Discussion of welding with JCP y09m06d23-26
UPDATE THIS PAGE, It has new entries all the time!
Home          

--------------------------------
y09m06d23

12:50pm
7:13 JCP asks why one wheel heats more, answer probably due to grounds.

1:10pm
6:07 Jeff measured with an ohmmeter, everything measures continuity to base plate of fixture. Need to discuss how 1 Ohm is not much of a path, 100 MicroOhms is, but looks same on a standard Ohmmeter. Need to define "continuity" and "resistance" measurements. Jeff will isolate ground on shock w tape, for a test.

5:10pm
8:05 Heat pattern still unbalanced even with shock insulated from tooling with tape. I recommended two Miyachis one on each shunt to each wheel, as the current seems imbalanced. JCP says the CT will not reach around the entire shunt, so I recommended taping it in place as far around as you can go, because we are after only a relative comparison between the two. After the two coils are in place, switch meters on the plug-in coils to make sure any imbalance indication is not due to wide variance in meter's calibration, which is not uncommon.

5:50pm
10:03 Joe and Jeff, wear is uneven & knurl pattern differs from one side to the other. Knurled wheels are not being changed in pairs, and you can see that in the welded seam, one will have a clear knurl pattern, the other is irregular with no knurl pattern. I recommend they do, because both wheels need to be treated exactly the same. Heat marks go toward cap on one wheel, and toward tube on other wheel, but shock is not being held off-axis.
JCP asks; "What if tube were longer on some, so that heat line is not right near end of overlapped tube/cap, would this make it colder? Answer, a little, but should not be noticeable.
Joe mentioned that "porosity exists, could it be the flood cooling?" (No), Oil? (Maybe) He will clean the oil and see if it stops.
Jeff will send photos of tooling for better discussions on the way the part is processed.
----------------------------------------31 mins 28 secs
y09m06d24

10:10am
4:27 Jeff wanted explanation of rim drive. First purpose was to drive any size wheel at a constant surface speed, therefore giving the same seam speed from wheel to wheel. The second purpose came along as an extra, the knurled drive dresses the wheel and keeps the inpurities to a minimum, keeping the surface resistance low.
A high surface resistance wastes heat rather than putting it in the faying surface where it belongs. The flood cooling is there to remove the unwanted surface heat, as it leads to expulsion, wheel contamination and wear.
Jeff will send photos now.

10:56am
15:15 Got two photos, transferred to graphics program to rotate and magnify. The upper and lower wheels and rim drives are apparent, even though rather blurry. There is a green cup that might hold the cap.
Q to Jeff: Is that cup isolated from the machine electrically?
There is a sliver tube that might hold the shock body, same Q to Jeff, is it isolated electrically?
Second picture, is this the loading operation? If that silver tube touches the shock, it cannot allow a path to the machine base or the mounting provisions, current will flow there rather than through the other wheel.

11:48am
8:20 Jeff says the wheels are cut at a 15 degree angle, even though the wheel is perpendicular to the shock. This means that the contact area changes drastically (1/4" to 1/2") as the wheel wears. This is not acceptable.
Also it is placed to overlap the cap (Which is the outside tube) as if they were trying to make a fillet weld with an arc welder. This is unacceptable, as it amounts to an edge weld.
This is not arc welding, you cannot make an edge weld, just like you cannot make one in spot welding.
It is OK to favor the seam placement to the side that is less likely to produce expulsion into the assy, but it is not OK to make an edge weld.
It is ideal to dress the wheels to give a flat surface, about 1/8" to 1/4" to contact the tube surface, but when this widens even by 10% it drastically reduces the current concentration, and must be either redressed or the schedule must be adjusted, usually with more current.
Cutting the face at an angle is just asking for trouble in this regard.
Seam wheels of different diameters is acceptable, that is the purpose of the rim drive, to eliminate that as a variable.

30:00 Time spent posting this.


-------------------------- +58:02 = 89min 30sec
y09m06d25

9:38am
7:45 Jeff needs info I sent him by email sent to Joe, asap. Discussion about the need for a correct turns ratio. It is not needed for % heat commands.
It is needed for kA heat commands, and it must be set by verifying a setting with an external meter, and adjusting the ratio until the current requested matches the actual current on the external meter.
"They are reading the current in the display" In that case it won't be accurate until you calibrate it with a proper turns ratio.

5:30 Sending info Jeff requested to Joe.

10:45am
8:05 Discussed weld schedule, wheels are lifting before weld is done, blowing holes, suggested inserting the indented items as below:

On valve 1
    On weld in progress
Squeeze 5
Impulse 3heat 2cool
72 imp at 85%
Process weld faults
Off weld in prog
    Off valve 1
On weld complete
Hold 3
Off weld complete

The length of the seam is controlled by the number of Impulses, shown as 72, above. Jeff says the sheet shows "45", so he will change that to 45.
Force was 350, was able to bump it up to 850, which should be OK. Jeff was able to find each heat command by using "Insert" and scrolling through the offerings.
One thing I didn't mention is that Jeff might bump the last couple of impulses up to accomplish a hotter finish as it approaches the beginning of the seam. This heat adjustment may be necessary because the previously welded area will shunt current from the seam in progress when it is closer than about 1/2" away.
To accomplish this, cut down on the number of impulses and add the following:
Impulse 3heat 2cool
02 imp at 95%
Add this right after the original impulse commands, after the existing line 5 and before "Process faults".

Jeff do you read this? You did, I got your email of 6-25, 7:43pm. Thanks
------------------------------------------------------------------------------------


Note for seam welds, use forces similar to this:
MILD STEEL
GMT RANGEFORCE
0.65-1.02480 LBS2.13 kN
1.03-1.44670 LBS2.98 kN
1.45-1.84950 LBS4.22 kN
1.85-2.451200 LBS5.33 kN
2.46-3.071660 LBS7.38 kN
3.08-3.501900 LBS8.44 kN

And the width of the seam should be about 5 to 6mm, any wider is excessive, causing a sloppy mess, just like too big of a spotweld. Cut the wheels down to that width and make them squarely meet the surface of the tube.

All the same things you would do for a spot weld, no angled tips, no edge welds, no huge wide tips, all that.
The wheel is to be treated as a weld tip, not as a hot iron melting the edge of the tube at an angle, sheesh...
Also the width is important, the wheel must be changed if it grows by much, say 1mm, just like spot weld tips must be changed.

BIG important tip --- Silver plate the surface where the wheel bolts to the hub, and silver plate the hub.
It is a copper-to-copper connection, and absolutely needs this. It is a standard in the automotive world to silver plate connections, equipment that does not have this is unacceptable.

You can look on the rest of this site for sources of the silver plating powder called Cool-Amp. Hit the "Home" button, then select "Links" from the index. They are in the middle of the page, row 7, on the right. No I don't get commission, thanks for asking... Home
Use the backspace key to return here.

The ingredients are harmless, most of them have been used to promote healing in wound dressings, how harmful is that?
VERY Important for consistent heats, especially since you have no current regulation.
VERY Important to keep from burning up the small mounting screws, which will try to carry the current around an oxidized connection.
Since they won't have Cool-Amp, be sure you scotch-brite all wheel mating surfaces until they get it. Make them take the wheel off and do this before you start on the machine, VERY Important.

(Did I mention that this was VERY Important?)

Here is a chart that shows tip size (same as spot size) based on thickness, and I'll bet the tubes are 1 to 1.5mm:
MIN. WELD SIZE
THICKNESSBUTTON DIA.
0.40 mm - 0.59 mm3.0 MM
0.60 mm - 0.79 mm3.5 MM
0.80 mm - 1.39 mm4.0 MM
1.40 mm - 1.99 mm4.5 MM
2.00 mm - 2.49 mm5.0 MM
2.50 mm - 2.99 mm5.5 MM
3.00 mm - 3.99 mm6.0 MM
3.50 mm - 3.99 mm6.5 MM
4.00 mm - 4.50 mm7.0 MM

It is OK to setup the tip size ("wheel width" in this case) 20% greater than these minimums, just like we do for spotweld tips.

1:45:00 spent compiling and posting this.
               

Talk to you Friday...

--------------------------- +2hr 6min 20sec = 3hr 4min 22sec
y09m06d26
3:40pm
14:54
Transformer of 250KVA keeps blowing up, using tap 8, 11.3kA 11.08 volts secondary, 3 on 3 off.
This is 125kVA, and is 50% duty cycle. Transformers are rated at 50% duty cycle also, so this is being used at half its thermal capacity. This means the problem is not thermal... hmm.
Could we be saturating the transformer? Impulse welding will saturate a transformer. It is a bit complicated but this is what Dietrick Roth of Roman Transformer used to say...
There is an imbalanced amount of time for the flux to relax, because each "plus" half cycle has a "minus" half cycle following it. But every other "minus" half cycle has a long time (cool time) to relax.
This causes a magnetic offset, adding up to a big offset, saturating the transformer.
This is why a seam welder traditionally uses a stacked core transformer with lots of iron. Wound cores saturate too easily.

The current measures 2000 amps higher on one wheel than the other, that is 20% difference, so current is going somewhere, rather than to the other wheel. A ground is a likely culprit, but complete disassembly of the secondary would be required to find it.
They insist on the 15 degree offset on the face of the wheel, to shape the material into the body. Jeff made a weld with the proper, flat face, and it was "phenominal", but they say, "everybody does it this way, so that's why we do it."
This is the same thing I was told about welding the high carbon rods with chrome on it. Some things are just wrong, no matter what the perception of the maintenance person, you can't negotiate with science, it is not right to fool Mother Nature.
I have no answer for how to enable the customer to do something against science.

Drawings I posted above show that an "off-angle" wheel will cause all kinds of problems, such as:
Edge welds, allowing expulsion.
Wheel touching the main body, shunting some unknown current around the seam.
Contact area of a 15 degree off-angle face on the edge of the tube causes extremely concentrated current density at the surface, where you don't want the heat.
If "everybody is doing it this way", then they are all making leakers. I ain't buying it...

6:03pm
10:03
Not getting anywhere, transformer only putting out about half the current it should, compared to another machine. I suggested to look at the "LV Center" or "Line Nominal" setting in the Medar. This is needed so it can regulate the % heat. If the nominal is 480, and it sees the actual voltage is higher, it subtracts from the percentage to give the same heat it would if the line voltage were at the "nominal" setting. It is limited to (probably) a 40% adjustment.
If the "Nominal" is set to "000" then it will subtract heat from all settings, as every voltage is higher than "nominal".
Jeff will check the setting.

Should also see what the primary current is, to see what power is going into the transformer, should be about the same KA as that coming out.

--------------------------- +24min 57sec = 3hr 29min 19sec
y09m06d30
10:09am
25:00
Discussed calibration of current. It is required if you want to trust the reading of secondary weld current after a weld.
Multiple attempts to adjust the reading with the turns ratio has failed to make much of a difference in the readings
on the Medar after a weld. The ratios are all setup at "100:1", but there were numbers entered in the schedules.
Any number for turns ratio entered in the schedule overrides the one in the Setup Parameters. So Jeff will change it in the
schedules. Must remember that each schedule will have to have the correct ratio entered for whatever TAP it
is intended to be used on. ALSO REMEMBER, if the tap switch is changed, the calibration is now wrong again.
It would be OK to calculate and post a set of turns ratio numbers for the available taps, then, through diligent
care in entering schedules, and turns ratios to match the tap settings, and vice-versa, we can keep the welder calibrated.

Discussed the meter required for secondary measurements, they were going to order one at $6000, which
had a large coil, printed out sinewaves and charts, loads of numbers too, but the BF Entron
meter can have a large coil ordered with it, and it can develop all the charts you want as it downloads to a computer
with a serial port. It also has a lot of storage for seam weld review. It is about 1/3rd the price.

Discussed settings from another plant with "Upslope" in the seam weld. I think this means
"Upsloped Impulse Welding" as it was still some cycles heat and some cool (I think we discussed 3 ON, 3 OFF)
If you had three impulses that were upsloped, it just means one impulse at the starting heat, the next at
a heat between starting and ending settings, then one at the ending heat. So multiple Heat-Cool entries could
be used to adjust heat as needed during a seam. The Medars will allow this, and the Entrons will do it through
"chaining", which allows jumping from the end of one schedule to the start of another.

BUT, the changes in heat require a lot of "fine-tuning", where it is noticed that most every time at a certain spot
in the seam, it seems too cold or too hot. A cut-and-etch of the seam longitudinally would be very helpful in
revealing hot and cold areas to fine-tune.

Some of the factors that cause the heat to be uneven are:
- The end of the seam might appear cold because the previously welded part will shunt current from the seam.
- Cold steel is a good conductor, with 1" typically being about 100 MicroOhms.
- Hot steel's resistance could be as much as 4 times higher, meaning that cold steel shunts more current.
- The unwelded spot between tips (or wheels) measures about 200 MicroOhms from tip to tip.
- When the nugget forms, the resistance drops to 50 MicroOhms, then rises as more heat is added to 100 MicroOhms.
- When the weld cools, the welds resistance drops to 10 MicroOhms.
Knowing all of this, it is easy to see that changing conditions will exist from the start of the seam to the end.
This is where the bumping up or down of the current is used to overcome these differences in heat.
Basically we want to be just under expulsion for the whole length of the seam.

--------------------------- +25min = 3hr 54min 19sec

y09m07d27
10:22am
2:55
No current on a Medar 3005. Just rebuilt secondary. One connection looks suspicious but a meter shows "continuity" through it.
BUT - You can't trust a meter to show the kind of continuity you need for welding current.
A meter may show only one ohm, but this value is 1000 times too high to pass welding current. You need a meter that can show a thousandth of an ohm, and it would have to show "zero" to insure you have a good connection, if it showed "one" that is too much resistance to pass current.

What you really need is a micro-ohm meter that shows millionths of an ohm, and it would have to show a complete path from one welding wheel, to the other of about 500 microOhms. Any more than that would seriously interfere with current flow to the point that you won't weld.

If you were to achieve 500 microOhms, that is one half of a thousandth of an Ohm. Since a meter doesn't read one ohm very good, and you need to read one thousandth of that, you are just out of luck reading it with an ohmmeter.
Here is one ohm:1.000 000 OhmsLowest your VOM goes
Normal secondary: .000 500 OhmsSame as 500 microOhms
One thousandth Ohm:.001 000 OhmsTwice as much resistance as you can have.


These small numbers are, unfortunately, beyond human imagination, but never-the-less absolutely true. I wish I could tell you how to measure the secondary with your ohmmeter, but you have about the same chance of doing that as you have of riding a bicycle to the moon!

10:34am
6:00
Discussed what would cause the current to be low, and the possibilities are:
1. Not firing the SCR
2. Open primary (The high voltage side of the transformer)
3. Open secondary (The low voltage side)

So we can read the Last Weld Data, and figure out what it is, so what was the last weld data?
Answer: Line = 480 something, Secondary current = 170 A, Power Facotor = 2
From this we know that the SCR fired, because we saw some current, and the low power factor shows that there was a lot of inductance, so the primary must not be open. So the secondary is open.
In order to verify this, we need to put a wire on the output lugs of the transformer, but you say you can't reach them so we will put the wire on the shunts, one of them to the other. The wire should explode and burn in half, and that shows power is geting this far.

10:50am
3:00
But the wire didn't blow open, so the test showed the secondary is open between the place you put the shorting wire and the transformer lugs. The most probable place is a connection where a shunt bolts to a lug, or something like that.
Important to see that the last weld data changed, because if it is the same, it means you didn't even initiate the welder. And the new info showed the same low PF = 2, but a new secondary current of 180. Small difference, but it does show that the control fired. Also you got a fault "No Current Limit".

11:13am
3:36
So we need to put a wire right on the transformer lugs, but you say you can't reach them.
What would I do? Well IF I had a microOhmmeter, I would measure the resistance shunt-to-shunt. Normal would be 200 to 500 microOhma, but I would expect to see 50 times that, because your 180 Amps is 1/50th of what normal current should be. So I would expect to read 15,000 microOhms, which by the way would be 0.015 Ohms, and there is no way a VOM or continuity checker could show the difference between this and 1.000 Ohms!
However, lacking such a meter reduces the choices down to finding a way to get to those lugs.

3:23pm
14:12
Does one shunt being larger than the other caise one wheel to have more current than the other?
NO!, there should be one path for current ONLY. Current should flow out of one lug on the transformer, through the shunt, the wheel through the part and through the other wheel, the other shun and the other lug of the transformer. The current would be the same through the entire circuit. It would be just like a hydraulic circuit, what ever the pump puts out, it must pull back in.

This is the way the current should flow, there is no way you would read a different current from one wheel to the other.
BUT, you say the lower wheel has more current than the upper wheel. Here is how this could happen, we have discussed this before, but this time I am trying to explain this with an illustration, because it just didn't leave a lasting impression, and it may be easier to explain to those you are working with.
Lets put a jumper cable between the part and the upper lug, like this:

You can see that the lower wheel is handling current for two paths, so it would handle more current. It would show a larger heat mark on the lower seam weld also.
But you don't have such a jumper, right? Maybe, what if the upper lug were grounded, as usually one of the two lugs are grounded, as it is a "safety" requirement, or at least thought to be...(There are much better ways to provide safety than this, another subject...)
Add to this, the fact that the tooling may be grounded, as it usually is, and nobody ever puts much of a thought to it. Often times, the ground is a poor connection, so it goes unnoticed. But lets say it becomes a better connection, it will act just like the hypothetical jumper we introduced above.
My point in all this is you MUST remove the grounds, no matter whether anybody you are working with understands or not. Obviously they don't because we discussed this last time, and here you are dealing with it again. They MUST come to understand these things, or we will never be able to get their machines running very reliably. This is the very reason all welding should be done in plants where there are experienced "WEMRS" so we don't have to retrain every time there is an unusual problem. Technical expertise is only recognized by the best of personnel managers, once you find that rare bird, we can start to be understood.

4:56pm
3:40
What does it mean when the fault "Initiate not present" is displayed?
Answer, the initiate signal started the control, or initiated it, but it was gone before the squeeze time was over. This is common where just a short pulse is used to initiate the control, and it does no harm.

The current you have after their work is 9000Amps, but you are at 95%, so it seems that there is still some "high" resistance. Remember that "high" is only just a few dozen microOhms, and without the meter, it will be difficult to find. So I will send meter info by email, normally these meters are $3000 to $6000 so you will be quite surprised that I found it for $200!
This summary took 2 hours to do. Now to research the meter and get you an email.

8:00pm
46:00
Research source for MicroOhmmeter and send info to JCP
--------------------------- +78min 43sec = 5hrs 13min 2sec
y09m08d18
5:00pm
46:00
Write up of MicroOhmmeter usage, emailed to JCP.
---------------------------- +46 min = 5hrs 59mins 2 sec
------------------------------------------------------------------------------------------------------------------------------

---------------------------
y09m08d31

3:12pm
6:15
Got big explosion when a bracket fell between the shunts or somewhere on the secondary. Now when the breaker is turned on, it seems to come on 100% power, what is wrong?
Could be a shorted SCR, remove the two pairs of signal wires going to the two SCRs, usually Red/White tristed pairs. Jeff says "Red Yellow", and they are on a connector. OK unplug the connector, and take off one of the big heavy wires on the SCR, this isolates it and you can make a resistance measurement across it.

You should read "open" with an ohmmeter across it, just like you haven't even touched the meter leads to anything. "Shorted" would be just like you put the two meter leads together. If you read "shorted" or low Ohms, it is a bad SCR.

If the SCR reads open, then the next problem could be the SCR getting turned on from a bad firing board. You could leave the signal leads off the SCR, and put the power cable back on, when you power the welder it should not pass any current because the SCR is not getting turned on.

If it does pass power, since the SCR is not shorted, it must be that power is going to ground from the primary. This could be a lot of power, so it would be safer if we had a 10 Amp fuse passing the power to the transformer. The test would not be spectacular, you would have to power back down and check the fuse, if it is not blown, then the transformer did not have a short to ground.

3:34pm
4:11
Checked SCR reads 9 something (Ohms?), so it is shorted. Read another one that is not in the circuit, it will read open.
Also check the transformer to ground, you should not have any continuity between a primary wire and ground. If you read "open", that is good, but not the best check, because your meter is using only 1.5 volts. using a Megger will check it at 500 volts, more like the 480 it runs on.
3:52pm
3:22
SCR read 9 "M" is that "milliOhms" No, not if you are using a VOM, it would be MegOhms, and a number tht high is about like reading humidity in the air. You need to force the autoranging meter to its lowest scale, no "M", or "K" shown after the reading. OK so it doesn't read anything on that scale, good, it is NOT shorted.
So what else is left? Could be a transformer with a primary to ground short. Do the megger test where you check for current flow from the megger between ground and one of the heavy cables going to the transformer, like the one you took off the SCR.
Make sure all the wires are off the SCR when you apply the megger, because it puts out 500 volts and can blow SCRs.
4:01pm
4:47
What if the meter reads just like the clips were together?
That would be a transformer with a short to ground, what is it reading?
5000 volts, and some other things...
Ye Gods, 5000 volts, you sure (yes) so I checked the Kyoritsu 3122, and by God it puts out 5000 volts. So you must immediately stop, this will puncture through the insulation on any transformer. That is a dangerous tool for the shop, shut it off and sell it on eBay. Do not put it across a transformer, ort anything else! It must be for checking high voltage sub stations, it is not for checking 480 volt transformers.
4:16pm
4:02
How to check for bad transformer without a megger?
Couple of checks:
1. Put a fuse in the cable going from the disconnect to the transformer primary, 10 amp or so, leave the other primary cable off.
Note that by "primary", I mean the primary winding of the transformer, not "primary power". It is too bad we use the same term for two things.
So now when you throw on power it should not blow the fuse, the only path for current flow will be from the primary to ground or through the secondary to ground, both are a dangerous problem, and should not exist.
4:23pm
2:34
Read ohms between primary wire and one of the shunts got ".001K" or "0.1" on regular ohms. "OL" is what this meter reads when open. So this is definitely a short primary to secondary, there should be no current flow or path for current flow between the 480 volt primary and the secondary. This is a very dangerous transformer putting line voltage on the secondary can zap someone. This transformer must be changed, it cannot be used.
I suspect that using a 5000 volt megger could cause this. I wonder who sold them that meter? That is the kind of thing that should require a license to own...

---------------------------
y09m11d19
11:06am
24:22 mins:secs
Discussed Medar firing:
The weld control "fires" once each half cycle, but does not give you the full half cycle, unless you are firing 99%. If you fire at 50%, it waits for half of each half-cycle to go by, then turns on the SCR, which shuts off by itself at the end of the half cycle. So you get a "chopped" half cycle, both positive side and negative side of the AC sine wave. This method of varying power is called "Phase Angle Control", and it is how the Medar controls power, whether it is in "Automatic Voltage Compensation" (AVC) or "Automatic Current Compensation" (ACC)

If you are setting the weld heat in "%", you are in AVC mode. To change to ACC mode, you must replace the Medar heat step in your schedule with another type heat command that uses "00000 Current", Jeff can show you that. Once you are using the ACC mode, you must calibrate the control. It reads Primary current, but it shows Secondary current after it welds. How is this possible? It uses the "turns ratio" that you enter in the set-up section of the control. This must be accurate, or the settings you enter will be meaningless. To get them accurate you must enter a current and do a test weld. On a spot welder, you can weld tip-to-tip if you like, and it can be just 5 or 10 cycles. On welders where tip-to-tip welding is not possible, like most projection welders, you must weld an actual part. It will probably produce a bad weld, hopefully you used a low enough setting so that you don't get expulsion, this messes up the test, lower it and try again.

Now, after the weld is done, see what your separate current meter said (Do not blank any cycles, the Medar doesn't, and you want to get the most accurate comparison) and compare it with the Medar "Last Weld Data", if they are not the same, your turns ratio is wrong. Remember if you get it within 200 Amps, that is great, don't even worry about getting closer than that.

If your Medar read higher than your Meter, your ratio was too high, lower it and try again. If you want to "zero in" quickly, then divide your Meter current by your Medar current, and multiply this by the turns ratio you used. You will get a new turns ratio, try that, it will be very close.
Now that you have it calibrated, remember that if you change the transformer tap, you just changed the turns ratio, you must recalibrate. Save these numbers, in case you go back to the other tap, this number you saved is still valid.


The advantage in running in ACC is that the control monitors and corrects for anything that changes the current. This is the most important ingredient in using electricity to make heat. The formula is "I squared R", meaning that any change in I (current) affects the heat by THE SQUARE of the change. This is a big factor, and it is what makes current regulation so successful.

If you are in AVC mode, then the control only makes up for line voltage problems, but there are so many more things that affect current, that this is just not good enough! This was a great feature when Pertron first offered it in their computerized weld control in the mid '70's, but in the late '80's, Automation Data offered it in their "Trans-Mate" control, and it caught on like wildfire!

There are a couple of disadvantages in the ACC mode;
1. You don't know how much extra power you have when you set "18,000 Amps". You do when you set "97%" and watch what Amps you get with a separate current meter. You know that you need to go to a higher tap if you are at 97%, because the smallest drop in line voltage will give you a drop in weld heat. How much of a drop? This is a big effect, let's consider a simple example of a welder running 1 Amp at 100 Volts. Multiply Volts x Amps for "heat" in Watts, you get 100 Watts. Now the line drops 10%, no big deal right? Not so! if you drop to 90% of the voltage, it will push only 90% of the current through the secondary. So multiply 0.9 Volts by 90 Amps and you get 81 Watts of power. This is almost a 20% drop in heat. See what a big deal this is?

2. If you get expulsion, it drives the regulator nuts. When you expel metal there is no current path until the tips fall down on the part again, so the current sample shows that the total current for that half cycle was low, and ACC will boost it up! Not the right thing to do! This happens with most projection welding, so use "%" for projection welding. We had a Welding Engineer that insisted on lots of expulsion, and we had to go to the plant and take all the controls out of ACC so that weld consistency was half decent. By the way, that Welding Engineer now hates expulsion, just like me.

So the Medar has a fault "Current compensation out of range" or something like that. It tells you that it cannot control the current to give you what you want. If you are in ACC, go look at the "last weld data" and see what % it used to fire the last weld. If it was very high (>95% or very low <25%) then you need a new tap setting.

Another related feature is the "C-Factor" which tells you what your maximum current would be if you fired at 100% power and the line voltage stayed at the nominal setting you have entered in the Set-Up mode. You do have a proper value there right? NO, Fix IT, very important.

With this C-Factor, you can see how much extra power is available. If the C-Factor says "200" just put two zeroes on the end "20,000" that is your max current, at nominal line voltage. I like to see 20% more than you need, to make up for things that may change in production. If you see that power is a little short of what you want, you need a higher tap. BUT REMEMBER to recalibrate or you will get false information back from the Medar.

The Medar allows you to put a Turns Ratio setting in each schedule also. You rarely need this, but if you do, it overrides the number in the Set-Up values, so don't use this if you don't have to, stick with the one setting in the Set-Up.

I also don't like to see too much power available. If you need 10,000 Amps but your C-Factor is "250", then your current regulation and the way your heat is applied is not optimum. The heat is parts of a sine wave, and if you fire at 63% to get 10,000 Amps, then your peak current at the peak of each half cycle is 25,000 Amps! This pounds the heat into the part at too great of a rate, just for an average heat of 10,000 Amps. This is bad for expulsion and changes in the chemistry of the steel. The new types of steel prefer DC welding, where the peak and the average are the same. So don't get carried away with too much power, it is tough to weld with.

About the line voltage, and how it affects C-Factor: it doesn't. The Medar control takes this into account, and computes what you would get at normal line volts. If it didn't do this it would be reporting all kinds of C-Factor numbers, and would be an unstable number. I think this was pretty clever. When I was at Pertron we thought Medar was awful, now that they are not a competitor, I see how wrong we were, Medar is wonderful, they have done many clever things...

So in summary:
Use ACC
Setup 'Nominal Volts" in Set-Up section.
Calibrate, be sure to not blank any cycles with your meter settings.
Read the % used in "Last Weld Data".
Use the "Current Regulation Fault" Which is automatic unless you override it in a schedule...Don't do that.
Make sure you are not overriding the Turns Ratio in a schedule, unless special conditions need it.
Don't change taps without recalibrating.
Be nice


---------------------------
y09m11d19
4:00pm
??:?? mins:secs
Discussed the wide variation in current on the "last weld" display.

Most applications have less than 1000 Amps difference between the highest and the lowest reported current, for a weld heat. When you have wider variations it means something is bad. In a spot welder it can be caused by expulsion, but in this seam welder, it is most likely a bad slip-ring connection in the rotating weld head.

This is the last thing anyone will want to hear, and there is always much resistance to the suggestion that it must be changed. In order to show it is bad, you can monitor across this sliding connection with a chart recorder or a memory scope. Any spike in voltage over 1 volt is a serious loss. Consider that at 18,000 Amps, a 1 volt drop is 18,000 Watts, enough to run three houses with central air!

To accomplish this you can mount one wire on the leaf shunt closest to the rotating connection, and the other wire has to be attached to the wheel. This wire will get twisted up as the wheel goes around, so the best thing to use is a piece of THHN 14 Ga wire, screwed to the wheel with one of the wheel mounting bolts. It will get ruined, big deal.

If you make another test after the slip-rings are replaced, find a way to preserve that image, so that you can more easily show people what it should look like.

With the shock absorber application, you have two of these rotating connections, so you will be subjected to twice the likelyhood of a bad slip-ring messing up your welding. In fact you will have to make two tests like just described to find which side is the problem.

---------------------------