The Pink Book documentation for the M68HC11 in expanded mode specifies
three bypass caps, a 10mF electrolytic, a 1 mF ceramic, and a 0.01 mF
ceramic. Why is so much capacitance required in expanded mode and not
in single chip mode?

What are the consequences of using the more typical 0.1 or 0.01 mF
bypass caps in expanded mode?

Thanks,
Dan



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

daniel62bull wrote:
> The Pink Book documentation for the M68HC11 in expanded mode specifies
> three bypass caps, a 10mF electrolytic, a 1 mF ceramic, and a 0.01 mF
> ceramic. Why is so much capacitance required in expanded mode and not
> in single chip mode?

It needs more in expanded mode, because it has to drive the bus, which
is highly capacitive at these frequencies, requiring charging current
spikes.

> What are the consequences of using the more typical 0.1 or 0.01 mF
> bypass caps in expanded mode?

Lotsa noise coupled into the power supply lines. Those bypass caps
need to be connected as closely to the PS pins of the chip as possible,
and not to a ground plane, otherwise you'll get ground bounce.

Also, don't return the clock crystal caps to a ground plane. Return
them to the Vss pin, for the same reasons.

> Thanks,
> Dan

Welcome.

Mike
--
p="p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
You have found the bank of Larn.
I can explain it for you, but I can't understand it for you.
I speak only for myself, and I am unanimous in that!



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

Dan,

It is because you are driving the multiplexed bus at internal bus speeds
(one could say the multiplexed bus portion is at twice the internal bus
speed). The longer wires connected to the external bus also have more
capacitance than the internal bus. These two factors together mean the
expanded-mode HC11 needs more energy when the switching occurs. I would
follow the recommendations (or more; you can almost never have too much
local energy storage for a microcontroller).

The biggest consequences of less capacitance would be more noise on the
power supply and ground conductors, unacceptable in a low noise
application. In some circumstances, the noise could be severe enough to
cause spurious errors and seemingly unexplained behavior.

Emmett Redd Ph.D. mailto:E...@missouristate.edu
Professor (417)836-5221
Department of Physics, Astronomy, and Materials Science
Missouri State University Fax (417)836-6226
901 SOUTH NATIONAL Dept (417)836-5131
SPRINGFIELD, MO 65897 USA

"When I was a child, I spake as a child, I understood as a child, I
thought as a child: but when I became a man, I put away childish
things," like being too embarrassed to ask questions which my peers
might think as stupid.

> -----Original Message-----
> From: m...@yahoogroups.com
> [mailto:m...@yahoogroups.com] On Behalf Of daniel62bull
> Sent: Wednesday, October 25, 2006 12:21 PM
> To: m...@yahoogroups.com
> Subject: [m68HC11] Bypass caps in expanded mode
>
> The Pink Book documentation for the M68HC11 in expanded mode specifies
> three bypass caps, a 10mF electrolytic, a 1 mF ceramic, and a 0.01 mF
> ceramic. Why is so much capacitance required in expanded mode and not
> in single chip mode?
>
> What are the consequences of using the more typical 0.1 or 0.01 mF
> bypass caps in expanded mode?
>
> Thanks,
> Dan
>
>



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

Redd, Emmett R wrote:

[a very nice explanation]

> expanded-mode HC11 needs more energy when the switching occurs. I would
> follow the recommendations (or more; you can almost never have too much
> local energy storage for a microcontroller).

I almost added to my message, and I'll add to yours, that this
recommendation is based on a few practical matters, along with
a caveat.

The rationale for putting a small cap in parallel with a large cap
is that the large cap works well for lower freq's, but often
(especially an electro) has a relatively high parasitic inductance
in series with it. So, replacing a 10uF cap with a 100uF cap
may actually result in *less* filtering/bypass capability
at the frequency of interest. If the frequency is above the
self-resonant freq of the cap, then it looks like an
inductor rather than a capacitor.

So a small cap in parallel works better for the higher freqs.
This may require more than one parallel cap. Usual ratios are
100:1, for example 10uF || 0.1uF || 1nF. Often, a better fix is
to use tantalum caps which have less parasitic inductance.

NB: I recommend against doing this in analogue high-frequency
circuits, where the active device may have gain at the resonant
frequency of the parasitic inductance of the larger cap in parallel
with the small cap. The practice of paralleling large electros with
smaller ceramics does not work well in these circuits, and can
result in very difficult to fix parasitic oscillations, which only
show up in certain of the manufactured articles, depending on
the specific characteristics of the particular electro which
happens to get installed on the board, with some of the boards
showing oscillation, while most do not. Murphy declares that
the prototype and the first 100 boards will not show the oscillation.
Only full production will show it up :-)

Unless this is a one-off home-built device, in which case
it will definitely show up, and be of a frequency which
the hobbiest has no means to detect, being too high for
his inexpensive 'scope and freq counter ;-)

> The biggest consequences of less capacitance would be more noise on the
> power supply and ground conductors, unacceptable in a low noise
> application. In some circumstances, the noise could be severe enough to
> cause spurious errors and seemingly unexplained behavior.

Yeah, like trying to use the A/D converter on the chip, for example.

Mike
--
p="p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
You have found the bank of Larn.
I can explain it for you, but I can't understand it for you.
I speak only for myself, and I am unanimous in that!



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

Hi Dan, Emmett, and Mike --

Help have I gone blind here or has everyone else???

According to my copy of the current "Pink Book" now known as the "White
Book" or more formally as "M68HC11RM/D" Rev 6, 4/2002, the power supply
bypass capacitors in Fig 2-22 "Expanded Mode Connections" are shown as
10uF, 1 uF, and .01 uF (In my opinion, quite normal values for such an
application) not the mF (.001 Farad units) you have cited. Are you sure
you have quoted the book correctly???

Best wishes,

Bob Smith

----- Original Message -----
From: "Redd, Emmett R"
To:
Sent: Wednesday, October 25, 2006 1:53 PM
Subject: RE: [m68HC11] Bypass caps in expanded mode
Dan,

It is because you are driving the multiplexed bus at internal bus speeds
(one could say the multiplexed bus portion is at twice the internal bus
speed). The longer wires connected to the external bus also have more
capacitance than the internal bus. These two factors together mean the
expanded-mode HC11 needs more energy when the switching occurs. I would
follow the recommendations (or more; you can almost never have too much
local energy storage for a microcontroller).

The biggest consequences of less capacitance would be more noise on the
power supply and ground conductors, unacceptable in a low noise
application. In some circumstances, the noise could be severe enough to
cause spurious errors and seemingly unexplained behavior.

Emmett Redd Ph.D. mailto:E...@missouristate.edu
Professor (417)836-5221
Department of Physics, Astronomy, and Materials Science
Missouri State University Fax (417)836-6226
901 SOUTH NATIONAL Dept (417)836-5131
SPRINGFIELD, MO 65897 USA

"When I was a child, I spake as a child, I understood as a child, I
thought as a child: but when I became a man, I put away childish
things," like being too embarrassed to ask questions which my peers
might think as stupid.
> -----Original Message-----
> From: m...@yahoogroups.com
> [mailto:m...@yahoogroups.com] On Behalf Of daniel62bull
> Sent: Wednesday, October 25, 2006 12:21 PM
> To: m...@yahoogroups.com
> Subject: [m68HC11] Bypass caps in expanded mode
>
> The Pink Book documentation for the M68HC11 in expanded mode specifies
> three bypass caps, a 10mF electrolytic, a 1 mF ceramic, and a 0.01 mF
> ceramic. Why is so much capacitance required in expanded mode and not
> in single chip mode?
>
> What are the consequences of using the more typical 0.1 or 0.01 mF
> bypass caps in expanded mode?
>
> Thanks,
> Dan



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

Robert Smith wrote:
> Hi Dan, Emmett, and Mike --
>
> Help have I gone blind here or has everyone else???
>
> According to my copy of the current "Pink Book" now known as the "White
> Book" or more formally as "M68HC11RM/D" Rev 6, 4/2002, the power supply
> bypass capacitors in Fig 2-22 "Expanded Mode Connections" are shown as
> 10uF, 1 uF, and .01 uF (In my opinion, quite normal values for such an
> application) not the mF (.001 Farad units) you have cited. Are you sure
> you have quoted the book correctly???

You are showing your (lack of) age here. Formerly, "mf" and "MF"
were used as abbreviations for "micro Farad", and "mmf" and "MMF"
for "micro micro Farad", which the europhiles have forced us to
abandon, and use "pF" for "pico Farad". Why we in the USA want to
kowtow to the ISO is beyond me. I'm not a fan of the United Nations
at all, not at all. I prefer "kc" and "mc" to "Hertz" any day.

Mike
--
p="p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
You have found the bank of Larn.
I can explain it for you, but I can't understand it for you.
I speak only for myself, and I am unanimous in that!



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

> -----Original Message-----
> [mailto:m...@yahoogroups.com] On Behalf Of Mike McCarty
> Sent: Wednesday, October 25, 2006 2:23 PM
> Subject: Re: [m68HC11] Bypass caps in expanded mode
>
> Redd, Emmett R wrote:
>
> [a very nice explanation]

Thanks Mike.



>
> I almost added to my message, and I'll add to yours, that this
> recommendation is based on a few practical matters, along with
> a caveat.
>
> The rationale for putting a small cap in parallel with a large cap
> is that the large cap works well for lower freq's, but often
> (especially an electro) has a relatively high parasitic inductance
> in series with it. So, replacing a 10uF cap with a 100uF cap
> may actually result in *less* filtering/bypass capability
> at the frequency of interest. If the frequency is above the
> self-resonant freq of the cap, then it looks like an
> inductor rather than a capacitor.
>
> So a small cap in parallel works better for the higher freqs.
> This may require more than one parallel cap. Usual ratios are
> 100:1, for example 10uF || 0.1uF || 1nF. Often, a better fix is
> to use tantalum caps which have less parasitic inductance.

Very nice addition.

>
> NB: I recommend against doing this in analogue high-frequency
> circuits, where the active device may have gain at the resonant
> frequency of the parasitic inductance of the larger cap in parallel
> with the small cap. The practice of paralleling large electros with
> smaller ceramics does not work well in these circuits, and can
> result in very difficult to fix parasitic oscillations, which only
> show up in certain of the manufactured articles, depending on
> the specific characteristics of the particular electro which
> happens to get installed on the board, with some of the boards
> showing oscillation, while most do not. Murphy declares that
> the prototype and the first 100 boards will not show the oscillation.
> Only full production will show it up :-)
>

I should probably look at this on my beta-test boards. Another
(practical)
way to look at this is that the digital high frequency that the small
cap is
attenuating comes from the (hopefully) square transitions. The analog
should not have such fast transitions lessening the need for the small
cap.



> Mike
> --

BTW, I am also old enough to not have been bothered by mF meaning uF.
Back
then-well before ultra-caps, everyone knew there no commonly available
milliFarad capacitors.

Emmett Redd Ph.D. mailto:E...@missouristate.edu
Professor (417)836-5221
Department of Physics, Astronomy, and Materials Science
Missouri State University Fax (417)836-6226
901 SOUTH NATIONAL Dept (417)836-5131
SPRINGFIELD, MO 65897 USA

"When I was a child, I spake as a child, I understood as a child, I
thought as a child: but when I became a man, I put away childish
things," like being too embarrassed to ask questions which my peers
might think as stupid.



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )

Redd, Emmett R wrote:
>>-----Original Message-----
>>[mailto:m...@yahoogroups.com] On Behalf Of Mike McCarty

[snip]

> Very nice addition.

Thanks.

>>NB: I recommend against doing this in analogue high-frequency
>>circuits, where the active device may have gain at the resonant
>>frequency of the parasitic inductance of the larger cap in parallel
>>with the small cap. The practice of paralleling large electros with

[snip]

> I should probably look at this on my beta-test boards. Another
> (practical)
> way to look at this is that the digital high frequency that the small
> cap is
> attenuating comes from the (hopefully) square transitions. The analog
> should not have such fast transitions lessening the need for the small
> cap.

Well, the issue is not so much the waveform, as the fact that
the active device is in its linear region, and may in fact
oscillate using the parasitic inductance and paralleled small
cap as the tank. A square wave which does not essentially
saturate and cut off the active device can still have parasitic
oscillations superimposed upon it. The device must be pushed to
the point that the gain is reduced at the resonant frequency
formed by the parasitic and the cap. This is what happens in
digital designs, and one reason why they are less susceptible
to electronic noise: reduced gain. An analogue design still
needs good bypass even though a square wave is passing through it.
Bypassing high frequency analogue devices is something of an art;
some would say a Black Art :-)

Mike
--
p="p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
You have found the bank of Larn.
I can explain it for you, but I can't understand it for you.
I speak only for myself, and I am unanimous in that!



(You need to be a member of hc11 -- send a blank email to hc11-subscribe@yahoogroups.com )