Sanguish development

Having read through this thread and looking at the schematics, I would highly recommend testing having an opto-coupler between the ATMega and the Heater FET's. From what I read in the specs of the IRFZ44N FET's at 5V they aren't all the way on yet, which helps account for some of the heat they are producing. By adding the opto-couplers, you can drive it easily with a full 12 Volts which will saturate the FET (they measure the full current capabilities @ 10V).

It shouldn't take a whole lot to jury rig one in with the supporting resistors instead of the FET resistor, and then jumper in the +12V & ground for a test. Then compare the FET temp after the same runtime to see if it's worthwhile to integrate it in in a future version. To save space, something like an LTV-827 can be used, but you still need to fit in 4 additional resistors as well to drive both.

reference [www.irf.com] with special attentiont to Continous Drain Current on pg 1, Figure 3 Typical Transfer Characteristics chart on page 3, and Figure 6 Typical Gate Charge vs Gate-to-Source voltage on page 4.

I agree it would be nice to drive the FETs at 12V, but with some of the newer FET's it doesn't seem necessary. I'll be using the IRLB 8743 which seems to do pretty well at 5V. It would be better at 12V but I'm not sure if I have space for the extra circuitry without a lot of changes to routing. I'll have to think about it for a future version.

The other consideration is that if you do that, then your board is limited to 12V on the heaters by the FET's as 24V would exceed the max gate voltage for most of them.

Agreed, with most FET's, you are limitted to 20V without having something to drop the voltage down. someone wanting to run 24V would need to supply a lower voltage to the FET driver from someplace else or provide a simple source such as from a 7812/7815 or resister + zener since current is minimal.

As I said in that posting, it's at least worthwhile to run a test with existing boards by wiring something in where the FET resistor is for testing purposes. That will help quantify the actual effect of driving a higher voltage with the existing FET, which is why I brought it up so you can test & decide if it is worth the effort.

Of course another circuit I didn't mentions is using a transsitor, since I'm biased to an opto-coupler and transistors don't save much space compared to an LTV-827.

Since I am looking at making a version of the board optimized for > 2A motors, and I don't have a major price jump if I'm greater then 10x10cm, I'll be playing with slightly larger layouts. I'm also biased to US PCB mfg instead of China, so the boards will be more expensive anyways, but I know they will be made to higher tolerances and no noticable extra cost for 2oz copper runs which will help for current and heat (I haven't checked the pricing on 3oz or 4oz from my favorite fab yet). This project is why I've been reviewing the design for ideas.

It's a good suggestion and something I've thought about before. I'll have to do some testing and see how much cooler the FET will run at 12V. If it's significant I'll have to try it.

I'm excited to see what you come up with.

From what I'm seeing, you're using sockets for the Servo Drivers. Any issues with those that anyone has seen at the 2A level? Any potential issues with those at 3A, or do you think in that case we're better off soldering the servo drivers in to get better contact and improve the heat dissapation through the PCB runs?

I've had no problems with it at 2 amps. The drivers do have to be ALL the way in to the sockets though. I've had issues with them if they are only most of the way in. These drivers are possibly more prone to die for various reasons than Pololu drivers though so I like the ease of replacement of a socket solution. For instance, wiring a coil between a and b outputs (cross wiring) will kill a driver pretty quick. As does incorrect power-on sequence. (enforced by power relay on Sanguish) Once you have wiring and stuff set up right, they seem to be pretty robust to me. I think the run time failures reported on the web at some CNC sites seem to be due to using too high a voltage. I would recommend no more than 24V. You want plenty of headroom between run voltage and max voltage.

I think the only real issue at 3 amps vs 2 would be cooling. I'd guess you could get by with just a fan, though some heat sinks might be advisable. The drivers will handle 2 amps in still air but they do get hot at that setting, the thermal protection doesn't kick in though. With a fan they stay pretty cool even without a heatsink.

I've got two boards built for Sanguish 1.0 that I am currently testing and it is going well. I'll update the wiki soon. The final 1.0 version is essentially unchanged from the 1.0 RC1 linked in the wiki except for a bit of cleanup on the USB connector's solder mask. I'll post final revision files as soon as I can.

If anyone is interested I have parts for 6 kits. They will sell for $50 for the full kit, or $45 without the USB to serial components (if you already have a USB-serial cable) Shipping is $5 to the US and $12 everywhere else. Shipping is by USPS 1st class and is only tracked in the US.

Glad to see the project is moving along nicely. Any plans to add an sd card slot?

I'm not sure if I can make the space for that. I did add the two pin header next to the 6 pin ICSP header in hopes that it would be compatible with Sanguinololu add on sd card boards. I haven't tested that yet but plan to soon. Once I have a working add on solution then I will look into integrating it if I can figure out where to put it. Perhaps on the back side? Also I'm pretty sure I'll be switching to auto style fuses. The polyfuse take too long to trip at too high a current for my comfort.

Edited 1 time(s). Last edit at 09/07/2013 01:49PM by bryanandaimee.

An addon board would be fine with me so long as its available

I've got one on order. The SDSL seems to be harder to get a hold of though so I might look at SDRamps too and see if I can get both to work.

SDSL works. Just edit pins.h sanguish section at the very bottom and replace

#define SDSS -1

with

#define SDSS 29

then you should be able to enable SD support and use it. I've used it with Repetier firmware and host. Sanguish is only tested with SDSL and pin compatible adapters. Many SD adapters are sold for Sanguinololu that are SDRAMPS pinout. These are not pin compatible with Sanguish, though I guess you could make a cable to arrange the pins as needed. Also since the header is not on the edge of the board you need a vertical mounting adapter.

So... finally managed to kill my Sanguish beta i got from you (;

Reverse polarity by mistake. fried voltage regulator - replaced. Fried atmega: replaced. Still not working again...

Peter

Hmm, the logic level side of the drivers might have blown too. That's bad luck. Often the 7805 dies but protects the other components. Do you get a connection after replacing the ATmega? Maybe I should put a polarity protection diode on the 7805.

Made a typo on the pin for SDSL. The proper pin assignment should be

#define SDSS 20

Pin 29 is the Y min pin.

Funny enough the SD card worked. I sent files to it, printed off it and everything. Then I started having strange issues with the Y min endstop.

A few days ago I added a config (config.sanguish.h) for Sanguish to Teacup, for now on the experimental branch. Also these few lines of code required to handle the power supply MOSFET. As I can't test this code, please give some feedback wether it works out of the box (it should!).

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

That's great! Thanks.

I'll test it out as soon as I get a minute or so.

Bryan

MOSFET code looks good. I got it up and running after changing pins to version 1.0. I'll send you tested versions for Beta and 1.0 as soon as I can.

Thanks
Bryan

I have a set of files up on the wiki for version 1.1. The main change is that I combined the last enable line to free up one more pin for an external LCD control panel. I think I'll add 5V back to the serial headers so the board can power bluetooth dongles. So You will likely see RC2 up sometime soon.

Sanguish 1.1 RC2 is up on the Wiki. This will be the version I am ordering boards of to test.

I have a couple left over Beta boards. They are fully built and tested. They are free to a good home. They will include drivers and ATMega 644P. Just let me know if you are interested. I've got some unpopulated beta PCB's as well.

I will take one or two off your hands but in only if you tell me what I owe you (: and no not free

That's cool.
I'll take one or two as well, not for free though! Looking for some brains for my 4x2meter laser cutter.

regards,
Muza

I thought it would be nice to drop a note:
I have been inspired by this design to make my own dual extrusion controller with TB6560's. Link

Got boards for first run of 1.1 RC2 which I'm guessing will be the official release of Sanguish 1.1. I'll build one up and do some testing then put together kits and sell the rest.

Sounds cool. Keep us posted
What's the kit price looking like?

Kit price will be the same as version 1.0, $50 plus S&H. Not much changed component wise. Just some changes to routing, placement, etc.

1.1 seems to be working fine. I'm still getting some interference between the mcp2200 and the other serial headers even though it's powered down when disconnected. Some usb-serial cables work and some don't unless you leave the mcp2200 off the board. Then it behaves as expected for external adapters.

I don't seem to be able to upload .h files anymore to the wiki. I'll get these on the sanguish wiki as soon as that gets fixed. Till then the config files I'm using for Repetier and Teacup are attached below.


config.sanguish1.1.teacup.h

Configuration.Sanguish1.1.repetier.h

pins.sanguish1.1.repetier.h

Hi

My markdown beta and bare 1.0 boards arrived (:

Been soldering up the 1.0 first though (:

Noticed some points - correct me if fixed in 1.1 - have not checked the board layout this morning

1. Mini USB - I have about 5 brands and all have two little nibs that go into holes in the footprint, keeps it lying still and cant move while soldering. In your footprint these holes are nonexistant. This means i had to cut off the nibs, and had a fun time sliding it around while soldering? gEDA cool and all, but this is where Eagle excels - proper libraries

2. Crystal: No offloading Caps? This wont run reliable USB??? I am tacking some on the back but it HAS to be fixed

3. Silkscreen has imnproved for the previous beta i bought (:

4. C19 is a little close to the ATmega if used with a quality socket. Cheap bent socket it kinda works, expensive milled socket it overlaps. Move C19 a bit closer to drivers

5. Too many components under the ATmega- again difference IC sockets maybe but i have ones with reinforcement in the middle - had to cut out the pastic with side cutter, not ideal

6. I assume AUX1 is a SD Select Pin?

7. Area with AUX1, two caps for the therm, and temp1+2 headers are very closely bunched together - space out a little


Peter

In some libraries the mounting tabs on the USB go into unplayed holes. Yes I know that's stupid. Libraries do stupid things sometimes. If you fix the USB, check the holes. They should be plated. I'm getting some completely unrelated boards re-done for those same stupid holes and an un-related stupid library. (Yes *someone* could have checked those holes ....I didn't ....).