TJ-067 Fan Control Mod for Precise Control

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
slickrock
Posts: 272
Joined: 13 years ago

#1: Post by slickrock »

With a little push from mj05, I took the plunge and modified the fan control potentiometer the TJ-067 that was originally discussed in the Load Sizes thread. I must admit, having done so has been a revelation in ventilation control on the North. I didn't fully realize just how limiting the stock fan control was until after I've spent some time with the fan mode on a few roasts. Now I'm able to set various degrees of low-flow fan speed, which has been great in the early phases of the roast, and then a wide medium fan range for the latter stages. OK, enough superlatives, here's the mod:

1. Totally forget what I said earlier: don't get a an audio/log taper potentiometer. I tried it and its actually more sensitive than the stock pot. It worked a little better in anti-log mode (switching leads and running the pot counter clockwise), but still overall more sensitive than the linear taper.
2. Rather, I took the 500K ohm audio pot (since I already bought it; a linear would work better) and soldered it in series with the existing 500K ohm linear pot for the fan control, basically creating 1M ohm of resistance at full off. Note that the North setup uses the fan pot in "series" rather than voltage tap mode. So soldering both in series makes sense with this setup.
3. I shunted this setup by soldering a 750K ohm resistor in parallel with the dual series pots per above, which creates something like 430K ohms of total resistance at full off. More importantly, what this does is effective modify the taper of the main linear pot just enough to get a wider fan control range from it.
4. Now, my 500K audio pot is used as a bias control to set the threshold by which the main linear pot starts the fan. Here's the procedure: Set the main fan linear pot full-off and bias pot full-off. Now slowly turn the bias pot clockwise until the fan just starts to move. Then back it off slightly until is just stops. This bias is now set.
5. Now you can turn the main fan control to control ventilation, with much, much better graduated control. Note, that there is some hysteresis with fan control when using the fan at very slow speeds. For my setup, I move the control to "20" to get the fan moving, but then back it off to "10" or less to get very slow fan speeds.

Enjoy!
07/11/1991, 08/21/2017, 04/08/2024, 08/12/2045

Dregs
Posts: 69
Joined: 10 years ago

#2: Post by Dregs »

Thanks for working out how to do this mod and for sharing the info. Excellent post!

User avatar
slickrock (original poster)
Posts: 272
Joined: 13 years ago

#3: Post by slickrock (original poster) »

Gary suggested that I provide some photos and other illustrations to help a novice person do this mod. I'll start with the schematic and follow up with some pics later.

Here is the basic setup of the stock fan control schematic (I simplified the drawing by depicting the rest of the fan circuitry as a power source, which it really isn't):

The fan control is 500K ohm pot placed in-line with the circuit (only two terminals are used - left and center terminal when viewed from the top; the right is left unconnected). When off full counter-clockwise resistance is maximum where all the voltage is dropped. Turning the pot clockwise reduces its in-line resistance so more voltage drop goes to the fan circuitry.

Here is the fan control mod schematic:

As you can see both pots are in-line and in series so they share the voltage drop together with the rest of the fan circuitry. The Taper load is resistor in parallel with the pots and creates a exponential-like taper for both the pots. Mere micro-amps would be flowing through the taper resistor so it can be a small 1/2 watt-er.

One final note: I arrived at this design/hack by trial-and-error and with what limited parts I had in my toolbox. One side-effect with this approach is that a little bit of top fan speed is sacrificed, since the resistance of the setup can't be taken down to zero, but since the North fan is so freaking strong, you won't really notice it. That said, after drawing the diagram, I can think of a few alternate approaches that might work as well, like using something like a 5M ohm pot as both the taper resistor and bias pot and reducing its resistance until just before the fan moves. With this approach, you can drive the circuit to zero resistance and hence get full fan speed, but I'm not sure the affect it would have on the taper profile, so one would have to test this to make sure it works and provides good control.
07/11/1991, 08/21/2017, 04/08/2024, 08/12/2045

User avatar
millcityroasters
Posts: 253
Joined: 9 years ago

#4: Post by millcityroasters »

Thanks for the schematic.

Once in place, is there any need to change the bias setting?

I'm wondering if you feel it's necessary to mount the second pot externally or can it be buried inside the panel enclosure?

Tourman
Posts: 16
Joined: 10 years ago

#5: Post by Tourman »

+1
Once in place, is there any need to change the bias setting?

I'm wondering if you feel it's necessary to mount the second pot externally or can it be buried inside the panel enclosure?
This goes to my question of physical layout - inside or out? If out, box mounted or ?
Dan

There is a crack in everything, that's how the light gets in. Leonard Cohen

ira
Team HB
Posts: 5497
Joined: 16 years ago

#6: Post by ira »

Seems like once you figure out a reasonable bias setting, you could either get a suitable sized linear pot or just use the current pot and 2 resistors. Has any one ever plotted resistance to fan speed or air flow so we'd know what the optimal pot would be?

Ira

User avatar
slickrock (original poster)
Posts: 272
Joined: 13 years ago

#7: Post by slickrock (original poster) »

millcityroasters wrote: ...Once in place, is there any need to change the bias setting?
I'm wondering if you feel it's necessary to mount the second pot externally or can it be buried inside the panel enclosure?
It's set and forget. Absolutely no need to externalize it, so it's buried in the enclosure. No need to box it as the existing fan control is rather exposed to begin with.
ira wrote:Seems like once you figure out a reasonable bias setting, you could either get a suitable sized linear pot or just use the current pot and 2 resistors. Has any one ever plotted resistance to fan speed or air flow so we'd know what the optimal pot would be?

Ira
You can only purchase pots in standard sizes, so substituting the stock fan control has its limitations; besides I like the "feel" of the stock control.
Also, YMMV for each individual fan unit in your roasters. So the exact resistance of bias for my setup may be different from yours, hence the pot approach would probably be best.
07/11/1991, 08/21/2017, 04/08/2024, 08/12/2045

ira
Team HB
Posts: 5497
Joined: 16 years ago

#8: Post by ira »

Yes, but Millcity asked the question and he needs to make it shippable and serviceable. Pots come in 470K and 300K without too much searching and one of those or one of those and a resistor might be good enough.

Ira

User avatar
slickrock (original poster)
Posts: 272
Joined: 13 years ago

#9: Post by slickrock (original poster) »

ira wrote:Yes, but Millcity asked the question and he needs to make it shippable and serviceable. Pots come in 470K and 300K without too much searching and one of those or one of those and a resistor might be good enough.
A different objective IMO. First, this a one-off mod (to existing roasters), not a design revision meant for future low cost production. Second, this is a simple cheap hack that's working well for me with limited parts on hand, not something optimized, as I've alluded earlier. Third, if you really want a design improvement, then I would go after the fan control circuitry as a whole, not just some simple taper tweak to an existing control pot. But, for most existing users, this simple mod should get you mostly there.
07/11/1991, 08/21/2017, 04/08/2024, 08/12/2045

User avatar
drgary
Team HB
Posts: 14345
Joined: 14 years ago

#10: Post by drgary »

Joel,

Where would I get those parts? And what does it look like soldered together?
Gary
LMWDP#308

What I WOULD do for a good cup of coffee!

Post Reply