50Hz vs. 60Hz?
Mike Ross
I'd appreciate some thoughts in general, and on some specific cases, on using
50Hz DEC equipment on 60Hz power, since I've started processsing and restoring a
large quantity of DEC equipment I've brought over from my UK storage to the USA:
1. Thinking about old DEC logic power supplies in general (pdp-8, -11, -12, -15
etc), is there liable to be any problem with using equipment sold in 50Hz
countries on US 60Hz supplies? Obviously output voltages would be checked as a
matter of course anyway!
Equipment which uses motors is, I would have thought, liable to be more of a
problem, for instance:
2. Tapes. Would 9-track (e.g. TU45, TU16 etc.) tapes be likely to be a problem?
It occurs to me the transport speed might be a little off...
3. DECtapes (& LINCtapes) : I would imagine they would be less of a problem;
they're self-clocking and practically indestructible - you can almost read a
DECtape with a toothbrush, some tinfoil, and a set of bicycle pedals! Thoughts?
4. Disks. Biggest problem I would imagine. RK05s I know have different motor
pulleys - they're stamped 50Hz or 60Hz. I would imagine running a 50Hz unit on
60Hz wouldn't work, the heads would never load because it would never be
detected going at the correct speed. And even if they did load there might be
air bearing issues due to incorrect speed. I have around 7 or 8 50Hz RK05s, but
only 3 60Hz, so... does anyone have any spare 60Hz RK05 pulleys, or motors with
60Hz pulleys attached? What about RL and RX drives? Similar problems?
Many thanks for any advice
Mike
--
http://www.corestore.org
'As I walk along these shores
I am the history within'
glen herrmannsfeldt
< I'd appreciate some thoughts in general, and on some specific
< cases, on using 50Hz DEC equipment on 60Hz power, since I've
< started processsing and restoring a large quantity of DEC
< equipment I've brought over from my UK storage to the USA:
< 1. Thinking about old DEC logic power supplies in general
< (pdp-8, -11, -12, -15 etc), is there liable to be any problem
< with using equipment sold in 50Hz countries on US 60Hz supplies?
Usually not. A 50Hz transformer requires a larger core for the
same rating, so you might have problems going the other way.
You don't mention voltage at all. If I remember, UK is 220, and
in the US you find closer to 240 for phase to phase in home wiring,
and 208 in industry (three phase) wiring. If the transformers
have enough taps to handle those voltages, you should do fine.
< Obviously output voltages would be checked as a matter of course anyway!
< Equipment which uses motors is, I would have thought, liable to
< be more of a problem, for instance:
< 2. Tapes. Would 9-track (e.g. TU45, TU16 etc.) tapes be likely to
< be a problem?
If they use DC motors, then it should not matter at all (assuming
the previous question is answered). If they use a constant speed
capstan, with a solenoid clutch and pinch roller, then the speed
has to be right. Most likely, as with disks below, there is a
way to change that. (Different capstan diameter, for example.)
< It occurs to me the transport speed might be a little off...
They might survive a 20% increase, if the processor can accept the
data fast enough. 1600 and 6250 are self clocking. 800 (NRZI)
might depend on a reference clock and would fail. Writing would
also be wrong, but would be readable on the drive that wrote it...
< 3. DECtapes (& LINCtapes) : I would imagine they would be less
< of a problem; they're self-clocking and practically indestructible
< - you can almost read a DECtape with a toothbrush, some tinfoil,
< and a set of bicycle pedals! Thoughts?
Do they use speed controlled DC motors, or clutch controlled
synchronous AC motors?
< 4. Disks. Biggest problem I would imagine. RK05s I know have
< different motor pulleys - they're stamped 50Hz or 60Hz.
< I would imagine running a 50Hz unit on 60Hz wouldn't work,
< the heads would never load because it would never be detected
< going at the correct speed. And even if they did load there
< might be air bearing issues due to incorrect speed. I have
< around 7 or 8 50Hz RK05s, but only 3 60Hz, so... does anyone
< have any spare 60Hz RK05 pulleys, or motors with 60Hz pulleys
< attached? What about RL and RX drives? Similar problems?
If you have one of the right pully, you might find a machine
shop that could make some for you just like it.
-- glen
dpi
> Folks,
>
> I'd appreciate some thoughts in general, and on some specific cases, on using
> 50Hz DEC equipment on 60Hz power, since I've started processsing and restoring a
> large quantity of DEC equipment I've brought over from my UK storage to the USA:
>
> 1. Thinking about old DEC logic power supplies in general (pdp-8, -11, -12, -15
> etc), is there liable to be any problem with using equipment sold in 50Hz
> countries on US 60Hz supplies? Obviously output voltages would be checked as a
> matter of course anyway!
As Glen said using 60hz on 50hz transformers should not be a problem.
The efficiency could even be a couple of percent better than when
running 50hz. Now if you were trying to use 400hz transformers like
Control Data used in all their older machines you would have big
problems. Those tiny transformers would look like a dead short to
50/60hz.
> Equipment which uses motors is, I would have thought, liable to be more of a
> problem, for instance:
>
> 2. Tapes. Would 9-track (e.g. TU45, TU16 etc.) tapes be likely to be a problem?
> It occurs to me the transport speed might be a little off...
I dont remember any of these being being AC motors.
> 3. DECtapes (& LINCtapes) : I would imagine they would be less of a problem;
> they're self-clocking and practically indestructible - you can almost read a
> DECtape with a toothbrush, some tinfoil, and a set of bicycle pedals! Thoughts?
TU55 and TU56 drives would not be a problem.
> 4. Disks. Biggest problem I would imagine. RK05s I know have different motor
> pulleys - they're stamped 50Hz or 60Hz. I would imagine running a 50Hz unit on
> 60Hz wouldn't work, the heads would never load because it would never be
> detected going at the correct speed. And even if they did load there might be
> air bearing issues due to incorrect speed. I have around 7 or 8 50Hz RK05s, but
> only 3 60Hz, so... does anyone have any spare 60Hz RK05 pulleys, or motors with
> 60Hz pulleys attached? What about RL and RX drives? Similar problems?
These used synchronous motors that would run at 360 rpm on 60hz and
300rpm on 50hz so a different pulley would be needed. Since you have
at least one of the correct pulley it would not be impossible for
someone to machine you a copy. Know anyone with a machine shop? Did
they use an idle wheel to keep tension on the belt or do they use
different length of belts? Or do you move the motor to tension the
belt?
Good luck!
Doug Ingraham
glen herrmannsfeldt
< As Glen said using 60hz on 50hz transformers should not be a problem.
< The efficiency could even be a couple of percent better than when
< running 50hz.
Just to be sure, the reason Europe started with 50Hz instead
of the 60Hz in US is that the transformer steel available in
Europe at the time had higher hysteresis loss. As silicon
steel improved, it was too difficult to change over.
I don't believe that is a problem for small transformers,though.
< Now if you were trying to use 400hz transformers like
< Control Data used in all their older machines you would have big
< problems.
< Those tiny transformers would look like a dead short to 50/60hz.
I don't know CDC so well, but the big IBM mainframes all ran
off motor/generators. That would seem to be the best way
to get 400Hz, also. Traditionally, 400Hz was used in airplane
power systems where weight was more important than core cost.
-- glen
Jonathan Knight
> You don't mention voltage at all. If I remember, UK is 220, and
> in the US you find closer to 240 for phase to phase in home wiring,
> and 208 in industry (three phase) wiring. If the transformers
> have enough taps to handle those voltages, you should do fine.
The UK was 240V during the time of those machines. It was parts of
Europe that was 220V.
The UK is moving to 230V gradually although there is a tolerance
requirement on electrical devices (10% I think) so that 220V or 240V
equipment can be sold and should work.
3-phase power used to be 415V, I think we moving to 400V
There's a Wikipedia article here:
http://en.wikipedia.org/wiki/Electrical_wiring_in_the_United_Kingdom
which has the details at the end for single phase voltage tolerances.
Jon.
jmfbahciv
and 60 for something (but I can't remember what). As long as all your
stuff runs with the same value, I don't think you'll have problems
with the software. The Hz value was GETTABable for user mode
programs but I don't recall seeing any that actually used it.
Perhaps some customers reading this might be able to say if they
ever had a need to know at user mode level.
/BAH
Techdas
>
> - Show quoted text -
The line clock used to interrupt at mains frequency, eg on 11/35s, so
it was useful to know what the frequency was. Maybe a sysgen
parameter or patchable location on some OSs. See "Line Time Clock
Interrupt Test" in http://www.psych.usyd.edu.au/pdp-11/hints.html,
"Hints on testing a dead PDP-11".
Rob Brown
On Thu, 24 Sep 2009 at 07:56 -0400, jmfbahciv wrote:
> The software used the values of 50 and 60 for something (but I can't
> remember what).
Good point.
All the old PDP-11s used line frequency for the clock interrupts.
During sysgen you specified 50 or 60 Hz so that your time-of-day ran
at the correct speed.
--
Rob Brown b r o w n a t g m c l d o t c o m
G. Michaels Consulting Ltd. (780)438-9343 (voice)
Edmonton (780)437-3367 (FAX)
http://gmcl.com/
dpi
> In alt.sys.pdp8 dpi <doug.ingra...@gmail.com> wrote:
> < Now if you were trying to use 400hz transformers like
> < Control Data used in all their older machines you would have big
> < problems.
>
> < Those tiny transformers would look like a dead short to 50/60hz.
>
> I don't know CDC so well, but the big IBM mainframes all ran
> off motor/generators. That would seem to be the best way
> to get 400Hz, also. Traditionally, 400Hz was used in airplane
> power systems where weight was more important than core cost.
Many of the large Mainframes used 400hz. The main advantages to this
were smaller transformers and smaller filter caps. This saved the
vendors a lot of money.
No way would I use a motor generator set today if I was going to power
up one of those big machines. I would build a converter and do it
electronically. A motor generator set would have poor efficiency
compared to an electronic converter. Additionally they were big and
noisy. I would guess efficiency would be no better than 70% with the
MG set. It should be possible to get to the 90% or maybe a little
more efficiency range with an electronic converter. With the amount
of power we are talking about this would be a lot of money savings in
electricity.
In the mid 1980's I mostly restored a CDC 3400 to life and did put a
CDC Cyber 174 back into service. These were run off of a Motor
generator that took 208 three phase and output 400hz. With no load on
the generator the current draw would exceed 100 amps on each phase
during startup. The service at this particular facility was 100 amp
with what they called demand billing for going over. The startup
would trip the demand and pretty much cost $50 just to turn on the
MG. It wasnt too awful to run the rest of the time.
Sorry about the off topic nature, I was reminded of this and got a
little nostalgic.
Doug Ingraham
Scott Dorsey
>
>No way would I use a motor generator set today if I was going to power
>up one of those big machines. I would build a converter and do it
>electronically. A motor generator set would have poor efficiency
>compared to an electronic converter. Additionally they were big and
>noisy. I would guess efficiency would be no better than 70% with the
>MG set. It should be possible to get to the 90% or maybe a little
>more efficiency range with an electronic converter. With the amount
>of power we are talking about this would be a lot of money savings in
>electricity.
The big advantage of the MG set is that you get immunity from short line
dropouts. And if you DO get a dropout longer than a second or two, you
have enough time for the thing to spin down that you can tell the computer
to flush caches and maybe even save machine state before it dies.
>In the mid 1980's I mostly restored a CDC 3400 to life and did put a
>CDC Cyber 174 back into service. These were run off of a Motor
>generator that took 208 three phase and output 400hz. With no load on
>the generator the current draw would exceed 100 amps on each phase
>during startup. The service at this particular facility was 100 amp
>with what they called demand billing for going over. The startup
>would trip the demand and pretty much cost $50 just to turn on the
>MG. It wasnt too awful to run the rest of the time.
I work in a place with some PC clusters that are on an old M-G set that
used to feed some big iron. It's not as solid as a modern UPS but it's not
bad.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
jmfbahciv
>
> On Thu, 24 Sep 2009 at 07:56 -0400, jmfbahciv wrote:
>
>> The software used the values of 50 and 60 for something (but I can't
>> remember what).
>
> Good point.
>
> All the old PDP-11s used line frequency for the clock interrupts. During
> sysgen you specified 50 or 60 Hz so that your time-of-day ran at the
> correct speed.
>
>
calculate CPU runtime used by the user. I can't think of any
other reasons. I betcha that somewhere somebody used the value
to make a decision that determined how some software would run.
One of the laws of computing is: If it's there, somebody will
use it. Since this value was a "constant" that couldn't be
changed, it might have been very convenient. :-)
/BAH
Rob Brown
On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
> Rob Brown wrote:
>>
>> On Thu, 24 Sep 2009 at 07:56 -0400, jmfbahciv wrote:
>>
>>> The software used the values of 50 and 60 for something (but I can't
>>> remember what).
>>
>> Good point.
>>
>> All the old PDP-11s used line frequency for the clock interrupts. During
>> sysgen you specified 50 or 60 Hz so that your time-of-day ran at the
>> correct speed.
>>
>>
> I think, but am not sure, the usage accounting used the value to
> calculate CPU runtime used by the user.
Anything that needs to convert ticks to seconds will be affected by
the number of ticks per second.
> Since this value was a "constant" that couldn't be changed, it might
> have been very convenient. :-)
That is what ZAP is for. :-)
jmfbahciv
>
> On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
>
>> Rob Brown wrote:
>>>
>>> On Thu, 24 Sep 2009 at 07:56 -0400, jmfbahciv wrote:
>>>
>>>> The software used the values of 50 and 60 for something (but I can't
>>>> remember what).
>>>
>>> Good point.
>>>
>>> All the old PDP-11s used line frequency for the clock interrupts.
>>> During sysgen you specified 50 or 60 Hz so that your time-of-day ran
>>> at the correct speed.
>>>
>>>
>> I think, but am not sure, the usage accounting used the value to
>> calculate CPU runtime used by the user.
>
> Anything that needs to convert ticks to seconds will be affected by the
> number of ticks per second.
Something is niggling my brain w.r.t. that. Maybe I can remember
why it's important when I'm sleeping.
>
>> Since this value was a "constant" that couldn't be changed, it might
>> have been very convenient. :-)
>
> That is what ZAP is for. :-)
>
>
I don't remember a ZAP. Did I just feel a <whoosh> go over my
head?
/BAH
Mike Ross
>Rob Brown wrote:
>>
>> On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
<snip>
>>> Since this value was a "constant" that couldn't be changed, it might
>>> have been very convenient. :-)
>>
>> That is what ZAP is for. :-)
>
>I don't remember a ZAP. Did I just feel a <whoosh> go over my
>head?
Isn't ZAP an IBMism? I certainly remember using SUPERZAP on IBM.
John Wallace
> On Sat, 26 Sep 2009 08:31:47 -0400, jmfbahciv <jmfbahciv@aol> wrote:
> >Rob Brown wrote:
>
> >> On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
>
> <snip>
>
> >>> Since this value was a "constant" that couldn't be changed, it might
> >>> have been very convenient. :-)
>
> >> That is what ZAP is for. :-)
>
> >I don't remember a ZAP. Did I just feel a <whoosh> go over my
> >head?
>
> Isn't ZAP an IBMism? I certainly remember using SUPERZAP on IBM.
>
> Mike
> 'As I walk along these shores
> I am the history within'
Isn't ZAP an RSXism?
Didn't RSX (or maybe IAS) SYSGEN use the HZ value to decide whether to
recommend a teabreak or a coffeebreak?
Just in case there's Qbus kit in this picture: lots of Qbus kit had
the 'LTC' generated locally by a crystal clock and was switchable
(jumperable?) between 50Hz and 60Hz, regardless of the actual line
frequency.
jmfbahciv
> On Sat, 26 Sep 2009 08:31:47 -0400, jmfbahciv <jmfbahciv@aol> wrote:
>
>> Rob Brown wrote:
>>> On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
>
> <snip>
>
>>>> Since this value was a "constant" that couldn't be changed, it might
>>>> have been very convenient. :-)
>>> That is what ZAP is for. :-)
>> I don't remember a ZAP. Did I just feel a <whoosh> go over my
>> head?
>
> Isn't ZAP an IBMism? I certainly remember using SUPERZAP on IBM.
>
/BAH
jmfbahciv
> On Sep 26, 4:19 pm, Mike Ross <m...@corestore.org> wrote:
>> On Sat, 26 Sep 2009 08:31:47 -0400, jmfbahciv <jmfbahciv@aol> wrote:
>>> Rob Brown wrote:
>>>> On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
>> <snip>
>>
>>>>> Since this value was a "constant" that couldn't be changed, it might
>>>>> have been very convenient. :-)
>>>> That is what ZAP is for. :-)
>>> I don't remember a ZAP. Did I just feel a <whoosh> go over my
>>> head?
>> Isn't ZAP an IBMism? I certainly remember using SUPERZAP on IBM.
>>
>> Mike
>> --http://www.corestore.org
>> 'As I walk along these shores
>> I am the history within'
>
> Isn't ZAP an RSXism?
>
> Didn't RSX (or maybe IAS) SYSGEN use the HZ value to decide whether to
> recommend a teabreak or a coffeebreak?
Even if it did, taking a tea break shouldn't affect the performance
of the system once you got it up and running.
>
> Just in case there's Qbus kit in this picture: lots of Qbus kit had
> the 'LTC' generated locally by a crystal clock and was switchable
> (jumperable?) between 50Hz and 60Hz, regardless of the actual line
> frequency.
/BAH
dpi
> 4. Disks. Biggest problem I would imagine. RK05s I know have different motor
> pulleys - they're stamped 50Hz or 60Hz. I would imagine running a 50Hz unit on
> 60Hz wouldn't work, the heads would never load because it would never be
> detected going at the correct speed. And even if they did load there might be
> air bearing issues due to incorrect speed. I have around 7 or 8 50Hz RK05s, but
> only 3 60Hz, so... does anyone have any spare 60Hz RK05 pulleys, or motors with
> 60Hz pulleys attached? What about RL and RX drives? Similar problems?
I ran into my DF32 manual over the weekend and noticed that the
transfer rate was a word every 66us at 60hz and 80us at 50hz. The
DF-32 uses direct drive to the platter. On 60hz the platters spin at
360rpm and on 50hz at 300rpm. So on 50hz power they simply run
slower. A little surprising that everything works with that much
variation.
Doug Ingraham
Rob Brown
On Sun, 27 Sep 2009 at 08:59 -0400, jmfbahciv wrote:
> John Wallace wrote:
>> On Sep 26, 4:19 pm, Mike Ross <m...@corestore.org> wrote:
>>> On Sat, 26 Sep 2009 08:31:47 -0400, jmfbahciv <jmfbahciv@aol> wrote:
>>>> Rob Brown wrote:
>>>>> On Fri, 25 Sep 2009 at 06:52 -0400, jmfbahciv wrote:
>>> <snip>
>>>
>>>>>> Since this value was a "constant" that couldn't be changed, it might
>>>>>> have been very convenient. :-)
>>>>> That is what ZAP is for. :-)
>>>> I don't remember a ZAP. Did I just feel a <whoosh> go over my
>>>> head?
>>> Isn't ZAP an IBMism? I certainly remember using SUPERZAP on IBM.
Yes, I remember SUPERZAP on IBM OS/360 or OS/370, but since I was a
student at the time I didn't get to use it. Remembering the other IBM
utility names, they probably really called it IEBSUPERZAP, or maybe
IEBSPRZP, or something.
>> Isn't ZAP an RSXism?
Yes, the ZAP in RSX was what I was referring to. It is an editor for
binary files. The user interface looks like ODT.
Obviously the ZAP name did not make it to every DEC PDP-11 OS of the
time.
>> Didn't RSX (or maybe IAS) SYSGEN use the HZ value to decide whether
>> to recommend a teabreak or a coffeebreak?
RSX certainly did. I never worked with IAS.
Just another bit of humour hidden in RSX. My most recent discovery
was the undocumented /ROMAN switch to the TIM command. This switch
caused the time and date to be displayed in Roman numerals, with Latin
month names. It also displayed the day of week and the phase of the
moon.
>TIME /ROMAN
September XXVIII, Anno Domini MMIX XI:VIII
The moon is at first quarter. Today is Monday.
> Even if it did, taking a tea break shouldn't affect the performance
> of the system once you got it up and running.
True.
>> Just in case there's Qbus kit in this picture: lots of Qbus kit had
>> the 'LTC' generated locally by a crystal clock and was switchable
>> (jumperable?) between 50Hz and 60Hz, regardless of the actual line
>> frequency.
I think this capability arrived later. I first saw it on an 11/84.
It was not on our 11/73 or on any of our earlier unibus equipment.
Way back in '79 we installed an 11/34 at a site which protected it
with a UPS. The system clock gained 8 minutes a day. Eventually we
figured out that the UPS was supplying 60.3 Hz (instead of 60.0 Hz).
Correcting the line frequency corrected the system time problem.
- Rob
Mike Ross
Ahhh now that is juicy and excellent news; I have a couple of DF32s in Scotland,
I'll drag them over to the USA!
Jonathan Knight
> Way back in '79 we installed an 11/34 at a site which protected it with
> a UPS. The system clock gained 8 minutes a day. Eventually we figured
> out that the UPS was supplying 60.3 Hz (instead of 60.0 Hz). Correcting
> the line frequency corrected the system time problem.
Way back in the mid-80's I went to a local hydro electric power station
on a tour. On the wall of the control room was an analogue clock that
was showing 11:56. Our guide explained that it recorded the offset that
clocks using mains frequency would currently have in seconds and minutes
(so it was showing 4 seconds slow).
What they were doing was running the mains at 50.1Hz (there was digital
display beside it showing the frequency) to catch up the 4 seconds and
get all the clocks back to the right time. Once they reached that point
they dropped the mains back to 50.0Hz.
This was all pre-pre-privatisation and I've no idea whether our current
private companies still do this. In these days of GPS, NTP and cheap
radio clocks there's probably little need.
Jon.
Mike Ross
wrote:
<snip>
>Way back in the mid-80's I went to a local hydro electric power station
>on a tour. On the wall of the control room was an analogue clock that
>was showing 11:56. Our guide explained that it recorded the offset that
>clocks using mains frequency would currently have in seconds and minutes
>(so it was showing 4 seconds slow).
Interesting - which power station was that? My dad worked for NOSHEB for 40
years, and I have a few of their power stations here:
http://www.corestore.org/hydro.htm
Mike
--
John Wallace
Yes that still goes on (at least in the UK). The mains frequency is
still required to average 50Hz over a 24 hour cycle, and deviations
below and above 50Hz indicate whether there is a slight excess of
demand over supply or vice versa.
For more info have a look at the National Grid's real time info on
demand and frequency, e.g. at
http://www.nationalgrid.com/uk/Electricity/Data/Realtime/Frequency/Freq60.htm
(60 as in minutes not as in Hz).
See also
http://www.nationalgrid.com/uk/Electricity/Data/Realtime/Demand/Demand60.htm
As I started to write this, just after 6pm BST, during the last hour
the frequency was mostly around 50.1 Hz, though just before 6pm it
did drop for a few minutes, going as low as 49.8 Hz for a minute or
two before recovering somewhat, and it's back under 50 again right now
(6:20pm). The frequency graph is a lot rougher than the demand graph.
This frequency variation can in theory be used for hints for demand
management - if you temporarily disconnect certain "interruptible"
loads when the frequency falls, you can smooth out some of the peaks
in demand. Fridges and aircon and water heating (domestic and
commercial) are often suggested as disconnectable loads. And vice
versa for when demand is low, turn the fridge motor (or hot water
heater) on even if the thermostat says there's no need. This can be
used to better match intermittent demand with not-very-controllable
generation (wind, nuclear, etc). Have a look at
http://www.dynamicdemand.co.uk/grid.htm
Power station control and regional electricity distribution network
control used to be safe solid PDP and VAX (and then Alpha) territory.
David Powell
<b45d9d58-4093-4369...@w36g2000yqm.googlegroups.com>,
John Wallace <johnwa...@yahoo.co.uk> in alt.sys.pdp10 wrote:
<Snip grid frequency stuff>
>Power station control and regional electricity distribution network
>control used to be safe solid PDP and VAX (and then Alpha) territory.
I used to attend a couple of the main EE / process control exhibitions
every year. Everything changed in a year, one year, most of the
stands had DEC hardware on display, the next, if they had DEC it was
hidden away at the back of the stand.
I certainly enjoyed myself with PDPs in that field, 1970-85ish, the
downside of dealing with CEGB was the never-ending series of technical
meetings with their engineers. Usually a few dozen on their side of
the table, two or three on our side. It mimicked the grid: supply of
their manpower exceded demand, so the frequency of meetings increased.
Pity about DEC, at one time I thought that I'd still be using their
kit in 2000, but the last jobs I did (turbine governors) were on
Intel hardware. I got out in 1989. At least I can get my
nostalgia-fix with the PDPs I have at home.
Regards,
David P.
David Powell
Earlier on the Qbus. The MXV11 multifunction board had 16kwords ram,
rom bootstrap, 2 serial line interfaces and a crystal clock back in
LSI-11 days. Choice of either LTC or crystal clock, just wire what
you wanted to BEVNT L.
>Way back in '79 we installed an 11/34 at a site which protected it
>with a UPS. The system clock gained 8 minutes a day. Eventually we
>figured out that the UPS was supplying 60.3 Hz (instead of 60.0 Hz).
>Correcting the line frequency corrected the system time problem.
>
We ran omnibus PDP-8s on inverter supply, but always phase locked the
inverter to the mains. It would survive an automatic changeover to
mains on inverter fail, but not the other way.
Regards,
David P.
John Wallace
> In article
> <b45d9d58-4093-4369-8543-6f8acc88e...@w36g2000yqm.googlegroups.com>,
Was there a specific event that made the PDPs and VAXes vanish? OS/2
and Windows NT and even SCO UNIX started allowing PC hardware to be
nearly credible competitors for SCADA systems, and on paper and in
Powerpoint they looked much more attractive than real computers, even
if in reality the software wasn't all it was made out to be?
Or was it more to do with privatisation and the change from engineers
being in charge of engineering to beancounters in charge of everything
in the industry (hence the increased number of times the Grid issues a
"Notice Of Insufficient Margin" these days) ?
Or (option 3) was it because some of the DEC-based SCADA players went
bust (Ferranti Syseca), changed hands (Ascada=>GE?
Westinghouse=>Schneider?), or got out of the business (Instem?)
Option 4: acquisition of DEC by HP, together with HP flogging off RTAP
(foundation for APMS) to a third party?
Option 5: none of the above?
Just interested on how it looked from the customer point of view (I
was watching from the inside).