realize it was the early '60s and all, but why does ECL generally use >>> 0V for VCC and -5.2V for VEE, rather than, oh, say... 5V for VCC and 0V >>> for VEE? Something related to how things were done when toobs ruled? (I >>> realize that you can almost always run ECL off of 5V/0V -- and >>> apparently this was popular practice at one time?) >>> >>> And why 5.2V anyway? (Granted, 5.2V is no stranger than 6.3V filament >>> transformers, I suppose...) >>> >> >>As Jim said, noise immunity. But also from itself. The upper transistors >> are the most prone to generate transients in ECL, so it helps if their >>collectors tie right into chassis. In the old days 4-6 layer boards were >>unheard of. All you had was 2-layer phenolic, and only if you were >>lucky. Nowadays that's not an issue anymore because the +5V plane in a >>PECL scenario is just about as good an RF sink as the ground plane. >> >>Why 5.2V I don't know but 6.3V is not an arbitrary voltage, just like >>12.6V isn't. That fit the typical car battery voltages just right. So >>you could hang the filaments straight onto the battery voltage and only >>had to generate the plate voltages. That was initially done with a >>mechanical switcher where the "buzzer cartridge" would wear out once in >>a while. > >Original ECL had no current mirrors, just resistors, thus -5.5V gave >the nicest (but still crappy) TC. > >Later on there was 10K/100K. My biggest challenge was making a >threshold specification bracket overlay that would display >automatically in PSpice simulations :-) > > ...Jim Thompson I forgot to mention: Some ECL line drivers featured a positive voltage for the collectors of the output followers. It was NOT low power, but it sure as hell was fast :-) ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | The only thing bipartisan in this country is hypocrisy
From: Joel Koltner on 7 Jun 2010 20:35 <krw(a)att.bizzzzzzzzzzzz> wrote in message news:4e3r065vr383q2hc9fiuhre2mu9opopbai(a)4ax.com... >>And why 5.2V anyway? (Granted, 5.2V is no stranger than 6.3V filament >>transformers, I suppose...) > > Stack up the voltages (don't forget the AND gate). Purportedly it'll run down to around VCC-VEE = 3V and up to about 8V before you start seeing massive performance differences. 5.2V is certainly pretty close to the center of those two... hmm... I wouldn't be surprised if that's how they came upon it!
From: John Larkin on 7 Jun 2010 20:37 On Mon, 7 Jun 2010 16:56:24 -0700, "Joel Koltner" <zapwireDASHgroups(a)yahoo.com> wrote: >I realize it was the early '60s and all, but why does ECL generally use 0V for >VCC and -5.2V for VEE, rather than, oh, say... 5V for VCC and 0V for VEE? >Something related to how things were done when toobs ruled? (I realize that >you can almost always run ECL off of 5V/0V -- and apparently this was popular >practice at one time?) > >And why 5.2V anyway? (Granted, 5.2V is no stranger than 6.3V filament >transformers, I suppose...) > >---Joel The output swing and receiver threshold are relative to the chip Vcc, the nominal logic threshold being about 1.35 volts below Vcc. That's basic to the circuit topology. At these speeds, and with 0.8 volts of swing, it makes sense to refer the swing and matched-impedance traces to a quiet solid ground plane, so Vcc=0 volts makes sense. ECL chips aren't very sensitive to Vee. 10KH type ECL does work fine at Vcc=5, Vee=0. That's called "PECL" mode, originally "pseudo ECL" and lately "positive ECL". One generally references all the signals to a nice 5-volt copper pour. Don't know why 5.2, and not just 5.0. Newer stuff, like EclipsLite, works at 3.3 volts, and some at 2.5. I do mixed-mode PECL and cmos/FPGAs off a +3.3 volt supply. John
From: Joerg on 7 Jun 2010 20:39 Joel Koltner wrote: > "Joerg" <invalid(a)invalid.invalid> wrote in message > news:875gebFf6oU1(a)mid.individual.net... >> As Jim said, noise immunity. But also from itself. The upper >> transistors are the most prone to generate transients in ECL, so it >> helps if their collectors tie right into chassis. In the old days 4-6 >> layer boards were unheard of. All you had was 2-layer phenolic, and >> only if you were lucky. > > Ah, gotcha -- that makes a lot of sense. "Ground is ground, but 5V (or > 5.2V or whatever) tends to 'wiggle around' rather more..." (Especially > in that the earlier ECL series where VBB wasn't actively regulated > against supply voltage variations.) > Those often didn't afford us the luxury of a VBB output even being there. >> Why 5.2V I don't know but 6.3V is not an arbitrary voltage, just like >> 12.6V isn't. That fit the typical car battery voltages just right. > > Hmm... 13.8V = car running w/alternator floating a 6-cell battery and > 12.6V = car turned off, battery powering radio then? > The first cars were all 3-cell and then there's wiring losses and all that. Back then dropping a voltage was easy, raising one was next to impossible. And cars had DC generators. In fact, my old Citroen did as well. Crankshaft-driven, no belts in the whole car. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Joel Koltner on 7 Jun 2010 20:40
<dagmargoodboat(a)yahoo.com> wrote in message news:95d4bd95-50ea-4935-a027-13f508cb01e2(a)u7g2000yqm.googlegroups.com... >Hypocrisy may be bipartisan, but the two camps--socialists vs: free >market--are critically different, and not even totally defined by >party. I think what might have once been largely a bimodal distribution is now converging towards a normal distribution, James... albeit with two additional little spikes close to either end as well. :-) I fully expect that given some centuries or perhaps even a few millennia we're going to have a single world government anyway, most people will be some shade of tan without very distinctive physical attributes, etc. Although apparently a lot of people back in the FDR's day felt that this would have already happened by 2000... ---Joel |