Going 24v

[Gr8Scott]

Got tired of waiting for the heated bed to warm up. Decided to order a 1000w 24v power supply. Is there a source for a really flat aluminum heat spreader out there? I tried to flatten a barely warped piece of aluminum sheet and only managed to turn a slight warp into an extreme warp.

[Polygonhell]

Got tired of waiting for the heated bed to warm up. Decided to order a 1000w 24v power supply. Is there a source for a really flat aluminum heat spreader out there? I tried to flatten a barely warped piece of aluminum sheet and only managed to turn a slight warp into an extreme warp.

To get really flat aluminum you have to start with cast plate, which is much more expensive than the cheap stuff.
The issue then is cutting it, most machine shops charge a large premium for a one off, but on the plus side it should be a single machining operation.
The alternative is to find someone with a large enough CNC router locally and have them do it.

[Gr8Scott]

My dad has a bridgeport 5 axis milling machine along with a "fly cutter". Would that do the trick?

[Polygonhell]

My dad has a bridgeport 5 axis milling machine along with a "fly cutter". Would that do the trick?

If it has enough travel, I couldn't CNC machine it on my mill because I only have 8 inches of travel in Y, I guess I could mount it to a rotary table and do it.

[cambo3d]

you wont need 1000w for the onyx, it only requires 18amps max at 24 volts. save your money.
http://forum.seemecnc.com/viewtopic.php ... =170#p7960

[Gr8Scott]

Too late. Already ordered the 24v supply. Two bridged 12v HP server power supplies. Not too terribly expensive. I plan to only run 24v into the top power feed that feeds the heated bed. The rest of my electronics will run off the standard 12v supply. I also found a piece of cast aluminum plate that has been machined flat. It's 5/16" thick, but I plan on mounting the Onyx flat to the base with no spacers and only a little aluminum tape as an insulator. I'll have to be sure to remove the aluminum tape that is immediately under where the soldering has taken place etc on the Onyx and route out a little belly room under that same spot to give enough clearance for the plate to sit flat on the base. Then it will all be 5/16" aluminum above that with 24v to heat it up. Given the larger thermal mass, It won't hurt to have the extra voltage to heat things up.

I also ordered some ceramic heaters and I'm trying to locate some Red Ice for the thermistor. That should give me a pretty good lock on getting temps under control with a quickness. That has been my biggest issue thus far.

[cambo3d]

you can't bridge two 12v power supplies, unless at least one has a floating ground or you will end up shorting them out.

check this by: metering your negative terminal to chassis ground, if they have continuity, then you don't have floating grounds and you'll need to isolate the ground from at least one of your power supplies, in order to get 24v without shorting.

another option: you can install high current diodes in your wiring to prevent from shorting.

[Gr8Scott]

I appreciate the info. I'll double check just to be sure, but these power supplies have apparently already been modified to be bridged to 24v. They come with a cord to connect them to get 24v serial. Here is the description on the auction. $86 plus shipping didn't seem too bad to me.

I have 24 volt 1000 watt power supply setups for sale. These are modified hp server power supplies pulled from working servers. They output very clean, very reliable power.
You will receive 2 power supply units as shown, with 24v jumper cable.



The outputs have 6" of 12awg w.s. deans ultra wire and 4mm banana jacks.
All heat shrink is high quality and adhesive lined.
Does not include AC power cord, units accept a standard computer power cable (2).


The DC output has been isolated from the exterior case, the exterior case still retains AC earth ground for maximum protection.
These units DO NOT require modified AC power cords.
These units are capable of touching one another.

[cambo3d]

no problem..

If your using 24v directly to the input of the heatbed circuit. another thing to be aware of is the 15amp fuse on the rambo, initial current draw for the onyx at 24v is 18amps, and its likely to blow that fuse.

[foshon]

You can get a fairly inexpensive 24v kapton heater from several places (trinity labs, ultibots) that will let you use the Rambo circuit at 8A or so.

[Gr8Scott]

So add a 5 ohm NTC thermistor in series with the Onyx and I'm good to go? Think maybe I should heat sink it to the aluminum plate to slow it down a touch?

[6bolt2g]

You can get a fairly inexpensive 24v kapton heater from several places (trinity labs, ultibots) that will let you use the Rambo circuit at 8A or so.

Have you used that type of heater?

If so, what has your experience been with it? Heat up times?

[foshon]

You can get a fairly inexpensive 24v kapton heater from several places (trinity labs, ultibots) that will let you use the Rambo circuit at 8A or so.

Have you used that type of heater?

If so, what has your experience been with it? Heat up times?

Yes, I use one on both of my machines. MAX's little brother a MM 1.5+ heats up in around 5-10 minutes, no longer then the extruder. MAX takes a bit longer in the 10-15 minute range, but I have not tweaked the settings at all, yet.

[Polygonhell]

Bear in mind that the Rostock Max bed is about 3x the area of the normal MendelMax bed.
What actually matters here is the power to the bed, V*I, or in terms of resistance (V*V)/R.
The existing Onyx is roughly 12*12/1.5 = 96W, if you want to heat as fast as that MendelMax you need more than that, probably ideally close to 300W.
At 24V that's roughly 2Ohms resistance, and will pull 12A of current, that's less than the 15A rating for the fuse on the Rostock BUT that's a 12V fuse, and fuses actually blow based on power through them which is why they are rated for both voltage and current, so you'd still likely have to replace the fuse.

The Trinity Labs round Kapton heater is slightly undersized, rated for 200W, so about 2x the existing Onyx which is probably fine, it will draw about 8.3A which will be marginal for the default heated bed fuse on the RAMBo, you can easily replace it with a 20A automotive fuse if you have a later RAMBO, if you have the earlier one with the poly fuse, it might work fine if you can get some air over the fuse, they tend to be a little erratic on exactly how much current before they stop conducting.

[foshon]

Bear in mind that the Rostock Max bed is about 3x the area of the normal MendelMax bed.
What actually matters here is the power to the bed, V*I, or in terms of resistance (V*V)/R.
The existing Onyx is roughly 12*12/1.5 = 96W, if you want to heat as fast as that MendelMax you need more than that, probably ideally close to 300W.
At 24V that's roughly 2Ohms resistance, and will pull 12A of current, that's less than the 15A rating for the fuse on the Rostock BUT that's a 12V fuse, and fuses actually blow based on power through them which is why they are rated for both voltage and current, so you'd still likely have to replace the fuse.

The Trinity Labs round Kapton heater is slightly undersized, rated for 200W, so about 2x the existing Onyx which is probably fine, it will draw about 8.3A which will be marginal for the default heated bed fuse on the RAMBo, you can easily replace it with a 20A automotive fuse if you have a later RAMBO, if you have the earlier one with the poly fuse, it might work fine if you can get some air over the fuse, they tend to be a little erratic on exactly how much current before they stop conducting.

The typical automotive blade fuse is actually rated for 32vdc.

[Gr8Scott]

Bear in mind that the Rostock Max bed is about 3x the area of the normal MendelMax bed.
What actually matters here is the power to the bed, V*I, or in terms of resistance (V*V)/R.
The existing Onyx is roughly 12*12/1.5 = 96W, if you want to heat as fast as that MendelMax you need more than that, probably ideally close to 300W.
At 24V that's roughly 2Ohms resistance, and will pull 12A of current, that's less than the 15A rating for the fuse on the Rostock BUT that's a 12V fuse, and fuses actually blow based on power through them which is why they are rated for both voltage and current, so you'd still likely have to replace the fuse.

The Trinity Labs round Kapton heater is slightly undersized, rated for 200W, so about 2x the existing Onyx which is probably fine, it will draw about 8.3A which will be marginal for the default heated bed fuse on the RAMBo, you can easily replace it with a 20A automotive fuse if you have a later RAMBO, if you have the earlier one with the poly fuse, it might work fine if you can get some air over the fuse, they tend to be a little erratic on exactly how much current before they stop conducting.

So I can change the 15 amp fuse on my very recent Rambo with a 20 amp fuse with no ill effects? If I can do that without causing other problems, it would be the least expensive and least complicated path.

[Polygonhell]

Depends on what the traces on that power path on the RAMBO are rated for.
100W is a lot of power when converted to heat in the wrong place, I've seen a loose wire on a heated bed connector melt the connector on a RAMPS board.
My guess is since it's rated for 24V and there are so many beds out there in the 200W range you would probably be fine.
If indeed automotive fuses are rated for 32V then you wouldn't need to change the fuse at all.

[Eaglezsoar]

How about letting the heatbed output from the Rambo control the input to a DC-DC solid state relay. The output of the
relay would handle the current going to the heatbed. This eliminates the possiblity of destroying the Rambo or blowing its
fuse. There would need to be an inline fuse on the heatbed. You can buy the solid state relay and a heatsink for it for
around $20-$25 on Ebay.

[Polygonhell]

How about letting the heatbed output from the Rambo control the input to a DC-DC solid state relay. The output of the
relay would handle the current going to the heatbed. This eliminates the possiblity of destroying the Rambo or blowing its
fuse. There would need to be an inline fuse on the heatbed. You can buy the solid state relay and a heatsink for it for
around $20-$25 on Ebay.

This is how most sanguino based solutions work and it seems to be fine.

[6bolt2g]

Yes, I use one on both of my machines. MAX's little brother a MM 1.5+ heats up in around 5-10 minutes, no longer then the extruder. MAX takes a bit longer in the 10-15 minute range, but I have not tweaked the settings at all, yet.

Do you have an aluminum plate with glass on top?

[Gr8Scott]

I have a SSR or two laying around here somewhere. If it covers the necessary amperage (it's rated for 110v), I should be able to use it for that purpose. I can then just drop a 20 amp inline fuse into the link between the power supply and the heat bed and then use the 12v from the Rambo as the trigger for the SSR.

[Eaglezsoar]

There are two types of solid state relays, one is meant for AC on the output the other is for DC on the output. If you are using 24VDC on the heating bed make sure the output of the SSR is made for DC.
It will say on it what type it is.

[Gr8Scott]

I have found that my ATX power supply is tripping the overload protection when trying to heat the bed with the aluminum heat spreader on it. I checked the resistance of the bed coming into the Rambo and it's 1.3 ohms as expected, so no dead shorts causing any issues. The PSU would overload before when turning on the heated bed and extruder heat at the same time. Now that the heated bed has more thermal mass, it can't heat the bed enough in the initial burst to allow me to use it again immediately afterward. The new power supply was needed after all. I will hopefully wire it in tonight.

[Gr8Scott]

Man what a letdown. Those SSR's I have were dc control, but rated for AC on the business end. They won't function with a DC power source. I'm a bit hesitant to run mains voltage 110AC to the Onyx even with a thick chunk of aluminum plate on it. I believe the traces would burn off in an instant. Now I have to go find a DC SSR instead. I had it all wired up and kept trying to figure out why the heat bed wouldn't receive voltage while the SSR LED kept lighting up and I could clearly see voltage when I bridged the two connections. LOL.

[Gr8Scott]

Heat up times look pretty good if I bypass the SSR and just hold the wires together until I get to 100 degrees. Took 5 minutes to get there. Even with the ATX power supply that came with it stock, it has been to 100 degrees C before, but it did take about 10 minutes to get there and the bed was warped pretty badly when it got there. I thought mounting the bed elevated on springs would allow the bed to spread out and would mitigate the warping, but I found that it still warped severely due to the temperature differential across the width of the material due to the uneven heating. Now with the aluminum heat spreader, the Rostock Max finally has a perfectly flat surface to print on.

[img]http://i152.photobucket.com/albums/s179 ... edTime.jpg[/img]