Jake_G wrote:The interesting thing is that using the tea water invariably leads to a boiler fill and sometimes that boiler fill overlaps the use of the tea wand...

Still scratching my head. Also, any thoughts on the warm xformer? My tray is out in the open without a top back behind the machine and it was very warm...

Aha. I think you may have deduced the common thread between your symptoms and the OP's -- i.e., running tea water invariably leads to an autofill, sometimes while tea water is still running.

If it's really the case that autofill is the culprit, why only autofill? It's unlikely that the autofill DC relay on your board and the OPs both happen to have some sort of defect, and I think that probably applies to the flyback diode as well. So what conditions prevail during autofill that don't prevail at other times? For example, it can't be the motor and autofill relays being energized at the same time because the motor and 3-way relays get energized together, as do the motor and tea water solenoid if pump is configured. One difference with tea water is that you could have the motor, autofill and tea water DC relays all energized at the same time. That would argue for a DC voltage sag due to a leaky C1 cap and the heavier DC load.

Along these lines, the intermittent nature of the problem suggest that maybe another load is coming into play, like one or both SSRs. Jake, you mentioned that the autofill turns off the heaters. Do you know that for sure? If that's true, it's a virtual certainty that the steam boiler SSR will come on when autofill completes due to the inrush of cold water. Maybe the CPU enables the SSR just before it de-energizes the motor and autofill relays, so there's a momentary spike in the DC power draw with all three loads going at the same time. Now suppose the brew boiler SSR just happens to firing at the same time. The total draw of the two relays and both SSRs, in conjunction with C1 leaking intermittently, may be enough to pull down the DC voltage and reset the CPU. An autofill with one or both SSRs firing along with the tea water DC relay being energized would be even worse.

Perhaps one way to test this is to see if the problem occurs in Half mode, which would eliminate the possibility of both SSRs to be on at the same time.

However, this doesn't explain why the problem doesn't occur when you brew a shot. There you can have the motor relay, the 3-way relay and both SSRs firing at the same time.

The only other thing I can think of that's unique to autofill is the steam boiler level probe. It transitions from ungrounded to grounded. Perhaps that transition sometime causes a momentary power supply short because there's a minute amount of mineral contamination in the area of the probe connector on the Gicar. In my experience, that area is prone to contamination. Or it could be a marginal semiconductor or IC that comes into play only during the probe transition.

As for the warm transformer, I've not noticed that when I've had my brain box pulled out. I'll check that next time . That said, I wonder if a leaky C1 might be increasing current draw through the transformer.

Jake_G wrote: Also, I measured the coils and diodes on all the relays, but I think the measurements are meaningless being in their respective circuits as I got 0.72 ohms on each one when biased against the diode and 0.56 ohms with forward bias. Could be my crappy meter, but maybe not. The open relay sockets read 0.56 ohms with forward bias and were open when backwards so I have zero idea where the 0.72 ohm reading coming from.

Mine reads 0.70 as well.

But not in Ohms. 0.70 in kilo-Ohm.

Are you sure you are not in the kOhm range?

I am almost willing to bet you are in kOhm range as reading a low 0.7 is far too low for a cheap DMM using cheap probes.
I use a Gossen and Probemasters and they read 1.25Ohms (when shorted). Obviously, I can use Kelvin connections when I want lower.....

Oh - and the 0.56 would be since your compliance voltage is enough to forward bias the diode.

Edit: I just measured the voltage on the capacitor and it is 13.3V. I think autofill is on as the board is disconnected from the machine so the autofill probe would be open.

Is your board powered on with the keypad?

Nope - just the 3d5. No connections. But the keypad doesn't consume anything (just switches) and the LCD is a few mA max...

(BTW - after 25 minutes the trafo is lukewarm at best. You have a short somewhere... Disconnect all connections and try.).

That's interesting because my power-on voltage across C1 is 17.5V with the board powered off.

sweaner wrote:I guess this is what happens when one buys a consumer appliance rather than a pro machine like a Breville Dual Boiler!!

I'm following this thread, because Jake and I are friends, my machine has the same "Jake-valve" profiling mod as Jake has on his, and I have always been a GS/3 fan But is it times like these that I am thankful that I only have to change an o-ring or PTFE seal every so often. This would be a real head scratcher to me.

-Peter

AssafL wrote:(BTW - after 25 minutes the trafo is lukewarm at best. You have a short somewhere... Disconnect all connections and try.).

If there's a short or even a partial one, that could explain the hot transformer, the fluctuating voltage and the resets.

I'm pretty sure the only DC outputs from the Gicar are to the SSRs, display and button pads. That should make it easier to determine if the short is on or off the board.

Is the transformer warm in standby?

If it is, that would be pretty solid evidence against C1, no?

I will check it when I get home.

Peppersass wrote:board.

Is the transformer warm in standby?

It's warmer than ambient, but only slightly so.

With warmup voltage at 13.5, using the tea spout drops it to 11.2 with no resets...