Hello Khalid, thank you for your valuable support to the astronomical DIY community. I made an Onstep Blue pill board but it worked strangely, I could only drive one axis. After a lot of research I came across a forum where you explained how many of these CPUs did not meet the stated memory amounts and how it could be remedied by replacing the chip with a RobotDyn STM32F303CCT6. Following this suggestion I replaced the chip but now I have difficulties with the Arduino IDE settings for programming; the error I get is "No debug probe detected". , the same error appears in the STM 32 CubeProgrammer. I am programming from the serial interface of the OnStep circuit ( USB UART CP2102 ) and not directly on the CPU. Thank you in advance for any suggestions you can give me. Ugo.
I don't have much to say here. A RobotDyn F303CC should work for both axes, at least with the original OnStep 4.x. In fact, the original Blue Pill should also work well with that version. The limitation is that slewing speed may be limited. Try to ask in the OnStep group, see if someone has more info to add.
Hello Khalid, I bought the Blue pill kit from Dave some time ago and am interested in the upgrade, but I can't find this Robodyne module anywhere. Has it been discontinued? Do you know of a source? And thank you for your efforts to continue to improve the OnStep system, it is really amazing what this does to bring Goto astronomy to the masses.
Unfortunately, the pandemic supply chain disruption has hit hard, and chip supply is one casualty. Before the pandemic, virtually all Blue Pills were 128K, and therefore worked with OnStep. After the pandemic, the 128K are rare and not easy to find. Same for RobotDyn. They seem to have trouble getting the chip. My only advice is to keep checking for when they have supply.
Hello Khalid, thank you for your efforts to make astronomy affordable. I am programming a second OnStep Blue pill controller (The first went on a GEM experimental mount that I designed using 3D printed parts). The current version of the Arduino programming software, 1.8.19, has a feature to overclock the STM32 to 128MHz. Is this safe for using with OnStep? I am asking because I would like to use faster microstepping than possible at the 72MHz base frequency. I am converting my Celestron Super C8+ (the one with the excellent Byers worm drive for the RA) to GoTo capability as it would help with positioning accuracy and tracking. FYI, a 400 steps/rotation is not available in the type of steppers I am using (a NEMA 14, hi torque for the RA and a NEMA17 with 50:1 drive reduction for the DEC). The beauty of using the NEMA 14 for the RA is that it fits inside the RA base without the need for a permanent modification (I would be happy to share some pictures and my procedure if you like). The DEC is very tricky due to the design of the Fork arms, but I have designed some 3D printed parts that I think will overcome this difficulty.. BTW, where can I find the wiring schematic for connecting the RJ45 connector wires to the stepper motor wires? I can't seem to find it. Also, have you seen the Black Pill available anywhere yet? Amazon has this for sale: www.amazon.com/gp/product/B09V7GCMRX/ref=ox_sc_saved_title_4?smid=A2OEGRG2ZM4R5C&psc=1 I don't suppose that this would work and at 100MHz it is not that significant of an upgrade. Can you confirm?
No idea if overclocking is safe or not. Instead of that, you can try MICROSTEPS_GOTO 2 levels from the one you use for tracking (e.g. if you are tracking at 1/16 or 1/32, then try 1/4 or 1/8 for MICROSTEPS_GOTO. That should speed up your slewing speed. For the RJ45, just combine each two adjacent wires by soldering them to one wire on each side of the connection (motor and controller)
Don't know if that module from Amazon would work. Compare the pin labels on it to the Blue Pill. If ALL OF THEM match, you can try it. If there is a difference, don't bother.
@@khalid_ba Thanks for the amazingly quick response. Regarding the RJ45, I know about the adjacent pairs being wired together, but when connecting to the stepper, which pair goes to A+, A-, B+, B- on the stepper wiring?
@@MikeHammer1 There are several ways of doing this. Some people go the trial and error route (since you can't harm the motor by wrong wiring, it will just be noisy and would stall). You can try Dupont female/female wires from the controller to the motor until you get the wires right, then solder accordingly. Others just wire first pair to the first motor wire, the second pair to second motor power, ...etc. and then if the motor is going the wrong direction, they would use the REVERSE parameter in the firmware to make it go the right way. Take your pick ...
I have converted eq5 elwith blue pill mount and im running it on 4°/sec and it seems considerably faster than your mount on this video. Payload is approx 6kg. Why is that so? Maybe it seems slow because there is no telescope on? Which stepoer drives do you use here? As soon i return home from a bussiness trip, i will make a video of my mount running. Btw, i have to thank you and Howard for this amazing project again! Here is the video with 2,5°/sec: ruclips.net/video/DpXOHIMBZkk/видео.html
I use LV8729, 1/16 microstep, and 18:1 + 180:1 reduction. Maybe it appears faster because there is no long object on it (optical tube, or counterweight shaft)? Or it may be that I need to wipe the EEPROM, as it is storing stuff from when I used it with the F103. Can you time a long slew, with the starting Alt/Az and the ending Alt/Az. I will do the same to see whether it is the same or not.
@@khalid_ba i will make some tests tommorow when im back at home. But despite that there is no longer object on the mount, it seems slower than mine at 4°. Will do a long slew tommorow and check the time needed to finish it and report it here.
@@leviner18 Make sure you not the Alt-Az of the start, and Alt-Az of the finish, and the length of time. Doing 3 tests like that will also help. I will repeat these on my mount to compare.
@@khalid_ba heres the video of slew east west ruclips.net/video/gInEiKPTiZg/видео.html It seems much faster than yours 4°/sec. Maybe is mine faster than what it says on the onstep app?
@@leviner18 Yes, mine was not fast in my video, but when I tested again yesterday, it was much faster. I don't know what I changed. Maybe pressed Slower then Faster, or something. So in the end this confirms that the F303 can go very fast, faster than what most people would use it at, given inertia, load on the mount, ...etc. I will stick to around 3 deg/sec.
![Natanael Cano - Amor Eterno [Official Video]](http://i.ytimg.com/vi/EO4prCs6WpM/mqdefault.jpg)
Interesting. Looking forward to more on this!
Hello Khalid,
thank you for your valuable support to the astronomical DIY community. I made an Onstep Blue pill board but it worked strangely, I could only drive one axis. After a lot of research I came across a forum where you explained how many of these CPUs did not meet the stated memory amounts and how it could be remedied by replacing the chip with a RobotDyn STM32F303CCT6. Following this suggestion I replaced the chip but now I have difficulties with the Arduino IDE settings for programming; the error I get is "No debug probe detected". , the same error appears in the STM 32 CubeProgrammer. I am programming from the serial interface of the OnStep circuit ( USB UART CP2102 ) and not directly on the CPU. Thank you in advance for any suggestions you can give me.
Ugo.
Hello Khalid,
maybe I didn't specify that I desoldered the chip from the board and soldered the STM32F303CCT6.
I don't have much to say here.
A RobotDyn F303CC should work for both axes, at least with the original OnStep 4.x.
In fact, the original Blue Pill should also work well with that version. The limitation is that slewing speed may be limited.
Try to ask in the OnStep group, see if someone has more info to add.
Hello Khalid, I bought the Blue pill kit from Dave some time ago and am interested in the upgrade, but I can't find this Robodyne module anywhere. Has it been discontinued? Do you know of a source? And thank you for your efforts to continue to improve the OnStep system, it is really amazing what this does to bring Goto astronomy to the masses.
Hi Mike, I am facing similiar problems. Seems like they sold out everywhere :(
Unfortunately, the pandemic supply chain disruption has hit hard, and chip supply is one casualty. Before the pandemic, virtually all Blue Pills were 128K, and therefore worked with OnStep. After the pandemic, the 128K are rare and not easy to find. Same for RobotDyn. They seem to have trouble getting the chip. My only advice is to keep checking for when they have supply.
Hello Khalid, thank you for your efforts to make astronomy affordable. I am programming a second OnStep Blue pill controller (The first went on a GEM experimental mount that I designed using 3D printed parts). The current version of the Arduino programming software, 1.8.19, has a feature to overclock the STM32 to 128MHz. Is this safe for using with OnStep? I am asking because I would like to use faster microstepping than possible at the 72MHz base frequency. I am converting my Celestron Super C8+ (the one with the excellent Byers worm drive for the RA) to GoTo capability as it would help with positioning accuracy and tracking. FYI, a 400 steps/rotation is not available in the type of steppers I am using (a NEMA 14, hi torque for the RA and a NEMA17 with 50:1 drive reduction for the DEC). The beauty of using the NEMA 14 for the RA is that it fits inside the RA base without the need for a permanent modification (I would be happy to share some pictures and my procedure if you like). The DEC is very tricky due to the design of the Fork arms, but I have designed some 3D printed parts that I think will overcome this difficulty..
BTW, where can I find the wiring schematic for connecting the RJ45 connector wires to the stepper motor wires? I can't seem to find it.
Also, have you seen the Black Pill available anywhere yet? Amazon has this for sale:
www.amazon.com/gp/product/B09V7GCMRX/ref=ox_sc_saved_title_4?smid=A2OEGRG2ZM4R5C&psc=1
I don't suppose that this would work and at 100MHz it is not that significant of an upgrade. Can you confirm?
No idea if overclocking is safe or not.
Instead of that, you can try MICROSTEPS_GOTO 2 levels from the one you use for tracking (e.g. if you are tracking at 1/16 or 1/32, then try 1/4 or 1/8 for MICROSTEPS_GOTO. That should speed up your slewing speed.
For the RJ45, just combine each two adjacent wires by soldering them to one wire on each side of the connection (motor and controller)
Don't know if that module from Amazon would work. Compare the pin labels on it to the Blue Pill. If ALL OF THEM match, you can try it. If there is a difference, don't bother.
@@khalid_ba Thanks for the amazingly quick response. Regarding the RJ45, I know about the adjacent pairs being wired together, but when connecting to the stepper, which pair goes to A+, A-, B+, B- on the stepper wiring?
@@khalid_ba Got it, thanks!
@@MikeHammer1
There are several ways of doing this.
Some people go the trial and error route (since you can't harm the motor by wrong wiring, it will just be noisy and would stall). You can try Dupont female/female wires from the controller to the motor until you get the wires right, then solder accordingly.
Others just wire first pair to the first motor wire, the second pair to second motor power, ...etc. and then if the motor is going the wrong direction, they would use the REVERSE parameter in the firmware to make it go the right way.
Take your pick ...
I found this board on Amazon. Would it also work as a drop in?
Songhe STM32F401 Development Board STM32F401CCU6 STM32F4
No. It is not pin compatible.
i tot can upgrade my blue pill to black pill, yet it seem no where to be found to purchase ...
The pandemic has cause both the original Blue Pill STM32F103 and the newer RobotDyn STM32F303 to be unavailable.
hello, do you have an address where we can buy this product directly as ready plug and play?
George Cushing sells them. See his web site. www.stmbluepillkits.com/p/order-kit.html
Onstep 1.5 or 1.i version support?
As I said in the video, support for the F303 is only in OnStep 4.20 and higher
@@khalid_bamy bad... I mean the board version.
@@alexchong1979 The RobotDyn F303CC should work with all PCB versions
I have converted eq5 elwith blue pill mount and im running it on 4°/sec and it seems considerably faster than your mount on this video. Payload is approx 6kg. Why is that so? Maybe it seems slow because there is no telescope on? Which stepoer drives do you use here? As soon i return home from a bussiness trip, i will make a video of my mount running. Btw, i have to thank you and Howard for this amazing project again!
Here is the video with 2,5°/sec:
ruclips.net/video/DpXOHIMBZkk/видео.html
I use LV8729, 1/16 microstep, and 18:1 + 180:1 reduction.
Maybe it appears faster because there is no long object on it (optical tube, or counterweight shaft)?
Or it may be that I need to wipe the EEPROM, as it is storing stuff from when I used it with the F103.
Can you time a long slew, with the starting Alt/Az and the ending Alt/Az. I will do the same to see whether it is the same or not.
@@khalid_ba i will make some tests tommorow when im back at home. But despite that there is no longer object on the mount, it seems slower than mine at 4°. Will do a long slew tommorow and check the time needed to finish it and report it here.
@@leviner18 Make sure you not the Alt-Az of the start, and Alt-Az of the finish, and the length of time. Doing 3 tests like that will also help. I will repeat these on my mount to compare.
@@khalid_ba heres the video of slew east west
ruclips.net/video/gInEiKPTiZg/видео.html
It seems much faster than yours 4°/sec. Maybe is mine faster than what it says on the onstep app?
@@leviner18 Yes, mine was not fast in my video, but when I tested again yesterday, it was much faster. I don't know what I changed. Maybe pressed Slower then Faster, or something. So in the end this confirms that the F303 can go very fast, faster than what most people would use it at, given inertia, load on the mount, ...etc.
I will stick to around 3 deg/sec.