Reliance Flexpak Plus (DC Drive)

[splicer480 0]

Hello all. I have a 300 HP Reliance DC Motor driven by a Reliance Flexpak Plus DC Drive. It is around 12 years old. It puts out 500 Vdc Armature and a 300 Vdc field. Last night around 3:00 am two of the three incoming 800 amp fuses blew and it shut down. Of course I don't have much info on the drive or Flexpak so I went to the Reliance website to see if I could find some motor and drive specs. I found some info about the motor which said there should be no less than 100,000 ohms of resistance to frame ground on either the (2) field leads or the (2) armature leads. This is to be tested with a standard ohm meter. I tested that and it didn't pick up anything to ground at all. Next I checked the field windings (300 V @ 15 Amp = 20 Ohms) Check also. Finally I checked resistance between the armature leads and it reads 0.2 or 0.3 ohms (This is the one I'm not too sure of) My next step was to check out the Flexpak itself. After a detailed visual inspection, checking circuit boards, wires, SCR's and other small fuses I decided to replace ALL 3 fuses and disconnect the motor field and armature leads. The only thing I left wired up was the tach and the over-temp switch. I powered on the Flexpak (with my fingers crossed ) and to my surprise no BOOM. Anyway, I read the field voltage output terminals: 300V dc. (Cool) and none of these (field) fuses were blown in the first place so that was good. I decided since the field in the motor itself read good I would go ahead and hook it back up to the flexpak. I wired that back up and powered the Flexpak back on. Once again I read the field voltage at the Flexpak and then the motor itself: 300 Vdc. (Cool again) Now for the armature. After power down I wired the armature leads back together, turned the drive speed pot all the way down and powered on the Flexpak. No boom yet I put my DC volt meter on the armature leads and slowly increased the speed pot and watched the percent current gauge and the motor shaft. I let the % current gauge get to about 25% and there was no movement from the motor shaft at all and no voltage registering on my meter, so I shut it down. (By the way, the motor shaft turns easy by hand) My last check of the day was to turn the system back on and use my AC amp meter on the incoming line leads (because I don't have a DC ammeter. Yet) thinking if I'm drawing DC current something should show up on the AC side. I turned everything back down and started the system again. Without increasing the speed pot yet, I was showing about 3 amps AC. I started to increase the pot and the amps followed accordingly up to my 25% current check position on the speed pot. The amps were around 160 on all 3 legs at the 25% current draw. There was still no shaft movement. So after this big rant my question is this: Does anyone know if this is the motor or the drive acting up? Or some kind of test I can do? I was thinking since the field seems fine, could I hook up a smaller DC motor, like a 1 hp 130 Vdc or so to the armature leads and just keep the speed pot throttled back to see if the smaller motor would turn or not? I read I would also have to unhook the tach and change a jumper over to an armature voltage feedback via the Flexpak itself to track a motor without a tach setup so it doesn't run away. Any help on this situation would be greatly appreciated. The plant super wants to know if we need a new motor or a new drive and I would like to be sure of the answer I give him. Thanks fellas

[JMK 0]

Hi Splicer - Under most circumstances, if you blow 2 of 3 line fuses, there was a short circuit across that phase... If the short circuit or ground fault originated in the motor, I would suspect that the field or armature fuses would have blown. "I found some info about the motor which said there should be no less than 100,000 ohms of resistance to frame ground on either the (2) field leads or the (2) armature leads. This is to be tested with a standard ohm meter." Don't let your standard DMM fool you... I strongly suggest using a megger in the above application. A standard multi meter energizes the circuit with 3-9 VDC. With a megger, you'll check resistance in the mega ohm range using 250V - 1 kV. This is the best way to verify insulation breakdown. "Next I checked the field windings (300 V @ 15 Amp = 20 Ohms) Check also. Finally I checked resistance between the armature leads and it reads 0.2 or 0.3 ohms (This is the one I'm not too sure of)" The armature resistance seems low... I would think that you wouldn't want to see < 1 ohm. "Now for the armature. After power down I wired the armature leads back together, turned the drive speed pot all the way down and powered on the Flexpak. No boom yet I put my DC volt meter on the armature leads and slowly increased the speed pot and watched the percent current gauge and the motor shaft. I let the % current gauge get to about 25% and there was no movement from the motor shaft at all and no voltage registering on my meter, so I shut it down. (By the way, the motor shaft turns easy by hand)" This is what really catches my attention. You had 160A draw on the AC line side with 0 VDC on the armature leads? Also, at 25% you did have field voltage, but no armature voltage? Field and armature voltage should increase proportionately. I would strongly suspect that you problem lies within the drive.

[paulengr 44]

These drives are nothing easy to troubleshoot. You need a DC amp clamp ASAP if you are going to be doing any more troubleshooting, and a meg-ohmeter. For instance, you don't know if the motor is conducting all the current or if it's the drive. If you had the DC amp clamp, you could have already isolated the motor/drive issue. Put the amp clamp on the armature lead and compare it to the drive reading. If the amp clamp is reading very low, you've got a drive problem. Both would serve you much better in this situation. You could still have a drive problem. The documentation isn't great but you can download full documentation (with troubleshooting guides) directly off AB's site under "Legacy" drives. They still have the info. Last I checked on the Reliance (now Baldor) site, the documentation was missing or scattered and very difficult to locate. In addition the total lack of proper coordination makes them that much harder to troubleshoot because the fact that you have an AC fuse blown suggests that you have a drive problem. However, because they are so poorly designed, it is just as likely or more likely to have a problem on the DC side. The last remaining problem that is easy to test is that you may have a blown SCR. I don't know about a 300 HP unit but at least with the 150 HP and smaller drives, unfortunately, Reliance went and soldered in the SCR's. The only way to do anything with them is to replace the drive entirely. If there's built-in diagnostics according to the manual (most drives have these now), use those. If not, then check if the SCR's are bolted in. If so, you are in luck. Remove them one-by-one being careful how they are placed (they have polarity) and test them in both directions (DC devices) with your megger. They should read about 100 kohms or more in both directions. If not, replace them. Should be <$200 for almost any size SCR.

[splicer480 0]

Thanks for all the info guys. I'm going to meg the armature leads and see what that gives me and I have a DC clamp on amp meter coming today which I will use also. I will keep you informed with what I come up with. Thanks again. Edited 6 Aug 2008 by splicer480

[splicer480 0]

Well yesterday I megged the motor to frame ground and got an infinity reading. (Good) The field was already OK. We decided to send the motor out because it needed new bearings and the end housings were worn from improper coupling alignment in the past. The repair shop called back later that day and told me electrically the motor was fine except for one weak brush spring, nothing that would have kept the motor from running. So back to the flexpak Plus. .... I was finally able to talk to someone REALLY knowledgeable at Reliance and told him that I had no info at all on this drive and could not find any info on their website for the size of the drive that I have. He researched it for me and found a complete manual for me in PDF format on and old hard drive somewhere. Unfortunately for a lot of the testing (per the manual) I've got to have the motor back and hooked up to take some readings on the regulator terminals. In the meantime I'm ordering 6 new hokey puck SCR's and I'm going to replace them before the motor gets back just to eliminate them from the trouble shooting guidelines. The SCR's are the first culprit on the list anyway plus to have blown the (2) 800 amp fuses it seems logical because these are mounted on a large heat-sink that is bolted to the Flexpak Plus frame, which is grounded. Sometime next week I'll post back with the results.

[paulengr 44]

Before you go too far, try this. It's happened to me once. Check the voltage between each phase leg plus the voltage from each phase leg to GROUND. Make sure that your phase relationship is reasonably well balanced. I ran into an issue once where a drive kept tripping out for "overvoltage" or blowing fuses before it tripped. It turned out in this case that the ground was never hooked up on the transformer (delta-wye), so the center tap just floated wherever. Unless the drive is designed for it, it is an absolute requirement to have a solid wye connection with a reliable connection to ground because on a 6 SCR rectifier bridge if you don't have that, then one of the SCR's will remain in conduction when the drive switches to the opposite phase and WHAM...line-to-line short across your SCR bridge. Checking SCR's is usually very easy. First, a quick little SCR theory. Check out this page: http://www.allaboutcircuits.com/vol_3/chpt_7/5.html With the size SCR's you have, they are wired into a 3 phase bridge configuration. See the circuit diagram at the top of this page for the circuit: http://services.eng.uts.edu.au/~venkat/pe_...s1/ch05s1p1.htm This one is very theory intensive but explains all the details of a basic DC drive: http://services.eng.uts.edu.au/~venkat/pe_...s6/ch05s6p1.htm Regardless, SCR's fail for one of two reasons. The N & P layers at the ends of the device are fairly large and can handle very high currents and voltages. The two N & P layers in the middle are relatively thin. Their purose is simply to act as a reverse-biased diode and block current flow until their either approach the reverse voltage breakdown voltage or you put a small bias current onto the gate. This small region of the SCR is not designed to tolerate high voltage spikes or excessive heat. When you get a failure in those two layers, then the SCR starts acting like a diode. It only takes a small failure for it to happen and because of the nature of a drive, the device goes into thermal runaway and virtually gaurantees that the entire SCR will be shorted out very quickly. The inner "diode" at the gate is a very thin layer. Chances are that you will NOT see any external damage whatsoever. Carefully remove each SCR making sure that you keep track of which is which, and especially be sure to keep track of polarity. Do not try to test them in place because quite often the driver circuitry will give you false readings. Then check the resistance all 6 ways. You should read at least 100 Kohms in all directions except from gate to anode which should be a very low resistance reading (acts like an ordinary diode which it is). 99% of the time, you will find that the anode/cathode path is dead shorted in a failed SCR. Identifying the 3 leads should be obvious. If not, the big fat connections are the anode and cathode. The trigger/gate port will be the one with a little connection. Sometimes the SCR will have some kind of internal wire path so that you get a "4" port device. The 4th port is just a second connection to the cathode. Resistance checks should pretty quickly clear up which is which. If you still don't find a bad one, that's the point where you have one of two choices. You can either rig up one of the test circuits shown in the above web pages and manually trigger it with a DC source external to the drive or rely on the on-board diagnostics. The only true way to test an SCR is under load but it's rare to find a bad one that tests good with a simple ohms check. More often this leads to a driver board failure, which is usually tested with the drive running. An SCR puts out about 1.5 watts per ampere. The heat sink grease that you may find on the SCR is a REQUIRED component. Even if not present, I tend to liberally grease them myself when I put new ones in. Check any external fans, make sure that the heat sinks are working properly, etc. Clean out everything and check if the cooling vents are blocked in any way. Heat is the #1 killer of SCR's. Thermal cycling over time will eventually destroy them even if nothing else is wrong. Also, if you have any upstream surge protectors and/or lightning arrestors, check these. Drives are notorious for getting zapped if you have a big transient come through, and older drives are more prone to this than newer ones in general. The driver/amplifier/regulator/firing board is there because logic-level voltages and currents are too small to fire a large power SCR. The driver board isn't even mentioned in the above web pages talking about the theory of SCR's. The driver board converts the logic level signal into a bigger one. Quite often the component on the driver board will be a big power MOSFET or IGBT. There are designs where a smaller SCR or TRIAC (back-to-back SCR's) is used. If anything on this board looks burned, it probably is. Usually you have to replace the entire driver board. Driver boards tend to be much more hardy than the main SCR's but they still can go bad over time. I know some of the information out there talks about the snubber circuits. These are important but they are just LC circuits and they tend to be very robust compared to power electronics. Once in a while though one will go bad. You've had one failure in a few years. I don't get nervous about the LC circuits in that case. If you have consistent and repeated failures, that's when you check out the LC circuits. Finally, a word about my comments to replace the whole DC system with AC. DC motors are very maintenance intensive, relatively speaking. They are also rare...you can't just call a motor shop and have them deliver a stock 300 HP DC motor. With AC motors, bigger motor shops often stock the most common design motor up to about 300-400 HP. If they don't have it, you can often get one from a factory warehouse within hours. Large DC motors are custom manufactured to order and take weeks to get. Because of the commutator, they are physically more complex and fail more often. At one time, DC motors dominated the variable speed control world for several reasons. AC drives were several times more expensive, even when considering the total package (DC motors are much more expensive than AC induction motors). AC drives were very noisy (audibly and electrically) and tended to be very unreliable. Before vector control came along in the 90's (especially sensorless vector control), AC motors have very weak starting torques and you usually couldn't start reliably under load unlike DC. AC drives and motors couldn't reliably turn below about 25-35% of full speed without overheating. Even with appropriate cooling, AC drives could not run at 0 (stall) speed, just holding a load. With two separate bridges to contend with (input and output), and some extremely complicated designs to get around the lack of a decent "switch" on the AC side, troubleshooting AC drives was up there with brain surgery in terms of difficulty. Since that time, sensorless vector and full vector control have come along. These have eliminated all of the torque issues. IGBT's and integrated drive design have eliminated the reliability issue, and driven the costs down to the point where DC drives are now frequently more expensive. Higher switching frequencies (another IGBT feature) have eliminated most of the noise complaints and provide much better control. At this point, the operating region of an AC drive exceeds that of a DC drive in every way. The drives are just as reliable and the motors are naturally more reliable because they have only two mechanical parts (the bearings). On the smaller drives, the technology is "throw-away". With the larger ones, field repair techniques are very similar to DC drives (simple checks with basic tools). Everything is modular so the parts are usually off-the-shelf. AC drives aren't quite perfect. With IGBT designs, etching bearings can be an issue sometimes with high switching frequencies and very cheap drives. If the cables are very long, standing waves can shred the cable insulation unless filtering is used to eliminate the problem. The only remaining area where DC drives still have an advantage is if you need to be able to creep at very low speeds (<50 RPM), an AC motor will tend to "cog" on you where a DC motor doesn't have this problem. In this operating region, a stepper motor is superior in any case. Edited 11 Aug 2008 by paulengr

[splicer480 0]

Thanks for all the help Paul. I would love to change this system out to an AC system but the powers that be just don't want to spend that kind of money right now. Heck they won't even spring for just a newer DC drive, something like a flexpak 3000 until they see if I can get this thing running by the end of the week. I had a heck of a time finding the (6) 410403-60AW SCR's that I need for this dinosaur unit. I found them for $90.00 US online. I only found two more sources that could get them and they were OUTRAGEOUS to say the least and had like a 3 week lead time. Personally, I don't like dumping that kind of money into this old unit but the higher ups want me to fix this old hunk of junk AND by the end of this week if possible. I've done some more checks on the (3) Power Modules and found (2) shorted to ground SCR's. The Gate reads full continuity to the anode and cathode. These (2) grounded SCR's were on the (2) phases where the fuses blew. I'm going to do the flashlight continuity test with all (6) tomorrow and see what I get. I've also got to check a couple of MOV's and capacitors that help with surge protection in each unit. I will be getting the 300 HP Motor back late tomorrow or early Wednesday. The (6) SCR's have been overnighted so I should get them Tuesday. One way or the other I'm gonna try to fire this beast up by late Wednesday or early Thursday and see what I get. Edited 14 Aug 2008 by splicer480

[paulengr 44]

It is quite common to find SCR's blown in pairs. Once one fails, the surge current destroys the other one until the fuses take it out of service. You may want to call Gould-Shawmut or Bussmann and change the fuses to something faster. Semiconductor protection fuse technology has been moving along in the last few years. Consider this lucky. I nursed a Southcon drive along for a couple years. As far as I know, I had the only remaining "print" and it was darned little to go on. This drive was built in the early 80's with old LSI chips like 555 timers and everything was socketed on the control board. With SCR's, you really only have to match up 3 specs. You don't need the exact SCR. First, it has to physically fit. Second, the current/voltage specs on the anode/cathode have to be rated for your use...this is almost never an issue and usually you can easily oversize. Third, the gate specs have to be close enough that the driver board can successfully trigger the SCR. Usually what I've found with "hockey pucks" is that the technology has advanced so much over time that the old stuff is so outdated that usually the physical size is the challenging part. If you can oversize the SCR's enough to make them physically fit in the heat sink framework without exceeding the trigger requirements, you can usually end up with an SCR that is very overpowered for your needs but will tend to outlast the originals. If someone is unwilling to pay for a new DC drive, then forget about AC. Checking some quick numbers, a new 300 HP DC drive is $22,000 list from Baldor. They usually give to 10-15% discount on these. A new Baldor 300 HP DC motor is $61,000. No idea how much discount you can get on that but the price sounds about right since the last time I checked quotes on a 150 HP DC motor about 3 years ago, it was close to $30,000. For comparison a 300 HP AC motor from Baldor is list at $32,000. Baldor is usually 3 times higher than competitors with comparable or better motors so the real price is going to be closer to $10K-$15K when you shop around. If you need just volts/Hz or sensorless vector, AB's Powerflex 400 is $18K list. Usually you can get this for about 60% of list or about $12K. The Powerflex 700 (full vector control) is $25K list, again probably 60% of that or about $15K in reality. If you are only using the drive for a soft start, then buy a real soft start and figure on cutting the drive costs in half. The rebuild cost for the AC motor will be about $3K and if you treat it right, it should last 5-10 years. The ultimate limiting factor will be the bearings. People get totally hung up on insulation degradation but if you treat it right and try to get the best insulation you can afford, I've always had bearing failures before insulation failures. You'll need to rebuild the DC motor every 2-3 years (commutators=more vibration) and the price will be about $5K. You have to put in about 12 pumps of grease on the AC motor every 6 months if you expect these life spans (the old rule of thumb for minimal greasing...1 pump per 25 HP, minimum of 1 pump). You need to open the DC motor up once every 2-4 weeks and inspect all the brushes for wear and tension as well as inspecting the commutator and cleaning as needed, plus greasing every 6-12 months if you expect this kind of life span. These are educated guesses on the numbers based on experience with paying the bills for motor rebuilds and buying motors in these size ranges.

[P Daniil 0]

In addition to the excellent posts above, a DC motor driven by a four-quadrant SCR drive still has an advantage when braking: All the generated braking energy flows back to the supply line during normal SCR commutation offering practically limitless braking capacity (effectively driving the grid´s infinite inertia) as well as superior power efficiency (no energy having to be dissipated in a resistor as in the case with AC drives). So if the application involves significant braking energies, DC drives have still one thing going for them.

[OkiePC 14]

My experiences with Flexpak drives are quite old but I remember them having (thyristors I think, but may have been SCRs) that if fired at the wrong time, could blow incoming fuses, bacically shorting two of the three phases together through the hockey pucks. This was caused by a problem with the replaceable gate driver board. It was also possible to test the thyristors in place to a limited degree, and also possible to view the gate driver pulses with an o-scope to compare the pattern with an example from Reliance, and look for missing pulses or pulse timing issues. I don't recall the specifics of how to perform those tests. I remember having a thyristor that tested good on the bench that would falter under higher loads, when viewed with a scope we could see it's pattern did not match the other pulses which resulted in a speed error fault, but no blown fuses... PC

[splicer480 0]

Success Well the SCR's did it. Got the motor back late Thursday afternoon wired everything back in and viola, up and running. Thanks for all the help guys. This fix basically cost about $3100.00. It was $2500.00 to get new brushes and brush holders, along with machining the end housings back into specs so the new bearings would fit right and of course the labor involved. The SCR's were roughly $600.00 for all (6) After installation, I took some DC amp readings with the motor running at about 70% current and got 324 amps on each leg A+ A-. The incoming AC line current was about 245 amps on all 3 legs. The DC field voltage is about 315 Vdc continuous. By the way this motor runs a very large plastic blow molding machine. It turns a large screw that moves 450 degree plastic through a barrel and into a 75lb head for shooting the plastic. It sure was nice to see this thing making parts again after a LONG week of downtime. Thanks again for all the help.