First Notes on a Pure Silver Heat Block

The 99.9% pure silver heat block which I cast and machined is a close copy of the E3D V5 aluminum block currently on my prefered printer. Installing the same 24V, 30 watt heater cartidge in to it, along with the PT1000 sensor, yielded some interesting results. I used Repetier to monitor the temperature in all cases, and to invoke PID tuning.

1) It gets to temperature roughly 4 times faster. It takes the original aluminum block about 60 seconds to go from 25C to 195C. The silver block, with the same 30 watt heater cartridge, took about 15 seconds. This surprised me. I had expected a difference, but not this kind of a difference. The same cartridge and sensor was used in both tests, and both were installed with minimal boron nitride paste.

2) A PID tune on the aluminum block gave roughly P20, I1, D80. The silver block with the same heater and sensor: P7, I1, D10. I usually tune by hand, but let Marlin "do its thing" (which I don't necessarily trust), but the difference, again, surprised me. The proportional gain is definitely in line with the time difference.

3) PT1000 RTD sensors suck, at least if you are limited to the usual 12 bit A/D input on the common controller board. You have a resolution of 4096 counts spread across a very shallow resistance slope, which gives a resolution of roughly 3C. Not good enough. The temperature plots are nasty. Without going to offset and gain control boards, and digitally communicating the result from a higher resolution A/D converter, the next best thing would be a PT10000 sensor. Such animals exist, and would give a resolution of about 0.3C. A PT100000 would be great, but aren't available in any reasonable form factor. I think a 650C thermistor is my best bet, although the response curve gets a bit flat at the high end, but still much, much better than any platinum RTD.

Silver blocks? OK. Platinum RTD? Nope. YMMV.

... I've used PT100 with an ADUC847 (what's a 24Bit-AD µC) with pretty high resolution in the 0.01C range

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

This got me thinking..

Oxidation would be a problem on a copper block.

Does anybody make/sell a nickel-plated copper heat block?

Plenty of brass ("hardened copper") ones for sale on eBay, but brass and aluminium have similar thermal conductivity, so, aside from appearance, there would be little difference. Pure copper (Cu11000) and Silver have similar thermal conductivity A Nickel plated copper block would achieve this same result at a fraction of the cost..(Copper is a lot harder to machine than brass, so it would cost more to manufacture.)

Hardened pure copper (work hardened is the only way it's sold) would lose its temper the first time you heated it up.

Regards,

Edited 2 time(s). Last edit at 09/21/2021 03:46AM by Andrew_F.

Quote
Andrew_F
This got me thinking..

Oxidation would be a problem on a copper block.

Does anybody make/sell a nickel-plated copper heat block?

Plenty of brass ("hardened copper") ones for sale on eBay, but brass and aluminium have similar thermal conductivity, so, aside from appearance, there would be little difference. Pure copper (Cu11000) and Silver have similar thermal conductivity A Nickel plated copper block would achieve this same result at a fraction of the cost..(Copper is a lot harder to machine than brass, so it would cost more to manufacture.)

Hardened copper (work hardened is the only way it's sold) would lose its temper the first time you heated it up.

Regards,

I believe E3D makes a nickle plated pure copper block.

Thanks rq3,

A local supplier had Nickel plated copper in C18150 (coefficient of thermal conductivity (k) ~ 324Wm^-1K^-1 at 20^oC) at around 12 euro each delivered. C18150 is the alloy used in soldering irons.

At the price, I bought one to trial.

For comparison:
Pure copper (Cu11000) has a k value of around 400 at room temp, with pure silver being 440 at room temp.
At a guess, the aluminium used is 2000 series, approx conductivity k = 120-190, depending upon the alloy, at room temperature. The exact value varies widely, dependent upon the temperature and alloy content..
Data that I found (which I'd trust) in a brief 20 minute search was an NIST journal article. [srd.nist.gov] <-- p10 and 11 - values stay pretty well constant between 0^oC.and 300^oC

Interestingly, however, your diamond nozzle has a thermal conductivity closer to 2000 at working temperature (at 200^oC) <-- p15, figure 6 of the above link. It gives (for type 1 diamond - clear) k=2000 at Room Temp (I'd trust your data better, as this data is 1972) but k essentially doubles at 500K to k ~3500 (it's on the nose of the conductivity curve.) Extrapolating, this gives you a k-value around 1600 to 2000.

Regards,

A

(this heater block is going to be heavy now, guessing the other reason why aluminium is used to minimise momentum.)

Edited 9 time(s). Last edit at 09/21/2021 05:58PM by Andrew_F.

Quote
Andrew_F
Thanks rq3,

A local supplier had Nickel plated copper in C18150 (coefficient of thermal conductivity (k) ~ 324Wm^-1K^-1 at 20^oC) at around 12 euro each delivered. C18150 is the alloy used in soldering irons.

At the price, I bought one to trial.

For comparison:
Pure copper (Cu11000) has a k value of around 400 at room temp, with pure silver being 440 at room temp.
At a guess, the aluminium used is 2000 series, approx conductivity k = 120-190, depending upon the alloy, at room temperature. The exact value varies widely, dependent upon the temperature and alloy content..
Data that I found (which I'd trust) in a brief 20 minute search was an NIST journal article. [srd.nist.gov] <-- p10 and 11 - values stay pretty well constant between 0^oC.and 300^oC

Interestingly, however, your diamond nozzle has a thermal conductivity closer to 2000 at working temperature (at 200^oC) <-- p15, figure 6 of the above link. It gives (for type 1 diamond - clear) k=2000 at Room Temp (I'd trust your data better, as this data is 1972) but k essentially doubles at 500K to k ~3500 (it's on the nose of the conductivity curve.) Extrapolating, this gives you a k-value around 1600 to 2000.

Regards,

A

(this heater block is going to be heavy now, guessing the other reason why aluminium is used to minimise momentum.)

Like all engineering, there is always a trade-off. My silver heat block is 99.99% pure, and is based upon the E3D V5 design, which is smaller than most current heat blocks at 16mm square and 12 mm thick. It weighs 23 grams.

I suspect that most aluminum heatblocks are 6000 series aluminum (6061 to be exact), since it is the most readily available extrusion alloy, and very easy to machine, finish, and anodize.

The DiamondBack nozzle I have on order is sintered polycrystalline diamond. The manufacturer claims a conductivity of 543 W/m-K, much lower than single crystal diamond, but way higher than anything else (ruby, for example, at 40; or standard brass at 114).

[www.championx.com]

My concern with the diamond nozzle is its hardness. I use a piezo nozzle contact sensor of my own design, and when things don't go well, the nozzle scrubs on the bed. This is not an issue with a brass nozzle, since I can just wipe the brass smear off of the bed with ferric chloride. The diamond nozzle will actively cut the glass bed. It will be interesting.