La Marzocco GS/3 Sometimes Reboots with Tea Button

Need help with equipment usage or want to share your latest discovery?
wiz561

#1: Post by wiz561 »

I've had my GS/3 for probably about 11 years. We've been though a lot over the years and have done a lot of work on it. Things have been working OK lately but I've noticed that the machine will sometimes reboot after I press the hot tea button and water stops coming out of the wand. It doesn't happen all the time, only sometimes. But the machine will click off and back on again and things will be working again.

Another thing I noticed that when the machine heats up to about the 140* mark, the breaker on the brain box (mains switch) will pop and the plug will be noticably warm. If I wait a few minutes, I can reset the breaker on the machine and things start up again. After this happens, the plug stays cool and the machine doesn't do it again. I'm guessing that maybe the steam boiler element is going bad. I've had a couple of strange things happen and when I tested the ohms last year, it was OK but sort of borderline. I figured this is the cause but thought I would maybe ask anyways, since it's another unusual thing going on with my machine.

Thanks!

User avatar
Jake_G
Team HB

#2: Post by Jake_G »

I thought I was the only one!

You can check the connections to the back of the thermal breaker switch. I had a loose connection on mine that had colored the insulation on the spade connector from the heat. I hoped that when I changed the connector that the random restarts would go away, but alas, they remain. Mine is primarily after a boiler refill call, though. Very rarely I get it, as you describe when the tea valve shuts off, but it is more likely to occur for me when the boiler fill solenoid clicks shut. Or when the pump kicks off, since they both happen concurrently when filling the boiler.

I wish I could help you but instead I'm just REALLY interested in what you find to be the solution. Mine is frequent enough that I often fire up my machine in the afternoon to trigger a boiler refill just to be sure that it will come up to temp in the morning.

A cold GS/3 makes me sad.

Cheers!

- Jake

wiz561

#3: Post by wiz561 »

Jake_G wrote:You can check the connections to the back of the thermal breaker switch. I had a loose connection on mine that had colored the insulation on the spade connector from the heat. I hoped that when I changed the connector that the random restarts would go away, but alas, they remain. Mine is primarily after a boiler refill call, though. Very rarely I get it, as you describe when the tea valve shuts off, but it is more likely to occur for me when the boiler fill solenoid clicks shut. Or when the pump kicks off, since they both happen concurrently when filling the boiler.

I wish I could help you but instead I'm just REALLY interested in what you find to be the solution. Mine is frequent enough that I often fire up my machine in the afternoon to trigger a boiler refill just to be sure that it will come up to temp in the morning.
Well, I'm hoping somebody knows and I'm also hoping that it doesn't get worst. Right now it just happens every so often, but it does seem to be happening more frequently as time goes on. I was thinking that maybe it was some gunk in the button area. Another thought I had was that it was somehow water damage in the button/display box. I've replaced the steam vacuum breaker with the retrofit kit, which included a "T" to splice into the drain box. Over time that plastic T cracks and steam gets out and fogs the display up. This has happened twice to me. It's thoroughly dried out but it still happens...but I don't think it would cause a reboot, but who knows.

Anyways, if I figure it out I'll post here.

User avatar
Peppersass
Supporter ❤

#4: Post by Peppersass »

Interesting problem.

I assume by "reboot" you both mean a restart like when you turn power off with the rear switch and turn it back on, as opposed to the machine going from On to Standby to On again. A restart suggests an intermittent short is momentarily interrupting the DC power supply.

Are you both saying that the reboot occurs when the solenoid (or solenoid and motor in Jake's case) turns off? Or is the reboot happening during the time the solenoid or solenoid/motor are activated and they turn off because off the reboot?

The former seems very odd. About the only thing I can think of that could cause something like that is an intermittent short in one or more of the DC relays on the logic board that switch AC to one of the solenoids or motor. What I'm thinking is that mechanical vibration from the relay opening somehow causes the coil to briefly short to ground, and the short causes the CPU to reboot. Or, if the Gicar has field suppression diodes across the coils, it could be a bad junction in one of the diodes that briefly shorts the 12VDC signal to ground when there's vibration from the relay contacts opening.

Of course, another possibility is that there's mineral contamination on the logic board. That can cause strange intermittent behavior. Jake, I assume you've thoroughly inspected your board. Mike, have you?

BTW, I had the same problem as Jake with a poor connection at the power switch causing a large current draw, heat from which triggered the internal circuit breaker in the power switch. Here's a photo:

Image

User avatar
Jake_G
Team HB

#5: Post by Jake_G »

Peppersass wrote:Are you both saying that the reboot occurs when the solenoid (or solenoid and motor in Jake's case) turns off? Or is the reboot happening during the time the solenoid or solenoid/motor are activated and they turn off because off the reboot?
The former for me.
Peppersass wrote:Or, if the Gicar has field suppression diodes across the coils, it could be a bad junction in one of the diodes that briefly shorts the 12VDC signal to ground when there's vibration from the relay contacts opening.
There are flyback diodes across all the relay coils. The supply voltage comes directly from the H-Bridge after the transformer. The Gicar switches the grounds through open drain DMOS transistors in a TPIC6C595 shift register. Here's block diagram of the shift register:
Image
Image: TPIC6C595 functional block diagram

The drains on each output go to each of the relays on the board. Here's shot showing the flyback diodes between the coil terminals of the relays and the shift register:
Image

And here is a shot showing the H-Bridge feeding the power rail to relays: (The rail runs along the other side after feeding the larger pump relay.)
Image

The H-Bridge feeds DC to the voltage regulator that feeds the AVR. I'm trying to figure out if kickback voltage from a relay coil would be sufficient to brown out the processor if a finicky diode were to blame or if there is another underlying issue likely.

In my case, it is probably one in 10 boiler fills that dumps the power. I can use tea water without discretion for the most part, but the boiler fill that follows is likely to cause trouble. It is instantaneous with the the solenoid and pump kicking off that the machine powers down and displays the firmware on the screen upon rebooting. I suspect it is the relay coil field causing the issue because when I power back up, the boiler does not finish the fill request. It is always a one time event. Fill the boiler then restart.

Cheers!

- Jake

User avatar
Peppersass
Supporter ❤

#6: Post by Peppersass »

What's the green stuff across the PCB traces in the last photo? Mineral deposits? Maybe more cleaning is in order?

That said, I can't think of a mode in which a mineral-deposit short would cause a reboot only when the solenoid and motor are turned off. Seems more likely to be caused by a relay coil field collapse in conjunction with a bad flyback diode. What bothers me about a bad diode is that I'd think it would be very rare. Not something that would show up in two GS/3s. Maybe a relay coil defect sometimes causes more flyback current to flow than the diode can handle?

If it were my machine, I'd get replacement relays and diodes and replace one at a time to try to isolate the cause.

User avatar
Jake_G
Team HB

#7: Post by Jake_G »

Peppersass wrote:What's the green stuff across the PCB traces in the last photo? Mineral deposits? Maybe more cleaning is in order?
Old photo from before I cleaned it :wink:

What about this being caused by EMF collapse in the coil of the solenoid itself instead of in the little relay coil? I've read of folks having microcontroller resets when the inductive load is connected to their processors through dc relays that go away when dry cycling...

Is there a failure mode in the solenoid coil that could cause it to cause an interruption in the AC mains feeding the 3d5 sufficiently to brown out the processor? I don't see any snubbers in the circuit to protect the 3d5 from collapsing fields in the motor or the solenoid coils, so maybe there is something inside the ODE coil that is degrading? At any rate, some solutions I've seen include ferrite cores between the relay contacts and the solenoid coil. It's an odd one, but easy to implement. I think it's also a bandaid for the root cause if it works, but I am scratching my head.

Even a faulty flyback diode makes no sense to me as the coils are connected directly to the DC bridge and the shift register is draining the current. Breaking that switch should trash the shift register, not blip the processor. I suppose we could also have poor filtering on the reset pin of the AVR but this, also seems highly rare and also odd that in my case it is 100% tied to a specific relay or load.

I'm inclined to swap coils between my fill and tea solenoids and see what there is to see. I suppose I could unhook the inductive load of the coil and make repeated boiler fill calls by unhooking the level probe connector and see if it still faults before swapping coils.

Cheers!

- Jake

User avatar
sweaner
Supporter ♡

#8: Post by sweaner »

I guess this is what happens when one buys a consumer appliance rather than a pro machine like a Breville Dual Boiler!! :wink:
Scott
LMWDP #248

User avatar
Jake_G
Team HB

#9: Post by Jake_G »

:lol: :lol: :lol:

Thank you Scott.

I needed that!

User avatar
Peppersass
Supporter ❤

#10: Post by Peppersass »

Jake_G wrote:I'm inclined to swap coils between my fill and tea solenoids and see what there is to see.
Good idea. I was going to suggest that, but you already thought of it!
Jake_G wrote:I suppose I could unhook the inductive load of the coil and make repeated boiler fill calls by unhooking the level probe connector and see if it still faults before swapping coils.
By "unhook the inductive load of the coil" do you mean disconnect the AC leads from the steam boiler fill solenoid coil? I guess that's OK because the bypass valve will protect the pump. If the machine still faults when you do that, you can try disconnecting the motor leads and reconnect the solenoid leads to see if it's related to the motor.

All that said, it seems unlikely to me that a field collapse could compromise the AC supply enough to make the CPU reboot. First of all, what condition of the coil would cause the field collapse to be greater than normal? Second, I think the primary issue with field collapse is that the spike can compromise the relay contacts on the logic board. If that happened, you'd probably see intermittent relay failure when the relay contacts get more charred or pitted. Third, from both descriptions it sounds more like the solenoid and motor turn off because the CPU is rebooting, rather than a glitch being caused by a field collapse when the solenoid and/or motor turn off.

I wonder if the problem is that the solenoid coil resistance has decreased for some reason, pulling enough AC current to glitch the power supply. After all, a common aspect of the failure in both machines is that the solenoid and pump are on at the same time, so it could be a combination of excess current being drawn by the bad coil and the motor. Of course, I'd expect AC current draw from the heaters would be much greater, but maybe the duty cycle of the SSRs reduces the peak draw.

Another thought occurs to me: what if the power supply cap (C1) is leaking? Perhaps the high current draw from the solenoid coil, which could be degraded, plus the motor, in conjunction with an intermittent power supply capacitor leak (short), sometimes causes the supply to glitch. Electrolytic caps in power supplies often degrade over time, and both machines are old.

Yet another thought: perhaps the combined current draw of the coils in the DC motor and fill (or tea) solenoid relays on the logic board is pulling the DC voltage below the threshold required by the CPU. Could be a bad or marginal coil in one or both relays, perhaps in conjunction with the 7805 regulator or C1 going bad. Something like that makes more sense to me than an AC spike or excess AC current draw.