16 Jan
2010
16 Jan
'10
5:48 p.m.
My Altair 680 has achieved the position of primary attention on
my workbench. All of the ICs are socketed. Not having seen
power in over 20 years, I did not trust the power supply as far
as I could throw a bus. I pulled all of the ICs, attached dummy
loads (automobile bulbs) and powered up. Sure enough, there was
AC all over the place (where there should have been DC). I
replaced all of the electrolytic caps and that got rid of the AC
but some of the voltage levels are very wrong, at least as compared
to what is written on the schematics. The worst offender is just
off the "plus" side of the full wave bridge (BR-1). It should be
nine volts but is actually at 11.5. This makes what should be the
main five volt rail a little more than 9 volts... the far side of
VR-1 (a 7805). The wave form coming off the transformer is really
ugly not a smooth sine wave. I hate power supplies. I don't really
understand, looking at the schematics, what this transformer should
be doing. My guess is that it should be making a nice nine volt,
60 cycle AC sine wave with each of the two outputs 90 degrees out
of sync? What could be wrong with a transformer that it would
produce something like:
__
/ \
| \
/ \_
/ \
| /
| /
\_/
bent and very squared off at spots? My scope shows nice clean 120
coming into the transformer. How do I figure out a part number for
a replacement transformer?
Thanks!
16 Jan
16 Jan
5:57 p.m.
On 16 Jan 2010 at 20:48, Bill Sudbrink wrote:
Bill, try putting a load on the output of the power supply and then
looking again. A 12V automotive tail or brake light bulb might do
the trick. These supplies were designed to run with a load.
--Chuck
9:31 p.m.
And you have some other problem if the output side of a 7805 is showing anything other
than five volts. You should be able to feed anything into it up to its maximum rating of
(if memory serves) 40v and get 5v out. If there isn't anything in the ground path of
the 7805, replace it; if there is (for instance, a pot or diode, or any sort of active
circuitry), check it carefully. -- Ian
________________________________________
From: cctalk-bounces at classiccmp.org [cctalk-bounces at classiccmp.org] On Behalf Of
Chuck Guzis [cclist at sydex.com]
Sent: Saturday, January 16, 2010 5:57 PM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Altair 680 power supply...
On 16 Jan 2010 at 20:48, Bill Sudbrink wrote:
Bill, try putting a load on the output of the power supply and then
looking again. A 12V automotive tail or brake light bulb might do
the trick. These supplies were designed to run with a load.
--Chuck
17 Jan
17 Jan
6:45 a.m.
Ian King wrote:
And you have some other problem if the output side of a 7805
is showing anything other than five volts.
Yea, that's what I figured was next. I just wasn't going to
go there until I understood why the input side of the 7805
was not at the schematic voltage.
If there isn't anything in the ground path of the
7805,
There isn't. You can grab a good PDF of the schematics here:
http://www.altair680kit.com/manuals.html
Thanks,
Bill
8:46 a.m.
Hi
I just came on an unusual problem. I'm fixing a
power supply with +12 and +5 outs. It uses 2
723 regulators. The one on the +12 was blown.
I bought another 723 regulator and it didn't
seem to be working.
I checked the voltages and noticed that the
5 volt one that is working fine, has a Vref of
about 3 volts while the new one has a Vref
of 7 volts.
Looking at the specs, it is suppose to be 7 Volts.
Looking at the resistors on the board, the
board expects 3 volts.
Has anyone seen these with different reference
voltages before?
I'll have to change the feedback resistors ( not
a big deal ) but this is a strange one.
Dwight
_________________________________________________________________
Hotmail: Trusted email with Microsoft?s powerful SPAM protection.
http://clk.atdmt.com/GBL/go/196390706/direct/01/
9:01 a.m.
New subject: power supply problem
Sorry, forgot to change subject!
From: dkelvey at hotmail.com
To: cctalk at classiccmp.org
Subject: RE: Altair 680 power supply...
Date: Sun, 17 Jan 2010 08:46:51 -0800
Hi
I just came on an unusual problem. I'm fixing a
power supply with +12 and +5 outs. It uses 2
723 regulators. The one on the +12 was blown.
I bought another 723 regulator and it didn't
seem to be working.
I checked the voltages and noticed that the
5 volt one that is working fine, has a Vref of
about 3 volts while the new one has a Vref
of 7 volts.
Looking at the specs, it is suppose to be 7 Volts.
Looking at the resistors on the board, the
board expects 3 volts.
Has anyone seen these with different reference
voltages before?
I'll have to change the feedback resistors ( not
a big deal ) but this is a strange one.
Dwight
_________________________________________________________________
Hotmail: Trusted email with Microsoft?s powerful SPAM protection.
http://clk.atdmt.com/GBL/go/196390706/direct/01/
_________________________________________________________________
Hotmail: Powerful Free email with security by Microsoft.
http://clk.atdmt.com/GBL/go/196390710/direct/01/
10:15 a.m.
New subject: power supply problem
From: cclist at sydex.com
To: cctalk at classiccmp.org
Date: Sun, 17 Jan 2010 09:54:55 -0800
Subject: Re: power supply problem
On 17 Jan 2010 at 9:01, dwight elvey wrote:
Hi Chuck
It has a trangle with a lamda in it.
The part number is: LAS 723
Date code: 7813
There is no question that the board
was designed for the lower referance
voltage.
Dwight
_________________________________________________________________
Your E-mail and More On-the-Go. Get Windows Live Hotmail Free.
http://clk.atdmt.com/GBL/go/196390709/direct/01/
I'll have to change the feedback resistors ( not
a big deal ) but this is a strange one.
That's a new one on me, Dwight. What are the markings on the old
one?
--Chuck
10:33 a.m.
New subject: power supply problem
On 17 Jan 2010 at 10:15, dwight elvey wrote:
It has a trangle with a lamda in it.
The part number is: LAS 723
Date code: 7813
A quick search turns up that the LAS723 does indeed have a Vref of
2.5v. datasheetarchive.com has the data sheet.
Never heard of this one before.
--Chuck
18 Jan
18 Jan
3:59 p.m.
New subject: power supply problem
From: cclist at sydex.com
To: cctalk at classiccmp.org
Date: Sun, 17 Jan 2010 10:33:35 -0800
Subject: RE: power supply problem
On 17 Jan 2010 at 10:15, dwight elvey wrote:
Hi
I suspect that Lambda wanted to take advantage of two things.
2.5 volts is a typical output of a band-gap reference and the
lower voltage meant that two less resistors were needed for
voltages below 7 volts.
I believe the original 723 design used zeners as references
and these are a little more process dependent than the band-gap
that depends more on physical geometry.
It would have been nice if they used another name, something
like 723-2.5.
Dwight
_________________________________________________________________
Hotmail: Free, trusted and rich email service.
http://clk.atdmt.com/GBL/go/196390708/direct/01/
It has a trangle with a lamda in it.
The part number is: LAS 723
Date code: 7813
A quick search turns up that the LAS723 does indeed have a Vref of
2.5v. datasheetarchive.com has the data sheet.
Never heard of this one before.
--Chuck
4:34 p.m.
New subject: power supply problem
On 18 Jan 2010 at 15:59, dwight elvey wrote:
I believe that the "stock" 723 uses a 6.2V zener as a reference.
Another possibility would be that they wanted a regulator that would
operate with lower supply headroom, particularly for low voltages. .
In any case, you're right about the part labeling. It is confusing.
--Chuck
I suspect
that Lambda wanted to take advantage of two things.
2.5 volts is a typical output
of a band-gap reference and the
lower voltage meant that two less resistors were needed for
voltages below 7 volts. 
5:30 p.m.
New subject: power supply problem
dwight elvey wrote:
I suspect that Lambda wanted to take advantage of two
things.
2.5 volts is a typical output of a band-gap reference and the
lower voltage meant that two less resistors were needed for
voltages below 7 volts.
I believe the original 723 design used zeners as references
and these are a little more process dependent than the band-gap
that depends more on physical geometry.
It would have been nice if they used another name, something
like 723-2.5.
Just as an addendum to the discussion, the 1981 IC Master lists Lambda
Semiconductor as producing a bunch of voltage regulator ICs. The LAS723 and
LAS723B are listed separately from the group list of manufacturers producing
the 723 (the difference between them is not specified, however).
Rather poor that they didn't modify the part number, as you say.
11:43 a.m.
New subject: power supply problem
On Jan 17, 2010, at 2:07 PM, Tony Duell wrote:
I've used lots and lots of Lambda power supplies, mostly bench
units. They are absolutely top-notch for "spare no expense"
projects. (thank heaven I can get them on the surplus market) I've
never seen one die, so I can't speak to the "odd IC" problem, but as
a frequent user over the years, I have nothing but good things to say
about them.
-Dave
--
Dave McGuire
Port Charlotte, FL
It has a
trangle with a lamda in it.
I've heard of Lambda brand PSUs, and I've heard they used 'odd' ICs
that
are now unobtainium. I could well believe a 723 with a difference
reference is the sortof trick they'd pull...
12:12 p.m.
New subject: 200V Zeners avail
While we're on the subject of power supplies, I came across about 5
1N2846 200V 50W zener diodes. These are TO-3 units, nicely gold-
plated, Motorola branded, probably from the 1960's.
I can't see myself ever having a use for these. Anyone care to make
an offer?
--Chuck
12:18 p.m.
New subject: 200V Zeners avail
While we're on the subject of power supplies, I
came across about 5
1N2846 200V 50W zener diodes. ?These are TO-3 units, nicely gold-
plated, Motorola branded, probably from the 1960's.
I can't see myself ever having a use for these. ?Anyone care to make
an offer?
Are they JAN? If yes, are they new in the packages?
--
Will
12:30 p.m.
New subject: 200V Zeners avail
On Sun, 17 Jan 2010, Chuck Guzis wrote:
Date: Sun, 17 Jan 2010 12:31:06 -0800
From: Chuck Guzis <cclist at sydex.com>
Reply-To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at
classiccmp.org>
Subject: Re: 200V Zeners avail
On 17 Jan 2010 at 15:18, William Donzelli wrote:
If nothing else they make fair high capacitance varicaps...
I had a 5V 50W stud mount zener (probably also Motorola) that had .01 uF of
capacitance!
Peter Wallace
Are they JAN? If yes, are they new in the
packages?
No and no. To the best of my recollection, they were removed from a
piece of Lenkurt gear.
--Chuck
5:15 p.m.
New subject: power supply problem
From: mcguire at neurotica.com
Subject: Re: power supply problem
Date: Sun, 17 Jan 2010 14:43:31 -0500
To: cctalk at classiccmp.org
On Jan 17, 2010, at 2:07 PM, Tony Duell wrote:
I've used lots and lots of Lambda power supplies, mostly bench
units. They are absolutely top-notch for "spare no expense"
projects. (thank heaven I can get them on the surplus market) I've
never seen one die, so I can't speak to the "odd IC" problem, but as
a frequent user over the years, I have nothing but good things to say
about them.
-Dave
--
Dave McGuire
Port Charlotte, FL
Hi
The only problem I've had is their " long life " electrolytics.
I've had to replace several of these on their LED readout
supplies. These are used for the low voltage supplies
used for the IC's of the display.
Still, they are the most robust supplies I've worked with.
Dwight
_________________________________________________________________
Hotmail: Free, trusted and rich email service.
http://clk.atdmt.com/GBL/go/196390708/direct/01/
It has a
trangle with a lamda in it.
I've heard of Lambda brand PSUs, and I've heard they used 'odd' ICs
that
are now unobtainium. I could well believe a 723 with a difference
reference is the sortof trick they'd pull...
5:27 p.m.
New subject: power supply problem
From: ard at p850ug1.demon.co.uk
Subject: Re: power supply problem
To: cctalk at classiccmp.org
Date: Sun, 17 Jan 2010 19:07:37 +0000
Hi Tony
It seems to be otherwise the same pins as the
generic 723. I'm just going to us one of these with
a slight change in feedback resistor divider.
It is in the power supply for my Nicolet Scope.
Dwight
_________________________________________________________________
Hotmail: Trusted email with powerful SPAM protection.
http://clk.atdmt.com/GBL/go/196390707/direct/01/
It has a trangle with a lamda in it.
I've heard of Lambda brand PSUs, and I've heard they used 'odd' ICs that
are now unobtainium. I could well believe a 723 with a difference
reference is the sortof trick they'd pull...
-tony
10:59 a.m.
Hi
I just came on an unusual problem. I'm fixing a
power supply with +12 and +5 outs. It uses 2
723 regulators. The one on the +12 was blown.
Just to conmfirm these are dfinietly supposed to be 723s (and not some
other regualtor chip that is pin compatible but has a different reference
voltage. I do recall other 'PSU kit chips' existing.
I bought another 723 regulator and it didn't
seem to be working.
I checked the voltages and noticed that the
5 volt one that is working fine=2C has a Vref of
about 3 volts while the new one has a Vref
of 7 volts.
Looking at the specs=2C it is suppose to be 7 Volts.
Every 723 I've ever used (and I've come across ones that are 40 years
old) have had a 7V reference or thereabouts.
Is this a metal-can or a DIP package? The reason I ask is I might be
convinced to look through databooks for a similar device with a 3V
reference./ And what is the origin (in parcilar what country) of the PSU?
-tony
6:40 a.m.
Chuck Guzis wrote:
On 16 Jan 2010 at 20:48, Bill Sudbrink wrote:
Bill, try putting a load on the output of the power supply and then
looking again. A 12V automotive tail or brake light bulb might do
the trick. These supplies were designed to run with a load.
Please reread my original post. That's exactly what I did.
Thanks,
Bill
16 Jan
16 Jan
6:06 p.m.
On Sun, 17 Jan 2010 12:18:56 pm Bill Sudbrink wrote:
What could be wrong with a transformer that it would
produce something like:
__
/ \
| \
/ \_
/ \
| /
| /
\_/
bent and very squared off at spots?
I've found that mains is hardly ever a nice sine wave. All of my transformer
outputs look a little wonky in some way.
Alexis.
17 Jan
17 Jan
5:27 a.m.
On Sunday (01/17/2010 at 12:36PM +1030), Alexis wrote:
On Sun, 17 Jan 2010 12:18:56 pm Bill Sudbrink wrote:
Are you looking at the output of the transformer while the bridge
rectifier is still connected? You may be seeing distortion due to a
bad diode in the bridge which may then explain some of the additional
problems downstream.
Chris
--
Chris Elmquist
What could be wrong with a transformer that it
would
produce something like:
__
/ \
| \
/ \_
/ \
| /
| /
\_/
bent and very squared off at spots?
I've found that mains is hardly ever a nice sine wave. All of my transformer
outputs look a little wonky in some way.
10:40 a.m.
My Altair 680 has achieved the position of primary attention on
my workbench. All of the ICs are socketed. Not having seen
power in over 20 years, I did not trust the power supply as far
as I could throw a bus. I pulled all of the ICs, attached dummy
RIght...
loads (automobile bulbs) and powered up. Sure enough,
there was
AC all over the place (where there should have been DC). I
Do you mean AC, or DC with excessive ripple. Did any of the lines go the
wrong side of ground? (my guess is they didn't, but it's best to be sure).
replaced all of the electrolytic caps and that got rid
of the AC
but some of the voltage levels are very wrong, at least as compared
to what is written on the schematics. The worst offender is just
off the "plus" side of the full wave bridge (BR-1). It should be
nine volts but is actually at 11.5. This makes what should be the
I wouldn't worry about that one. It's an unregulated supply, and will
drop on load. Assuming it's the input to a regulator (like that 7805 you
mention), it'll be fine.
main five volt rail a little more than 9 volts... the
far side of
VR-1 (a 7805). The wave form coming off the transformer is really
Now that does worry me, for obvious reasons. You need to correct that
before you put the ICs back (but you knew that ;-)).
There several reasons why this voltage could be high
1) The ground connection to the 7805 'common' pin is open. The 7805 treis
to get the output pin 5V _above that common pin_, so if the latter is not
ground, the output will be wrong wrt groud too. A few manufactuers used
the mouting screw of the TO220 package as the common conneciton and
didn't connect the middle pin anywhere. Others used the heatsink of a TO3
packaged regulator as part of the common connection (I've seen a version
of the latter in an HP instrument!). Check the ground connection, if it's
made through the mounting screws, check they're clean and tight.
2) THe 7805 is defective. They do fail, alas, which is why you chack the
PCB on a dummy load :-))
3) There is resistor in parallel with the regulator. Some manufacturers
connected a high-wattage resistor between the input and output pins of
the regulator, of a suitable value to carry most of the load current at
the expected voltage drop, thus leaving the reguator to carry (and
control) a much smaller current. This is an evil trick because the output
voltage will rise (and the rewgualtor can't do anything about it) if the
load is too small. Again, I've come across this in an HP device, it was
fitted with a crowbar circuit, and I wondered why it tripped with the
supply on dummy load. The reason was my load was not large enough. Do you
have scehamtics? If so, that'll show if the resistor is present or not.
ugly not a smooth sine wave. I hate power supplies.
I don't really
understand, looking at the schematics, what this transformer should
be doing. My guess is that it should be making a nice nine volt,
60 cycle AC sine wave with each of the two outputs 90 degrees out
No, assuming it's a centre=-tapped winding, the ends will be 180 degrees
out of phase.
With no load on the transforemr, the outputs should be approximately
sinusoidal. With the sort of load you're giving it (rectifiers and
smoothing capacitors), the waveform may look horrible.
Provided the transofrmer is not getting exessively hot, I'd not worry
about that too much.
-tony
11:26 a.m.
On 17 Jan 2010 at 18:40, Tony Duell wrote:
3) There is resistor in parallel with the regulator.
MITS did this routinely and it wouldn't surprise me if they did this
on the 680. In particular, I remember that the +5 supply for the 4K
DRAM boards had this dirty little trick. Anything, I suppose, to
shave a few dimes off the manufacturing cost.
Everything's fine with a full load, but if the load should drop for
any reason, any IC connected to the regulated (?) line were toast.
--Chuck
11:42 a.m.
On 17 Jan 2010 at 18:40, Tony Duell wrote:
Oh, right... The OP says he has the schematics, do they show any such
resistor.
3) There is resistor in parallel with the
regulator.
MITS did this routinely and it wouldn't surprise me if they did this
on the 680. In particular, I remember that the +5 supply for the 4K DRAM boards had this dirty little trick. Anything, I
suppose, to
shave a few dimes off the manufacturing cost.
I came across it in the HP59500 HPIB interface for the multiprogrammer.
This is basically a unit which send/receives ASCII strings over HPIB and
has 16-out, 13-in TTL level interface to the multiprogrammer.
It's a large board of TTL (something like 70 chips) and a little PSU
board giving the necessary 5V (that's the only supply it needs). The PSU
board has 2 TO3 componnents on it, one is a regulator (I forget it it's a
fixed of adjustable one) and the other is an SCR [1] for the crowbar
circuit. Anyway, there is a reistor shunted across the regulator to carry
most of the load current. If you try to test that supply without a dummy
load oe with too small a load, the crowbar will fire and the resistor
gets hot and bothered.
[1] We (UK) often call that component a 'thyristor' by analogy with
'thyratron' (a gas-filled triode or more complex valve, line the 2D21
(EN91), etc). Is that name recognised elsewhere?
Everything's fine with a full load, but if the load should drop for
any reason, any IC connected to the regulated (?) line were toast.
Indeed.
-tony
1:31 p.m.
On Sunday (01/17/2010 at 07:42PM +0000), Tony Duell wrote:
Yes. There is a 15 ohm, 2W resistor from the input to the output of VR-1,
a 7805 regulator.
Danger Will Robinson.
--
Chris Elmquist
On 17 Jan 2010 at 18:40, Tony Duell wrote:
Oh, right... The OP says he has the schematics, do they show any such
resistor. 3) There is resistor in parallel with the
regulator.
MITS did this routinely and it wouldn't surprise me if they did this
on the 680. In particular, I remember that the +5 supply for the 4K
18 Jan
18 Jan
11:47 a.m.
3) There is
resistor in parallel with the regulator.
MITS did this routinely and it wouldn't surprise me if they did this
on the 680. In particular, I remember that the +5 supply for the 4K Danger Will Robinson.
I could understand mentioning Heath Robinson here...
-tony
2 p.m.
On Monday (01/18/2010 at 07:47PM +0000), Tony Duell wrote:
15 ohm, 2W is what my original schematics say. It's entirely possible they
changed this value over time. If need be, I can take the lid off the
machine and see what is really in there... but I think we've converged
on the basic idea.
Yes. There is a 15 ohm, 2W resistor from the input to the output of VR-1,
a 7805 regulator.
OK. Let's analyse that circuit. A quick look at a 7805 data sheet will
convince you it can only pull the output toiwards the input voltage, not
towards ground. So if the output is already higher than 5V (in this
case), the 7805 wioll do nothing. Consider the resistor on its own. The OP said there
was an input voltage
of 11.5V. SO there's 6.5V (11.5-5) to be dropped between the input and
output -- without the 7805, that means the resistor is droping 6.5V and
carrying 6.5V/15R = 0.4333A.
If the load drwas less than 0.4333A, the output voltage will be higher
than 5V. If it drwas more, then part of that current will be carried by
the 7805, and the output will remain at 5V (without the 7805, of course,
the output votlage would be less than 5V in this case).
Is it fair (as in rational, smart, non-stupid) to put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
Light bulbs are an approximatley constant-current
device. A 12V car bulb
that draws 0.433A (at 12V) is a 5.2W bulb. If the load is significantly
less than that, then of course the output voltage will be too high.
Yes.. of course! That works too.
http://en.wikipedia.org/wiki/Danger,_Will_Robinson
--
Chris Elmquist
Danger Will Robinson.
I could understand mentioning Heath Robinson here...
19 Jan
19 Jan
12:34 a.m.
On Jan 18, 2010, at 5:00 PM, Chris Elmquist wrote:
--
Dave McGuire
Port Charlotte, FL
Is it fair (as in rational, smart, non-stupid) to put
multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
What about balancing problems? They won't be identical from part
to part..
-Dave
4:44 a.m.
New subject: [personal] Re: Altair 680 power supply...
----- Original Message -----
From: "Dave McGuire" <mcguire at neurotica.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, January 19, 2010 8:34 AM
Subject: [personal] Re: Altair 680 power supply...
On Jan 18, 2010, at 5:00 PM, Chris Elmquist wrote:
What's the current ?, 78S05 will give 2A
7805 cannot be directly paralleled due to the small difference in regulation
points part to part. You would need a small value series resistor per part
to load share.
Best regards,
Mike Hatch
soemtron.org
Is it fair (as in rational, smart, non-stupid) to
put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
What about balancing problems? They won't be identical from part to
part..
-Dave
--
Dave McGuire
Port Charlotte, FL
5:37 a.m.
New subject: [personal] Re: Altair 680 power supply...
Mike Hatch wrote:
----- Original Message ----- From: "Dave
McGuire" <mcguire at neurotica.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, January 19, 2010 8:34 AM
Subject: [personal] Re: Altair 680 power supply...
You cannot parallel fixed regulators such as 7805. If you need more
current either go
to a device that can handle the current or use a regulator and
transistor combination.
I've used the rather low power 7805(to220 part) with an external pass
transistor
to get 5A, the design is straight out of National Semi (and others) apnotes.
Allison
On Jan 18, 2010, at 5:00 PM, Chris Elmquist
wrote:
What's the current ?, 78S05 will give 2A
7805 cannot be directly paralleled due to the small difference in
regulation points part to part. You would need a small value series
resistor per part to load share.
Best regards,
Mike Hatch
soemtron.org
Is it fair (as in rational, smart, non-stupid) to
put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
What about balancing problems? They won't be identical from part
to part..
-Dave
--
Dave McGuire
Port Charlotte, FL
6:22 a.m.
New subject: [personal] Re: Altair 680 power supply...
Date: Tue, 19 Jan 2010 08:37:58 -0500
From: ajp166 at verizon.net
To:
Subject: Re: [personal] Re: Altair 680 power supply...
Mike Hatch wrote:
Hi
While it is incorrect to parallel regulators, one can always split
the power rails to balance the loads. I've seen this done on
a S-100 memory board.
In general, well heat sinked 7805s can be paralleled if the power
drop in is not too large. They do have a built in current limiter
that is designed to handle a continuous constant current without
failure.
It does require good bypass capacitors or the board can make
a nice 1MHz oscillator.
Dwight
_________________________________________________________________
Hotmail: Trusted email with Microsoft?s powerful SPAM protection.
http://clk.atdmt.com/GBL/go/196390706/direct/01/
----- Original Message ----- From: "Dave
McGuire" <mcguire at neurotica.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, January 19, 2010 8:34 AM
Subject: [personal] Re: Altair 680 power supply...
You cannot parallel fixed regulators such as 7805. If you need more
current either go
to a device that can handle the current or use a regulator and
transistor combination.
I've used the rather low power 7805(to220 part) with an external pass
transistor
to get 5A, the design is straight out of National Semi (and others) apnotes.
Allison
>
>>
>> -Dave
>>>
>>
>>
>> --
>> Dave McGuire
>> Port Charlotte, FL On Jan 18, 2010, at 5:00 PM, Chris Elmquist
wrote:
What's the current ?, 78S05 will give 2A
7805 cannot be directly paralleled due to the small difference in
regulation points part to part. You would need a small value series
resistor per part to load share.
Best regards,
Mike Hatch
soemtron.org
Is it fair (as in rational, smart, non-stupid) to
put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
What about balancing problems? They won't be identical from part
to part..
5:38 a.m.
New subject: [personal] Re: Altair 680 power supply...
On Tuesday (01/19/2010 at 12:44PM -0000), Mike Hatch wrote:
----- Original Message ----- From: "Dave
McGuire" <mcguire at neurotica.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, January 19, 2010 8:34 AM
Subject: [personal] Re: Altair 680 power supply...
The discussion was about a way to get more current from a 7805 design
that uses a bypass resistor to cheat the regulator. A number of TO-220
packaged options were suggested but I was curious if one could get there
with a different sort of cheat-- since plain old 7805 are so cheap and
plentiful (still).
On Jan 18, 2010, at 5:00 PM, Chris Elmquist
wrote:
What's the current ?, 78S05 will give 2A Is it fair (as in rational, smart, non-stupid) to
put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
What about balancing problems? They won't be identical from part to
part.. 7805 cannot be directly paralleled due to the small
difference in
regulation points part to part. You would need a small value series
resistor per part to load share.
Right. I saw one other suggestion that in addition to the series resistor
you would also bring them into a Schottky diode summing network which
presumably keeps them from back driving each other.
In any case, you need enough other stuff to accomplish this that
retrofitting with an LM323, 78S05 or any of the drop-in switchers is
probably a much better option.
Chris
--
Chris Elmquist
5:28 a.m.
On Tuesday (01/19/2010 at 03:34AM -0500), Dave McGuire wrote:
On Jan 18, 2010, at 5:00 PM, Chris Elmquist wrote:
Then that means only one of them carries most of the current... True.
I've had very few power supply failures in my life but those that I
have had, were shorted 2N3055's used as pass transistors around 78xx
regulators. So, I have grown to not appreciate that approach either.
Chris
--
Chris Elmquist
Is it fair (as in rational, smart, non-stupid) to
put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
What about balancing problems? They won't be identical from part to
part..
20 Jan
20 Jan
10:34 a.m.
I've had very few power supply failures in my life
but those that I
have had, were shorted 2N3055's used as pass transistors around 78xx
regulators. So, I have grown to not appreciate that approach either.
The normal cause of that is not providing some current limiting circuit.
The 7805 includes one (of course), but adding an external pass transistor
will negate this unless you add some more components (how to do it is
shown in the better data sheets for said 3 terminal regualtors). If you
don't and accidentally short the output to ground (or give it too heavy a
load in some other way), the 2N3055 will most likely fail. And it may
fail shorted
-tony
19 Jan
19 Jan
12:09 p.m.
Is it fair (as in rational, smart, non-stupid) to put
multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
It's a bad idea to parallel regualtors like that, since the output
votlages will not be indentical and one will supply the lion's share of
the current (and will try to feed current back into the others).
What you can do is have separae regualtors for different bits of the
circuit (the fact that all the 5V rails are not _quite_ the same won't
matter). The S100 bus did this, of course (regulators on each card, the
5V line was not bussed between cards). I believe very early versions of
the BBC Micro PSU (the linear one that 'took the paint of tables' because
it ran so hot) used 3 7805s with 3 feed wires to the logic board. On
later boards, designed for use with the SMPSU, the 5V terminals were all
linked, I beliueve they may not have been on very early boards.
-tony
1:10 p.m.
Tony Duell wrote:
Couldn't you get around at least part of the problem with isolation diodes?
Peace... Sridhar
Is it fair (as
in rational, smart, non-stupid) to put multiple 7805 in
parallel? Can't say I've ever seen that done. Only the more typical
design of adding a pass transistor to carry more current.
It's a bad idea to parallel regualtors like that, since the output
votlages will not be indentical and one will supply the lion's share of
the current (and will try to feed current back into the others).
1:27 p.m.
On 19 Jan 2010 at 16:10, Sridhar Ayengar wrote:
Couldn't you get around at least part of the
problem with isolation
diodes?
No sir. Consider that the output of one 7805 might be 5.05 volts and
of the other, 4.95 (manufacturing tolerances being what they are,
that's not unreasonable). The bulk of the current will be conducted
by the 7805 with the higher voltage output. One might put a resistor
in series with the output of each one to balance the load a bit, but
that would play hob with the voltage regulation. (On the other hand,
this technique is used when paralleling rectifiers to balance the
current.)
But, as someone mentioned, the bigger problem is that linear
regulators tend to dissipate a lot of heat, especially when the input
voltage is substantially higher than the otuput voltage.
Better to use a buck regulator in this application, methinks, and
save some watts.
--Chuck
1:39 p.m.
Chuck Guzis wrote:
But I thought part of the problem was that the regulator with the
slightly higher output voltage would drive the voltage difference
against the output of the lower-voltage regulator? Wouldn't isolation
diodes prevent that from happening?
Peace... Sridhar
Couldn't
you get around at least part of the problem with isolation
diodes?
No sir. Consider that the output of one 7805 might be 5.05 volts and
of the other, 4.95 (manufacturing tolerances being what they are,
that's not unreasonable). The bulk of the current will be conducted
by the 7805 with the higher voltage output. One might put a resistor
in series with the output of each one to balance the load a bit, but
that would play hob with the voltage regulation. (On the other hand,
this technique is used when paralleling rectifiers to balance the
current.) 
1:44 p.m.
Sridhar Ayengar wrote:
Chuck Guzis wrote:
But I thought part of the problem was that the regulator with the
slightly higher output voltage would drive the voltage difference
against the output of the lower-voltage regulator? Wouldn't isolation
diodes prevent that from happening?
How about plan B -- use lower power chips if possible?
Plan C ... use Alien Technology.
Couldn't you get around at least part of the problem with isolation
diodes?
No sir. Consider that the output of one 7805 might be 5.05 volts and
of the other, 4.95 (manufacturing tolerances being what they are,
that's not unreasonable). The bulk of the current will be conducted
by the 7805 with the higher voltage output. One might put a resistor
in series with the output of each one to balance the load a bit, but
that would play hob with the voltage regulation. (On the other hand,
this technique is used when paralleling rectifiers to balance the
current.) Peace... Sridhar
Ben.
PS. I would like to remove the power resistor ... ripple on the 5+.
2:48 p.m.
Sridhar Ayengar wrote:
Chuck Guzis wrote:
But I thought part of the problem was that the regulator with the
slightly higher output voltage would drive the voltage difference
against the output of the lower-voltage regulator? Wouldn't isolation
diodes prevent that from happening?
Diodes will effectively leave you with only one regulator supplying the
current, as only one diode will switch on.
Using Chuck's numbers as an example, the one reg supplies 5.05V into one diode
anode, the output at the cathode is 5.05-0.7=4.35V. The diode from the other
reg (cathodes connected together) now has 4.95-4.35=0.6V dropped across it,
which isn't enough to even send it into conduction, that reg supplies nothing
to the load (..until the first reg over-temps and burns up or shuts down).
If one looks at high-current regulators they have multiple pass transistors
connected 'almost' in parallel, the catch is a small resistor is placed in the
emitter lead of each so the CE & BE circuits take on some linear
characteristics and distributes the current between the transistors. The
regulation sense point is then placed after all the resistors join together for
the output, so the resistors are inside the feedback loop and hence accounted
for in the regulation.
Couldn't you get around at least part of the problem with isolation
diodes?
No sir. Consider that the output of one 7805 might be 5.05 volts and
of the other, 4.95 (manufacturing tolerances being what they are,
that's not unreasonable). The bulk of the current will be conducted
by the 7805 with the higher voltage output. One might put a resistor
in series with the output of each one to balance the load a bit, but
that would play hob with the voltage regulation. (On the other hand,
this technique is used when paralleling rectifiers to balance the
current.)
20 Jan
20 Jan
10:43 a.m.
If one looks at high-current regulators they have
multiple pass transistors
connected 'almost' in parallel, the catch is a small resistor is placed in the
emitter lead of each so the CE & BE circuits take on some linear
characteristics and distributes the current between the transistors. The
regulation sense point is then placed after all the resistors join together for
the output, so the resistors are inside the feedback loop and hence accounted
for in the regulation.
I've seen at least one PSU where one of those eqalisation resistors was
also used as the current sense resistor (the circuit assumed the load
current was shared approximatley equally between the pass transistors).
It looks neat on paper, but the problem comes if the pass transsitor
feeding the resistor also used for sensing opens. Then the circuit thinks
the PSU is delivering no current, and won't shut down even if the load is
shorted. Not what I would call a good design.
-tony
1 Aug
1 Aug
9:42 p.m.
New subject: Qantel minicomputer available.
I'm trying to find a home for my Qantel minicomputer. It's not the most
exciting system, but it may be the last Qantel system around. The CHM
turned it down do to lack of space so I thought someone here might be
interested.
It's a desk sized unit and includes a CRT terminal, two printers,
a bunch of disk packs, and very little documentation. You can see
some photos here:
http://www.dvq.com/qantel/qantel.htm <http://www.dvq.com/sale/sale.htm>
It's built out of TTL IC's and uses semiconductor memory.
I'm not looking for money, just to find a nice home for it as I just
need the space
and it's now sitting outside!
Located in beautiful Bonny Doon, CA in the Santa Cruz mountains. Pickup
only,
unless you can arrange shipping.
Bob Rosenbloom
2 Aug
2 Aug
12:39 p.m.
New subject: Qantel minicomputer available.
Nice... I was about to respond with information that the father of one
of my colleagues was head of R&D at Quantel in the early days. Then I
spotted that there's a Quantel and a Qantel. How confusing is that?!
J
On 02/08/2011 05:42, Bob Rosenbloom wrote:
I'm trying to find a home for my Qantel
minicomputer. It's not the most
exciting system, but it may be the last Qantel system around. The CHM
turned it down do to lack of space so I thought someone here might be
interested.
It's a desk sized unit and includes a CRT terminal, two printers,
a bunch of disk packs, and very little documentation. You can see
some photos here:
http://www.dvq.com/qantel/qantel.htm <http://www.dvq.com/sale/sale.htm>
It's built out of TTL IC's and uses semiconductor memory.
I'm not looking for money, just to find a nice home for it as I just
need the space
and it's now sitting outside!
Located in beautiful Bonny Doon, CA in the Santa Cruz mountains.
Pickup only,
unless you can arrange shipping.
Bob Rosenbloom
5 Aug
5 Aug
3:01 p.m.
New subject: Qantel minicomputer available.
On 02/08/11 20:39, James Wilson wrote:
Nice... I was about to respond with information that
the father of one
of my colleagues was head of R&D at Quantel in the early days. Then I
spotted that there's a Quantel and a Qantel. How confusing is that?!
The same Quantel that designed (and sold) the Paintbox graphics editing
systems?
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
4:43 p.m.
New subject: Qantel minicomputer available.
In article <4E3C6841.6050802 at philpem.me.uk>,
Philip Pemberton <classiccmp at philpem.me.uk> writes:
On 02/08/11 20:39, James Wilson wrote:
Quantel made the paintbox.
Qantel made a business minicomputer.
Definately confusing.
I'd love to have paintbox stuff, but I've never seen it anywhere.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
Nice... I was about to respond with information
that the father of one
of my colleagues was head of R&D at Quantel in the early days. Then I
spotted that there's a Quantel and a Qantel. How confusing is that?!
The same Quantel that designed (and sold) the Paintbox graphics editing
systems? 
7 Aug
7 Aug
10:55 p.m.
New subject: Qantel minicomputer available.
Quantel are local to me here in Newbury UK. They are still very much in
business although they don't make the headlines these days. Some years ago I
ran a wiring business and Quantel were a major customer.
We used to (literally) wire up their prototype boards. We used Kynar wire
wrap wire to connect the IC pads. They had elongated pads with extra drilled
holes and the wiring went on the component side.
Lots and lots of TTL.!!
Regards
?
Rod Smallwood
?
-----Original Message-----
From: cctech-bounces at classiccmp.org [mailto:cctech-bounces at classiccmp.org]
On Behalf Of Richard
Sent: 06 August 2011 00:43
To: cctalk
Subject: Re: Qantel minicomputer available.
In article <4E3C6841.6050802 at philpem.me.uk>,
Philip Pemberton <classiccmp at philpem.me.uk> writes:
On 02/08/11 20:39, James Wilson wrote:
Quantel made the paintbox.
Qantel made a business minicomputer.
Definately confusing.
I'd love to have paintbox stuff, but I've never seen it anywhere.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
Nice... I was about to respond with information
that the father of one
of my colleagues was head of R&D at Quantel in the early days. Then I
spotted that there's a Quantel and a Qantel. How confusing is that?!
The same Quantel that designed (and sold) the Paintbox graphics editing
systems?
19 Jan
19 Jan
3:37 p.m.
Date: Tue, 19 Jan 2010 16:39:49 -0500
From: ploopster at gmail.com
To:
Subject: Re: Altair 680 power supply...
Chuck Guzis wrote:
But I thought part of the problem was that the regulator with the
slightly higher output voltage would drive the voltage difference
against the output of the lower-voltage regulator? Wouldn't isolation
diodes prevent that from happening?
Peace... Sridhar
No
It would still be a supplied by the regulator with the largest voltage
out.
I did mention that these regulators are designed to run in the
current shut-down mode, indefinitely unless the power spec is
exceeded. That means they can be paralleled with care.
The regulator, with the slightly higher voltage, will current limit,
and then start loading up the next regulator.
It's not the best for the life of the regulator but still within the
design spec of the regulator.
Dwight
_________________________________________________________________
Hotmail: Powerful Free email with security by Microsoft.
http://clk.atdmt.com/GBL/go/196390710/direct/01/
Couldn't you get around at least part of the problem with isolation
diodes?
No sir. Consider that the output of one 7805 might be 5.05 volts and
of the other, 4.95 (manufacturing tolerances being what they are,
that's not unreasonable). The bulk of the current will be conducted
by the 7805 with the higher voltage output. One might put a resistor
in series with the output of each one to balance the load a bit, but
that would play hob with the voltage regulation. (On the other hand,
this technique is used when paralleling rectifiers to balance the
current.) 
18 Jan
18 Jan
12:04 p.m.
On Sun, 17 Jan 2010 19:42:48 +0000 (GMT)
ard at p850ug1.demon.co.uk (Tony Duell) wrote:
[1] We (UK) often call that component a
'thyristor' by analogy with
'thyratron'
[...]
Is that name recognised elsewhere?
It is called
'thyristor' in Germany too.
--
tsch??,
Jochen
Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/
12:16 p.m.
On 18 Jan 2010 at 21:04, Jochen Kunz wrote:
On Sun, 17 Jan 2010 19:42:48 +0000 (GMT)
ard at p850ug1.demon.co.uk (Tony Duell) wrote:
Both "thyristor" and "SCR" are used in the U.S., but "SCR"
is easier
to spell. "Thyristor" is more of a generic term that can include
TRIACs, DIACs, SIDACs, LASCRs, and all manner of other 4-layer
devices.
--Chuck
[1] We (UK) often call that component a
'thyristor' by analogy with
'thyratron'
[...]
Is that name recognised elsewhere?
It is
called 'thyristor' in Germany too.
1:21 p.m.
On 18 Jan 2010 at 21:04, Jochen Kunz wrote:
:-)
On Sun, 17 Jan 2010 19:42:48 +0000 (GMT)
ard at p850ug1.demon.co.uk (Tony Duell) wrote:
Both "thyristor" and "SCR" are used in the U.S., but "SCR"
is easier [1] We (UK) often call that component a
'thyristor' by analogy with
'thyratron'
[...]
Is that name recognised elsewhere?
It is
called 'thyristor' in Germany too. to spell. "Thyristor" is more of a generic
term that can include
TRIACs, DIACs, SIDACs, LASCRs, and all manner of other 4-layer
devices.
Right. Over here, AFAIK, a 'thyristor' was a 4-layer, 3 lead,
unidirectional device, I've neer head it being used for any of the other
devices. The name, I beliece, is derrived from the (ancient?) Greek
'Thura' meaning a door.
I grew up calling them thyristors, but now normally call them SCRs since
it causes less confusion. Similarly I learnt the German symbols for logic
gates (the AND gate looks like what you'd expect, the OR gate looks like
an AND gate -- it has a straight back -- but the input lines are
continuted across the symbol to the curved face), but now use the normal
US sysmbols when I'm drawing schematis, since everybody else does. I can
still read a scehamtic using the German symbols, of course (nad often
have to ;-)).
I also learnt the zig-zag lien for a resistor, and nothing is going to
get me to change that to the narrow rectangle used in the rest of Europe.
Again I have no problem reading schemaitcs drawn with either symbol.
-tony
2 p.m.
On 18 Jan 2010 at 21:21, Tony Duell wrote:
Similarly I learnt the German symbols for logic gates
(the AND gate
looks like what you'd expect, the OR gate looks like an AND gate -- it
has a straight back -- but the input lines are continuted across the
symbol to the curved face), but now use the normal US sysmbols when I'm
drawing schematis, since everybody else does. I can still read a
scehamtic using the German symbols, of
course (nad often have to ;-)).
I'dve thought that the EU would have changed over to those dreadful
IEC symbols by now (BS/EU EN 60617-12:1999). I stumble through those
with my nose held. It's like reading a musical score that's been
reduced to text with note names and durations. There's no visual
"flow".
--Chuck
19 Jan
19 Jan
12:05 p.m.
On 18 Jan 2010 at 21:21, Tony Duell wrote:
Oh, officially we probably have. But I've not met anyone who uses them.
Similarly I learnt the German symbols for logic
gates (the AND gate
looks like what you'd expect, the OR gate looks like an AND gate -- it
has a straight back -- but the input lines are continuted across the
symbol to the curved face), but now use the normal US sysmbols when I'm
drawing schematis, since everybody else does. I can still read a
scehamtic using the German symbols, of
course (nad often have to ;-)).
I'dve thought that the EU would have changed over to those dreadful
IEC symbols by now (BS/EU EN 60617-12:1999). I stumble through those with my nose held. It's like reading a musical
score that's been
reduced to text with note names and durations. There's no visual
"flow".
Oh, and don't get me started on the related subject of schematics that
have not been drawn by electronic engineers. A good schematic should be
instantly recognisable for what it is. Not a 'pretty' arrangement of
symbols on the page.
-tony
2:18 a.m.
On Mon, 18 Jan 2010, Chuck Guzis wrote:
Both "thyristor" and "SCR" are
used in the U.S., but "SCR" is easier
to spell.
Maybe for you, but not for me ;-)
"Thyristor" is more of a generic term that
can include
TRIACs, DIACs, SIDACs, LASCRs, and all manner of other 4-layer
devices.
No! A thyristor is a semiconductor version of a thyratron. A triac is
something different and has *never* been called a thyristor even though it
may be seen as two anti-parallel thyristors (which is not correct).
Christian
7:15 a.m.
On Tuesday 19 January 2010, Christian Corti wrote:
On Mon, 18 Jan 2010, Chuck Guzis wrote:
Yes! It appears that some people on this list call a Triac a thyristor,
and it's in use elsewhere too:
http://www.allaboutcircuits.com/vol_3/chpt_7/6.html
Sorry to spoil the fun,
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
Both "thyristor" and "SCR"
are used in the U.S., but "SCR" is
easier to spell.
Maybe for you, but not for me ;-)
"Thyristor" is more of a generic term
that can include
TRIACs, DIACs, SIDACs, LASCRs, and all manner of other 4-layer
devices.
No! A thyristor is a semiconductor version of a thyratron. A triac is
something different and has *never* been called a thyristor even
though it may be seen as two anti-parallel thyristors (which is not
correct).
9:41 a.m.
On 19 Jan 2010 at 10:15, Patrick Finnegan wrote:
On Tuesday 19 January 2010, Christian Corti wrote:
Must be a EE cultural thing. I think "A DIAC is a type of
thyristor...", I should check my old GE databook and see what, if
any, distinction is made.
This reminds me of my introduction to Hoosier English as she is spoke
outside of the Calumet Region. At my first meal at Purdue, I asked
the guy tending the steam table for some of the "string beans". "We
don't have string beans; those is green beans." Actually, he said
"strang" rather than "string", and yes--he really did say "those
is".
I learned to adapt---and developed a taste for onion rangs and
learned that "grits" has two syllables.
Cheers,
Chuck
No! A thyristor is a semiconductor version of a
thyratron. A triac
is something different and has *never* been called a thyristor even
though it may be seen as two anti-parallel thyristors (which is not
correct).
Yes! It appears that some people on this list call a Triac a
thyristor, and it's in use elsewhere too:
http://www.allaboutcircuits.com/vol_3/chpt_7/6.html
17 Jan
17 Jan
2:32 p.m.
Tony Duell wrote:
<snip>
Do you mean AC, or DC with excessive ripple. Did any
of the lines go the
wrong side of ground? (my guess is they didn't, but it's best to be sure).
DC with excessive ripple. Nothing on the wrong side of ground but 60 and
120 Hz waveforms almost down to ground.
replaced all
of the electrolytic caps and that got rid of the AC
but some of the voltage levels are very wrong, at least as compared
to what is written on the schematics. The worst offender is just
off the "plus" side of the full wave bridge (BR-1). It should be
nine volts but is actually at 11.5. This makes what should be the
I wouldn't worry about that one. It's an unregulated supply, and will
drop on load. Assuming it's the input to a regulator (like that 7805 you
mention), it'll be fine.
main five volt rail a little more than 9 volts...
the far side of
VR-1 (a 7805). The wave form coming off the transformer is really
Now that does worry me, for obvious reasons. You need to correct that
before you put the ICs back (but you knew that ;-)). 3) There is resistor in parallel with the regulator.
Some manufacturers
connected a high-wattage resistor between the input and output pins of
the regulator, of a suitable value to carry most of the load current at
the expected voltage drop, thus leaving the reguator to carry (and
control) a much smaller current. This is an evil trick because the output
voltage will rise (and the rewgualtor can't do anything about it) if the
load is too small. Again, I've come across this in an HP device, it was
fitted with a crowbar circuit, and I wondered why it tripped with the
supply on dummy load. The reason was my load was not large enough. Do you
have scehamtics? If so, that'll show if the resistor is present or not.
Yes, there is an 18 Ohm 2 Watt resistor between the +9 and +5 (the input
and output pins of the 7805). I was wondering what that is supposed to do.
Is this resistor causing the +5 side to "look higher" than it really is?
Is everything actually OK, even though my VOM and scope both agree the +5
rail is at +9.2 _WITH_ automobile light dummy loads?
A PDF of the schematics is here:
http://www.altair680kit.com/manuals/Altair_680-Schematics-03-Beta_1-Searcha…
The power supply is on page 2.
Thanks,
Bill
3:01 p.m.
On 17 Jan 2010 at 17:32, Bill Sudbrink wrote:
Yes, there is an 18 Ohm 2 Watt resistor between the +9
and +5 (the
input and output pins of the 7805). I was wondering what that is
supposed to do. Is this resistor causing the +5 side to "look higher"
than it really is? Is everything actually OK, even though my VOM and
scope both agree the +5 rail is at +9.2 _WITH_ automobile light dummy
loads?
Back to what I said. Your load isn't heavy enough. At an 11 V high-
side supply, an 18 ohm resistor will produce 9V at a load current of
100 mA, even if the 7805 is completely kaput.
Put a load that pulls about 2A at 5V (a 10W 3 ohm resistor should
do) and see what your readings are.
Better yet, ditch the 7805 and resistor and install an LM323 with
suitable heatsink. Or use one of the switching buck regulators.
MITS was about shaving pennies from their designs.
---Chuck
18 Jan
18 Jan
3:26 p.m.
Chuck Guzis wrote:
Put a load that pulls about 2A at 5V (a 10W 3 ohm
resistor should
do) and see what your readings are.
I built a test load by putting three one Ohm 10W resistors (available
at the local Radio Shack) in series. With the test load, the +9 rail
goes down to a little more than 8 and the +5 rail goes to about 4.8,
smooth as glass on my scope. With that test, I decided to go ahead
and reinstall the ICs. I've applied power and the +5 rail is between
5.1 and 5.2. The system seems to function as far as I can test it with
my limited operations knowledge. I can deposit values in memory and
they remain as long as power is applied. Now I need to learn a little
6800 machine code and try a short program.
Thanks for all of the help.
4:26 p.m.
On 18 Jan 2010 at 18:26, Bill Sudbrink wrote:
I built a test load by putting three one Ohm 10W
resistors (available
at the local Radio Shack) in series. With the test load, the +9 rail
goes down to a little more than 8 and the +5 rail goes to about 4.8,
smooth as glass on my scope. With that test, I decided to go ahead
and reinstall the ICs. I've applied power and the +5 rail is between
5.1 and 5.2.
I strongly urge you to replace the 7805 and parallel resistor with an
LM323T unit. The pinout and package are the same, but the current
rating is 3A, not 1. This would give you a measure of protection
that the resistor + 7805 does not afford.
There are, of course, other regulator option, but the LM323T is the
simplest.
--Chuck
4:40 p.m.
Chuck Guzis wrote:
I strongly urge you to replace the 7805 and parallel
resistor with an
LM323T unit. The pinout and package are the same, but the current
rating is 3A, not 1. This would give you a measure of protection
that the resistor + 7805 does not afford.
But where do you get 3 amp heat sinks? All I ever see are the wimpy
heat sinks.
There are, of course, other regulator option, but the
LM323T is the
simplest.
--Chuck
5:13 p.m.
On 18 Jan 2010 at 17:40, Ben wrote:
But where do you get 3 amp heat sinks? All I ever see
are the wimpy
heat sinks.
Well, at Radio Shack, to start with:
http://www.radioshack.com/product/index.jsp?productId=2102857
But you could also just pull out your Aavid catalog and see what will
work. For TO-220, how about this one:
http://www.aavidthermalloy.com/cgi-bin/stdisp.pl?Pnum=568000b00000g
If you had to leave the resistor/7805 in, I'd add a fuse and
overvoltage crowbar across the output.
Alternatively, one could use one of the switching regulator ICs, such
as the Micrel MIC4576-5. Still a TO-220 package, but 5 leads and
70% efficiency requiring a couple of extra components.
--Chuck
6:07 p.m.
On Jan 18, 2010, at 8:13 PM, Chuck Guzis wrote:
Alternatively, one could use one of the switching
regulator ICs, such
as the Micrel MIC4576-5. Still a TO-220 package, but 5 leads and
> 70% efficiency requiring a couple of extra components.
I like the TI/PowerTrends PT78ST105. >85%, 1.5A, no external
components. About the size of a TO220+heatsink. Very, very nice
modules.
-Dave
--
Dave McGuire
Port Charlotte, FL
6:32 p.m.
On 18 Jan 2010 at 21:07, Dave McGuire wrote:
I like the TI/PowerTrends PT78ST105. >85%, 1.5A,
no external
components. About the size of a TO220+heatsink. Very, very nice
modules.
I agree, but I'm not sure that a 1.5A rating is enough for this
application. The Micrel part is rated for 3A.
It's all good--my idea of a "voltage regulator IC" used to be a 741
op amp feeding a darlington. :)
--Chuck
6:52 p.m.
On Jan 18, 2010, at 9:32 PM, Chuck Guzis wrote:
Oh, duh. I wasn't paying attention to the application. I do
really like those parts though!
--
Dave McGuire
Port Charlotte, FL
I like the
TI/PowerTrends PT78ST105. >85%, 1.5A, no external
components. About the size of a TO220+heatsink. Very, very nice
modules.
I agree, but I'm not sure that a 1.5A rating is enough for this
application. The Micrel part is rated for 3A. It's all good--my idea of a "voltage
regulator IC" used to be a 741
op amp feeding a darlington. :)
Hey, there you go. ;) I'm a 2N3055 man, myself. You can arc weld
with those things.
-Dave
19 Jan
19 Jan
12:43 p.m.
> It's all good--my idea of a "voltage
regulator IC" used to be a 741
> op amp feeding a darlington. :)
Of course the 723 is a referece supply and an op-amp (which you can use
to feed a darlington) ;-)
Hey, there you go. ;) I'm a 2N3055 man, myself. You can arc weld
with those things.
Ah yes. Good solid transistors. The PSU board of the plotter I am
currently repairing has 5 of them on it. One for the -15V logic supply
regulaotr, and 2 for each of the motor drivers (I would have expected an
NPN/PNP pair here, but I guess there wasn't a suitable PNP transsitor
available in 1959)
IIRC the 'modern' 2N3055 is actually a 2N3055H, you can't get the original
one any more. It doesn't matter for most applications, though. Either Bob
Pease or Jim Williams makes a comment in one of their books.
-tony
18 Jan
18 Jan
12:08 p.m.
Tony Duell wrote:
You would expect ot see 60Hz waveforms from a half-wave rectifier (one
diode) with no smoothing, and 120Hz waveforms from a full-wave rectifier
(either 2 diodes with a centre-tapped transformer winding or a 4-diode
bridge rectifier) with no smoothing. The open-circuit capacitors would
account for 'no smoothing', but if you have what appears to be a
full-wave circuit giving 60Hz ripple, I would suspect faulty diodes too.
Do you mean AC, or DC with excessive ripple. Did
any of the lines go the
wrong side of ground? (my guess is they didn't, but it's best to be sure).
DC with excessive ripple. Nothing on the wrong side of ground but 60 and
120 Hz waveforms almost down to ground. Yes, there is an 18 Ohm 2 Watt resistor between the +9
and +5 (the input
and output pins of the 7805). I was wondering what that is supposed to do.
Is this resistor causing the +5 side to "look higher" than it really is?
No, the voltmeter is genuinely measuring the voltage on the 5V rail. It
is 9V. The circuit _may_ be working -- if you incres the load, the
output voltage may come down to 5V and then _stay there_ for increasing
loads (that's more current, so lower resistance load, of course), until
you get to tyhe point where the 7805 goes into current limiting.
I explained the action of said resistor in another message, but the
poster there said it was 15 ohms. If it's 18 Ohms, the principle is the
same, you jkust have to re-do the calculations.
Is everything actually OK, even though my VOM and
scope both agree the +5
rail is at +9.2 _WITH_ automobile light dummy loads?
What sort of car bulbs? What wattage?
-tony
4647
days inactive
5215
days old
68 comments
23 participants
participants (23)
-
ajp166＠verizon.net -
ard＠p850ug1.demon.co.uk -
bfranchuk＠jetnet.ab.ca -
bobalan＠sbcglobal.net -
cc＠informatik.uni-stuttgart.de -
cclist＠sydex.com -
chrise＠pobox.com -
classiccmp＠philpem.me.uk -
dkelvey＠hotmail.com -
hilpert＠cs.ubc.ca -
IanK＠vulcan.com -
james＠machineroom.info -
jkunz＠unixag-kl.fh-kl.de -
legalize＠xmission.com -
mcguire＠neurotica.com -
mike＠brickfieldspark.org -
pat＠computer-refuge.org -
pcw＠mesanet.com -
ploopster＠gmail.com -
rodsmallwood＠btconnect.com -
thrashbarg＠kaput.homeunix.org -
wdonzelli＠gmail.com -
wh.sudbrink＠verizon.net