GICAR RL30 Autofill Schematic

Need help with equipment usage or want to share your latest discovery?
westland
Posts: 20
Joined: 15 years ago

#1: Post by westland »

I've been trying to make sense of the electronics in my Quickmill Anita, and did an analysis and photo shoot on the GICAR controller ($100-200 ... there must be something in there). I couldn't find much at http://www.gicarsrl.com/ so looked under the hood .... hmm ...not much. I have posted this to http://ids313.info/gicar.pdf for those who are interested, with a very rough sketch of what I believe to be the electrical circuit (minus a cap, a few signal diodes for polarity, and a bridge rectifier which I think just doubles up). I'd welcome comments from those more knowledgeable than I (or less, for that matter)

User avatar
erics
Supporter ★
Posts: 6302
Joined: 19 years ago

#2: Post by erics »

While I don't believe this is what you are looking for, here is the electrical schematic for Anita:
http://users.rcn.com/erics/QM_ELEC_1.jpg
Skål,

Eric S.
http://users.rcn.com/erics/
E-mail: erics at rcn dot com

westland (original poster)
Posts: 20
Joined: 15 years ago

#3: Post by westland (original poster) »

Hi Eric:

I have a copy of this from some of your earlier posts. I have one of David Blaine's PID systems installed (and I think the controller might be going bad), so my layout is slightly different (BTW, I purchased one of your grouphead thermometers, which is awesome ... should be standard on the HX machines IMHO).

Anyway, you should start building these auto-fill controllers (again IMHO) as the markup must be phenomenal. I bet you could build these for $25 ea in quantity (use PCBExpress for the boards ... I've had success with them). At $75, that's 200% markup and you'd have the aftermarket.

Chris

User avatar
erics
Supporter ★
Posts: 6302
Joined: 19 years ago

#4: Post by erics »

. . . and I think the controller might be going bad
Well, this is getting a little OT, :) but . . . on a fully warmed up PID'ed machine, the boiler pressure and the grouphead temperature should remain essentially constant as the machine is simply idle.

Now, I'm not going to say that I even begin to understand the rationale around this but it has been posted previously that a repeat of the autotune process helped cure PID controller woes. Initiate your autotune at a temperature 10% below your normal setpoint.
Skål,

Eric S.
http://users.rcn.com/erics/
E-mail: erics at rcn dot com

jpboyt
Posts: 220
Joined: 14 years ago

#5: Post by jpboyt »

Not sure what your question is. As far as the component funtions on the board you are looking at a transformer to reduce the AC line voltage, bridge rectifiers and capacitors to complete the AC to DC conversion. Almost all electronics run on DC although your signal may be AC, DC, pulsed or a combination. Most of the time the extra diodes you see are to shunt off voltages of the wrong polarity (spikes) or to bleed off the stored energy in the relay coils. Simple box. Looks at the level probe to open the auto fill valve and run the pump. Main failures are burnt relay contacts ,either welded shut or burnt and unable to carry current on closing, overheated transformer causing an open circuit or short internally, and dried out power supply caps. I have also found a few switching transistors die on the more complex control boxes with integrated touch pads.
jpboyt

westland (original poster)
Posts: 20
Joined: 15 years ago

#6: Post by westland (original poster) »

Eric:

Actually David had me run through the autotune process (similar to what you have on your PID writeup). This I did a couple of times (letting it hunt for an hour or so. It never did settle in to a constant temperature, and the first time hunted around 138F, the second time around 77F. The thermistor might be a problem, but one other thing is a little fishy ...the output to the controller (which is a Taiwanese Delta DBT4824 ... very similar to the Chinese PID controllers you have done your writeup on) is 24.6 vdc (versus the 14 vdc 40ma that the specs say it should be). I notice your Chinese controllers have a 10vdc output. Since the SSR takes 3-32vdc, this will trip it properly (and is) but it still doesn't work properly.

Chris

westland (original poster)
Posts: 20
Joined: 15 years ago

#7: Post by westland (original poster) »

Hi jpboyt

I believe you ... this looks pretty straightforward, and much of the circuitry is there to control transients. I would think that this whole control unit could be built around cheap, low power solid state relays (they are just powering the solenoids for the fill operations). I'm going to see if I can come up with a design.

Chris

User avatar
erics
Supporter ★
Posts: 6302
Joined: 19 years ago

#8: Post by erics »

. . . this looks pretty straightforward . . .
For an electrical engineer with a background in controls, this is likely so. BUT, for the uninitiated, it is not. In this thread, Giotto not maintaining boiler pressure, I said that I had two Gicar controls with identical model numbers in two different machines. The newer machine, a Bezzera BZ07, had a controller having about 75% of the footprint of Anita's controller. For sure, both controllers had the transformer and DPST relay but the other "stuff" appeared reduced in the Bezzera, IIRC. The transformer powers the relay (a lot of us know that) and also provides the power for the level control circuitry. According to Bill Crossland (billc), this is a "capacitance driven" circuit - Astoria SAE Junior (CMA/Laurentis/RIO) Refurbish Log . At one point in time, about 40 years ago, I was very familiar with RLC circuits but to reverse engineer this today from both a technical and financial standpoint would be intimidating.

As a sidenote, there have been numerous posts on this forum dealing with Gicar/Giemme controller problems and there have been a large number of successful repairs by very resourceful individuals BUT nowhere will you find "ABC Electronics" that stocks rebuilt/tested boards for various machines.
Skål,

Eric S.
http://users.rcn.com/erics/
E-mail: erics at rcn dot com

jpboyt
Posts: 220
Joined: 14 years ago

#9: Post by jpboyt »

Chris
I do circuit repair and I think that you would have a hard time with the economics of the endeavor. There is a lot of thought and reasoning behind every component that goes onto a board. Reduced component count is always a factor as the lower the number the fewer the failure points. A lot of the parts are on the board to give the components a chance when some other part fails. I have many boards that are blown up in spite of being fused.The other part is getting your stuff to market through as few hands as possible. Everbody has to eat. And speaking of product liability. UL testing and the like... Commercial endeavors bring a lot of baggage. As far as switching line voltages, which is what we are talking about, this can be done in hardware or silicon. If you go hardware we are talking mechanical relays and control circuit normally consisting of a Darlington transistor pair to engergize the coil in the relay. Solid state for low current is normally handled by triacs fired by opto couplers. As a side note, I started a business repairing espresso circuit boards and recently tested four different Gicar 3D5 boards. All boards appear to be identical but exhibit different tea button funtions and auxilary circuit functions. Group fabbed boards for hobbyists is not uncommon but is not designed to generate any income for anyone. Too much coffee,,,starting to ramble.
jpboyt

westland (original poster)
Posts: 20
Joined: 15 years ago

#10: Post by westland (original poster) »

Thanks for the input jpboyt and eric ... and I'm sure these are wise words. I've looked around the forum too for GICAR posts (and in particular was interested in understanding the nature of the sensor ... is it binary, resistance, capacitative, or something else?). I did come across this:

"The Gicar autofill controller (RL30) used on the Duetto has a timed delay built in....when the water level drops below the probe, for around 15-20 seconds it will not call for water....when it does call for water the fill does not stop immediately the probe is grounded as it fills. There is another delay so that it keeps filling for another 5-8 seconds after the probe is grounded. This results in a filled water level above the level of the probe tip. This is great because as you used water for steaming, it doesn't immediately kick the pump in and ruin the shot. The end result is you can steam quite a bit before it tries to cut in again, perhaps 2 or 3 sessions of steaming. It also avoids cool water immediately entering and reducing steam power. These timed delays also happen within the electronic circuit used on the internal tank."

So some of the circuitry is delay circuitry. So it's probably best to keep my hands out of teh GICAR (and mine works fine right now anyway).

Post Reply