Prev: Thermal Epoxy
Next: paypal wholesale Y-3 (paypal payment)( www.jordanonline06.com)
From: Andrew Holme on 20 Mar 2010 11:02
The MC100EP142 has "default LOW" inputs with internal 75k pull-down
resistors to VEE. Default HIGH inputs also have internal 36.5k pull-ups to
VCC which bias them between VIH (min) and VIH (max) but the default LOW
(=VEE) is way outside the "operational" VIL range. One might infer it's OK
to bias inputs down at VEE, provided they're static; but I wish the
datasheet was more explicit on this!

I'm not confident about leaving inputs open. I would rather tie them off to
VEE, if that's OK. In the case of inputs defaulting high, where an
application requires a permanent high state on the input, and a suitable low
impedance source is not conveniently available to drive it externally, I
would be tempted to at least add an external decoupling capacitor to help
noise immunity.



From: John Larkin on 20 Mar 2010 13:46
On Sat, 20 Mar 2010 15:02:55 -0000, "Andrew Holme" <ah(a)nospam.com>
wrote:

>The MC100EP142 has "default LOW" inputs with internal 75k pull-down
>resistors to VEE. Default HIGH inputs also have internal 36.5k pull-ups to
>VCC which bias them between VIH (min) and VIH (max) but the default LOW
>(=VEE) is way outside the "operational" VIL range. One might infer it's OK
>to bias inputs down at VEE, provided they're static; but I wish the
>datasheet was more explicit on this!

Yup, ECL is happy with inputs at Vee. Most are also happy with inputs
at Vcc (strapped high) but some older parts aren't.

>
>I'm not confident about leaving inputs open. I would rather tie them off to
>VEE, if that's OK.

That's OK, but it doesn't matter. You'd need an insane amount of noise
to pull a pin from Vee up to the logic threshold. I've never seen it
happen.

In the case of inputs defaulting high, where an
>application requires a permanent high state on the input, and a suitable low
>impedance source is not conveniently available to drive it externally, I
>would be tempted to at least add an external decoupling capacitor to help
>noise immunity.

That case has less noise margin, but I doubt you'd have trouble there
either. The '142 datasheet says that the diff input common mode range
includes Vcc, so you could just pull it high.

The '142 differential inputs have one pin pulled down to Vee and the
other weakly held at about -1.6 in a 5-volt ECL circuit. That's not
technically proper "ECL" levels, but it's a solid diff input given the
common-mode range of the input circuit. Base currents are pretty high
in ECL.

John


From: krw on 20 Mar 2010 14:20
On Sat, 20 Mar 2010 10:46:12 -0700, John Larkin
<jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:

>On Sat, 20 Mar 2010 15:02:55 -0000, "Andrew Holme" <ah(a)nospam.com>
>wrote:
>
>>The MC100EP142 has "default LOW" inputs with internal 75k pull-down
>>resistors to VEE. Default HIGH inputs also have internal 36.5k pull-ups to
>>VCC which bias them between VIH (min) and VIH (max) but the default LOW
>>(=VEE) is way outside the "operational" VIL range. One might infer it's OK
>>to bias inputs down at VEE, provided they're static; but I wish the
>>datasheet was more explicit on this!
>
>Yup, ECL is happy with inputs at Vee. Most are also happy with inputs
>at Vcc (strapped high) but some older parts aren't.

Strapping ECL inputs at VCC saturates that side of the current switch. If the
input is on the inverting side of the switch it can affect the output level,
as well. It's better to strap them a diode down from VCC.

>>I'm not confident about leaving inputs open. I would rather tie them off to
>>VEE, if that's OK.
>
>That's OK, but it doesn't matter. You'd need an insane amount of noise
>to pull a pin from Vee up to the logic threshold. I've never seen it
>happen.
>
> In the case of inputs defaulting high, where an
>>application requires a permanent high state on the input, and a suitable low
>>impedance source is not conveniently available to drive it externally, I
>>would be tempted to at least add an external decoupling capacitor to help
>>noise immunity.
>
>That case has less noise margin, but I doubt you'd have trouble there
>either. The '142 datasheet says that the diff input common mode range
>includes Vcc, so you could just pull it high.
>
>The '142 differential inputs have one pin pulled down to Vee and the
>other weakly held at about -1.6 in a 5-volt ECL circuit. That's not
>technically proper "ECL" levels, but it's a solid diff input given the
>common-mode range of the input circuit. Base currents are pretty high
>in ECL.
>
>John
>
From: John Larkin on 20 Mar 2010 14:37
On Sat, 20 Mar 2010 13:20:54 -0500, "krw(a)att.bizzzzzzzzzzzz"
<krw(a)att.bizzzzzzzzzzzz> wrote:

>On Sat, 20 Mar 2010 10:46:12 -0700, John Larkin
><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Sat, 20 Mar 2010 15:02:55 -0000, "Andrew Holme" <ah(a)nospam.com>
>>wrote:
>>
>>>The MC100EP142 has "default LOW" inputs with internal 75k pull-down
>>>resistors to VEE. Default HIGH inputs also have internal 36.5k pull-ups to
>>>VCC which bias them between VIH (min) and VIH (max) but the default LOW
>>>(=VEE) is way outside the "operational" VIL range. One might infer it's OK
>>>to bias inputs down at VEE, provided they're static; but I wish the
>>>datasheet was more explicit on this!
>>
>>Yup, ECL is happy with inputs at Vee. Most are also happy with inputs
>>at Vcc (strapped high) but some older parts aren't.
>
>Strapping ECL inputs at VCC saturates that side of the current switch. If the
>input is on the inverting side of the switch it can affect the output level,
>as well. It's better to strap them a diode down from VCC.

As I said, some older parts do that. Most newer ECL is happy with a
commmon-mode input range that includes Vcc. The datasheets usually
specify.

There's usually a Vbb available somewhere, so there's no reason to
risk it.

John

From: Andrew Holme on 20 Mar 2010 15:58

"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message
news:r81aq51jobb2athvc55u5eqijb752r4p6f(a)4ax.com...
> On Sat, 20 Mar 2010 15:02:55 -0000, "Andrew Holme" <ah(a)nospam.com>
> wrote:
>
>>The MC100EP142 has "default LOW" inputs with internal 75k pull-down
>>resistors to VEE. Default HIGH inputs also have internal 36.5k pull-ups
>>to
>>VCC which bias them between VIH (min) and VIH (max) but the default LOW
>>(=VEE) is way outside the "operational" VIL range. One might infer it's
>>OK
>>to bias inputs down at VEE, provided they're static; but I wish the
>>datasheet was more explicit on this!
>
> Yup, ECL is happy with inputs at Vee. Most are also happy with inputs
> at Vcc (strapped high) but some older parts aren't.
>
>>
>>I'm not confident about leaving inputs open. I would rather tie them off
>>to
>>VEE, if that's OK.
>
> That's OK, but it doesn't matter. You'd need an insane amount of noise
> to pull a pin from Vee up to the logic threshold. I've never seen it
> happen.
>
> In the case of inputs defaulting high, where an
>>application requires a permanent high state on the input, and a suitable
>>low
>>impedance source is not conveniently available to drive it externally, I
>>would be tempted to at least add an external decoupling capacitor to help
>>noise immunity.
>
> That case has less noise margin, but I doubt you'd have trouble there
> either. The '142 datasheet says that the diff input common mode range
> includes Vcc, so you could just pull it high.
>
> The '142 differential inputs have one pin pulled down to Vee and the
> other weakly held at about -1.6 in a 5-volt ECL circuit. That's not
> technically proper "ECL" levels, but it's a solid diff input given the
> common-mode range of the input circuit. Base currents are pretty high
> in ECL.
>
> John
>
>

The '142 has a lot of single-ended inputs. Can these be tied to VCC?

Thanks.


 |  Next  |  Last
Pages: 1 2
Prev: Thermal Epoxy
Next: paypal wholesale Y-3 (paypal payment)( www.jordanonline06.com)