One of my two hobbyist DS10L systems has bit it; the main logic board
is dead and it just isn't cost feasible to repair/replace right now.
So I have a full set of working spares (less the MLB, all tested in
the second DS10L) including two extra 512MB DIMMs, sadly not the rare
low profile DataRAM models.  And I have some tools (and access to
others) and I'm not afraid to use them. especially on spare parts.

I was thinking about cutting a nice rectangle in the lid right over
the memory slots.  Then making a sheet metal cap that would provide
enough clearance so that I could use the DIMMs vertically instead of
with the right-angle adapters.  That would allow me to take my working
DS10L up to 2GB.  The cap would get riveted to the case lid (I don't
have access to a welder, nor the skills to non-destructively weld
sheet metal).

I don't expect any problems except for possibly the frontmost case
screw mount, which might partially overlap with the area needing to be
cut.  Given continuous metal still shielding the system/RAM, am I
likely to run into any EMI problems caused by the system? I don't
think cooling will be affected enough to matter (as long as there is
some turbulence to prevent the metal cap from becoming a dead-air
hotspot).

Fun to think about while I wish  (and keep an eye out for) a free/
cheap DS15 or free/dirt cheap DS10...

Rich

On Thursday, March 15, 2012 1:09:03 PM UTC-4, Rich Jordan wrote:
> One of my two hobbyist DS10L systems has bit it; the main logic board
> is dead and it just isn't cost feasible to repair/replace right now.
> So I have a full set of working spares (less the MLB, all tested in
> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
> low profile DataRAM models.  And I have some tools (and access to
> others) and I'm not afraid to use them. especially on spare parts.
>=20
> I was thinking about cutting a nice rectangle in the lid right over
> the memory slots.  Then making a sheet metal cap that would provide
> enough clearance so that I could use the DIMMs vertically instead of
> with the right-angle adapters.  That would allow me to take my working
> DS10L up to 2GB.  The cap would get riveted to the case lid (I don't
> have access to a welder, nor the skills to non-destructively weld
> sheet metal).
>=20
> I don't expect any problems except for possibly the frontmost case
> screw mount, which might partially overlap with the area needing to be
> cut.  Given continuous metal still shielding the system/RAM, am I
> likely to run into any EMI problems caused by the system? I don't
> think cooling will be affected enough to matter (as long as there is
> some turbulence to prevent the metal cap from becoming a dead-air
> hotspot).
>=20
> Fun to think about while I wish  (and keep an eye out for) a free/
> cheap DS15 or free/dirt cheap DS10...
>=20
> Rich

The biggest concern with modding the cabinet is air flow. As long as the "b=
ox" remains sealed, you should be OK.

On another note:  What appears to be the problem with the MB?  It is rare f=
or these to fail.  If you are in the US, perhaps I can repair for a reasona=
ble fee.  Contact me offline.  I may also have a few spares around.  Shoot =
me some details.  Let me know also if you find a cheap/free DS15 as I could=
 use one too...

Dan

In article <10808166.1468.1331834038332.JavaMail.geo-discussion-forums@vbbfy7>, abrsvc <dansabrservices@yahoo.com> writes:
>On Thursday, March 15, 2012 1:09:03 PM UTC-4, Rich Jordan wrote:
>> One of my two hobbyist DS10L systems has bit it; the main logic board
>> is dead and it just isn't cost feasible to repair/replace right now.
>> So I have a full set of working spares (less the MLB, all tested in
>> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
>> low profile DataRAM models.  And I have some tools (and access to
>> others) and I'm not afraid to use them. especially on spare parts.
>>=20
>> I was thinking about cutting a nice rectangle in the lid right over
>> the memory slots.  Then making a sheet metal cap that would provide
>> enough clearance so that I could use the DIMMs vertically instead of
>> with the right-angle adapters.  That would allow me to take my working
>> DS10L up to 2GB.  The cap would get riveted to the case lid (I don't
>> have access to a welder, nor the skills to non-destructively weld
>> sheet metal).
>>=20
>> I don't expect any problems except for possibly the frontmost case
>> screw mount, which might partially overlap with the area needing to be
>> cut.  Given continuous metal still shielding the system/RAM, am I
>> likely to run into any EMI problems caused by the system? I don't
>> think cooling will be affected enough to matter (as long as there is
>> some turbulence to prevent the metal cap from becoming a dead-air
>> hotspot).
>>=20
>> Fun to think about while I wish  (and keep an eye out for) a free/
>> cheap DS15 or free/dirt cheap DS10...
>>=20
>> Rich
>
>The biggest concern with modding the cabinet is air flow. As long as the "b=
>ox" remains sealed, you should be OK.
>
>On another note:  What appears to be the problem with the MB?  It is rare f=
>or these to fail.  If you are in the US, perhaps I can repair for a reasona=
>ble fee.  Contact me offline.  I may also have a few spares around.  Shoot =
>me some details.  Let me know also if you find a cheap/free DS15 as I could=
> use one too...

I did it with mine.  First, I "nibbled" just enough space to allow the memory
to poke through.  Then, I built a cap for it out of two pieces of clear plex-
iglass. One piece serves as a spacer, the other as the cover.  I toyed with
the idea of modifying the sockets but decided against that.  

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

Well I speak to machines with the voice of humanity.

On Mar 15, 2:20=A0pm, VAXman-  @SendSpamHere.ORG wrote:
> In article <10808166.1468.1331834038332.JavaMail.geo-discussion-forums@vb=
bfy7>, abrsvc <dansabrservi...@yahoo.com> writes:
> >On Thursday, March 15, 2012 1:09:03 PM UTC-4, Rich Jordan wrote:
> >> One of my two hobbyist DS10L systems has bit it; the main logic board
> >> is dead and it just isn't cost feasible to repair/replace right now.
> >> So I have a full set of working spares (less the MLB, all tested in
> >> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
> >> low profile DataRAM models. =A0And I have some tools (and access to
> >> others) and I'm not afraid to use them. especially on spare parts.
> >>=3D20
> >> I was thinking about cutting a nice rectangle in the lid right over
> >> the memory slots. =A0Then making a sheet metal cap that would provide
> >> enough clearance so that I could use the DIMMs vertically instead of
> >> with the right-angle adapters. =A0That would allow me to take my worki=
ng
> >> DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don't
> >> have access to a welder, nor the skills to non-destructively weld
> >> sheet metal).
> >>=3D20
> >> I don't expect any problems except for possibly the frontmost case
> >> screw mount, which might partially overlap with the area needing to be
> >> cut. =A0Given continuous metal still shielding the system/RAM, am I
> >> likely to run into any EMI problems caused by the system? I don't
> >> think cooling will be affected enough to matter (as long as there is
> >> some turbulence to prevent the metal cap from becoming a dead-air
> >> hotspot).
> >>=3D20
> >> Fun to think about while I wish =A0(and keep an eye out for) a free/
> >> cheap DS15 or free/dirt cheap DS10...
> >>=3D20
> >> Rich
>
> >The biggest concern with modding the cabinet is air flow. As long as the=
 "b=3D
> >ox" remains sealed, you should be OK.
>
> >On another note: =A0What appears to be the problem with the MB? =A0It is=
 rare f=3D
> >or these to fail. =A0If you are in the US, perhaps I can repair for a re=
asona=3D
> >ble fee. =A0Contact me offline. =A0I may also have a few spares around. =
=A0Shoot =3D
> >me some details. =A0Let me know also if you find a cheap/free DS15 as I =
could=3D
> > use one too...
>
> I did it with mine. =A0First, I "nibbled" just enough space to allow the =
memory
> to poke through. =A0Then, I built a cap for it out of two pieces of clear=
 plex-
> iglass. One piece serves as a spacer, the other as the cover. =A0I toyed =
with
> the idea of modifying the sockets but decided against that.
>
> --
> VAXman- A Bored Certified VMS Kernel Mode Hacker =A0 =A0VAXman(at)TMESIS(=
dot)ORG
>
> Well I speak to machines with the voice of humanity.

I'm definitely not touching the sockets; I only have one working
system right now.

Plexi or plastic would be easier to work with, I just don't want to
create RF interference to my neighbors (or my clock radio, which I
tragically still use ;)

On Mar 15, 5:31=A0pm, VAXman-  @SendSpamHere.ORG wrote:
> In article <c9081d4e-59d1-4ab7-9c78-c62e1ab03...@i5g2000yqo.googlegroups.=
com>, Rich Jordan <jor...@ccs4vms.com> writes:
>
>
>
>
>
>
>
>
>
> >On Mar 15, 2:20=3DA0pm, VAXman- =A0@SendSpamHere.ORG wrote:
> >> In article <10808166.1468.1331834038332.JavaMail.geo-discussion-forums=
@vb=3D
> >bfy7>, abrsvc <dansabrservi...@yahoo.com> writes:
> >> >On Thursday, March 15, 2012 1:09:03 PM UTC-4, Rich Jordan wrote:
> >> >> One of my two hobbyist DS10L systems has bit it; the main logic boa=
rd
> >> >> is dead and it just isn't cost feasible to repair/replace right now=
..
> >> >> So I have a full set of working spares (less the MLB, all tested in
> >> >> the second DS10L) including two extra 512MB DIMMs, sadly not the ra=
re
> >> >> low profile DataRAM models. =3DA0And I have some tools (and access =
to
> >> >> others) and I'm not afraid to use them. especially on spare parts.
> >> >>=3D3D20
> >> >> I was thinking about cutting a nice rectangle in the lid right over
> >> >> the memory slots. =3DA0Then making a sheet metal cap that would pro=
vide
> >> >> enough clearance so that I could use the DIMMs vertically instead o=
f
> >> >> with the right-angle adapters. =3DA0That would allow me to take my =
worki=3D
> >ng
> >> >> DS10L up to 2GB. =3DA0The cap would get riveted to the case lid (I =
don't
> >> >> have access to a welder, nor the skills to non-destructively weld
> >> >> sheet metal).
> >> >>=3D3D20
> >> >> I don't expect any problems except for possibly the frontmost case
> >> >> screw mount, which might partially overlap with the area needing to=
 be
> >> >> cut. =3DA0Given continuous metal still shielding the system/RAM, am=
 I
> >> >> likely to run into any EMI problems caused by the system? I don't
> >> >> think cooling will be affected enough to matter (as long as there i=
s
> >> >> some turbulence to prevent the metal cap from becoming a dead-air
> >> >> hotspot).
> >> >>=3D3D20
> >> >> Fun to think about while I wish =3DA0(and keep an eye out for) a fr=
ee/
> >> >> cheap DS15 or free/dirt cheap DS10...
> >> >>=3D3D20
> >> >> Rich
>
> >> >The biggest concern with modding the cabinet is air flow. As long as =
the=3D
> > "b=3D3D
> >> >ox" remains sealed, you should be OK.
>
> >> >On another note: =3DA0What appears to be the problem with the MB? =3D=
A0It is=3D
> > rare f=3D3D
> >> >or these to fail. =3DA0If you are in the US, perhaps I can repair for=
 a re=3D
> >asona=3D3D
> >> >ble fee. =3DA0Contact me offline. =3DA0I may also have a few spares a=
round. =3D
> >=3DA0Shoot =3D3D
> >> >me some details. =3DA0Let me know also if you find a cheap/free DS15 =
as I =3D
> >could=3D3D
> >> > use one too...
>
> >> I did it with mine. =3DA0First, I "nibbled" just enough space to allow=
 the =3D
> >memory
> >> to poke through. =3DA0Then, I built a cap for it out of two pieces of =
clear=3D
> > plex-
> >> iglass. One piece serves as a spacer, the other as the cover. =3DA0I t=
oyed =3D
> >with
> >> the idea of modifying the sockets but decided against that.
>
> >> --
> >> VAXman- A Bored Certified VMS Kernel Mode Hacker =3DA0 =3DA0VAXman(at)=
TMESIS(=3D
> >dot)ORG
>
> >> Well I speak to machines with the voice of humanity.
>
> >I'm definitely not touching the sockets; I only have one working
> >system right now.
>
> >Plexi or plastic would be easier to work with, I just don't want to
> >create RF interference to my neighbors (or my clock radio, which I
> >tragically still use ;)
>
> No more radio noise than your neighbor is causing you by running that
> PeeCee of theirs in the plastic box or with the cover off.
>
> FWIW, and I 'm sure I mentioned this here before, the DS10s will NOT
> run with the 1GB DIMMs that work in their bigger bretheren DS20/ES4x.
> I wish it would have worked as then I could have 2 GB without having
> to modify the case.
> --
> VAXman- A Bored Certified VMS Kernel Mode Hacker =A0 =A0VAXman(at)TMESIS(=
dot)ORG
>
> Well I speak to machines with the voice of humanity.

I wish I could find those mythical DataRam modules that were short
enough to run in a DS10L without the adapters!  2GB would be very nice
to have.  But at this point a DS15 is the dream machine (anything else
is too noisy or too big).  But they sure still attract a lot of money
in the aftermarket... too much for a hobby box.

Rich Jordan wrote 2012-03-15 22:43:
> On Mar 15, 5:31 pm, VAXman-  @SendSpamHere.ORG wrote:
>> In article<c9081d4e-59d1-4ab7-9c78-c62e1ab03...@i5g2000yqo.googlegroups.com>, Rich Jordan<jor...@ccs4vms.com>  writes:
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>> On Mar 15, 2:20=A0pm, VAXman-  @SendSpamHere.ORG wrote:
>>>> In article<10808166.1468.1331834038332.JavaMail.geo-discussion-forums@vb=
>>> bfy7>, abrsvc<dansabrservi...@yahoo.com>  writes:
>>>>> On Thursday, March 15, 2012 1:09:03 PM UTC-4, Rich Jordan wrote:
>>>>>> One of my two hobbyist DS10L systems has bit it; the main logic board
>>>>>> is dead and it just isn't cost feasible to repair/replace right now.
>>>>>> So I have a full set of working spares (less the MLB, all tested in
>>>>>> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
>>>>>> low profile DataRAM models. =A0And I have some tools (and access to
>>>>>> others) and I'm not afraid to use them. especially on spare parts.
>>>>>> =3D20
>>>>>> I was thinking about cutting a nice rectangle in the lid right over
>>>>>> the memory slots. =A0Then making a sheet metal cap that would provide
>>>>>> enough clearance so that I could use the DIMMs vertically instead of
>>>>>> with the right-angle adapters. =A0That would allow me to take my worki=
>>> ng
>>>>>> DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don't
>>>>>> have access to a welder, nor the skills to non-destructively weld
>>>>>> sheet metal).
>>>>>> =3D20
>>>>>> I don't expect any problems except for possibly the frontmost case
>>>>>> screw mount, which might partially overlap with the area needing to be
>>>>>> cut. =A0Given continuous metal still shielding the system/RAM, am I
>>>>>> likely to run into any EMI problems caused by the system? I don't
>>>>>> think cooling will be affected enough to matter (as long as there is
>>>>>> some turbulence to prevent the metal cap from becoming a dead-air
>>>>>> hotspot).
>>>>>> =3D20
>>>>>> Fun to think about while I wish =A0(and keep an eye out for) a free/
>>>>>> cheap DS15 or free/dirt cheap DS10...
>>>>>> =3D20
>>>>>> Rich
>>
>>>>> The biggest concern with modding the cabinet is air flow. As long as the=
>>> "b=3D
>>>>> ox" remains sealed, you should be OK.
>>
>>>>> On another note: =A0What appears to be the problem with the MB? =A0It is=
>>> rare f=3D
>>>>> or these to fail. =A0If you are in the US, perhaps I can repair for a re=
>>> asona=3D
>>>>> ble fee. =A0Contact me offline. =A0I may also have a few spares around. =
>>> =A0Shoot =3D
>>>>> me some details. =A0Let me know also if you find a cheap/free DS15 as I =
>>> could=3D
>>>>> use one too...
>>
>>>> I did it with mine. =A0First, I "nibbled" just enough space to allow the =
>>> memory
>>>> to poke through. =A0Then, I built a cap for it out of two pieces of clear=
>>> plex-
>>>> iglass. One piece serves as a spacer, the other as the cover. =A0I toyed =
>>> with
>>>> the idea of modifying the sockets but decided against that.
>>
>>>> --
>>>> VAXman- A Bored Certified VMS Kernel Mode Hacker =A0 =A0VAXman(at)TMESIS(=
>>> dot)ORG
>>
>>>> Well I speak to machines with the voice of humanity.
>>
>>> I'm definitely not touching the sockets; I only have one working
>>> system right now.
>>
>>> Plexi or plastic would be easier to work with, I just don't want to
>>> create RF interference to my neighbors (or my clock radio, which I
>>> tragically still use ;)
>>
>> No more radio noise than your neighbor is causing you by running that
>> PeeCee of theirs in the plastic box or with the cover off.
>>
>> FWIW, and I 'm sure I mentioned this here before, the DS10s will NOT
>> run with the 1GB DIMMs that work in their bigger bretheren DS20/ES4x.
>> I wish it would have worked as then I could have 2 GB without having
>> to modify the case.
>> --
>> VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG
>>
>> Well I speak to machines with the voice of humanity.
>
> I wish I could find those mythical DataRam modules that were short
> enough to run in a DS10L without the adapters!  2GB would be very nice
> to have.  But at this point a DS15 is the dream machine (anything else
> is too noisy or too big).  But they sure still attract a lot of money
> in the aftermarket... too much for a hobby box.
>

Correct. I asked my local used DEC/CPQ/HP gear brooker for a DS15,
but he could not fix one at a "hobbyist-price", the demand was
(and still seems to be) to high for them. Got a dual-CPU DS25 in
cabinet package instead (since everyone wants rack mounting). :-)

In article <c9081d4e-59d1-4ab7-9c78-c62e1ab03874@i5g2000yqo.googlegroups.com>, Rich Jordan <jordan@ccs4vms.com> writes:
>On Mar 15, 2:20=A0pm, VAXman-  @SendSpamHere.ORG wrote:
>> In article <10808166.1468.1331834038332.JavaMail.geo-discussion-forums@vb=
>bfy7>, abrsvc <dansabrservi...@yahoo.com> writes:
>> >On Thursday, March 15, 2012 1:09:03 PM UTC-4, Rich Jordan wrote:
>> >> One of my two hobbyist DS10L systems has bit it; the main logic board
>> >> is dead and it just isn't cost feasible to repair/replace right now.
>> >> So I have a full set of working spares (less the MLB, all tested in
>> >> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
>> >> low profile DataRAM models. =A0And I have some tools (and access to
>> >> others) and I'm not afraid to use them. especially on spare parts.
>> >>=3D20
>> >> I was thinking about cutting a nice rectangle in the lid right over
>> >> the memory slots. =A0Then making a sheet metal cap that would provide
>> >> enough clearance so that I could use the DIMMs vertically instead of
>> >> with the right-angle adapters. =A0That would allow me to take my worki=
>ng
>> >> DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don't
>> >> have access to a welder, nor the skills to non-destructively weld
>> >> sheet metal).
>> >>=3D20
>> >> I don't expect any problems except for possibly the frontmost case
>> >> screw mount, which might partially overlap with the area needing to be
>> >> cut. =A0Given continuous metal still shielding the system/RAM, am I
>> >> likely to run into any EMI problems caused by the system? I don't
>> >> think cooling will be affected enough to matter (as long as there is
>> >> some turbulence to prevent the metal cap from becoming a dead-air
>> >> hotspot).
>> >>=3D20
>> >> Fun to think about while I wish =A0(and keep an eye out for) a free/
>> >> cheap DS15 or free/dirt cheap DS10...
>> >>=3D20
>> >> Rich
>>
>> >The biggest concern with modding the cabinet is air flow. As long as the=
> "b=3D
>> >ox" remains sealed, you should be OK.
>>
>> >On another note: =A0What appears to be the problem with the MB? =A0It is=
> rare f=3D
>> >or these to fail. =A0If you are in the US, perhaps I can repair for a re=
>asona=3D
>> >ble fee. =A0Contact me offline. =A0I may also have a few spares around. =
>=A0Shoot =3D
>> >me some details. =A0Let me know also if you find a cheap/free DS15 as I =
>could=3D
>> > use one too...
>>
>> I did it with mine. =A0First, I "nibbled" just enough space to allow the =
>memory
>> to poke through. =A0Then, I built a cap for it out of two pieces of clear=
> plex-
>> iglass. One piece serves as a spacer, the other as the cover. =A0I toyed =
>with
>> the idea of modifying the sockets but decided against that.
>>
>> --
>> VAXman- A Bored Certified VMS Kernel Mode Hacker =A0 =A0VAXman(at)TMESIS(=
>dot)ORG
>>
>> Well I speak to machines with the voice of humanity.
>
>I'm definitely not touching the sockets; I only have one working
>system right now.
>
>Plexi or plastic would be easier to work with, I just don't want to
>create RF interference to my neighbors (or my clock radio, which I
>tragically still use ;)

No more radio noise than your neighbor is causing you by running that 
PeeCee of theirs in the plastic box or with the cover off.

FWIW, and I 'm sure I mentioned this here before, the DS10s will NOT
run with the 1GB DIMMs that work in their bigger bretheren DS20/ES4x.
I wish it would have worked as then I could have 2 GB without having
to modify the case.
-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

Well I speak to machines with the voice of humanity.

On Mar 15, 12:09=A0pm, Rich Jordan <jor...@ccs4vms.com> wrote:
> One of my two hobbyist DS10L systems has bit it; the main logic board
> is dead and it just isn't cost feasible to repair/replace right now.
> So I have a full set of working spares (less the MLB, all tested in
> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
> low profile DataRAM models. =A0And I have some tools (and access to
> others) and I'm not afraid to use them. especially on spare parts.
>
> I was thinking about cutting a nice rectangle in the lid right over
> the memory slots. =A0Then making a sheet metal cap that would provide
> enough clearance so that I could use the DIMMs vertically instead of
> with the right-angle adapters. =A0That would allow me to take my working
> DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don't
> have access to a welder, nor the skills to non-destructively weld
> sheet metal).
>
> I don't expect any problems except for possibly the frontmost case
> screw mount, which might partially overlap with the area needing to be
> cut. =A0Given continuous metal still shielding the system/RAM, am I
> likely to run into any EMI problems caused by the system? I don't
> think cooling will be affected enough to matter (as long as there is
> some turbulence to prevent the metal cap from becoming a dead-air
> hotspot).
>
> Fun to think about while I wish =A0(and keep an eye out for) a free/
> cheap DS15 or free/dirt cheap DS10...
>
> Rich

On further study of the service guide it looks like the DS10 series do
have SROM access like older boxes but so far I haven't found info on
accessing it.  There's a 10-pin IDC-style port labeled SROM (factory
only), and the dip switches have a couple of interestingly named
options:  SW3#8 apparently select for fail safe boot; on an AS200 I
blieve that caused it to boot a firmware floppy to load SRM (assuming
corrupted flash).  SW8#5 can run the SROM mini-debugger, and SW1#2
"SROM flash select spare".  I tried a failsafe boot with an old floppy
download and still got the 6 beeps; I'm going to see if I can get a
new floppy built to try just in case.

But I still suspect the MLB is dead.

On 3/15/2012 12:09 PM, Rich Jordan wrote:
> One of my two hobbyist DS10L systems has bit it; the main logic board
> is dead and it just isn't cost feasible to repair/replace right now.

B-Cache errors?

Not confirmed as applicable to a DS10, but they are similar boxes.

http://labs.hoffmanlabs.com/node/441

Regards,
-John
wb8tyw@qsl.network
Personal Opinion Only

On Mar 15, 5:52=A0pm, "John E. Malmberg" <wb8...@qsl.network> wrote:
> On 3/15/2012 12:09 PM, Rich Jordan wrote:
>
> > One of my two hobbyist DS10L systems has bit it; the main logic board
> > is dead and it just isn't cost feasible to repair/replace right now.
>
> B-Cache errors?
>
> Not confirmed as applicable to a DS10, but they are similar boxes.
>
> http://labs.hoffmanlabs.com/node/441
>
> Regards,
> -John
> wb8...@qsl.network
> Personal Opinion Only

Actually I've been perusing Hoffman Labs because I finally remembered
FSL instead of SROM for the AS200 (I did one once and it worked).  So
in case its corrupted flash I tried an FSL with a valid freshly made
firmware floppy.  UNfortunately no change.  Environment light
(leftmost) flashes 16 times during POST, then 6 beeps, then
environment light on solid yellow, power light solid green, others
dark.  The floppy drive never twitched so I don't think its getting to
the point of checking for firmware corruption.

I don't get any feedback at all from the machine other than the 6
beeps but I don't know how to connect to the SROM port yet either.

Rich Jordan wrote:
> One of my two hobbyist DS10L systems has bit it; the main logic board
> is dead and it just isn't cost feasible to repair/replace right now.
> So I have a full set of working spares (less the MLB, all tested in
> the second DS10L) including two extra 512MB DIMMs, sadly not the rare
> low profile DataRAM models.  And I have some tools (and access to
> others) and I'm not afraid to use them. especially on spare parts.
> 
> I was thinking about cutting a nice rectangle in the lid right over
> the memory slots.  Then making a sheet metal cap that would provide
> enough clearance so that I could use the DIMMs vertically instead of
> with the right-angle adapters.  That would allow me to take my working
> DS10L up to 2GB.  The cap would get riveted to the case lid (I don't
> have access to a welder, nor the skills to non-destructively weld
> sheet metal).
> 
> I don't expect any problems except for possibly the frontmost case
> screw mount, which might partially overlap with the area needing to be
> cut.  Given continuous metal still shielding the system/RAM, am I
> likely to run into any EMI problems caused by the system? I don't
> think cooling will be affected enough to matter (as long as there is
> some turbulence to prevent the metal cap from becoming a dead-air
> hotspot).
> 
> Fun to think about while I wish  (and keep an eye out for) a free/
> cheap DS15 or free/dirt cheap DS10...
> 
> Rich

Don't use rivits.  Too permanent.  I'm not sure you need to worry much about the 
emissions, but if it matters to you, a Faraday cage does NOT have to be solid.  It does 
have to be grounded.

It really depends upon what you're decent at working with.  For some people, it's metal, 
for others, it could be plastic, or fiberglass, or ....

Even some aluminum foil, wire, whatever, as long as it's grounded, works.

If you don't need to ground a cover, I'd stick it on with silicon.  A single sheet metal 
screw (keep it short and test it!) would ground any cover.

Just don't over-engineer it ....

Save me the time to look it up, what is a DS15.  I seem to remember it being a DS10 with a 
faster processor, but it's been some time, and I'm not sure I'm right.

In article <f5b6cea0-e96c-49c9-b3f6-35f615264621@h12g2000yqi.googlegroups.com>, Rich Jordan <jordan@ccs4vms.com> writes:
>{...snip...}
>I wish I could find those mythical DataRam modules that were short
>enough to run in a DS10L without the adapters!  2GB would be very nice
>to have.  But at this point a DS15 is the dream machine (anything else
>is too noisy or too big).  But they sure still attract a lot of money
>in the aftermarket... too much for a hobby box.

That's what this is for?
-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

Well I speak to machines with the voice of humanity.

On 3/15/2012 6:22 PM, Rich Jordan wrote:
> On Mar 15, 5:52 pm, "John E. Malmberg"<wb8...@qsl.network>  wrote:
>> On 3/15/2012 12:09 PM, Rich Jordan wrote:
>>
>>> One of my two hobbyist DS10L systems has bit it; the main logic board
>>> is dead and it just isn't cost feasible to repair/replace right now.
>>
>> B-Cache errors?
>>
>> Not confirmed as applicable to a DS10, but they are similar boxes.
>>
>> http://labs.hoffmanlabs.com/node/441
>
> Actually I've been perusing Hoffman Labs because I finally remembered
> FSL instead of SROM for the AS200 (I did one once and it worked).  So
> in case its corrupted flash I tried an FSL with a valid freshly made
> firmware floppy.  UNfortunately no change.  Environment light
> (leftmost) flashes 16 times during POST, then 6 beeps, then
> environment light on solid yellow, power light solid green, others
> dark.  The floppy drive never twitched so I don't think its getting to
> the point of checking for firmware corruption.

To quote HoffmanLabs.com:

11-Nov-2008 � AlphaServer DS15 series B-Cache errors can be tied back to 
a failed Dallas and to the validity of the NVRAM EEPROM contents of the 
FRU tables within the system; a magic sequence that can rebuild the FRU 
tables (jumper J22 to 15/16 to allow the box to power up, then >>> 
buildfru -S HMB 00 80 08 02 01 E8 03 01 00) is discussed in ITRC 
threadId=1256015.

Unfortunately the links to ITRC all got broken with the move to the new 
forums.

There is also some things to try on a DS10/DS10L, with a note that if 
this is the issue, a sequence from a working DS10/10L is probably needed 
before attempting to fix the broken one.

Nothing involved a boot floppy, but did involve installing a jumper to 
bypass some POST tests to allow you to get to a console prompt.

-John
wb8tyw@qsl.network
Personal Opinion Only

On Mar 15, 6:49=A0pm, David Froble <da...@tsoft-inc.com> wrote:
> Rich Jordan wrote:
> > One of my two hobbyist DS10L systems has bit it; the main logic board
> > is dead and it just isn't cost feasible to repair/replace right now.
> > So I have a full set of working spares (less the MLB, all tested in
> > the second DS10L) including two extra 512MB DIMMs, sadly not the rare
> > low profile DataRAM models. =A0And I have some tools (and access to
> > others) and I'm not afraid to use them. especially on spare parts.
>
> > I was thinking about cutting a nice rectangle in the lid right over
> > the memory slots. =A0Then making a sheet metal cap that would provide
> > enough clearance so that I could use the DIMMs vertically instead of
> > with the right-angle adapters. =A0That would allow me to take my workin=
g
> > DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don't
> > have access to a welder, nor the skills to non-destructively weld
> > sheet metal).
>
> > I don't expect any problems except for possibly the frontmost case
> > screw mount, which might partially overlap with the area needing to be
> > cut. =A0Given continuous metal still shielding the system/RAM, am I
> > likely to run into any EMI problems caused by the system? I don't
> > think cooling will be affected enough to matter (as long as there is
> > some turbulence to prevent the metal cap from becoming a dead-air
> > hotspot).
>
> > Fun to think about while I wish =A0(and keep an eye out for) a free/
> > cheap DS15 or free/dirt cheap DS10...
>
> > Rich
>
> Don't use rivits. =A0Too permanent. =A0I'm not sure you need to worry muc=
h about the
> emissions, but if it matters to you, a Faraday cage does NOT have to be s=
olid. =A0It does
> have to be grounded.
>
> It really depends upon what you're decent at working with. =A0For some pe=
ople, it's metal,
> for others, it could be plastic, or fiberglass, or ....
>
> Even some aluminum foil, wire, whatever, as long as it's grounded, works.
>
> If you don't need to ground a cover, I'd stick it on with silicon. =A0A s=
ingle sheet metal
> screw (keep it short and test it!) would ground any cover.
>
> Just don't over-engineer it ....
>
> Save me the time to look it up, what is a DS15. =A0I seem to remember it =
being a DS10 with a
> faster processor, but it's been some time, and I'm not sure I'm right.

DS15 - Black DS10 style case, 21264C (EV68) Alpha processor at 1GHz,
2MB cache.  4GB max memory (4 slots), onboard VMS supported U160 SCSI.

We've got two out at customer sites and they are awesome machines.

On Mar 16, 10:57=A0am, koeh...@eisner.nospam.encompasserve.org (Bob
Koehler) wrote:
> In article <b94e5514-4f87-4d63-ac02-c26072468...@k24g2000yqe.googlegroups=
..com>, Rich Jordan <jor...@ccs4vms.com> writes:
>
>
>
> > I was thinking about cutting a nice rectangle in the lid right over
> > the memory slots. =A0Then making a sheet metal cap that would provide
> > enough clearance so that I could use the DIMMs vertically instead of
> > with the right-angle adapters. =A0That would allow me to take my workin=
g
> > DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don't
> > have access to a welder, nor the skills to non-destructively weld
> > sheet metal).
>
> =A0 =A0I'm about ready to do the same thing. =A0I have a couple of new DI=
MMs
> =A0 =A0and with the lid open, everything works, once I found the proper
> =A0 =A0order for two different size DIMMs.
>
> =A0 =A0I just have to remember to remove the paint from a piece I have th=
at
> =A0 =A0will form part of the bump. =A0Ironically the system is already si=
tting
> =A0 =A0on top of a VAX 4000 Model 105 with it's Qbus bump, balanced by a
> =A0 =A0scrap piece of wood that happens to be about the same thinkness
> =A0 =A0as the bump.
>
> =A0 =A0And yes, despite thier being 4 working slots, the documentation fo=
r
> =A0 =A0the DS10L cleary calls out for a maximum of 2 DIMMs. =A0Modifying
> =A0 =A0the cabinet specifically designed to fit into a 1U rack slot just
> =A0 =A0isn't a supported configuration. =A08-)

It'd be like the 'turbo bulge' on some wannabe sports car hoods.

Even if I got a second unit working again, it will never be
rackmounted; I can just put taller feet on the top unit to keep clear
of the memory 'turbo bulge' ;)

On Mar 16, 7:10=A0am, "John E. Malmberg" <wb8...@qsl.network> wrote:
> On 3/15/2012 6:22 PM, Rich Jordan wrote:
>
>
>
>
>
>
>
>
>
> > On Mar 15, 5:52 pm, "John E. Malmberg"<wb8...@qsl.network> =A0wrote:
> >> On 3/15/2012 12:09 PM, Rich Jordan wrote:
>
> >>> One of my two hobbyist DS10L systems has bit it; the main logic board
> >>> is dead and it just isn't cost feasible to repair/replace right now.
>
> >> B-Cache errors?
>
> >> Not confirmed as applicable to a DS10, but they are similar boxes.
>
> >>http://labs.hoffmanlabs.com/node/441
>
> > Actually I've been perusing Hoffman Labs because I finally remembered
> > FSL instead of SROM for the AS200 (I did one once and it worked). =A0So
> > in case its corrupted flash I tried an FSL with a valid freshly made
> > firmware floppy. =A0UNfortunately no change. =A0Environment light
> > (leftmost) flashes 16 times during POST, then 6 beeps, then
> > environment light on solid yellow, power light solid green, others
> > dark. =A0The floppy drive never twitched so I don't think its getting t=
o
> > the point of checking for firmware corruption.
>
> To quote HoffmanLabs.com:
>
> 11-Nov-2008 =97 AlphaServer DS15 series B-Cache errors can be tied back t=
o
> a failed Dallas and to the validity of the NVRAM EEPROM contents of the
> FRU tables within the system; a magic sequence that can rebuild the FRU
> tables (jumper J22 to 15/16 to allow the box to power up, then >>>
> buildfru -S HMB 00 80 08 02 01 E8 03 01 00) is discussed in ITRC
> threadId=3D1256015.
>
> Unfortunately the links to ITRC all got broken with the move to the new
> forums.
>
> There is also some things to try on a DS10/DS10L, with a note that if
> this is the issue, a sequence from a working DS10/10L is probably needed
> before attempting to fix the broken one.
>
> Nothing involved a boot floppy, but did involve installing a jumper to
> bypass some POST tests to allow you to get to a console prompt.
>
> -John
> wb8...@qsl.network
> Personal Opinion Only

UNfortunately I have a DS10L-466, not a DS15.  I'll check to see if
there's a comparable sequence available though, now that you've
brought this one out.  Thanks!

Rich

On Mar 16, 7:10=A0am, "John E. Malmberg" <wb8...@qsl.network> wrote:
> On 3/15/2012 6:22 PM, Rich Jordan wrote:
>
>
>
>
>
>
>
>
>
> > On Mar 15, 5:52 pm, "John E. Malmberg"<wb8...@qsl.network> =A0wrote:
> >> On 3/15/2012 12:09 PM, Rich Jordan wrote:
>
> >>> One of my two hobbyist DS10L systems has bit it; the main logic board
> >>> is dead and it just isn't cost feasible to repair/replace right now.
>
> >> B-Cache errors?
>
> >> Not confirmed as applicable to a DS10, but they are similar boxes.
>
> >>http://labs.hoffmanlabs.com/node/441
>
> > Actually I've been perusing Hoffman Labs because I finally remembered
> > FSL instead of SROM for the AS200 (I did one once and it worked). =A0So
> > in case its corrupted flash I tried an FSL with a valid freshly made
> > firmware floppy. =A0UNfortunately no change. =A0Environment light
> > (leftmost) flashes 16 times during POST, then 6 beeps, then
> > environment light on solid yellow, power light solid green, others
> > dark. =A0The floppy drive never twitched so I don't think its getting t=
o
> > the point of checking for firmware corruption.
>
> To quote HoffmanLabs.com:
>
> 11-Nov-2008 =97 AlphaServer DS15 series B-Cache errors can be tied back t=
o
> a failed Dallas and to the validity of the NVRAM EEPROM contents of the
> FRU tables within the system; a magic sequence that can rebuild the FRU
> tables (jumper J22 to 15/16 to allow the box to power up, then >>>
> buildfru -S HMB 00 80 08 02 01 E8 03 01 00) is discussed in ITRC
> threadId=3D1256015.
>
> Unfortunately the links to ITRC all got broken with the move to the new
> forums.
>
> There is also some things to try on a DS10/DS10L, with a note that if
> this is the issue, a sequence from a working DS10/10L is probably needed
> before attempting to fix the broken one.
>
> Nothing involved a boot floppy, but did involve installing a jumper to
> bypass some POST tests to allow you to get to a console prompt.
>
> -John
> wb8...@qsl.network
> Personal Opinion Only

Thanks, Hoff, for already updating the broken ITRC link!

On Friday, March 16, 2012 10:17:50 AM UTC-4, Rich Jordan wrote:
> On Mar 16, 10:57=A0am, koeh...@eisner.nospam.encompasserve.org (Bob
> Koehler) wrote:
> > In article <b94e5514-4f87-4d63-ac02-c26072468...@k24g2000yqe.googlegrou=
ps.com>, Rich Jordan <jor...@ccs4vms.com> writes:
> >
> >
> >
> > > I was thinking about cutting a nice rectangle in the lid right over
> > > the memory slots. =A0Then making a sheet metal cap that would provide
> > > enough clearance so that I could use the DIMMs vertically instead of
> > > with the right-angle adapters. =A0That would allow me to take my work=
ing
> > > DS10L up to 2GB. =A0The cap would get riveted to the case lid (I don'=
t
> > > have access to a welder, nor the skills to non-destructively weld
> > > sheet metal).
> >
> > =A0 =A0I'm about ready to do the same thing. =A0I have a couple of new =
DIMMs
> > =A0 =A0and with the lid open, everything works, once I found the proper
> > =A0 =A0order for two different size DIMMs.
> >
> > =A0 =A0I just have to remember to remove the paint from a piece I have =
that
> > =A0 =A0will form part of the bump. =A0Ironically the system is already =
sitting
> > =A0 =A0on top of a VAX 4000 Model 105 with it's Qbus bump, balanced by =
a
> > =A0 =A0scrap piece of wood that happens to be about the same thinkness
> > =A0 =A0as the bump.
> >
> > =A0 =A0And yes, despite thier being 4 working slots, the documentation =
for
> > =A0 =A0the DS10L cleary calls out for a maximum of 2 DIMMs. =A0Modifyin=
g
> > =A0 =A0the cabinet specifically designed to fit into a 1U rack slot jus=
t
> > =A0 =A0isn't a supported configuration. =A08-)
>=20
> It'd be like the 'turbo bulge' on some wannabe sports car hoods.
>=20
> Even if I got a second unit working again, it will never be
> rackmounted; I can just put taller feet on the top unit to keep clear
> of the memory 'turbo bulge' ;)

Rich,

Spare working MB on its way once you shoot me a shipping address.

Dan

In article <b94e5514-4f87-4d63-ac02-c2607246884d@k24g2000yqe.googlegroups.com>, Rich Jordan <jordan@ccs4vms.com> writes:
> 
> I was thinking about cutting a nice rectangle in the lid right over
> the memory slots.  Then making a sheet metal cap that would provide
> enough clearance so that I could use the DIMMs vertically instead of
> with the right-angle adapters.  That would allow me to take my working
> DS10L up to 2GB.  The cap would get riveted to the case lid (I don't
> have access to a welder, nor the skills to non-destructively weld
> sheet metal).

   I'm about ready to do the same thing.  I have a couple of new DIMMs
   and with the lid open, everything works, once I found the proper
   order for two different size DIMMs.

   I just have to remember to remove the paint from a piece I have that
   will form part of the bump.  Ironically the system is already sitting
   on top of a VAX 4000 Model 105 with it's Qbus bump, balanced by a
   scrap piece of wood that happens to be about the same thinkness
   as the bump.

   And yes, despite thier being 4 working slots, the documentation for
   the DS10L cleary calls out for a maximum of 2 DIMMs.  Modifying
   the cabinet specifically designed to fit into a 1U rack slot just 
   isn't a supported configuration.  8-)

In article <00ABE63F.4355D19D@SendSpamHere.ORG>,   VAXman-  @SendSpamHere.ORG writes:
> 
> I did it with mine.  First, I "nibbled" just enough space to allow the memory
> to poke through.  Then, I built a cap for it out of two pieces of clear plex-
> iglass. One piece serves as a spacer, the other as the cover.  I toyed with
> the idea of modifying the sockets but decided against that.  

   Plexiglass, of course, won't provide ESD protection.  It just might
   increase ESD.

   But one side of my kid's tower is clear plastic and I don't think ESD
   is the biggest issue.

On Mar 16, 9:58=A0am, abrsvc <dansabrservi...@yahoo.com> wrote:
> On Friday, March 16, 2012 10:17:50 AM UTC-4, Rich Jordan wrote:
> > On Mar 16, 10:57=A0am, koeh...@eisner.nospam.encompasserve.org (Bob
> > Koehler) wrote:
> > > In article <b94e5514-4f87-4d63-ac02-c26072468...@k24g2000yqe.googlegr=
oups.com>, Rich Jordan <jor...@ccs4vms.com> writes:
>
> > > > I was thinking about cutting a nice rectangle in the lid right over
> > > > the memory slots. =A0Then making a sheet metal cap that would provi=
de
> > > > enough clearance so that I could use the DIMMs vertically instead o=
f
> > > > with the right-angle adapters. =A0That would allow me to take my wo=
rking
> > > > DS10L up to 2GB. =A0The cap would get riveted to the case lid (I do=
n't
> > > > have access to a welder, nor the skills to non-destructively weld
> > > > sheet metal).
>
> > > =A0 =A0I'm about ready to do the same thing. =A0I have a couple of ne=
w DIMMs
> > > =A0 =A0and with the lid open, everything works, once I found the prop=
er
> > > =A0 =A0order for two different size DIMMs.
>
> > > =A0 =A0I just have to remember to remove the paint from a piece I hav=
e that
> > > =A0 =A0will form part of the bump. =A0Ironically the system is alread=
y sitting
> > > =A0 =A0on top of a VAX 4000 Model 105 with it's Qbus bump, balanced b=
y a
> > > =A0 =A0scrap piece of wood that happens to be about the same thinknes=
s
> > > =A0 =A0as the bump.
>
> > > =A0 =A0And yes, despite thier being 4 working slots, the documentatio=
n for
> > > =A0 =A0the DS10L cleary calls out for a maximum of 2 DIMMs. =A0Modify=
ing
> > > =A0 =A0the cabinet specifically designed to fit into a 1U rack slot j=
ust
> > > =A0 =A0isn't a supported configuration. =A08-)
>
> > It'd be like the 'turbo bulge' on some wannabe sports car hoods.
>
> > Even if I got a second unit working again, it will never be
> > rackmounted; I can just put taller feet on the top unit to keep clear
> > of the memory 'turbo bulge' ;)
>
> Rich,
>
> Spare working MB on its way once you shoot me a shipping address.
>
> Dan

Done, thanks!

On Thu, 15 Mar 2012 22:31:12 GMT
VAXman-  @SendSpamHere.ORG wrote:

> FWIW, and I 'm sure I mentioned this here before, the DS10s will NOT
> run with the 1GB DIMMs that work in their bigger bretheren DS20/ES4x.

Have you had success with driving a DS20/DS20E w/ 1GB modules?
Officially supported are only the 256MB modules and I wouldn't want to
spend a lot of money in a non-working combination.
ES40/ES45 work with 1GB, as do the DS15 and the DS25.