Losing too much heat.

I've got my Prusa working and the prints are nice.

However, on larger prints where the head has to move longish distances, I'm losing too much heat.

If my tip gets below 175 it stops extruding.

Currently, I'm babysitting the prints and pausing when head temp gets below 180. waiting until it is at least 185 to resume.

This isn't good for quality prints.

The bed also drops from 110 (with blanket) to 88-87 during prints. I can get 90-97 without blanket and motion.

How can I improve this?

The hot end is using a heater block. The heat bed is above the aluminum with insulation in-between.

How warm should the room be?

Thanks,

If your head is dropping in temperature that much it sounds like your power supply is not powerful enough to run everything. Using a heated bed usually requires a 30 amp 12 volt supply.

You could try insulating the hot-end similar to what maker bot (I think) does with theirs to reduce heat loss.

Komb' Wrote:
-------------------------------------------------------
> The bed also drops from 110 (with blanket) to
> 88-87 during prints. I can get 90-97 without
> blanket and motion.

I insulated the bottom of my print bed with two 4mm layers of flexible insulation tape meant for insulating pipes. That got me about 20 degrees more heat than before. It didn't have a temperature rating but it seems to take 110C just fine. I don't remember the brand, but it was the fancier one available at the local Bauhaus store.

crispy Wrote:
-------------------------------------------------------
> If your head is dropping in temperature that much
> it sounds like your power supply is not powerful
> enough to run everything. Using a heated bed
> usually requires a 30 amp 12 volt supply.
>

Hum...

I'm using a PC supply 350wat. ... 17A on 12V... I think I see a problem...

> You could try insulating the hot-end similar to
> what maker bot (I think) does with theirs to
> reduce heat loss.

If I knew what to get and where locally, (Toronto Canada) I'd look into it.

ttsalo Wrote:
-------------------------------------------------------

> I insulated the bottom of my print bed with two
> 4mm layers of flexible insulation tape meant for
> insulating pipes.

There is insulation between the aluminum bed and the heat bed.

Theoretically, a heated bed would draw less than 10 A, a hot-end just over 2 A, and the steppers less than 2 A in total (the other electronics negligible)
So "theoretically" - a 17 A PSU should be plenty,

You can check your PSU by using a meter on the 12v supply **AT to the PSU terminals**. With everything running, if it drops by more than a few points of a volt, then it's struggling.

If voltage does not drop significantly at the PSU end, ensure you have heavy enough wiring and good connectors, and check the voltage close to the heater ends. If that is much lower, there is a problem in-between. (possibly
This could get trickier with PID/PWM, but if they're getting cool, even PID should be turning the heaters hard on, allowing a realistic voltage reading to be made.

My hotend started cooling last night - very puzzling until I noticed the screw connector I used 1/2 way back to the PCB glowing.
Replaced, tightened properly, all good. (I love my new J-head)

Do you have any fans on the machine, near the hot-end & bed?


(re "steppers less than 2 A in total"
- although you may have 5 steppers, each rated at 1 to 2 A, they are usually rated at about 3V or less. (and possibly the Z's in series)
Running them from 12v, the driver uses PWM to control power delivered, and acting like a switching stepdown converter with the inductance of the coils, this means that 1 to 2A through the stepper coils will actually only draw an average PSU current of about 0.2 to 0.5 A (my NEMA17s draw 0.25A ea)

nb99 Wrote:
-------------------------------------------------------
...
> You can check your PSU by using a meter on the
> 12v supply **AT to the PSU terminals**. With
> everything running, if it drops by more than a few
> points of a volt, then it's struggling.
>

11.04v at PSU terminals with both bed and hot end on.

> If voltage does not drop significantly at the PSU
> end, ensure you have heavy enough wiring and good
> connectors, and check the voltage close to the
> heater ends. If that is much lower, there is a
> problem in-between.

11.33 at hot end with bed off.
10.96 at hot end with bed on.

10.02 at bed with hot end off.
9.96 at bed with hot end on.

As to heavy enough wiring, see pic of hot end.

The bed is using two wires each side from a flat cable. I just noticed they are warm. Obviously some power loss there.

If you are using a pc power supply consider put a load resistor about 5R on the 5 volt rail. it usually inproves the 12 V rail. Also if your bed cables are getting warm you will be getting a volt drop there as well. You can check the volts at the heated bed terminals.

To me those voltage readings indicate you need a more powerful power supply. Mine never drops below 12v at full load.

He should first try loading the 5V rail (if he hasn't already).

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

Komb' Wrote:
-------------------------------------------------------
> There is insulation between the aluminum bed and
> the heat bed.

Ok, I see. I have just the aluminum build plate with power resistors bolted straight into it. I was losing quite a bit of heat downwards until I added the insulation to the bottom of the plate.

Shouldn't virtually all modern switch mode power supplies go into overload protection shutoff if they're loaded so much that the voltages start significantly dropping?

Hadn't Loaded the 5v rail.

As I don't have any resisters hanging around, I grabbed an old hard drive.

With both bed and hot end on I now get 11.28v from psu.

Hot end 11.22v with bed on.

Bed 10.08 with hot end on. (10.87 at Ramps)

That's an improvement.
I'll try a print with this.

May add some lines to the heat bed later.

Thanks all! Will report back after some prints.

ttsalo,
A single rail PSU yes, but a cheap multi-rail PSUs like a PC PSU only have a single feedback loop for all the rails. If the 12V load is high it has to open the throttle but if there is no load on the 5V rail that would go over voltage so it can't. If you measure the 5V rail you will see it is a little high. The more load you add the higher the 12V rail will go.

There is some cross rail regulation but it is done with clever magnetics rather than a control loop.

The 12V rail isn't low because it it is overloaded. If you did overload it it should shut down. Also it may shut down if the 5V gets too high due to the load imbalance.

---

[www.hydraraptor.blogspot.com]

First print was successful! It touched 179, but generally stayed in the safe zone.

After reading Nophead's post I added a second hard drive.

Now the hot end actually stays near where I set it.

May have to back-off the settings. Too hot I think.

If you don't have any resistors you could put an auto car bulb across the 5 V rail, say about 25 W

With the extra current, I ran into the problem of the PTFE getting hot and grabbing the filiment.

Attempted to attach a 40mm fan to cool the PTFE and a baffle to keep it away from the hot end/bed.

The baffle worked, but the spalsh back from the local walls prevented proper heating.

When I turned the fan off, I found the aluminum baffle I installed acted like a heat sink/heat deflector and kept the PTFE cool.

Removed fan and baffle and made a better heat shield/sink.

Printing small objects worked again.

Printing medium sized objects still showed heat swings generally trending down...

Large print test Failed.

I'm off to buy a better PSU.

... Already have a better hot end on order. But I'm still going to see if I can't get this one to work reliably.

For PSUs, go to an electronics recycler or computer repair shop, and bring your multimeter to test for broken ones. I got a 66W for free out of a broken server
You could also make a heated build chamber out of MDF, it would keep the heat from dissipating and lower the load on your bed and hot end.

Edited 1 time(s). Last edit at 02/29/2012 07:16PM by shardbearer.

Update

Installed a PSU with 2 12v rails at 22a each.

The head and bead heat up faster now.

A test print was working then the head started losing heat anyway. Poked the wire to the heat resister, heard a crunch, then the heat came back!

So, I swapped the resister with a spare I had. It had visible bulges in the caseing and was hard to remove. The new one whent in easily.

While I was doing that I wrapped the heatblock with fiberglass I had just picked up.

Removed the heat sink/shields from the brass and ...

It's working like a charm. Holds its temp within 4 degrees during an entire medium sized print.

I forsee no problems with the large print now.

Thanks to all for the info and help.

I have some leftover glass wool insulation if you would like some.

I use it to insulate the hotend and underside of the HBP.