A modified pop-corn popper with separate control circuits

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
orwa

Postby orwa » Sep 01, 2009, 4:18 pm

My brother and I had a busy night yesterday. We worked continuously for 12 hours to assemble a modified pop-corn popper that we have been designing for a while (a whole week; planning, buying parts/tools/wires, etc.). The modified popper is wholly enclosed in the original casing, which is a little bit unusual given the complexity of the modification and the size of the original pop-corn popper. The roaster has a small fan on the side to cool the electronics inside (namely the 1000W dimmer circuit and the 4-13.5VDC fan control circuit). The modification was tedious, and I had times where I was about to give up due to some complexities, but it worked after all, and I am very happy that it worked.

To briefly describe the roaster, it is provided with a small, fan-speed-control PCB that has a knob attached to it to control the fan speed. The circuit is designed to ensure a minimum operating voltage of 4 volts for the DC motor of the fan, and a maximum operating voltage of 13.5 volts. This minimum bound on the fan speed/voltage is intended to secure the heating element from being damaged, or from damaging the surrounding wires and electronics. The DC voltage is provided to the DC motor using a laptop's switching power supply (a.k.a. charger). This type of power supplies is different from a conventional transformer in the sense that it's very efficient and automatic with regard to the input voltage (i.e. will always produce a regulated 15.5 volts regardless of the input AC voltage). The fact that I used this type of power supplies for the fan control circuit means that the whole roaster can still be operated using a Variac with only the heating circuit being effected by the varying input AC voltage.

That is for the fan circuit. For the heater circuit, on the other hand, I used a 220V/1000W dimmer switch (made in China, sold for 35 S.R. -less than $10). I pulled the dimmer circuit out of its casing, attached a small heat sink to the back of the output transistor (the electronic component that dissipates most of the power), and added a small 12V cooling fan to the side to aid in its cooling. This active cooling of the cheap 1000W dimmer seems to allow me to use this dimmer to control the 220V/5A heating element without damage. I also used a panel current meter to watch the current being passed to the heating element, since the relationship provided by the dimmer switch is highly non-linear. Note that in the case of the fan control circuit, the circuit is highly linear by design (half way = half power), and thus needs no such addition.

The modification didn't cost much in total if we only calculated the cost of the parts that made it into the roaster, but it was very tedious. Note that one of the most expensive parts was already available, which the the 15.5V/5A TOSHIBA laptop charger. I bought a replacement charger 3 years back for my laptop when the original charger stopped working due to a mechanical shock. I was fortunate enough to examined the old charger today to figure out that one of the inductors was broken off the board, and by only soldiering the inductor back to its correct place on the PCB, the charger was working again.

I said that I was going to describe the roaster in brief but I seem to have given an exhaustive explanation of it, which I think is good thing.

Image
It's generally not a good idea to work on the mechanical and the electronic aspects at once, it can lead into a big mess in my opinion.
Image
From the fan motor back (clockwise in circular fashion): the simple fan control circuit with a heat sink attached to it, the original heating chamber of the pop-corn popper, the heating element assembly (to be put inside the heating chamber), the 15.5V switching power supply (a.k.a. laptop charger), the 10A panel current meter I bought (50 S.R.), and finally, the small 12VDC cooling fan I bought (10 S.R.).

Image
Extremely tired...

Image
Image
The inside of the roaster is very crowded. I don't believe we were able to put everything inside. Note the (out-of-focus) Aluminium tube spacer we used to clear some space at the bottom, this lead to increasing the height of the roaster and not being able to put the lid on in its place, but had the great advantages of allowing us to put everything inside and making the electronics far from the heat of the chamber.

Image
The solution I used to solve the height problem, note that the springs were ideal since they were in balance with the push-away spring attached to the heating assembly, which in turn prevents the assembly from touching the roasting chamber (if you don't get this, it's not an important detail).

Image
Without the lid, springs still on.

Image
Image
The very first roast, hard bean, not well dried, not well developed, chlorine-pool tastes, etc.

Image
In action, not enough light for photography.

Image
The panel current meter, it has 10A written on it but in actuality it shows up to 20A. Note that for this roaster I only use the range from 0A to 5A.

Image
The electronics cooling fan. You may think this is a little bit over-engineered but I think the electronics needs it. I know that the roaster fan already provides enough airflow to cool the electronics but I wanted to be sure... The fan is positioned to cool the side of the dimmer which contains the power transistor.

Image
The switching power supply did not fit wholly inside so I had to add this extension, which is cut from the plastic of the charger casing. Note that the cable of the machine can now be plugged and unplugged.

Image
We also moved the power switch to become closer to the knobs.

Image
The upper knob is for the fan, whereas the lower knob is for the dimmer (i.e. the heating element).

I am very happy that I finally succeeded in modifying a popper, which I think will make me delay the QUEST M3 roaster purchase I already discussed in this forum. It's always better to be exposed more to something similar to the thing you want to buy.

User avatar
gyro

Postby gyro » Sep 01, 2009, 5:02 pm

Good job, I love mad scientist stuff! Hope it works well for you.

randytsuch

Postby randytsuch » Sep 01, 2009, 6:39 pm

Cool project, I like the built in current meter.

I currently use a variac and a popper. Pretty much the same funcionality, except I don't have control over the fan speed right now, but your setup is all self contained, which would be much easier to set up. I already had the variac, so I just used it

Randy

billm3

Postby billm3 » Sep 02, 2009, 1:39 pm

Great job!

I have been thinking about a project like this for my Gene Cafe roaster (only 110V in the US). You may just be the inspiration that I need.

-Bill

User avatar
orwa

Postby orwa » Sep 02, 2009, 3:28 pm

Thanks. The fan control circuit we used is intended to control a load that needs a DC voltage up to 13.5 volts (two volts below the power supply, which in my case is 15.5 volts DC). I will be more than happy to post a schematic in case someone can benefit from this circuit (it costs around $5), however, my guess is that a design such as the Gene Cafe will use an AC motor instead (which simplifies the circuit a lot). The only part of my modification that can be beneficial to you is the little discovery that a cheap 1000W dimmer (Chinese, sold for around $10 in my place) can actually control a 220V/5A load (or, similarly, a 110V/10A load) in case the dimmer was provided with adequate cooling. I can illustrate to you the steps I followed to ensure an adequate cooling for the dimmer circuit. The reason I am telling you this is that there is a considerable jump in the dimmer prices as you go above the 1000W rating. I was able to find another, cheap 1500W Chinese dimmer (was only available for 110V) in my place for 85 S.R. ($23) but in general, the price of a decent 1500W dimmer can reach up to 350 S.R. ($93). I don't have any problem with spending money, but like to act efficiently, which means benefiting from cheap parts when it's possible, and investing more thought in case it would allow it.

noah

Postby noah » Sep 02, 2009, 4:40 pm

orwa wrote:I will be more than happy to post a schematic in case someone can benefit from this circuit (it costs around $5)


I would love to see this, along with the illustrated steps you took to cool the dimmer circuit. Thanks for taking the time to post all this!
LMWDP #263

billm3

Postby billm3 » Sep 02, 2009, 7:00 pm

As far as I can tell on the Gene, the fan that blows into the heater, then the roast chamber is a computer type DC fan and the motor that tumbles the drum is also DC, both possibly 12V. I havent't completely disassembled yet because I want to use it as long as possible until such time as I NEED to re-wire.... or until I have nothing else to do :D

Post your schematics if possible, just so I (and others) can prepare.

User avatar
orwa

Postby orwa » Sep 03, 2009, 2:55 am

I disassembled the roaster today and took many photos, so I will guide you through the photos, show you the humble circuit we made and then show you the schematic for it (which is nothing but the example application circuit mentioned in the data sheet of the LM150 component). The amount of detail someone is interested in will necessarily vary depending on the person, but I tried to cover as much detail as possible, so that more people can benefit from this post. For the length of time I have been going through roaster modifications made by others, I always wished if those people could provide as much detail as possible, so I guess this is what I am trying to do here.

Image
Image
Front and back views of the machine before disassembly.

Image
Image
Image

Image
This is the fan control circuit, with a heatsink attached to the LM150 component.

Image
Next to the fan control circuit is the 1000W dimmer switch, with a heatsink added to it as well (will be shown clearly later when I take the dimmer out).

Image
The cooling fan, which sucks air right off between the plates of the dimmer's heatsink, and also aids in the cooling of the fan control circuit on the other side (the cooling fan, the dimmer and the fan control circuit are all on the same line).

Image
Note that the type of terminals used for motor wires can still be inserted successfully even though there is a capacitor welded into the terminals of the DC motor (this 100nF capacitor was advised in the data sheet for the linear voltage control circuit).

Image
The original heating assembly (not shown, enclosed inside) is arranged into two heating elements rather than one, which both serve as a simple voltage divider that can be used to produce an unregulated 12V supply. To those who don't know what a voltage divider is, there are two heaters inside the pop-corn popper, one much smaller than the other. By passing electricity through the path composed of the big and the small heaters, the manufacturers of this cheap appliance could extract a low voltage from in-between the heaters, which can be used to operate the low-voltage (12VDC) motor. In my case, I deal with the two heaters as one big heater, pass the overall current through them both, and neglect the middle wire which was originally used to extract the low voltage from in-between the two heaters (this is the wire insulated with red tape).

Image
These are the terminals of the AC panel current meter. Given that the current to be measured is alternating (AC), there is no point of trying to connect the terminals in a crossed fashion like this. I was too tired at the time of assembling the roaster so I did this stupid mistake. This simple mistake lead into one of the wires getting above the other, which in turn caused it to touch the heating chamber. When I examined the machine, the thermal sleeve protecting the wire had a hole at the point of contact with the heating chamber, and the wire was about to start deteriorating at that point. However, I was lucky enough to open the machine up for this photography session and to replace the thermal sleeve/jacket before this happens.

Image
The crowded interior of the machine.

Image
There is also a black dot at the thermal sleeve that protects the wires of the fan speed potentiometer. This also has been a point of contact with the perimeter of the heating chamber.

Image
Oops, too crowded, about to touch the insulation of the switching power supply (switching power supplies are noisy units which is why they are sometimes covered with metal sheets to shield electromagnetic radiation).

Image
The dimmer switch taken out (front panel and original knob removed of course).

Image
We had also to bend the end of the dimmer's metal plate (which serves as a heatsink) to make the dimmer switch fit in the roaster.

Image
I sawed the side of the plastic casing, namely the side where the power component is located (which is the component enclosed in a TO-92 package and screwed to a metal plate that supposedly acts as a heatsink). After sawing the side to expose the back of the power component, I attached an additional heatsink as shown next.

Image
This is the type of heatsink I used. I used a paper clip to mount it and a high density thermal compound to enhance the thermal conductivity between the back of the power component and the added heatsink. Note that the cooling fan is directed to this added heatsink. The continuous stream of air will continuously change the air between the plates of the added heatsink, achieving a very good cooling capability.

Image
The fan control circuit taken out of the roaster, along with the terminals I used to connect the motor to the case of the LM150 component (in the case of the TO-3 package shown, the package case serves as the output voltage terminal which necessitates a reliable connection such as the one shown here).

Image
The interior of the roaster without the dimmer switch.

Image
Due to the fact that my DC power supply provides 15.5 volts whereas the cooling fan requires 12 volts, I used a resistor in series with the fan to pass only 11.5 volts to it. The resistor is covered by a heat shrink tube and thus cannot be seen in this photo. We determined the value of the resistor by connecting resistors of different values and measuring the voltage across the fan using a multi-meter.

Image
Needless to say, someone can use any knob he pleases for his own invention, which I believe is one of the most amusing choices in the process. I originally used the classic knob shown at the back but decided recently to pick the one made of aluminium. This knob is much better because it is larger and thus allows me to control the heat more easily (due to the use of a cheap dimmer switch, very small moves can sometimes lead into big differences in the current).
Image
Image
The new knobs, much easier to handle though the roaster looks now like a stereo.

Image
Very crowded...

Image
Mistake corrected.

Image
All inside again, can't believe it.


Image
The application circuit I used, which is very simple (only the LM150/LM350 component is needed along with a heatsink and two resistors). Note that to obtain the voltage range someone wants (such as the range from 4 to 12 volts for example), some simple calculations needs to be performed to induce the values needed for the resistors. In general, R1 will be 240 ohms whereas R2 will be broken into two parts: a fixed-value resistor in series with a potentiometer (a variable resistor). The values for the fixed resistance and variable resistance will necessarily depend on the minimum and maximum voltages someone wishes to obtain. Note that this circuit doesn't allow you to shut the device you are controlling completely (i.e. it doesn't allow you to go below 1.25VDC), which has not been a problem in the case of the fan control circuit because I needed a considerable minimum voltage to secure the heating element from damage. Beyond this, however, the circuit is very linear and hence is a pleasure to deal with. The LM150/LM350 components can pass up three amps of current, but there are similar components that can pass a guaranteed five. Note that in either case, there are also some calculations that need to be done to determine the type/size of the heatsink that needs to be used. In my case, the component was dissipating 4W at the worst case which means that a small heatsink such as the one shown could do, however, this could change based on the dissipation.

I am very sleepy, so I think I will go to sleep. In case someone needs any help with this circuit, I will be more than happy to help him with the calculations.

billm3

Postby billm3 » Sep 03, 2009, 8:43 am

WOW! Above and beyond!

Thank you kindly.

-Bill

johndoe

Postby johndoe » Sep 04, 2009, 10:51 am

excellent work and dialog. It nice to see a real craftsperson.

As for roast - try a bit darker next time to beginning of "rice krispy" crackle pop some smoke.

I find this level of roast, the one you picture, the same, light, somewhat ashy and bright = tart or even bitter...

Good luck and thanks for the posts.