Sanguish development

1. My USB connectors don't have those but I'll look into it and try to add it to the next version.

2. The USB oscillator is designed to be a crystal resonator for space reasons.

3. Thanks

4, 5, I agree, I'll adjust the components a bit to be more compatible.

6. Yes, Aux is for SD adapter. The SDSL adapter should be pin compatible.

7. Yes, I spaced them out a bit on 1.1, but it might still need more.

And one question: would it be possible to use Marlin on a TB6560 design board?

It's possible. You'll need to modify the firmware to lengthen the step pulse on time, and invert the enable line as explained in the wiki section on Sanguish firmware. As is, the step pulse is too short for these drivers. It's been done with other TB6560 boards but not yet as far as I know with Sanguish. Someday I'll get motivated and take a look at it and do some testing with Marlin.

repetier is all you need (: marlin fans will burn me lol... But I love Repetier...

Back to Betas: ceramic oscillators on a usb chip is looking for trouble. Trust me, been down that road before. Even with expensive oscillators

I'm on a bit of a break lately but I'm still puttering along slowly. The latest puzzle is the interference from the MCP 2200. I can run a board with external USB to serial converters only if I don't install the mcp2200 on board usb-serial chip. But if it's soldered on then the external serial adapters rarely work at all. I thought tying the MCP 2200's power to USB 5V would help,(V1.1RC2) as the chip would power off completely when disconnected but that didn't seem to make a difference. I'll be doing some more tests to isolate the source of trouble (ceramic oscillator, cap, chip, etc.)

Any insight would be helpful.

Merry Christmas

Sanguish 1.1 is out! I just sold a few and ran out of a couple of components so I won't have more kits to sell for a few days. The design is still either/or for USB-Serial. If you install the MCP2200 on board usb-serial converter then your external usb-serial cables likely won't work even though the header is available. So if you are planning on using some combination of cable or bluetooth(untested) then don't install the MCP2200. I figured this was a small problem since the vast majority will just use on board usb. That's pretty standard with many other boards like Sanguinololu and Gen7, so I'm not too concerned about it, but I will still be doing some testing to see if they can be made to coexist.

RC2 is the final release candidate. Files are on the wiki. You'll find a gerber folder in the release files. You can just send that folder to your favorite board house to get it if you want to go it alone. I have used both Seeedstudio, and iteadstudio and both have been fine. Very cheap, and not too long a wait. BOM and build instructions are a work in progress as yet.

Just wondering, I was busy with your design and TB6560 chips. And my chips continuously are heating up extremely fast when I apply the 5V. (I haven't tried putting the Vmot also on them, I found the heat during 5V startup alarming enough to stop experimenting.) What do you suggest? Have you ever encountered similar problems?

---

Eppur si muove!

Double check that the enable line is being held low until you want motors enabled. The main thing if I recall correctly is to have 5V up and stable before supplying Vmot, and then enable the drivers. If you let enable go high before 5V (and possibly Vmot) it can burn out the chip. The enable line is active high, which is opposite from Pololu drivers which have an active low enable line. Also make sure your current limiting resistors are sized appropriately. Other than that the only other thing I've run accross that will burn them out is a cross connected motor. If you short either coil A output to a coil B output it will fry.

For the cannonical power on sequence see the datasheet. But I think the enable line is the critical one. Make sure you don't enable till the chip is powered up and ready.

Other things to check would be cold solder joints. A floating ground pin or some other pin might cause something like that.

Bryan

Edited 1 time(s). Last edit at 01/26/2014 10:37PM by bryanandaimee.

I have a few more Sanguish 1.1 kits available from the current batch. I'm trying out the Emakershop.com website so they will be sold through that website.

Link to emakershop listing

Found something interesting! Remember my TB6560AHQ trouble? None of the suggestions helped. It drove me quite crazy. And then I discovered somebody else with similar problems: [giantpong.wordpress.com]. And it gave me a clue. But what caused it?

Actually I was punished for being greedy. I bought 5 TB6560AHQ’s from eBay (from GC_Supermarket). After reading the website above, I checked all my TB6560AHQ’s. They all gave a zero ohm reading between pin 1 and 17. Something which isn’t possible with genuine parts. (Always check the data sheet!)

After my discovery I bought 5 pcs from greenhouse408 on eBay. Directly after buying, I checked the pin 1/pin 17 reading, and it showed nearly unlimited resistance. Got it! Genuine TB6560AHQ’s!

Now I can finally continue testing my board! In the mean time I made the board a little more rugged, by adding some very cheap decoupling capacitors and some other stability stuff (connections to earth of unused connections etc.).

---

Eppur si muove!

Nice catch! I'll have to put that up on the troubleshooting page for Sanguish as soon as I get around to making one.

Then I also have another tip for all the users with trouble out there: check the sockets you are using to mount the TB6560AHQ's. It is very likely that the sockets don't make proper contact with the leads of the IC. If this is the case, you get strange brown-out errors and resets. In the worst case, you can damage your TB6560AHQ's due to improper powering. So be warned...

---

Eppur si muove!

To make it a little easier, these sockets work perfect and are more reliable: [www.sparkfun.com] (wich you can also find on eBay cheaper...)

---

Eppur si muove!

bryanandaimee - tried to send an order to [dirtypcbs.com], but their script tells me one of the sides is too large - 10.032!
Is there any way to reduce it another 32 hundredths of a centimeter? I must admit a complete lack of Geda knowledge, unfortunately..

Board size is defined by the four lines on the "outline" layer. Changing them slightly should help.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Yep, just use the move tool and drag the outline in a bit. I can send you some modified board files if you can't figure it out easily. I'll sneak it in a bit for the next version to avoid that issue. Also for those wanting to self source. Use the Parts list on the wiki. The Geda files don't necessarily have the correct parts and values yet. It's on the list of things to do. (eventually)

Edited 1 time(s). Last edit at 05/22/2014 03:27PM by bryanandaimee.

Traumflug, thanks for the suggestions. Unfortunately my ignorance is such I don't even know where to start looking for these lines...
bryanandaimee - next version, what next version? Can we expect it anytime in the nearest future?

Necessity is mother of education
Got it - the "vertical" size of the board was a bit larger than 10 cm indeed. Had to move the two horizontal paths (one to the U20, and other that runs through one side of all the jumpers) a bit lower and reduce the overall size of board there as well. The U20 name on top silkscreen sticks out as well
Now to see if dirtypcbs really work.. Would be great to etch it by myself, but drilling so many holes would take me - for ever!

Glad to hear you got it figured out. As for the next version it will be a minor revision to make the board more compatible with different style sockets and stuff. Nothing major. For this version though keep in mind the ATmega socket shouldn't have a bar in the middle. Also the USB socket is the type without the little alignment posts. And don't trust the GEDA files for parts list. Take the parts list from the Wiki.