New MKS SBASE Smoothieware-compatible board from China

If it wasn't plugged into the right pins, it wouldn't read anything at all. But yeah there is definitely some bad stuff going on here and I'm trying to figure out if is a problem in the board design itself or it my combination of parts is causing the problem. To do that I'd need to pull the RAMPS controller and power supply from my other printer and use that to power the hotend to see if it behaves the same way, and I really really don't want to do that.

After looking at the EPCOS 100k thermistor, even it is only 5 ohms per degree C at 230C so I guess the Semitec thermistor is ok. Not sure why 4.7k seems to be the defacto standard for the other half of the resistor divider on every printer control board though when a lower value would push the temperature range higher. With 4.7k, the peak sensitivity is at 95C. With a 1k resistor, that could be pushed up to 148C. But I'm sure there's a reason I'm overlooking.

In any instance I think maybe this SBASE board may be unusable for me. I'm interested if others have the same experience with their thermistors being very noisy at hotend temperatures.

If you had the one side of the thermistor connected to the pin that is connected to the 4K7 resistor and ADC input, and the other side connected to the *wrong* ground pin, that could explain the problem. But if you are using a 2 pin connector that the board provides for the purpose, then you won't have done that.

RepRapPro originally used a 1K resistor instead of 4K7 on the Duet. The trouble was that with a 100K hot end thermistor, the error at room temperature could be high enough to trigger a low temperature error. So they changed to 4K7 like everyone else uses in later Duets.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I was using the 2 pin connector, yeah. It would be difficult to tell if they've got all the grounds running together because I don't have a schematic or layout file and a multimeter is just going to tell you they're connected. I'm waiting on my bonafide Smoothieboard for the comparison so now I am really anxious to get it.

Interesting tidbit about the 1k on the Duet! That does make sense because at room temperature you're pretty close to bottoming out the ADC (or topping it out actually). For those curious, I was just looking at the curves in a tool I use for comparing thermistors for my other project Semitec Thermistor Resolution. Since you're a wealth of useful information, do you know why there isn't a resistor between the thermistor input and the 10u capacitor? The capacitor I am guessing is there to be a heavy low-pass filter, but the way it is configured it is only filtering out noise on the AVCC, and the thermistor can return noise on its wire (which is likely considering it is usually strapped to the heater PWM power line).

Edited 1 time(s). Last edit at 07/06/2015 04:12PM by CapnBry.

I've had no such issues in maintaining hot end temp, and so far tested it with a Pico, a e3d v4 and a makergear v3b. Also auto tune seems to work fine for me.

i hooked up a generic red board type 12864 "reprap GLCD" to the EXT1 and EXT2 headers and it didnt work
and i set the config appropriately. any ideas?

I've figured it out, the connectors are backwards. pin 1 is pin 10 on both exp1 and exp2, so you have to cut / file off the tab off the connector and turn it around, and it works.

Quote
innkeeper
I've had no such issues in maintaining hot end temp, and so far tested it with a Pico, a e3d v4 and a makergear v3b. Also auto tune seems to work fine for me.

That's very odd. I removed the SBASE and connected the exact same printer in the exact same configuration to a Smoothieboard and the temperature is rock solid. My SBASE is almost unusable as a 3D printer controller due to the noise making the nozzle temperature jump all around. I'm going to try updating the firmware because the smoothieboard has a much newer version of the firmware, which maybe adds some software filtering to the ADCs.

I didn't have the issue with the LCD connectors being backward, the tab was on the correct side for me (to match the slot which faced the outside edge of the SBASE PC.

Edited 1 time(s). Last edit at 07/13/2015 10:44AM by CapnBry.

FWIW, I have gotten cables with the LCD panels with cables both ways, it appears that the Chinese have refused to accept anything as a standard . . .

- Tim

Quote
tadawson
FWIW, I have gotten cables with the LCD panels with cables both ways, it appears that the Chinese have refused to accept anything as a standard . . .

SainSmart shipped me a GLCD ribbon cable with one end crimped backwards. I popped the 5V voltage regulator on my Megatronics board because I didn't notice the red stripe was on the wrong side of one plug.....

Quote
CapnBry
My SBASE is almost unusable as a 3D printer controller due to the noise making the nozzle temperature jump all around. I'm going to try updating the firmware because the smoothieboard has a much newer version of the firmware, which maybe adds some software filtering to the ADCs.

I tried installing the newer firmware from the Smoothieboard to the SBASE (which is awesomely simple to do) to see if it worked any better but it does not. The temperature still jumps around by more than a degree, which leaves the software struggling to keep the temperature stable. For fun I tried hooking it to the bed temperature input and it jumps around as well so I think it is an inherent flaw in this specific board or the layout in general. The variation can be kept small by selecting conservative PID constants, I used P=30 I=1 D=15.

Quote

FWIW, I have gotten cables with the LCD panels with cables both ways, it appears that the Chinese have refused to accept anything as a standard . .

My GLCD was missing a surface mount resistor so the backlight didn't work reliably. Because it is an open source design I just looked at the schematics and tracked it down and said "Wait a second, there's no resistor here". It was a size I happened to have on hand so I just soldered a new one in. As I did with the head-mounted magnifier, I found the resistor that was supposed to be there a couple of inches away stuck to the LCD pins. Good thing it had the holographic sticker that said "QC" on it so I know it passed all the tests!

Edited 1 time(s). Last edit at 07/14/2015 09:36AM by CapnBry.

To CapnBry ,
Thanks for your trying our MKS SBASE Smoothieware-compatible board and find the problem in maintaining hotend temperature.After we compare the origenal Smoothieware oard and our MKS Sbase. We find that when we use the DC-DC converter R-78E5.0-1.0 as origenal Smoothieware, it is OK. And our DC-DC conver L5970D is using in our MKS series Board for 2 years with Marlin and other firmware, it is OK. So we think maybe the Smoothieware firmware cause the problem. We are trying to check the firmware further and solve the problem.

---

Guangzhou Qian hui MDt InfoTech Ltd
www.makerbase.com.cn
leiochen@gzmqe.com.cn

I am also having this same issue with the temperature jumps. Mine is also + or - 3 degrees. I am using the most recent version of Smoothieware firmware. I have E3D v6 Hotend which uses a Semitec 100k thermistor.

Thanks for the tip on the tuning CapnBry...I will give that a try. MakerBase seems to be much more conscientious than other Chinese companies I have dealt with. Hopefully they can sort this out.

UPDATE 7/17: Auto tuning didn't help. I tried your values CapnBryn...still wild.

UPDATE 7/18: I hooked up the heated bed using some 100k thermistors I ordered here [www.aliexpress.com]. The temperature is holding solid with no problem. It does jump up/down one degree every once in a while but no big jumps like on the hotend.

Config for the bed:

temperature_control.bed.enable true #
temperature_control.bed.thermistor_pin 0.24 #
temperature_control.bed.heater_pin 2.5 #
temperature_control.bed.thermistor Semitec # see [smoothieware.org]
#temperature_control.bed.beta 4066 # or set the beta value

temperature_control.bed.set_m_code 140 #
temperature_control.bed.set_and_wait_m_code 190 #
temperature_control.bed.designator B #

No Special PID values. I will play with the hotend a bit more to see if I can get stable temps.

Edited 4 time(s). Last edit at 07/18/2015 12:05PM by GeoDave.

I also got my board yesterday, printing like a champ (E3D V6 + Volcano 1.0MM + EZStruder). Also did some PID Auto tuning and jump 2/3 +/- degrees about. Running the latest smoothieware firmware. Wanted to try 626 Pilot's auto calibration smoothieware firmware but it had issues with the head hitting to hard on the bed or just rubbing against it. Smoothieware Edge calibration works amazingly fine however with my FSR's. Dupont connectors fit fine however the JST XH headers are probably best if you plan on permanently mounting this somewhere. The web interface is same as smoothie as usual and works fine. Soldering job is good and overall the quality is good. Package was actually pretty beat up however it was packed really well. Might buy another for my smaller printer but i might just get a duet and see how that is as ive heard great things about it.

Quote
flyway97
To CapnBry ,
Thanks for your trying our MKS SBASE Smoothieware-compatible board and find the problem in maintaining hotend temperature.After we compare the origenal Smoothieware oard and our MKS Sbase. We find that when we use the DC-DC converter R-78E5.0-1.0 as origenal Smoothieware, it is OK. And our DC-DC conver L5970D is using in our MKS series Board for 2 years with Marlin and other firmware, it is OK. So we think maybe the Smoothieware firmware cause the problem. We are trying to check the firmware further and solve the problem.

It's not the firmware : this problem has been reported only with your board, not with the several other Smoothie-compatible boards.

I wanted to check if you hadn't routed the analog ground incorrectly, but can't find any of the source files for your board anywhere. Is your board Open-Source ? If so where are the files ?

Edited 1 time(s). Last edit at 07/22/2015 05:23AM by arthurwolf.

Quote
GeoDave
UPDATE 7/18: I hooked up the heated bed using some 100k thermistors I ordered here [www.aliexpress.com]. The temperature is holding solid with no problem. It does jump up/down one degree every once in a while but no big jumps like on the hotend.

There's two reasons those 100K thermistors work better. At 100C (typical bed temperatures), there's a lot more resistance than at 230C hotend temperatrures. This means less variation in the temperature. They're also 3950 beta compared to 4265 beta for the Semitec, which means they have less resistance change per degree C. If you measure the resistance at 100C you're probably getting like 6500 ohms, but at 230C it would be closer to like 400 ohms which makes a big difference in the noise.

@flyaway97: It definitely is not a software problem, the issue is in hardware. I can take the SBASE board and hook it to the same everything and the temperature jumps around on the hotend. The same hardware switched for a real smoothieboard is perfectly stable. This is true with both the firmware that came on the SBASE board originally, or the latest Smoothieware. Both firmwares work fine on a Smoothieboard, both exhibit the problem on the SBASE.

My Smoothieboard's 5V supply comes from an Murata OKI-78SR-5 for the record.

to arthurwolf ,CapnBry : We have reseach the problem for some day. you are right,this problem is not about firmware. The most possibility come from voltage Interference. we will change the DC-DC and pay attention to analog ground and test if can solve this problem. Our board Circuit diagram can be found at here

flyway97: if there is a retrofit we can do to reduce the temperature fluctuation issue please kindly post it. Other then the temperature measurement fluctuations, and a preference for wishing the connectors were DuPont type, I like the board design

Any updates? I am waiting to purchase a board but need to know if the problem is resolved.
thanks

Quote
lmclaren
Any updates? I am waiting to purchase a board but need to know if the problem is resolved.
thanks

Probably won't be fixed till v1.1 of the board gets made. Which I believe will include a reserved spot for thermocouples, tempurate fix, heatsinks, and some other stuff. I think someone said sometime in August but if you need a board now, the current version still works great, I have one and use it on my daily printer with an MKS tft28 touch screen LCD.

Edited 1 time(s). Last edit at 07/30/2015 08:20AM by frodriguez96.

thanks for the update

thanks frodriguez96 , lmclaren: we have found the reason of temperature problem come from DC-DC, L5970D 's Switching Frequency is not enough , we are testing new DC-DC now . This problem will be solve on on MKS SBASE V1.1 , and will add pins used for Max31855 thermocouples ,add heatsinks on Drv8825, just as frodriguez96 said. . MKS SBASE V1.1 will be saled about August or Setpember.

Quote
flyway97
add heatsinks on Drv8825, just as frodriguez96 said. .

Those chips are not designed to be cooled by heatsinks. Read the datasheets.

They should be cooled by the PCB itself, which should have enough layers/copper weight to properly cool them.

In any case, if you end up having to use heatsinks due to bad PCB design, they shouldn't be used on the top of the chip ( where they have little effect, and can even in some cases cause the chip to bend and un-solder ) but on the bottom of the board, bellow the chips.

Thanks Flyway97, I will be keeping an eye out for it.

Quote
flyway97
we have found the reason of temperature problem come from DC-DC, L5970D 's Switching Frequency is not enough , we are testing new DC-DC now . This problem will be solve on on MKS SBASE V1.1 , and will add pins used for Max31855 thermocouples ,add heatsinks on Drv8825, just as frodriguez96 said. . MKS SBASE V1.1 will be saled about August or Setpember.

flyway97: is there any modification that we can do to the 1.0 boards to fix this issue or if not, anything that can be done for the existing 1.0 board owners that have the issue?

Quote
innkeeper

Quote
flyway97
we have found the reason of temperature problem come from DC-DC, L5970D 's Switching Frequency is not enough , we are testing new DC-DC now . This problem will be solve on on MKS SBASE V1.1 , and will add pins used for Max31855 thermocouples ,add heatsinks on Drv8825, just as frodriguez96 said. . MKS SBASE V1.1 will be saled about August or Setpember.

flyway97: is there any modification that we can do to the 1.0 boards to fix this issue or if not, anything that can be done for the existing 1.0 board owners that have the issue?

I'd like to second this. Although I can still print with the 1.0 board, would be nice if there was a quick fix maybe something to replace or solder on or anything you guys have in mind for current owners. I mean even some sort of credit to spend on a v1.1 would be nice but any quick fix on the current board would be better.

innkeeper , frodriguez96 : change the DC-DC from L5970D to R-78E-5.0 can solve this problem on V1.0.

arthurwolf : yes, DRV8825 and A4988 's radiating is on the chip bottom, so the best method to cool them is by the PCB, so we have carefuly deal with it on our PCB. But if the current up to 1.5a or higher, the temperature is also too high in spite of good PCB design. So add heatsink is useful(not include those small heatsinks ) .The best way to cool is use fan to blow the board, at this time heatsinks is useful too.

Quote
flyway97
innkeeper , frodriguez96 : change the DC-DC from L5970D to R-78E-5.0 can solve this problem on V1.0.

arthurwolf : yes, DRV8825 and A4988 's radiating is on the chip bottom, so the best method to cool them is by the PCB, so we have carefuly deal with it on our PCB. But if the current up to 1.5a or higher, the temperature is also too high in spite of good PCB design. So add heatsink is useful(not include those small heatsinks ) .The best way to cool is use fan to blow the board, at this time heatsinks is useful too.

Could you please show and or explain how the component is replaced? According to the BOM and circuit diagram L5970AD is located at U2 on the circuit diagram. I found it just fine and I can solder SMD components but that's as far as my knowledge goes. How do I replace L5970AD, which has 8 legs, with R-78E5.0-1.0 which only has 3 legs?

1. take off component D5 and F3;
2. Connect R-78E-5.0's 3 pins to the 3 solder joints as shown in the picture,
(pin1 to label 1, pin2 to label 2, pin3 to label 3),then it can work well.

Quote
flyway97
1. take off component D5 and F3;
2. Connect R-78E-5.0's 3 pins to the 3 solder joints as shown in the picture,
(pin1 to label 1, pin2 to label 2, pin3 to label 3),then it can work well.

Thank you! My MKS SBASE should show up in a week or so. I ordered a few R-78E5.0-1.0 from Digikey so I'll report back once I have everything setup.

Quote
Boxcar

Quote
flyway97
1. take off component D5 and F3;
2. Connect R-78E-5.0's 3 pins to the 3 solder joints as shown in the picture,
(pin1 to label 1, pin2 to label 2, pin3 to label 3),then it can work well.

Thank you! My MKS SBASE should show up in a week or so. I ordered a few R-78E5.0-1.0 from Digikey so I'll report back once I have everything setup.

Post a how to for us old timers with feeble eye sight. Please?