Custom Quantum/Hadron PCB Heated Build Plate

Electronics related to CNC

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby BenJackson » Fri Feb 17, 2012 6:33 pm

I just did a bunch of work to validate that the RAMPS 1.4 PCB's heater MOSFET could withstand more than the 11A (at 12V) that the polyfuse is rated for. Doing the math on a 150W heated bed using 5V @ 30A that FET would almost certainly overheat. You would also be bypassing the polyfuse in that configuration.

(fwiw I worked out that 20A @ 12V is feasible with that FET provided you bypass the polyfuse, which is trivial)
BenJackson
 
Posts: 522
Joined: Fri Apr 15, 2011 6:13 pm

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby frob » Fri Feb 17, 2012 6:51 pm

BenJackson wrote:I just did a bunch of work to validate that the RAMPS 1.4 PCB's heater MOSFET could withstand more than the 11A (at 12V) that the polyfuse is rated for. Doing the math on a 150W heated bed using 5V @ 30A that FET would almost certainly overheat. You would also be bypassing the polyfuse in that configuration.
(fwiw I worked out that 20A @ 12V is feasible with that FET provided you bypass the polyfuse, which is trivial)

Cool thats great input!

Its looks like i'll support both 5V and 12V using separate connectors - @ 12V i'm shooting for 12-15A, (for a minimum 150W cold) and probably will use 170C rated PCB so pre-heating the plate to 150C at first to accelerate warming up the glass will be possible. The "local" fet / temp controller will work at 5V only and will likely move to a separate mini-board for practical reasons (i can bolt it directly to the makerslide to avoid over-stressing the electronics) - it will be a separate orderable item. Will include an ATX PSU connector and other neat features 8-)

There will be smaller 100W board tailored to the full size of the Quantum. Cost is about the same or may be slightly higher due to lower anticipated demand.
frob
 
Posts: 260
Joined: Tue Jan 31, 2012 6:24 pm
Location: Montreal, Quebec, Canada

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby Enraged » Fri Feb 17, 2012 9:28 pm

a great idea, hot swappable build plate, using magnets: http://prusadjs.cz/
Enraged
 
Posts: 439
Joined: Fri May 06, 2011 3:13 pm

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby frob » Fri Feb 17, 2012 10:49 pm

Enraged wrote:a great idea, hot swappable build plate, using magnets: http://prusadjs.cz/

Cool!
My approach is similar but different - You use pins or screws coming up from the bottom plate - using smooth rounded pins would mean you can quickly swap out the hot plate from one machine to another if you want to do that- no magnets required. there will be threaded standoffs on the bottom side instead of magnets , so screws won't protrude past the surface of the hot plate (and nothing else either) thus:
- you can put a glass plate on top the full size of the plate- maximum usable build area 214 x 214mm
- probably the best way to keep the glass in place is to use a 0.5mm thin silicone thermal gap pad like those used for heatsinks;
it would help insure even heating and acts like a suction cup as its soft silicone - glass wont budge in use but easily lifts off when its done.
frob
 
Posts: 260
Joined: Tue Jan 31, 2012 6:24 pm
Location: Montreal, Quebec, Canada

Anyone interested in a 24V option ?

Postby frob » Fri Feb 17, 2012 10:54 pm

That would mean only 6A to get 150W - lighter wiring and cooler mosfet when using RAMPS etc.
If you use 24V on the steppers you also get faster max speed from the motors, as a side benefit :)

I think supporting all 3, i.e.: 5-6V, 12V, and 24V is possible, if marginally more complicated.
I have some large sheets of FR4 laminate with 1oz copper so i'll prototype this "old school" before ordering a large batch.

Also just got some initial pricing estimates -
bare board with special high-temp 170deg.Tg FR4: i can keep the $25 intro pricing, as long as i get enough orders to justify getting a batch of 100 or more made in Hadron size.
If i get less than 50 requests it will gave to be regular 140deg.Tg FR4.
Aluminum core boards - increases the cost quite a bit, to ~ $40 - @ same qty's.
I like the idea of Aluminum boards, they don't need extra top plate and could heat much faster, to a higher temp, and use much less power- but the extra cost is probably not justified.
- If anyone really would prefer Aluminum drop me a PM, i wont rule it out just yet - i'm open to doing a "deluxe" version in smaller batches.
- maybe i could even mount the V-wheels / timing belt directly to it and avoid the need for the carriage plate, and get a little more Z height in the bargain 8-)

Sizes:
216 x 216 matching the carriage on the Hadron;
130 x 130 matching the carriage on the Quantum.
10 mounting points (1 in the center, 4 corners, 4 sides, use as many as you like) on a 104.5mm grid (Hadron) or 61mm (Quantum); the center hole is really meant for a thermistor/thermocouple, but it will be same size hole as the other so its your choice :)

Going forward for reservation PM's please specify Quantum in the title if that's what you need, otherwise Hadron is assumed :)
Last edited by frob on Fri Feb 24, 2012 5:37 am, edited 2 times in total.
frob
 
Posts: 260
Joined: Tue Jan 31, 2012 6:24 pm
Location: Montreal, Quebec, Canada

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby BenJackson » Fri Feb 17, 2012 11:07 pm

If you exceed 12V on the input you can't install D1 on RAMPS to power Arduino from the daughter card. You have to provide separate 12V for Arduino or use the USB. If you plan to run standalone the USB won't be present. For some people that is not an issue.

Current designs using <11A @ 12V (so < 130W) take a long time to hit 115C+. I use a resistor-heated aluminum plate which is currently connected to a bench PSU. With a glass plate on top (which provides some amount of insulation) it takes 4 minutes to hit 60C and then 36W continuous to hold 60C. Since that's 40C over ambient and 115C is 95C over ambient and the thermal resistance is probably the same for both it will probably take 2.4x as much power to hold the 95C temperature differential. That's 85W. That's why you need a lot of excess power to get quick bed heating to ABS temperatures.

(My heated bed is about 230W although originally I planned to use mains voltage and get about 320W I ended up with a surplus of 12V PSUs)
BenJackson
 
Posts: 522
Joined: Fri Apr 15, 2011 6:13 pm

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby frob » Fri Feb 17, 2012 11:38 pm

Thanks Ben again that's great info!
BenJackson wrote:If you exceed 12V on the input you can't install D1 on RAMPS to power Arduino from the daughter card.

Maybe you could substitute a 12V Zener for D1 in that case? do you know what the average running current of the Arduino board is ?
Current designs using <11A @ 12V (so < 130W) take a long time to hit 115C+. ....it takes 4 minutes to hit 60C and then 36W continuous to hold 60C. Since that's 40C over ambient and 115C is 95C over ambient and the thermal resistance is probably the same for both it will probably take 2.4x as much power to hold the 95C temperature differential. That's 85W...

Nice that kinda confirms my initial estimates, though they might be a bit low, i'm still in the right ballpark.
frob
 
Posts: 260
Joined: Tue Jan 31, 2012 6:24 pm
Location: Montreal, Quebec, Canada

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby frob » Thu Feb 23, 2012 5:57 pm

Just a quick update on this :

Before going too far in laying down tracks i wanted to validate and refine the math for accurately working out the required trace width & length to get the resistance and power output needed - and as close to dead-on as possible in the first try -

Its actually a tricky problem because we're starting with a fixed area and want a to optimize the track design for a given power area density and thus resistance.

So i've been playing around with equations in Excel to do this and i think i finally have something that works well and correlates closely with published numbers and equations / calculators you can find for working out the track width for a given current /length/ temp rise as is the more common scenario.

So a consequence of this is that i believe it wont be hard to not only optimize the layout to support several voltages as i'd hoped,
but also to do a single layout for both the Hadron size and the Quantum size: in other words, the Quantum heater forms the center section of the full size Hadron heater -
To convert a Hadron heater to Quantum size you just "cut along the dotted lines".

This is a good thing considering the lower demand for the Quantum; this way the risk and variability in providing a Quantum-optimized version is eliminated.

Now since the Quantum is about 36% the area of the Hadron, to get the same power density across the surface,
if the Quantum size is 100W (at room temp), the the Hadron will be 280W.
In theory that's what you need to get a reasonably quick warm up time and the ability to reach over 100C temperatures.
Now 280W at 12V is over 23A - which does limit your power supply options - so there is some tradeoff required here.
However, if you use the 24V terminals at 12V instead, then you end up at close to 100W again with only about 8-9 amps needed, so that may be the best way to support a lower current mode for those that don't have a 23A power supply.
Another possibility for the more technically adventurous is to hook up the 5V and 12V in series (requires hacking into your ATX supply a little ) so 17V into the 24V terminals gets you 140W at the same 8-9A.
Any comments on this?

The other cool feature i'm probably doing here is to provide and option to connect these through a single piece of flat flex cable, like you would find in just about every printer -
The benefit here is that it has for all intents an infinite fatigue life and more importantly zero drag on the fast-moving build carriage mechanism and very little chance of snags.

The connector will be on the underside of the plate, at the right front corner of the Quantum heater section, pointing towards the back of the unit.
A small board that can be bolted to the makerslide beneath the build carriage fixes the other end of the flex cable, and provides heavy duty screw terminals and the optional temperature controller (with an ATX power supply connector as well).
It will also be designed so the flex cable exits towards the back, so it loops 180 degrees allowing a very neat and tidy installation with the flex cable mostly out of view.

The flex cable and adapter board would be a separate optional add-on; all the boards will include 1/4" "Quick-Fit" style lugs for the power connections and something more appropriate for the thermistor /led connections - not sure what yet though, open to suggestions there. :?
frob
 
Posts: 260
Joined: Tue Jan 31, 2012 6:24 pm
Location: Montreal, Quebec, Canada

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby bdring » Thu Feb 23, 2012 7:33 pm

I would suggest making it work with 12V and not to exceed 10-11 AMP. That makes it RAMPS compatible. I am not interested in anything that makes my Bot more complicated. With glass on top and an insulator underneath, this works fine.

Try not to have the flat flex cable be a heater too ;)
Bart
"If you didn't build it, you will never own it."
bdring
Site Admin
 
Posts: 2966
Joined: Sun Nov 22, 2009 7:33 pm
Location: Chicago, IL, USA

Re: Custom Quantum/Hadron PCB Heated Build Plate

Postby Zat German » Thu Feb 23, 2012 10:55 pm

bdring wrote:I would suggest making it work with 12V and not to exceed 10-11 AMP. That makes it RAMPS compatible. I am not interested in anything that makes my Bot more complicated. With glass on top and an insulator underneath, this works fine.

I have to agree. I do not want to run a separate power supply just for the heat bed when I can get something that runs off the 30A power supply for RAMPs that I already have.
Zat German
 
Posts: 141
Joined: Sat Feb 18, 2012 4:48 am

PreviousNext

Return to Electronics

Who is online

Users browsing this forum: No registered users and 3 guests

cron