Sign in to follow this  
Followers 0
ssommers

Anyone up for a resistance welder mystery?

37 posts in this topic

I apologize in advance for the length of this post, but I'm in a quandry. This problem's been kicking around for over 2 years, so there's a lot of background information... I have a tube mill with a resistance welder. The welder has a large motor-generator set that is controlled by a very old DC control board - Meltronics Focus 2. The Focus 2 runs on single phase 230VAC, has a 0-10VDC potentiometer input and puts out a DC signal from an SCR to a 3Hp DC motor that governs the generator. There is no tach feedback. I have a datalogger wired to the Focus 2 because the weld voltage occasionally drops down low for about 4 seconds, creating about 20 feet of unwelded tube. The operators thought that it was related to stopping the beadwinder motor to empty the spool, but it happened so rarely that no one was ever able to confirm it... until now. I've recorded it happening twice in the last month including last night on 3rd shift. Originally, I thought it might be a noisy 230VAC line. Our plant power is noisy on a normal basis, but the AC voltage into the board is steady during the minute before & after the drop. I've ruled out someone playing with the potentiometer based on hard data & operators that I trust. The Focus 2 board output signal to the DC motor drops first. The potentiometer voltage & welder voltage follow it like they're reacting to a feedback loop. Then everything recovers about 3 seconds later. A few more pieces to the puzzle... There are 2 beadwinder motors called Upper & Lower (480VAC, 3 PH) and a trim tool that is held down with a solenoid (120VAC). The excess weld on the outside of the tube is trimmed off and wound onto a beadwinder. When the beadwinder gets full, the operator turns off the full winder, turns on the empty winder, moves the trimmed weld bead to the empty one and then dumps the full winder. The other time a beadwinder will shut off while the mill is running is when the trim tool is raised to change to a new tool edge. Either one of these operations is accomplished in about 4 seconds. The Upper beadwinder motor was moved to a separate 460VAC disconnect a few months ago when the operators felt that it was only happening when they changed the trimmed bead from the Lower winder to the Upper. The problem didn't go away, it has just happened less frequently because the operators refused to use the Upper beadwinder anymore. Today's event is directly attributed to turning off the Lower beadwinder motor - I got that from the 3rd shift supervisor, will get the real story from the operator tomorrow morning. The event a couple weeks ago happened with a trim tool change which also momentarily turned off the Lower beadwinder - that happened on 1st shift & that info is straight from the operator. What management wants me to do is... 1. Move both beadwinder motors & trim tool controls off to a separate control cabinet, disconnect & plant transformer. We have three 13800/480 VAC transformers in the plant so this is possible although highly inconvenient. 2. Reroute all conduits so there will be absolutely no wires or conduit for the motors or solenoid running near the welder control conduits. What I want to do in addition... 1. Put a constant voltage transformer on the 230 VAC input to the Focus 2 board to eliminate stray noise. (This costs major $$$ so management wants put this one on hold.) 2. Put surge suppressors on the beadwinder motors & trim tool solenoid. 3. Put dry contact relays so the main machine e-stops & run contactors still shutdown the motors for safety but all signals are isolated. Now what I need help from you on... 1. Are there other DC motor controllers are on the market for resistance welders? 2. Is there anything else that I can do for electrical noise suppression on the 230 VAC power line besides a constant voltage transformer? I still think this is going to improve weld quality in the long run. 3. Could this whole thing be caused by a low resistance path between the beadwinder motor and the welder through the weld bead that's being cut off? Pretty far fetched, but I'm looking at every possibility. If you got this far... Thanks for hanging in there! Any and all comments, questions & suggestions will be highly appreciated. Thanks, Susan Edited by ssommers

Share this post


Link to post
Share on other sites
What size transformer are we talking here, and what are the units of major $$$? I like the Sola units myself, it all depends on the size, they don't come in very large capacity. http://www.solaheviduty.com/products/power...vs.htm#Features

Share this post


Link to post
Share on other sites
Dear Susan: Yes I made it thru but I am sorry I am not going to be much help. I have little background with welding, but if I get this right the 0 to 10 DC input controls the 230 AC out which runs the generator which creates the actual welding voltage. Is this correct? Did I also read that your data logger shows no change in the 0 to 10 DC but the 230 output from the control board does drop. THis means to me that either your board is experiencing a thermal condition or the inputs are telling it to cut the voltage back. My only input would be on your search for a constant AC transformer. Controlled Power company in Troy MI makes units for NASA and has a great product at a reasonable price. And no I don't work for them, jsut used their units before.

Share this post


Link to post
Share on other sites
hi, I got to the end.... (and English is only my 3rd language) Have you been able to log if you get spikes on the power network when shutting down the beadwinders? Are the beadwinder motors direct on line? If so you could maybe try to put an inverter in front of those motors. This would create less problems with spikes (but maybe more problems with high frequency noise) Can you check if the control board has a self resetting fuse? The fact that the output goes down while the 0-10V stays the same could mean the power board goes into overcurrent/overvoltage.... CU Beegee

Share this post


Link to post
Share on other sites
I spec'd out the Sola Hevi-Duty 7.5 KVA to be able to handle 3HP at full load. It's approx $5000 which is just the level where capital requests kick in and the general manager has to scrutinize the expense to death. The plant manager realizes that it should improve weld quality, but the GM doesn't believe it will solve the problem of the welder dropping out for those 4 seconds. I have to admit I have my doubts too, but it would be one more layer of noise insulation.

Share this post


Link to post
Share on other sites
I haven't seen any spikes on the power line due to the beadwinders going on or off (they're on 480), but I'll check it on the datalogger this morning. I doubt the line supervisor will let me try to induce the fault, but he might if I catch the mill starting up for a new tube size. I know there's a bunch of HF noise on the 480VAC power line. I have induction welders & annealers in the plant (including right behind this resistance welder) so noise is a constant concern. The Focus 2 has a 480/240 transformer without any line conditioning in front of it and all the power noise goes straight through it to the output to the DC motor. That alone shouldn't make the output drop so drastically. I'll put up more info when I get to work.

Share this post


Link to post
Share on other sites
Here's a graph showing when the welder drops out for 4 seconds. You can see that the Focus 2 output voltage drops first and actually spends some time at 0. The potentiometer voltage, which is within the Focus 2, bottoms out about 0.1 seconds later - the stair steps seen are the finest resolution I have. The weld voltage bottoms out about 0.25 seconds later as expected given the delay through the MG set. Now I'm off to hook up more inputs to the datalogger. I'm going to add the lower beadwinder & the trim tool solenoid to see if there's any noise caused by them. 070306_71.pdf

Share this post


Link to post
Share on other sites
Hello Susan. With a time of 4 seconds for the whole "incidence" then it has nothing to do with noise or spikes (*). The graph shows that the voltage from the potentiometer indeed does drop. So you should focus on this. As the potentiometer is a relatively simple passive electrical device, there shouldnt be any mysteries to it. 1. Datalog the supply voltage on the input side of the potentiometer. 2. Exchange the potentiometer (this option is not in your list). *: edit. At least not directly. I cannot rule out that noise or spikes can cause the electronics that supply the 10VDC to the potentiometer to freak out. Edited by JesperMP

Share this post


Link to post
Share on other sites
I don't know if this helps, but here is Melltronics: http://www.scrcontrols.com/products.asp?pr...&page=specs

Share this post


Link to post
Share on other sites
The potentiometer itself has been changed. In fact, I think every piece in this part of the mill - beadwinder motors, contactors, solenoids, switches, pots, Focus drives, and wires - has been changed at least twice in the last 6 months. The electricians & supervisors were taking a "shotgun" approach to this problem long before I arrived here almost year ago. I just had enough time this morning to add monitoring for the trim tool solenoid & the beadwinder motor voltages. I'll add another data point for the potentiometer supply voltage next time the machine is down for a change over. Question: Could the pot voltage going down be a result of something else happening inside the drive? The pot supply voltage is generated from inside the Focus 2 drive which has many, many op-amps & resistor-capacitor networks. I tried doing some analog feedback analysis from the schematic on p.53 in the drive book and managed to make my head hurt alot. I keep thinking that there's some kind of cascade going on in the control board that has something to do with the SCR output to the DC motor, but I can't figure out what starts the whole thing off. It has to be an impulse type of thing. I'm sorry if I sound frustrated, but I am. Thanks for hanging in with me!

Share this post


Link to post
Share on other sites
I have some experience with resistance welding, but only with spot and projection welding. Never did any continuous seam welds. Nonetheless though, I feel compelled to chime in :) The first thing that strikes me as odd is this motorized generator you're using. Is that common in your industry, or is it just very, very old? I've never seen anything like that in an auto plant.. The only thing I've seen that comes close is an old Lincoln arc welder at the community college I attend. It has a huge DC generator as the power source. It was from the mid-60s. Would it be feasable for your company to install a modern mid-frequency DC inverter welding control? That is what is considered commonplace in the auto industry, and would probably eliminate all your problems. (again though, I'm not familiar with seam welding.. not sure if MFDC offers the same advantages to your process) Here's some information on what I'm talking about: http://www.medar.com/welding_concepts/controls_mfdc.html Of course, this would be a major redesign of your process, so it may not be practical or possible for your company to do it. Just a thought though!

Share this post


Link to post
Share on other sites
While you're at it adding measurement points, can you also check the GND of the control board. Many of the problems in that area I encountered were due to bad grounding.... I have learned to check this the hard way.

Share this post


Link to post
Share on other sites
This is just very, very old - mid 70's approximately. There are 3 resistance welders in the plant with 100HP (or bigger) MG sets. The rest of the tube mills are induction welders. The company is heading toward induction welders - cleaner non-contact welds, less daily maintenance, etc - but it will take some serious shoehorning to get a new induction welder into this spot. I've been working on the planning the changeover for 6 months and it may be another 2 years before it actually happens. Until then, I have to keep plugging on this. Beegee - Thanks. I've already checked the control board GND. I'm a stickler for shielding & grounding so I know that's not it. But it's a good thought.

Share this post


Link to post
Share on other sites
caveat : I'm no welding expert. If I understand correctly, the control board regulates the speed of a DC motor that is driving a DC generator that supplies the welding current. I've studied the schematic and the plot. If the scaling is correct, you seem to be operating at the low extreme of the speed pot - approx. 140 millivolts in a 10 volt range. The dip is approx 80 millivolts. Perhaps other gain settings on the board are too high, giving you no span to work with. It would appear from the plot that the motor is attempting to stop (control o/p goes to zero). The generator output doesn't reach zero (coasting perhaps?). With the "power cube" supplying the motor in parallel with the control board's supply, any load disturbance at the motor that strains the 230V supply transformer will be seen by the controls. The reference for the speed pot comes from a simple zener regulator and unregulated 24V. If installing a CVT, I would be inclined to only supply the control board from it, therefore it would be small.

Share this post


Link to post
Share on other sites
I can personally vouch for removing every last one of those constant voltage transformers from my plant. All you are buying is an autotransformer. So it fails to provide even the small protection that an isolation transformer gives you since one leg is tied directly to the primary side. In theory it will help cut down on some sags and the collapsing field MAY help it muddle through mild brownouts that last for less than a cycle. That's about it. I can demonstrate case after case after case of strange electrical failures and voltage sags that I've tracked down to CVT's. However, it will do a wonderful job of overheating anything around it, including itself. It also does an excellent job of failing on a regular basis, requiring replacement. You may as well pay for some cheap line reactors or something like that because it's going to give you just as much of an improvement as those overpriced heaters that Sola sells.

Share this post


Link to post
Share on other sites
Correct, there's no brake on the generator so it coasts for a while. The question I have is... Why did the control voltage output to the DC motor make it all the way to zero and then turn itself back on? And why has it happened the same way on at least 3 different Focus drives over the last 6 months? I could buy an intermittent problem with one, but not all 3. I'm off to the floor to gather more data...

Share this post


Link to post
Share on other sites
Mybe some ideas. If you can measeure voltag on potenciometer bitvin pin 8 and 10. It mast bee steade at 10V. How old is drive thet is replaced, ther is some problem with elektrolit if it's older then 10 year.If so tray to replace it with new, I have some result with this.

Share this post


Link to post
Share on other sites
Update time... I haven't gotten much in the way of data since my last report. It's union contract renewal time and some little gremlins have been stopping or unplugging my datalogger several times per day. I don't want to babysit the recorder, but I may be forced to do just that until the union votes next weekend. I have found out that both the 10 VDC & 24 VDC on the Focus 2 board look good & solid during a run. The 10 VDC is developed from a voltage divider that depends on the Start signal being present. This leads me to believe that the Start signal (relay) may actually be the one dropping out. I'm off to see if I can keep the gremlins away from my logger today...

Share this post


Link to post
Share on other sites
Do your 10/24VDC stay ok during the drop of speed or haven'tv you been able to log that?

Share this post


Link to post
Share on other sites
I haven't been able to log a weld failure this week due to gremlins. I've had an electrician in that area since mid-morning running some conduit & the logger is running fine - saving data without problems just like always. This is a very occasional event. It took 2 months for me to see it twice under normal operation. I asked the production & plant managers if I could try inducing the fault. They got that "deer in the headlights" look and said NO, so I have to wait for it to fail naturally. I did get a chance to look at the 10 & 24 VDC signals during a normal shutdown & a normal speed change. They're both solid in a small speed change & both decay at the same rate when stopping. Now I'll know what to look for when the failure happens. Back to warding off the 2-legged gremlins...

Share this post


Link to post
Share on other sites
the Hamlin relays used in focus drives start logic on the main pcb, can be a bit of a pain. I have replaced too many. If you are looking for a better solution, look seriously at a 3 hp VFD and a marathon black max motor. I recommend the V7 or 3G3 MV Yaskawa/Omron for your situation. At the 3 hp level you should be able to get both for around a grand. The drive can have external controls and speed pot or use the built in operator if so equipped. Using a 3% reactor on the VFD input helps with noise both ways. I have seen many single phase 5 hp and less DC drives on welders and on the ones converted to VFD, they run much steadier and smoother too. The DC bus in a VFD helps with line voltage variations. A DC drive with no tach like the Focus 2, have 5% speed regulation characteristics. The VFD has about a 0.5% or better speed regulation characteristic. If you do not want to convert to a VFD, I have seen using a regualr single phase transformer, rated for at least 50% bigger than the drive/motor help out quite a bit. The xfmr helps smooth out noise and eliminates line grounds from getting to the drive and motor. It also helps when the motor grounds or drive has issues by limiting what gets put back to the line.

Share this post


Link to post
Share on other sites
Ok, which relay is this? Is it the one labelled CRR? And I'm going to look into the rest of your suggestion, but it's hard to know what will & won't work with this old MG set. Management definitely won't change from the MG right now, but it's always good to have options. As for the gremlins... My 2-legged gremlins went away on Friday as soon as they heard that a new union contract is ready to vote on, so the mill operators finally helped me find the REAL gremlin. The datalogger internal power supply had been slowly dying and rebooting all on it's own. It finally went into terminal reboot-itis yesterday afternoon. Thankfully, I'm only renting this unit so they're sending me a new one and it will be here this morning. Off to see if it's here yet... Thanks, Susan

Share this post


Link to post
Share on other sites
If you change over to an inverter with a normal AC motor, take care to take a look to the startup torque and the torque at low speed. I you need high and steady torque at very low or zero speed, you should take a closed loop vector inverter (for instance F7 staying with Omron Yaskawa)

Share this post


Link to post
Share on other sites
That is why I said V7 as it is a vector drive and has excellent low speed torque especially when wired to a Black Max motor. Typically these welders do not require tremendous starting torque as they tend to have a large gear reduction. ssommers, Glad to hear some of your "headaches" have been cured. BTW, by going to an inverter, you might be able to eliminate the MG set. Unless it powers multiple units. Even at that, you will me SSSSSOOOO far ahead to dump the MG set and get some newer equipment in there. Cost to operate alone by going to the VFD's drops dramatically.

Share this post


Link to post
Share on other sites
I guess I'm still curious about the whole using-a-motor-for-welding concept. Is that still the state of the art for this type of weld? It seems to me that if we're talking inverters, we should be directly converting line voltage to welding current and leaving the mechanical stuff out of it! That's how all the spot, projection, and MIG welds are done in all the plants I've been to.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0