Are my Leadshine DCS810 brushed Servo Drives fake

Are my DCS810 the real thing or low quality rip off copies? They all have the Leadshine name on them but Leadshine appears to suggest on their web site that rip off copies are being sold on the internet. The issue for me is a reliability problem with the ones I have purchased and I have no way of knowing if they are the real thing or fake copies. For reference I have more than 40 years’ experience behind me as an electronics technician and many of those years involved large servo control systems. Thus I do my own repairs when parts needed are available and have all required test equipment.

I have purchased a total of 6 Leadshine DCS810 brushed servo drives on Ebay, 80VDC, 20A peak with differential encoder inputs of which only one has survived so far. Some have failed shortly after power on with no motion while others failed during motion. The failed ones have been repaired more than once now by me and the same thing occurs again after a period of time. At least two that arrived appear to have been repaired before I received them due the presence of thermal compound. Usually the failures that I have run into required replacing all four output MOSfets, failed 33 ohm mosfet driver resistors and in some cases the 2003 half bridge drivers. The chip 0.02 ohm current sensing resistor blows up on some, and the 0.05 ohm resistor that is on the DC power input opens on occasion while melting part of the connector next to it and part of the plastic cover. Some of the chips in these units have had their tops ground down a lot to hide their identification markings. Some look like this was done very deep with a dremel tool which likely caused a lot vibration on the inside of the chips. On others it looks like the chip ID was blasted off with blasting media in some fashion..

Three of these were purchased about a year ago and never used them until recently, and three more units a month or so ago to replace a damaged one that failed after a unplanned high velocity motion, the others were spares for a while. None of the units have never seen a single full day of operation before they failed.

Today was the latest failure. It was on my Z axis and it failed with shorted MOSfets, burned open 0.05 resistor and an internally shorted FPGA chip which effectively shorted out the 3.3v internal regulated power supply distribution section and placed the on board power supply in a over current condition. With this kind of failure it is only good for untrustworthy parts or the trash can. This failure occurred shortly after power up and just sitting still.

The DCS810 is stated as having internal current limiting at 20 amps but I’m not sure that it works very well when activated in part because I can’t see its factory programmed values for the programmable current modes due to no password being made available to do that. I would actually like to lower the factory setting since I do not need a lighting fast acceleration or speed.

After the first failure occurred I added 1 ohm series resistors to my drive motors to reduce the chance of future failures due to over current but that did not help. My Motors run off a 48vdc 25 amp regulated switch mode power supply with protection against voltages that may be produced by motors acting as generators. The power supply is fed by a 230VAC power source. So far I have had an 84% failure rate with the things marked DCS810. I think the last one has at least a 84% change of failing if I were continue to use it.

I tried to contact Leadshine on their web site, but it appears to indicate that if I did not buy directly from them or their selected outlets they could not help with support due to many being sold on the internet that are not the real thing. I suspect there is truth in that especially when I consider my unexplained failures.

I’m not sure if I have fake copies or not but reliability is the prime concern so my larger milling machine is back in full manual mode for a while longer. Due to these disappointing failures that are potentially being caused by fake drivers with low quality parts I’m looking to replace the these drivers with units that I can depend on 100% of the time and with 20 amp or higher ratings. I have another machine that is Using RUTEX drivers with about capabilities for many years and have not had a failure since they were installed. I'm looking for something different for this machine.

Suggestions would be greatly appreciated.

After some thought I have decided to replace all the DCS810 drives with Rutex 2020 units. These will fit in my enclosure with only some cabling routing and connector changes being required.

I will also need to change the machines logic power supply to one that can supply the required 24vdc for the logic in the Rutex drives. The new logic power supply will need to supply 5vdc for the Mesa boards and 24vdc logic power for the drives. The 24vdc will also be dropped down to 12vdc via separate three terminal regulator for the limit, home switches and indicators on the control panel.

It will be so nice to have a drive reset function so power will not need to be cycled to reset a drive that may have entered a alarm condition.

If I had known about the secret pass word requirement for tuning current related variables for the DCS810 I would never have purchased them. I feel that access to those variables is absolutely required for proper and thus optimum tuning.

It looks like the Rutex 2020 drives will not be my choice due to having discovered that the US supplier that I have dealt with in the past stopped distributing Rutex items several years back.

My alternative to the Rutex 2020 was the newer CNCdrive model DG4S-16035. So it is going to be my choice for replacing the drives It will take a little more work to make changes to my enclosure and rewire but it wont require a logic power supply change.

The Rutex units I have on my 4 axis gantry mill have done well but are not mounted on the original mounting plates and the Mosfets and other items on the bottom of the board are attached to the new plate with screws and shoulder washers and individual silicon pads. They are no longer held against the mounting plate by circuit board pressure. Its not an easy change but it does prevent circuit board flexing and lets the Mosfets and voltage regulators have much improved contact for heat dissipation with the mounting plate and thus increases reliability.

I had purchased some blown Rutex 2020 drives several years back for potential repair and use. One of the things that surprised me on these was the very thin and flexible mounting bracket which is also the heat dissipation plate. Since I never repaired or used these I did not get around to make something that is not so flimsy like i did for the 2010's in my gantry mill. Usually I will machine a 1/4 inch thick mounting plate, drill and tap as required.

I ran into this particular model of drive while I was doing research on drives that were rated at about 35 amps, could handle motors with low DC resistance, run closed loop and also use the old step/dir interface. I do not know if these drives will suit my needs 100% but I did read some reviews of their drives that were replaced by this newer model. Those reviews indicated that the drives being reviewed worked well and that the company provided good support. The company responded rapidly to my inquiry which was nice. Its hard to judge how these drives will work for me until I get some in my hands and wired in.

I'm actually more used to working with drives that take a analog input with tachometer or quadrature feedback in addition to error correction between commanded and actual positions. But prior investments make me stick to step and direction which can also work fine when speed is kept within reason. None of my systems have stepper motors.

I had looked at the other sources mentioned for drives.

I just purchaced three of these DCS810 drives and one failed after 15 minutes into tuning it and another after about two hours. Im yet to and affraid to even power up the third yet. After opening them up the only thing i can visually see that got fried is the .05ohm output resistor on each, but am going to order new IRF540N mosfets at the same time. Would using a 5w resistor instead of the 3w thats there present any potential problems? That resistor is easy to change, but im not really equipped to replace SMD components. What variants of the IRF540 are acceptable to use? I am actually hoping that i can use a higher power FET... Any ideas / sugestions?

So yeah, my 75% failure rate seems about inline with your experience, and am also thinking of going to the DG4S-16035 drives.

To the best of my knowledge the 0.05 ohm 3 watt resistor (R96) is meant to act as a fuse. Increasing the wattage may make it more likely that the current sense resistor (RS1) will blow. If you heard a pop when the drives failed it is likely that RS1 is already open. If you replace the IRF540N devices you should be prepared to also replace some surface mounted parts. Most all my failures have caused R59 (33 ohms) to open up. It is near the on board logic power supply transformer and may be in a area of the circuit board that is more prone to flexing which can cause surface mounted parts to crack and thus fail. If a 33 ohm resistor (R56, R57,R58 and R59) has failed to a open condition or measures a higher in value then there is a good chance that a S2003 half bridge driver associated with such a failed resistor may also need to be replaced.

If you do decide to repair your boards I suggest that you use a bench current limited power supply after parts are replaced to make sure all is OK before attempting to use it in a machine. Based on my findings I think you will be disappointed after a repair since I have repaired each board a number of times and had them fail again for no apparent reason. I'm leaning towards the idea that the failures may be associated with bad firmware for which there will likely not be a cure. It is difficult to determine if a alarm condition that is poorly handled by internal processes is the culprit that starts the self destructive process. Such a error could potentially cause timing errors that allow the two sides of the half bridge outputs to both conduct at the same time which would lead to instant failure.

Based on the failures that I have seen I suspect that repairing these failed drives is not worth the effort and expense. I have ordered the replacement drives for my failed drives. The replacement drives are the DG4S-16035 drives that I mentioned before.

So i replaced the resistors in the two drives and they both blew again immediately, so something else is also bad. Finding and replacing it is beyond my abilities at this point. Looks like I'm gonna be writing these things off, unless you feel like taking a look at these things for me?

I couldnt help but power up my last working driver, and did get it working seemingly well while waiting for the resistors to show up. After testing the two with new resistors, I plugged the working one back in. I knew better than to do it, but I plugged it in to a live power source. It sparked at the plug and died... It appears to be RS1 the blew. Now the drive simply delivers full power to the motor at all times.

Seems its time to save up another $350 for some decent drivers.

I suspected that you would have the same results as I had. Unfortunately I feel it is not worth the time to attempt to repair these since I repaired some of mine multiple times. They will work for a while but always fail again.

I have attached a picture of my replacement drives (Yellow) for reference and size comparison to the DCS810 drives. The software for programming these drives looks a lot more professional than the DCS810 tuning software. The replacement drives also require a external programming module that interfaces the drive to a USB port. The programming module plugs into a small 4 pin connector on the drive. My replacement drives shipped from the manufactures distributor in Florida.

I'm still in the process of building a new enclosure which attaches to the side of my milling machine and hope to have it all operational by the end of April so that I can continue with the remaining configurations to be completed in LinuxCNC.

You happen to have any working ones left you wanna offload? I figure one of these would hopefully be good enough for my Z axis at least.

While I'm sure they'd be powerfull enough, did those new drives happen to tell you what thier maximum wattage is? Every spec sheet I can seem to find only outlines the 160v / 35a...but I'm certain there is no way they can actuallly handle 5600 watts.