Tubes in broken spotwelder & other questions

[et...@whidbey.com]

All,
I know almost nothing about tubes(AKA valves). My son bought a
capacitor discharge spotwelder that uses tubes. It doesn't work. Parts
of it do but on the whole it appears that it doesn't charge the caps.
There are three types of tubes in the thing.
3 each OA2WA voltage regulators
2 each 2D21 thyratrons
1 each 5U4GB rectifier
There are also 6 each physically large electrolytic caps that are,
I believe, rated for 450 volts. They are made by Sprague and are
marked 200-450. No uf markings though. Apparently a couple rattle when
shaken-probably not a good sign.
There are two transformers. The first has many taps and the second
is connected through a relay to the bank of caps. The secondary of
this xmfr is the spotwelder output.
As near as I can tell the circuit basically operates like so: The
first xmfr provides 450 volts AC which goes through the 5U4GB tube to
be rectified and then through three OA2WA tubes which are connected
in series. From there the power goes to the cap bank and charges the
caps. And from the caps through a big relay to the output xmfr.
The OA2WA tubes have only two of the pins connected. So power goes
in one pin and then out the other to the next OA2WA tube and so on. It
looks like there is no output from the last tube in the series. So
maybe this is why the caps won't charge.
At least one 2D21 thyratron seems to work because the relay that
connects the cap bank works when the machine is triggered.
I would like to know how to test the tubes without a tube tester.
Maybe this means I will need to make a tester. I also would like to
know how to test the big caps. My meter has a cap test function but it
doesn't seem to work with these big caps. Maybe this is because the
caps are high voltage types. Is it (fairly) safe to attempt charging
the caps with line voltage (120 volts) by first rectifying and then
passing the current through a resistor and then to the caps?
Here is a link to pics my son took of the guts of the spotwelder:
https://goo.gl/photos/LF4JKeGyA4YWiCwv9
Any comments about the spotwelder would be much appreciated.
Thanks,
Eric

[pf...@aol.com]

First, where are you? All the tubes you mention are quite common and not very expensive. I probably have all of them in my rather random connection. The bigger issue are those capacitors.

Please note the interpolations.

On Monday, November 14, 2016 at 1:34:55 PM UTC-5, et...@whidbey.com wrote:
> All,
> I know almost nothing about tubes(AKA valves). My son bought a
> capacitor discharge spotwelder that uses tubes. It doesn't work. Parts
> of it do but on the whole it appears that it doesn't charge the caps.
> There are three types of tubes in the thing.
> 3 each OA2WA voltage regulators
> 2 each 2D21 thyratrons
> 1 each 5U4GB rectifier
> There are also 6 each physically large electrolytic caps that are,
> I believe, rated for 450 volts. They are made by Sprague and are
> marked 200-450. No uf markings though. Apparently a couple rattle when
> shaken-probably not a good sign.

200uF @ 450V. You could pay anywhere from $15 to $100 for those caps. Given the application, do not cheap out, but get as high a quality as you can afford. You also have the choice to put caps in parallel to make the capacity - and series/parallel to make a higher working voltage. There is plenty of room given the size of modern caps to go this route.

> There are two transformers. The first has many taps and the second
> is connected through a relay to the bank of caps. The secondary of
> this xmfr is the spotwelder output.
> As near as I can tell the circuit basically operates like so: The
> first xmfr provides 450 volts AC which goes through the 5U4GB tube to
> be rectified and then through three OA2WA tubes which are connected
> in series. From there the power goes to the cap bank and charges the
> caps. And from the caps through a big relay to the output xmfr.
> The OA2WA tubes have only two of the pins connected. So power goes
> in one pin and then out the other to the next OA2WA tube and so on. It
> looks like there is no output from the last tube in the series. So
> maybe this is why the caps won't charge.

It is probably a blessing that your rectifier (5U4) tube does not work. Were you to be passing B+ - well, the results may be unhappy - see below.

> At least one 2D21 thyratron seems to work because the relay that
> connects the cap bank works when the machine is triggered.
> I would like to know how to test the tubes without a tube tester.

Other than for filament continuity and gross shorts (shorts on a cold tube), no you have no certain way to test tubes without a tester.

> Maybe this means I will need to make a tester. I also would like to
> know how to test the big caps. My meter has a cap test function but it
> doesn't seem to work with these big caps. Maybe this is because the
> caps are high voltage types. Is it (fairly) safe to attempt charging
> the caps with line voltage (120 volts) by first rectifying and then
> passing the current through a resistor and then to the caps?
> Here is a link to pics my son took of the guts of the spotwelder:
> https://goo.gl/photos/LF4JKeGyA4YWiCwv9
> Any comments about the spotwelder would be much appreciated.
> Thanks,
> Eric

The big caps are toast - full stop. DO NOT EVEN TRY to operate the tester without them as it is possible to do permanent damage to the transformers if the caps are shorted.

If you are anywhere near the Melrose Park section of Cheltenham Township (Philadelphia area) let me know. Despite the size and industrial application, this is a fairly basic issue to troubleshoot.

Now, showing my age, I used one of those ancient beasts some 40 years ago when I was working as a machinist. When set up properly, they are amazingly versatile. I used one to weld various thickness of inconel wave washers and pimple washers onto stainless steel seal backs for aircraft and nuclear applications. From stuff that was almost as thin as paper (0.004 thick) to stuff as thick as index card stock - dial in the machine, pass the pull-tests and it would hold the setting, effectively, until changed. There was a small stash of tubes above the machine - but no 2D21s. When one of them died one way, I found it at Leon Fertik's establishment for $1.50. I replaced them both and purchased two spares. The machine shop is still with us and thriving. Leon, sadly gave up his shop after his wife passed.

Best of luck with it - I can think of a lot of uses.... Especially if your son is an artist.

Peter Wieck
Melrose Park, PA

[et...@whidbey.com]

I'm north of Seattle on Whidbey Island so a visit would be unlikely.
My son has tried to power the thing up and even though parts do work
it won't do any welding. There is no odor of magic smoke so I think
the xmfrs are both OK. I told my son not to power it up any more until
we figure out what's wrong with the thing. The welder has two ranges:
0 to 20 watt seconds and 0 to 200 watt seconds. And my son has uses
for both ranges. Since I sorta talked him into getting a spot welder
like this one I kinda need to make sure we get it up and running.
Thanks for the advice and cap ID.
Eric

[John Robertson]

I suggest you consider PHOTOFLASH rated caps - they have a very fast
discharge (low internal ESR - you might want to get a Blue ESR meter to
test these caps) and aren't too expensive...

http://ca.mouser.com/Passive-Components/Capacitors/_/N-5g7r?Keyword=photoflash&FS=True

Look up REFORMING capacitors. Even old ones like yours may be saved,
there was a recent thesis published that showed how large value caps can
last a long time if they are reformed carefully.

2D21's are very old design - there are many circuits online to replace
them with SCRs if you want to bring the machine up to the 21st century...

Likewise your 5U4s, this can be replaced with a solid state bridge (dual
diode) rectifier. As the 5U4 has its own transformer then you need to
make sure it is fused correctly on the primary side in case of a shorted
diode.

OA2 tubes are 150VDC regulators, so with three of them the output is
regulated to 450VDC. If the OA2s are glowing purple then they are at
their regulated voltage (more or less) - all three need to be glowing
for your 450VDC to be regulated. I would expect them to be across the
output of the 5U4 to ground to regulate the output voltage...

John :-#)#

>
> It is probably a blessing that your rectifier (5U4) tube does not work. Were you to be passing B+ - well, the results may be unhappy - see below.
>
>> At least one 2D21 thyratron seems to work because the relay that
>> connects the cap bank works when the machine is triggered.
>> I would like to know how to test the tubes without a tube tester.
>
> Other than for filament continuity and gross shorts (shorts on a cold tube), no you have no certain way to test tubes without a tester.
>
>> Maybe this means I will need to make a tester. I also would like to
>> know how to test the big caps. My meter has a cap test function but it
>> doesn't seem to work with these big caps. Maybe this is because the
>> caps are high voltage types. Is it (fairly) safe to attempt charging
>> the caps with line voltage (120 volts) by first rectifying and then
>> passing the current through a resistor and then to the caps?
>> Here is a link to pics my son took of the guts of the spotwelder:
>> https://goo.gl/photos/LF4JKeGyA4YWiCwv9
>> Any comments about the spotwelder would be much appreciated.
>> Thanks,
>> Eric
>
> The big caps are toast - full stop. DO NOT EVEN TRY to operate the tester without them as it is possible to do permanent damage to the transformers if the caps are shorted.
>
> If you are anywhere near the Melrose Park section of Cheltenham Township (Philadelphia area) let me know. Despite the size and industrial application, this is a fairly basic issue to troubleshoot.
>
> Now, showing my age, I used one of those ancient beasts some 40 years ago when I was working as a machinist. When set up properly, they are amazingly versatile. I used one to weld various thickness of inconel wave washers and pimple washers onto stainless steel seal backs for aircraft and nuclear applications. From stuff that was almost as thin as paper (0.004 thick) to stuff as thick as index card stock - dial in the machine, pass the pull-tests and it would hold the setting, effectively, until changed. There was a small stash of tubes above the machine - but no 2D21s. When one of them died one way, I found it at Leon Fertik's establishment for $1.50. I replaced them both and purchased two spares. The machine shop is still with us and thriving. Leon, sadly gave up his shop after his wife passed.
>
> Best of luck with it - I can think of a lot of uses.... Especially if your son is an artist.
>
> Peter Wieck
> Melrose Park, PA
>

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
(604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."

[pf...@aol.com]

On Monday, November 14, 2016 at 2:48:20 PM UTC-5, et...@whidbey.com wrote:


> I'm north of Seattle on Whidbey Island so a visit would be unlikely.
> My son has tried to power the thing up and even though parts do work
> it won't do any welding. There is no odor of magic smoke so I think
> the xmfrs are both OK. I told my son not to power it up any more until
> we figure out what's wrong with the thing. The welder has two ranges:
> 0 to 20 watt seconds and 0 to 200 watt seconds. And my son has uses
> for both ranges. Since I sorta talked him into getting a spot welder
> like this one I kinda need to make sure we get it up and running.
> Thanks for the advice and cap ID.
> Eric

Yeah - that is a bit distant, for now. I did get a chance to visit one of the participants herein in his native habitat (central Texas) recently, so, stranger things have happened.

At this point:

a) Focus on the caps. Get them replaced with caps rated, either singly or in combination at 200uF @ 450V minimum. Some basic rules of thumb: Caps in series are additive for capacity, but not for voltage. 10 x 20uF @ 450V (IN PARALLEL) = 200uF @ 450V.

Two caps in series must be calculated, but keeping it very simple, if two 20uF @ 450V are connected in series, the actual capacitance will be 5uF @ 900V. You may see the calculations here: https://www.kitronik.co.uk/blog/how-to-calculate-capacitors-in-series-and-parallel/

b) Test the tubes by checking the filaments and for any dead shorts internally when cold. A base diagram will tell you which are the filament pins, and then any pin to any pin not internally connected (also given on the base diagram) for shorts.

There are lots of sources for tubes - if you get stuck, let me know and I will root through my spares box. I have 0A2s and 5U4s, most likely in some quantity. As to 2D21s I am not sure, but they are stocked at Antique Electronics Supply (AES) for $5 or so.

Then, your best friend is cleaning anything that can be cleaned without damage. Dust removal, correct lubrication if applicable, rust removal (Beware of steel wool!) and similar.

Best of luck with it - given that you are in the PNW, and, worse, on an island, make sure that you keep the system dry and free of salt. Given the operating parameters of that beast - you DO NOT want stray currents floating around.

[pf...@aol.com]

Forgive the top posting but WARNING: if a big honking electrolytic capacitor that is more than 60 years old is rattling when shaken, it is absolutely and irredeemingly, irrevocably TOAST. As are its siblings. Reforming is an exercise in futility, and possibly very dangerous, as if by some chance you do actually reform one (or more) for the moment, and it should fail under actual use, all the work previous would be for naught. **POW**.

Agree on the photoflash caps in any case. However, I suspect that the OEM caps are vastly oversize to make up for their rather slow chemistry.

Peter Wieck
Melrose Park, PA

[John Robertson]

On 201611/14/ 12:20 PM, pf...@aol.com wrote:
> Forgive the top posting but WARNING: if a big honking electrolytic capacitor that is more than 60 years old is rattling when shaken, it is absolutely and irredeemingly, irrevocably TOAST. As are its siblings. Reforming is an exercise in futility, and possibly very dangerous, as if by some chance you do actually reform one (or more) for the moment, and it should fail under actual use, all the work previous would be for naught. **POW**.
>
> Agree on the photoflash caps in any case. However, I suspect that the OEM caps are vastly oversize to make up for their rather slow chemistry.
>
> Peter Wieck
> Melrose Park, PA

I hear what you are saying Peter, but those caps are industrial grade
caps, not what you would find in a radio. Agreed the shake test will
tell if they are really bad (no rattles!), however attempting to reform
them is not an exercise in futility unless there are obvious signs of
venting around the rubber plug on the top of the can.

If you have access to an ESR meter that will give you a pretty good idea
of the capacitors condition. We sell the Bob Parker kit (as do others)
and using that with an ohmmeter to see if the low ESR is an actual short
circuit or not will give you a pretty good idea of the condition of
these electrolytic caps.

These would have been photoflash rated caps originally and they were
common in photographic gear from the 50s and on. My dad's architectural
photography studio had many flash boxes that used racks of those caps,
back in the 50s/60s and 70s.

John :-#)#

PS, I forgive the top-posting (ducking)

[pf...@aol.com]

On Monday, November 14, 2016 at 3:36:25 PM UTC-5, John Robertson wrote:
> On 201611/14/ 12:20 PM, pf...@aol.com wrote:
> > Forgive the top posting but WARNING: if a big honking electrolytic capacitor that is more than 60 years old is rattling when shaken, it is absolutely and irredeemingly, irrevocably TOAST. As are its siblings. Reforming is an exercise in futility, and possibly very dangerous, as if by some chance you do actually reform one (or more) for the moment, and it should fail under actual use, all the work previous would be for naught. **POW**.
> >
> > Agree on the photoflash caps in any case. However, I suspect that the OEM caps are vastly oversize to make up for their rather slow chemistry.
> >
> > Peter Wieck
> > Melrose Park, PA
>
> I hear what you are saying Peter, but those caps are industrial grade
> caps, not what you would find in a radio. Agreed the shake test will
> tell if they are really bad (no rattles!), however attempting to reform
> them is not an exercise in futility unless there are obvious signs of
> venting around the rubber plug on the top of the can.
>
> If you have access to an ESR meter that will give you a pretty good idea
> of the capacitors condition. We sell the Bob Parker kit (as do others)
> and using that with an ohmmeter to see if the low ESR is an actual short
> circuit or not will give you a pretty good idea of the condition of
> these electrolytic caps.
>
> These would have been photoflash rated caps originally and they were
> common in photographic gear from the 50s and on. My dad's architectural
> photography studio had many flash boxes that used racks of those caps,
> back in the 50s/60s and 70s.
>
> John :-#)#
>
> PS, I forgive the top-posting (ducking)

I keep "that" ESR meter, and it is extremely useful. I also have a neat little cap tester that actually tests caps at full operating voltage (up to 450V, that is). It is quite interesting to see caps that seem to test OK in the short term, but over time start to get *warmer* and *warmer* until that **POW** actually happens. It may take an hour or two... When I find one of these, I will hang it (connected) in a Home Depot bucket outside and wait ;-)>

My point is that a capacitor, even half-a-dozen capacitors are usually pretty cheap if measured against the alternative failing at a critical moment. These are not the typical cap found in the typical radio - there is a LOT of energy stuffed inside those beasts. Too many for me to trust a 50+ year old cage for containment.

[jurb...@gmail.com]

>"Two caps in series must be calculated, but keeping it very >simple, if two 20uF @ 450V are connected in series, the >actual capacitance will be 5uF @ 900V. "

That's not what I "heard". It should be 10 uF.

[pf...@aol.com]

On Monday, November 14, 2016 at 4:13:07 PM UTC-5, jurb...@gmail.com wrote:
> >"Two caps in series must be calculated, but keeping it very >simple, if two 20uF @ 450V are connected in series, the >actual capacitance will be 5uF @ 900V. "
>
> That's not what I "heard". It should be 10 uF.

(C1 x C2) / (C1 + C2) = Working Capacitance is the equation I have used these many years.

Example 1:

10 x 10 = 100
10 + 10 = 20
100 / 20 = 5

Example 2 (from the link):

10 x 22 = 220
10 + 22 = 32
220 / 32 = 6.875

At least where I come from.

[Ralph Mowery]

In article <6fc1c49e-9ee2-4471...@googlegroups.com>,
pf...@aol.com says...

If equal value capacitors are in series (same for resistors in parallel)
all you have to do is devide the value of one of them by the number.
That would be 20 devided by 2 for 10.

You can also take the long hard way and do the product over the sum as
in your example. You just need to LOOK AT THE VALUE IN THE EXAMPLE AS
BEING 20 INSTEAD OF THE 10 FOR EACH ONE THAT YOU USED.

For equal values of capacitance the voltage will add, if the voltages
are not equal then there are problems. Resistors are needed across the
capacitors when equal values of capacitors are used. This is mainly due
to the wide variation of capacitance in most of them. Helps to bleed
off the stored charge also.

[pf...@aol.com]

OK - I have enough stray caps at home to do a lash-up and I have enough very accurate meters to test the actual results of that lash-up. But, I just rebuilt the power-supply of my Dynaco ST-70 and my actual working results were as the equation implied, as I remember.

Nor would I ever run caps of differing voltage in series. Or parallel for that matter.

I will report my results.

Again, here is the link. I am NOT making this up.

https://www.kitronik.co.uk/blog/how-to-calculate-capacitors-in-series-and-parallel/

[Adrian Caspersz]

On 14/11/16 21:58, pf...@aol.com wrote:

> Nor would I ever run caps of differing voltage in series. Or parallel for that matter.
>
> I will report my results.
>
> Again, here is the link. I am NOT making this up.
>
> https://www.kitronik.co.uk/blog/how-to-calculate-capacitors-in-series-and-parallel/
>

Read Ralphs post again?

Example 1:

C1 = 20, C2 = 20

not 10 ...

--
Adrian C

[John Robertson]

The previous poster was simply pointing out that you made an error in
your math. You had a 10uf and a 20uf cap as your math components based
on the link, but were speaking of two 20uf caps in series in the
preceding text.

20*20=400
20+20=40
400/40 = 10uf

For two identical value caps in series divide be two for the resulting
capacitance. Voltage is added. If there are differences in the
capacitance between the two caps then the formula is needed.

Hey, we all make mistakes, even I've made one or two (a minute)...

John ;-#)#

[et...@whidbey.com]

I found 220 mfd 450 caps at JustRadio.com. I'm not sure if these will
discharge fast enough. I emailed the site and am waiting for an
answer. A quick search for photoflash caps turns up 300 volt caps, no
450 volt caps. So I'll wait to see if Just Radio thinks its caps are
up to the job. I have decided to just change out all the caps in the
thing if the correct main big caps are available at a reasonable
price. I don't think I want to go through the trouble of coming up
with solid state solutions and then rewiring the unit just yet. If it
worked that well in the past with tubes and if they have reasonable
lifetimes I'll just keep it the way it is. On the other hand, it is my
son's machine and he can do whatever he wants with it, no matter what
kind of help dad wants to give.
Eric

[Ralph Mowery]

In article <q5fk2cppqo2i0v69g...@4ax.com>,
et...@whidbey.com says...

>
> >> "Old pinballers never die, they just flip out."
> I found 220 mfd 450 caps at JustRadio.com. I'm not sure if these will

The 220 mfd will work fine where the 200 mfd caps are used. Many of
them have a large tolerance of 20 to 50 %or more.

[John Robertson]

Did you look at the Mouser link I included a few posts ago? They list
400 and 500VDC Photoflash caps. Mind you, you need to order one to two
hundred.

http://ca.mouser.com/Passive-Components/Capacitors/_/N-5g7r?Keyword=photoflash&FS=True

Here is a surplus place with photoflash caps, not cheap, but they are in
stock:

https://www.surplussales.com/capacitors/Electrolytics/eaf.html

However, if you use the part number you will probably find a stocking
distributor...

JustRadio sells good caps, but they aren't photoflash rated - pretty
safe bet on that!

John :-#)#

[mike]

Might be helpful to take a peek below the obvious.

Take a close look at the pix of the capacitors at the mouser link.
Compare the solder pins to the size of the buss bars in the welder.
At least you'll be able to say you thought about the consequences
of tiny pins and dubious solder connections. That compounds if you
go with fewer larger caps. All that energy into the weld
came from the caps. It's just as easy to melt the solder at that
end.

Don't go too crazy with the cap values. My spot welder is calibrated
to deliver energy based on the voltage and cap value.
If the cap is different, the calibration can be off by that ratio.
If your meter has a square-law scale, that can make a significant
difference in calibration at one end or the other. You can't
just add a resistor to fix it. But you can put a post-it note
on the panel and do the scale factor in your head.

If you don't have a schematic, create one. It only takes one
wrong assumption to make a lot of smoke.

Beware saturation in the output transformer. Mine has a core reset
pulse between firings. If that part of your system is broke,
you can saturate the core. At best, that can take your diagnostics
down the wrong path. At worst, more smoke that takes out the parts
you just fixed.

Are we having fun yet?

[whit3rd]

On Monday, November 14, 2016 at 2:44:55 PM UTC-8, et...@whidbey.com wrote:

> I found 220 mfd 450 caps at JustRadio.com. I'm not sure if these will
> discharge fast enough.

For a welding purpose, 100l milliseconds is very fast (no functioning
capacitor you are likely to find would have too high ESR for this job).

ESR gets to be important with low voltage high current, where a fraction of
a volt of ripple is a malfunction. Fraction of a volt difference on a 450V capacitor
charge is negligible.

[Clifford Heath]

On 15/11/16 12:45, whit3rd wrote:
> On Monday, November 14, 2016 at 2:44:55 PM UTC-8, et...@whidbey.com wrote:
>> I found 220 mfd 450 caps at JustRadio.com. I'm not sure if these will
>> discharge fast enough.
> For a welding purpose, 100l milliseconds is very fast (no functioning
> capacitor you are likely to find would have too high ESR for this job).

I agree that the speed doesn't matter, but low ESR caps have wiring that
will survive higher currents. No sense having a whopping cap
with internally-fused connections.

I also have a bench-top spot welder, from my father who was an
orthodontist. It has a 110V input (so we needed a transformer
from 240V), and that feeds via a small Variac into a selenium
rectifier. I suspect the caps need replacing (again - they were
last done 25 years ago) and I have a bunch of 300V photoflash
caps from disposable cameras that I hope will suffice.

The welding contacts are an anvil and an upper contact connected
to the foot pedal, via an adjustable pressure-operated switch
that fires the contacter. That leaves both hands free to hold and
position the work. When you press down hard enough, the thing
fires, very cute.

What I don't know is how to determine the maximum energy I can
dump through the output transformer without saturating it.
Anyone know how I can tell (other than just keeping the
capacitance below the original value)?

Clifford Heath.

[mike]

That's not a simple question.
The first order approximation is that CURRENT is what
creates the magnetic field that saturates the transformer.
If you put a scope current transformer on the primary,
load the secondary with the resistance and inductance of
your welding setup, crank up the voltage until you see
a sharp rise in the slope of the current pulse. That's
the too-much point. What happens when you add more caps
is dependent on the transformer characteristics.
Energy is linear in caps but quadratic in volts.

25 years ago, I was tasked with fixing an OEM forward
converter that would randomly self-destruct.
I built a fixture to synchronize the load transient with
the switching frequency. I could watch the primary current
head for the sky on the scope as the load transient crossed the
switching point.
Took it to the vendor site and proceeded
to blow up power supplies until they conceded that they had
a board layout problem to fix.

After the weld, the field will be sitting somewhere on the
B-H curve of the core. The next weld might be very dependent
on where you left it last time.

This is a page from my Unitek CD spot welder. Shows how they
reset the core.

http://imgur.com/ZeZerGx

I experimented with a microwave oven transformer battery tab
welder. I was hitting it with a timed pulse. Repeatability
was horrible.
When I synchronized the pulse with the line and gave it an
integral number of full cycles, the starting point on the
B-H curve was consistent and the welds got much more repeatable.

There was considerable discussion on exactly when you should terminate
the pulse, but mine was constrained to somewhere near zero current by
the triac.

[Clifford Heath]

On 15/11/16 17:56, mike wrote:
> On 11/14/2016 8:54 PM, Clifford Heath wrote:

snip

>> What I don't know is how to determine the maximum energy I can
>> dump through the output transformer without saturating it.
>> Anyone know how I can tell (other than just keeping the
>> capacitance below the original value)?
>>
>> Clifford Heath.
> That's not a simple question.
> The first order approximation is that CURRENT is what
> creates the magnetic field that saturates the transformer.
> If you put a scope current transformer on the primary,
> load the secondary with the resistance and inductance of
> your welding setup, crank up the voltage until you see
> a sharp rise in the slope of the current pulse. That's
> the too-much point. What happens when you add more caps
> is dependent on the transformer characteristics.

Thanks, that all makes sense. I'd like a bit more capacity
if I can get it.

I'll have a look and see if it has a core reset current too.

Clifford Heath.

[Chris Jones]

Be careful, remember there is a secondary current and its magnetic field
will largely cancel that of the primary, so looking at the primary
current is a red herring unless the secondary is open-circuit.

The rate of change of flux in the core will be proportional to the
primary voltage. What you need to look at is actually the integral of
primary voltage with respect to time, as this will give you the change
in the flux in the core. The longer the pulse lasts, the less voltage it
will take to cause saturation. (For similar reasons, 60Hz mains
transformers will saturate at a lower voltage if used on 50Hz mains.)
You could get more energy into the weld if you increse the voltage of
the primary pulse and decrease the duration, because the dissipation in
the weld is proportional to the integral of v sqared with respect to
time, whereas the change in the flux in the core is proportional to the
integral of v (not squared) with respect to time. Also, make sure that
the core starts out with maximum flux in the opposite direction. This
might be achieved with a reset pulse as others have pointed out,
provided the core material is one that will retain the magnetisation. If
not, you could perhaps feed a reverse current through the primary using
a low voltage power supply, to build up reversed flux just before the
main welding pulse. That might make the switching a bit more complicated.

[et...@whidbey.com]

On Mon, 14 Nov 2016 17:21:30 -0800, John Robertson <sp...@flippers.com>
wrote:

I can't use 100 caps. I did find photoflash caps but the leads are
kinda thin. I'm still waiting to see if Just Radio thinks their caps
will do.
Eric

[et...@whidbey.com]

On Mon, 14 Nov 2016 17:45:28 -0800 (PST), whit3rd <whi...@gmail.com>
wrote:

So photoflash rating is probably not necessary?
Thanks,
Eric

[et...@whidbey.com]

Greetings Mike,
The machine I am working with is a Unitek model 1-048-03. Is this the
same as yours? If not do you know if it is close enough to use for
trouble shooting and operating the machine I am trying to get working?
If it would be useful is there some way you could get a copy to me?
Maybe scan and email? I don't mind paying you for your trouble. I
wouldn't even ask but so far I have not been able to find a manual of
any type for the model my son has.
Thanks,
Eric

[mike]

It's a unitek 125.
All I have is a very poor schematic and three pages of
part of the cal procedure.

http://i.imgur.com/vSg8agC.png
http://i.imgur.com/mrQIq1f.png
http://i.imgur.com/lE2DroY.png
http://i.imgur.com/8Nv2KpU.png
http://i.imgur.com/TZyfuLp.png

If you have a way to archive them permanently, please do so.
They won't stay on my page for long.

This is my DIY battery tab welding fixture made out of an arbor press.

http://i.imgur.com/er1BqSb.jpg
http://i.imgur.com/OeQZdWH.jpg

I emailed Unitek years ago. At that time, they were very helpful.

[Geo]

On Mon, 14 Nov 2016 10:38:12 -0800, et...@whidbey.com wrote:

<snip>

> The OA2WA tubes have only two of the pins connected. So power goes
>in one pin and then out the other to the next OA2WA tube and so on. It
>looks like there is no output from the last tube in the series. So
>maybe this is why the caps won't charge.

<snip>

I cannot see properly from the photos - but is there a connection to
the last (bottom) regulator cathode?
I would guess it might be to chassis so "no output" is correct.

If all three glow then the voltage across each should be fine.

[pf...@aol.com]

OK!! EVERYBODY!!

Step back a minute and have a look at what is being discussed here:

a) This is a TUBE-based spot welder designed probably more than 60 years ago.
b) The bean counters (a scourge even then) would NEVER have specified photo-flash capacitors when just a bit more of conventional capacitance would address 'slow chemistry' at a fraction of the cost.
c) A healthy gap on the transformer would eliminate sateuration issues in this application - 'at no additional cost' and without complications, and entirely permanently.
d) This is a piece of industrial equipment designed to be used by production worker who very likely have neither the time nor the inclination to do anything other than follow directions, set the various knobs and switches then GO for hours at a time.
e) And, as with any decent tool from any decent manufacturer, it is designed to be serviced in a simple and straightforward manner by techs who need do no more than follow directions from a standard troubleshooting schedule.
f) And, finally on this list, replacement parts, especially generics (capacitors, tubes) will be entirely conventional - no silliness about "matched pairs" for instance, or boutique caps. If one has a small machine shop in Bucksnort, Tennessee ( https://en.wikipedia.org/wiki/Bucksnort,_Tennessee ) one does not have the time to search down something exotic. One is going to get on the phone to Graybar and order whatever it says on the can, and Graybar will have it in stock. http://www.graybar.com/ And, back in the day Graybar stocked industrial tubes as well.

Meaning that there is NOTHING esoteric about this device, and it will not require ANY special 'stuff' to make it go. The caps in this device are toast. If even one of them "rattles" then all of them, being of similar age, living in similar environments and doing similar work, are also done. Not replacing all of them is just plain stupid. Full Stop.

Purchase physically rugged caps that either are, or add up to, 200uF & 450 VDC per each, at least. Make sure that the soldering points and connections can handle anticipated voltages and currents. Replace all the tubes, even if they are not necessarily done, as now you will be starting fresh, nor is the investment all that much. If I have 2D21s, I would send the whole shebang to Oregon/Washington State for US$30 including shipping and think I was making too much. Otherwise, AES will do the same for perhaps half-again as much.

Sheesh, guys and gals - it is almost as if this were rocket science. It ain't nohow.

[et...@whidbey.com]

On Wed, 16 Nov 2016 09:25:37 +0000, Geo <nhhu...@dea.spamcon.org>
wrote:

There is a connection. It does not go to the chassis. I think the tube
is bad. All tubes are gonna be replaced anyway.
Eric

[et...@whidbey.com]

That's it, I'm convinced. I'm gonna order the caps from JustRadio.
They are the correct rating for voltage and capacitance, different
connections-not screws, but big tabs instead. Easily soldered to the
bus bars. I have been thinking about those big bus bars and am
wondering if the size is either for mechanical reasons or lower
resistance reasons. I can't see the original caps dumping enough
current to need such robust bars. Especially considering the relay
contacts connected to the cap bank and the much lighter wires running
from the bus bars and relay to the xmfr. Lemme know if you have all
the tubes and I'll send a check.
Thanks,
Eric

[pf...@aol.com]

On Wednesday, November 16, 2016 at 12:06:15 PM UTC-5, et...@whidbey.com wrote:

Please note the interpolations:

> That's it, I'm convinced. I'm gonna order the caps from JustRadio.
> They are the correct rating for voltage and capacitance, different
> connections-not screws, but big tabs instead. Easily soldered to the
> bus bars. I have been thinking about those big bus bars and am
> wondering if the size is either for mechanical reasons or lower
> resistance reasons.

Both. Mechanical as this machine moves, vibrates and is subject to various stresses thereby. And, consider that the shop I worked in a _long_ time ago used their welder two 10-hour shifts in a row, about 2 days per week. That is a lot of welds.

Electrical as the enemy of these machines is heat. Heat increases resistance which can throw off settings - and with 0.003 inconel pimple washers the difference between the weld not taking and a burn-through isn't much, and on aircraft and/or nuclear rated parts that is unacceptable.

I can't see the original caps dumping enough
> current to need such robust bars. Especially considering the relay
> contacts connected to the cap bank and the much lighter wires running
> from the bus bars and relay to the xmfr. Lemme know if you have all
> the tubes and I'll send a check.

You will have to wait into the weekend for me to get an edgewise to check my stock. Our youngest cat (3 years old, 18 pound Maine Coon mix) has just discovered that he can use the dog door. So, my primary task is to cat-proof the dog yard so he may go outside, but not any further. We may be suburban, but we have foxes, raccoons, other cats and cars to contend with.

As to sending a check, that is not how it works. If I have the tubes, I will send them to you - and then *if* they work out, you may send the check.

Peter Wieck
Melrose p

[John Robertson]

Keep your old tubes until you can test them and be sure they are bad.
These old tubes are not getting any more common! 2D21s were always a bit
fussy and were replaced by 2050 tubes not long after they came
out...even new out of the box there were failures. Seeburg jukeboxes
used 2D21s for their early tormat design and quickly changed the design
to 2050s part way through the second year of production due to read-out
and write-in issues. I have schematics for 2D21 testers that I can dig
up (probably find similar schematics easily enough online) and post of
my tech ftp pages if there is a demand/request. The 2D21 is pretty easy
to test, if you have access to a good tube tester it will usually show
them on the Seldom Used auxiliary tube list. The 2D21s are go/no-go
kinda things.

OA2s are also pretty basic tubes. They use a conductive gas to regulate
to 150VDC. If the tube is not broken or gassy then it should be just
fine. Check the Getter flash in the tube - it should not be showing any
white powder, that is a tube that has lost its seal and oxygen has
entered...

John :-#)#

[pf...@aol.com]

On Wednesday, November 16, 2016 at 1:04:24 PM UTC-5, John Robertson wrote:

> Keep your old tubes until you can test them and be sure they are bad.
> These old tubes are not getting any more common! 2D21s were always a bit
> fussy and were replaced by 2050 tubes not long after they came
> out...even new out of the box there were failures. Seeburg jukeboxes
> used 2D21s for their early tormat design and quickly changed the design
> to 2050s part way through the second year of production due to read-out
> and write-in issues. I have schematics for 2D21 testers that I can dig
> up (probably find similar schematics easily enough online) and post of
> my tech ftp pages if there is a demand/request. The 2D21 is pretty easy
> to test, if you have access to a good tube tester it will usually show
> them on the Seldom Used auxiliary tube list. The 2D21s are go/no-go
> kinda things.
>
> OA2s are also pretty basic tubes. They use a conductive gas to regulate
> to 150VDC. If the tube is not broken or gassy then it should be just
> fine. Check the Getter flash in the tube - it should not be showing any
> white powder, that is a tube that has lost its seal and oxygen has
> entered...

2D21s are/were used commonly in dental X-ray and Fluoroscope machines, and when, also back in the day, I was working for a GC doing some serious rebuilding at HUP in Philadelphia, one of the docs needed to resurrect such a machine for historical purposes. I found him the tubes he needed locally, including two 2D21s. This was "Before Internet", so not an instantaneous process.

I also keep a very good tester (Hickok 539B), so no tube would leave this house untested.

Yes, of course, keep any undamaged old tube.

[Clifford Heath]

On 17/11/16 00:35, pf...@aol.com wrote:
> OK!! EVERYBODY!!
> Step back a minute and have a look at what is being discussed here:

> Sheesh, guys and gals - it is almost as if this were rocket science.

You're responding to Chris' answer to *my* question about how I
can make *my* welder do more than it was designed to. I'm thankful
to him for that, and I don't think he needs to "step back a minute".

[pf...@aol.com]

No, I made a generic suggestion that we not over-analyze what is, in actuality, a very simple issue. A discussion about cap-reforming has no place in this particular discussion, nor do photo-flash caps except as a perhaps interesting sidebar. A "you might try this if practical" type thing. And at that level, I agree, one might try it.

This discussion, for the OP in any case, is about restoring a piece of industrial equipment back to reliable and optimal function. Emphatically Not rocket science.

Now, here is the issue with industrial tools with the goal of making it 'do more than it was designed to do'. Well designed tools have a level of resilience already built in. Going beyond those parameters puts the tool, operator and surroundings at-risk either a little bit, or perhaps quite a bit depending on the tool. It is, in every case, under every condition, an exceedingly bad idea. If you choose to put yourself and your property at-risk, that is your privilege - and may you never be in line for a Darwin Award.

I have worked in an industrial production environment with very fast, very powerful, very hot machines, materials and moving parts in close proximity. I would NOT want to think that some lashed-up tool operating outside its design parameters had my life or health dependent upon its smooth operation.

[Chris Jones]

On 18/11/2016 02:31, pf...@aol.com wrote:
> Well designed tools have a level of resilience already built in.
> Going beyond those parameters puts the tool, operator and
> surroundings at-risk either a little bit, or perhaps quite a bit
> depending on the tool. It is, in every case, under every condition,
> an exceedingly bad idea.

meh. That is quite a generalisation. Sometimes upgrades and
modifications make things safer, especially when starting with older
equipment that, in its original condition, would not meet modern safety
standards. Of course one has to apply sensible engineering, and not
overstress parts, and consider any safety consequences of parts failing,
but that does not necessarily preclude improving performance in some
cases, as the original designers of the equipment did not have the
components available to them that we do now.

[pf...@aol.com]

On Friday, November 18, 2016 at 7:02:02 AM UTC-5, Chris Jones wrote:

> meh. That is quite a generalisation. Sometimes upgrades and
> modifications make things safer, especially when starting with older
> equipment that, in its original condition, would not meet modern safety
> standards. Of course one has to apply sensible engineering, and not
> overstress parts, and consider any safety consequences of parts failing,
> but that does not necessarily preclude improving performance in some
> cases, as the original designers of the equipment did not have the
> components available to them that we do now.

Sure. But, look at what you just stated. Sensible engineering. Not willy-nilly speculation by individuals with no or very limited basic knowledge of the design parameters of the tool/machine involved.

"more than they were designed to do" is not sensible engineering.

Better bearings, better capacitors, better lubricants, more effective switches, better snubbers (rubber parts), better insulation, better electrodes, better shielding - all of the above are perfectly valid (and sensible) improvements for about anything. One does not have to use 30w non-detergent oil in a vintage automobile, or propylene glycol antifreeze today.

But, at the same time, running a stock Model T on nitro-methane is, perhaps, inadvisable.

[Ron D.]

I totally agree. I rebuilt a nearly 40 YO gas dryer:

I upgraded the drum lighting to LED. I upgraded the kind of thrust bearings. I added two thrust bearings. I bought a new old stock bearing assembly for the fan and upgraded the grease. I painted and replaced the new nylon hinges on the filter door. Nylon and heat don't like each other. I painted the rusty side that was next to the washer. I replaced the 40 YO drum belt. I replaced the blower belt. I upgraded/replaced, made from scratch a duct gasket. The thrust bearings were $24.00 not including postage.

I changed a set screw to brass tipped so the pulley would slip if the bearing housing froze. Now a cool upgrade would be a belt slip detector.

All of these are minor upgrades, but they make it better.

I have a few more planned:

1) Make a new lint filter
2) Make it easier to do preventative maintenance by adding Threadserts instead of sheet metal screws.

The upgrades were costly. The grease was about $30.00. The material to make the filter was about $70.00, but I'll use that also to make the lint filter. The threadsert kit was about $80.00 USD.

I will try to measure and determine the type of wire in the ignitor and maybe eventually make a replacement. I have two brand new ones.