Announcement

Collapse
No announcement yet.

Another smoked control box!

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Another smoked control box!

    Hi all,
    We are now into my third machine with a destroyed control card. First I had to replace the main and keypad cards in a 2007 NS Appia due to malfunction, and two weeks after installation half the LEDs in the NEW keypad are dead even observing proper ESD protocol!, then the control card in the 1996 Faema Due I rebuilt refused to manual dose and a miracle eBay card cured it, and now a VFA-Expres volumetric machine I rebuilt four months ago dropped the Giemme control box in a puff of smoke. As I am in North America, the unit is a 240V 60Hz power configuration.

    For some reason, the control box went into thermal runaway and fried some resistors on one of the perpendicular boards inside. The box is not protected by a fuse. As luck would have it, a similar machine came in for scrap that had an identical control box. I have installed it, but the autofill does not work and occasionally and randomly, the box shuts down and all lamps on the doser keypads flash in unison.

    The box can be reset by switching the machine off and back on again, but again eventually shuts down at random times. Autofill does not work, though the call-for-water lamp remains lit. Are there any other control boxes that will interface with this machine's membrane keypads? The unit is 230V. The keypads use 12-conductor amphenol plugs, the flowmeters are standard Gicar three-wire units, and all these devices including the autofill probe plug into one side of the box: the keypads into amphenol plugs at the top, and the flowmeters and level probe into spade terminals at the bottom. The bottom of the opposite card has eight spade terminals at the bottom, which connect to the ground wire, neutral wire, three-way valves for group one and two, and autofill solenoid. The autofill lamp piggybacks on the wires for neutral and autofill solenoid. Sounds simple...

    My only observations about the electrical service to the machine are these:

    -The machine takes over twice as long to heat-up at the customer's cafe as it does in my shop. The client says this is because the machine is run off a "pony" panel under the counter. According to him, an electrician gave his shop's 200A service a clean bill of health recently. I measured 240V across hot and neutral, and 120V from each leg to ground.

    -The electrical service is not surge-protected.

    If I have to install new membrane keypads from another manufacturer to interface with another control box, I can do that, but hacking-in remote keyboards with mechanical switches would be beyond my financial capabilities. I need a cost-effective solution. Any ideas?

    Thanks for any advice.
    Attached Files

  • #2
    Have you measured supply voltage under load yourself?

    Comment


    • #3
      Skydragondave, I assume you are in Australia. If that is the case you have a supply issue. You should see 240v across active and neutral, 240 v active to earth and zero neutral to earth. You should not be able to pickup a 120v reading in a standard installation.

      Comment


      • #4
        Originally posted by skydragondave View Post
        . As I am in North America, the unit is a 240V 60Hz power configuration.
        Quote for clarity.

        They use a weird twin active system.

        Comment


        • #5
          Originally posted by skydragondave View Post
          Hi all,
          We are now into my third machine with a destroyed control card. First I had to replace the main and keypad cards in a 2007 NS Appia due to malfunction, and two weeks after installation half the LEDs in the NEW keypad are dead even observing proper ESD protocol!, then the control card in the 1996 Faema Due I rebuilt refused to manual dose and a miracle eBay card cured it, and now a VFA-Expres volumetric machine I rebuilt four months ago dropped the Giemme control box in a puff of smoke. As I am in North America, the unit is a 240V 60Hz power configuration.

          For some reason, the control box went into thermal runaway and fried some resistors on one of the perpendicular boards inside. The box is not protected by a fuse. As luck would have it, a similar machine came in for scrap that had an identical control box. I have installed it, but the autofill does not work and occasionally and randomly, the box shuts down and all lamps on the doser keypads flash in unison.

          The box can be reset by switching the machine off and back on again, but again eventually shuts down at random times. Autofill does not work, though the call-for-water lamp remains lit. Are there any other control boxes that will interface with this machine's membrane keypads? The unit is 230V. The keypads use 12-conductor amphenol plugs, the flowmeters are standard Gicar three-wire units, and all these devices including the autofill probe plug into one side of the box: the keypads into amphenol plugs at the top, and the flowmeters and level probe into spade terminals at the bottom. The bottom of the opposite card has eight spade terminals at the bottom, which connect to the ground wire, neutral wire, three-way valves for group one and two, and autofill solenoid. The autofill lamp piggybacks on the wires for neutral and autofill solenoid. Sounds simple...

          My only observations about the electrical service to the machine are these:

          -The machine takes over twice as long to heat-up at the customer's cafe as it does in my shop. The client says this is because the machine is run off a "pony" panel under the counter. According to him, an electrician gave his shop's 200A service a clean bill of health recently. I measured 240V across hot and neutral, and 120V from each leg to ground.

          -The electrical service is not surge-protected.

          If I have to install new membrane keypads from another manufacturer to interface with another control box, I can do that, but hacking-in remote keyboards with mechanical switches would be beyond my financial capabilities. I need a cost-effective solution. Any ideas?

          Thanks for any advice.
          When I lived in Colorado, every 240V device I owned with high turn on current fouled up on standard home wiring connected to the American grid. If you are in Canada, it may be the same, however this time I am assuming US. Luckily, I had designed my own power filters years ago in Australia (protecting computer servers & a wine bottling plant) so I had a few sent over and solved the destructive side of the problem. After 4 or 5 disasters my "fairly strong", although not entirely certain conclusion is as follows:-
          1) The US grid itself automatically & fully compensates for load variations with a small delay.
          2) When you "hit the switch" on a device with a high turn on load*, the 240v side voltage drops down massively and any inverter type device will exacerbate the drop by attempting to maintain 240v at the device and passing even more load (i.e. plus any inefficiency) and therefore drop to the 110v grid. Note: the home / US grid is basically incapable of handling high transient loads, and that much I did prove at the time.
          3) The grid sees a much lower voltage than the raw load would normally generate.
          4) The grid finally compensates (complete with lag) by massively increasing its voltage into the inverter.
          5) The inverter then doubles the increased voltage and passes it onto the device. Note: A number of poorer quality inverters actually overshoot at this point, which does not help at all.
          6) The device fries unless it has really good, rapid overvoltage protection. I actually measured just over 350v during this stage in Colorado.
          7) The grid returns to normal. The device doesn't.
          * High turn on load is relative here, most European espresso makers connect to a grid with several times the transient load capability, as does Australia. They would not consider the US grid as an issue there, as they make special 110v machines to cope with the lower capability. They are not just lower voltage conversions... Mark Prince has an article somewhere about La Marzocco trying to get a GS3 to work on 110v, with limited success at the time. The final shipped machine was rubbish compared to a 240v one, and yes, I have used both. Like disconnecting half your spark plugs in a car. In my own case, my highly modified Tilbrook 6000 mono power amps (1.8kva transformers per channel) would actually pulse my house lights in reverse synch to the music if I wired the second channel in. I didn't ever try the other 4, I just gave up and sent them back to Oz.
          There are three practical solutions to this.
          Get a 240v surge protector with a reset, get a slow turn on unit or get a 240v 2200KVA+ line interactive UPS. I can only comment on Liebert / Emmerson UPS's as my original testing (too many years ago to be relevant now) failed every other brand at the time.
          The former is far cheaper and easier, and can even be purchased as part of a 240v powerboard. However it will trip too often and be a general PITA.
          The slow turn ons vary from device to device, and may be worth pursuing. Making your own is easy if you are handy with electrical / electronic gear.
          The Liebert / Emmerson line interactive UPS is way more expensive, however it effectively buffers the mains from the device by providing its own “mini power station” via its batteries and inverters. Mind you, I am unsure if the US can even provide enough oomph to run one unless you go to three phase over there. Smaller than that is a waste of time, and even a 2200 is marginal for most “high end” 240v espresso machines.
          FYI, my own filter used a corcom power filter that would absorb a fair amount of surge before blowing my sacrificial varistors. I mainly chose it at the time because it was immune to the numerous & constant frequency shifts on the Australian power grid at that time and would withstand most lightning strikes (a real problem in West Australia as most of the state is either iron ore or bauxite, and lightning loves it)... These days with a number of devices using the mains to generate signals, it is not a realistic option for most users. Today I would use a few large MOV's and some high speed circuit breakers if I felt the need... and no, I am not volunteering.



          Anyway, hope that helps.



          Cheers

          TampIt

          Comment


          • #6
            Originally posted by TampIt View Post
            When I lived in Colorado, every 240V device I owned with high turn on current fouled up on standard home wiring connected to the American grid. If you are in Canada, it may be the same, however this time I am assuming US. Luckily, I had designed my own power filters years ago in Australia (protecting computer servers & a wine bottling plant) so I had a few sent over and solved the destructive side of the problem. After 4 or 5 disasters my "fairly strong", although not entirely certain conclusion is as follows:-
            1) The US grid itself automatically & fully compensates for load variations with a small delay.
            2) When you "hit the switch" on a device with a high turn on load*, the 240v side voltage drops down massively and any inverter type device will exacerbate the drop by attempting to maintain 240v at the device and passing even more load (i.e. plus any inefficiency) and therefore drop to the 110v grid. Note: the home / US grid is basically incapable of handling high transient loads, and that much I did prove at the time.
            3) The grid sees a much lower voltage than the raw load would normally generate.
            4) The grid finally compensates (complete with lag) by massively increasing its voltage into the inverter.
            5) The inverter then doubles the increased voltage and passes it onto the device. Note: A number of poorer quality inverters actually overshoot at this point, which does not help at all.
            6) The device fries unless it has really good, rapid overvoltage protection. I actually measured just over 350v during this stage in Colorado.
            7) The grid returns to normal. The device doesn't.
            * High turn on load is relative here, most European espresso makers connect to a grid with several times the transient load capability, as does Australia. They would not consider the US grid as an issue there, as they make special 110v machines to cope with the lower capability. They are not just lower voltage conversions... Mark Prince has an article somewhere about La Marzocco trying to get a GS3 to work on 110v, with limited success at the time. The final shipped machine was rubbish compared to a 240v one, and yes, I have used both. Like disconnecting half your spark plugs in a car. In my own case, my highly modified Tilbrook 6000 mono power amps (1.8kva transformers per channel) would actually pulse my house lights in reverse synch to the music if I wired the second channel in. I didn't ever try the other 4, I just gave up and sent them back to Oz.
            There are three practical solutions to this.
            Get a 240v surge protector with a reset, get a slow turn on unit or get a 240v 2200KVA+ line interactive UPS. I can only comment on Liebert / Emmerson UPS's as my original testing (too many years ago to be relevant now) failed every other brand at the time.
            The former is far cheaper and easier, and can even be purchased as part of a 240v powerboard. However it will trip too often and be a general PITA.
            The slow turn ons vary from device to device, and may be worth pursuing. Making your own is easy if you are handy with electrical / electronic gear.
            The Liebert / Emmerson line interactive UPS is way more expensive, however it effectively buffers the mains from the device by providing its own “mini power station” via its batteries and inverters. Mind you, I am unsure if the US can even provide enough oomph to run one unless you go to three phase over there. Smaller than that is a waste of time, and even a 2200 is marginal for most “high end” 240v espresso machines.
            FYI, my own filter used a corcom power filter that would absorb a fair amount of surge before blowing my sacrificial varistors. I mainly chose it at the time because it was immune to the numerous & constant frequency shifts on the Australian power grid at that time and would withstand most lightning strikes (a real problem in West Australia as most of the state is either iron ore or bauxite, and lightning loves it)... These days with a number of devices using the mains to generate signals, it is not a realistic option for most users. Today I would use a few large MOV's and some high speed circuit breakers if I felt the need... and no, I am not volunteering.



            Anyway, hope that helps.



            Cheers

            TampIt
            Thank you, this is a very comprehensive response. I am in Canada. Not sure of any differences between countries other than electrical code.
            I have not tried measuring voltage under load and that is a very good idea. I have learned that the power cable from the main panel to the pony panel is only 8ga. It is possible there is a voltage drop because the machine has a 3500W heating element. The former machine it replaced four months ago was only 3000W and was a manual dosing machine.
            I was not aware that voltage could spike after the load switched off. Do I have that right?
            I don't think the customer would be receptive to an expensive electrical upgrade like a UPS, though they have agreed to a surge protector. Would an auto-reset type be available?
            A slow turn-on device sounds more practical but I wonder if either increasing the gauge of the pony panel supply wire or swapping in a lower power heating element might be a better solution.
            Thanks to all for your thoughts.

            Comment


            • #7
              Pulling 15A over 8mm^2 cable shouldn't be an issue unless the cable run is particularly long.

              Supply seems to be identical to the US, though your infrastructure might be better designed, I don't know.

              The startup and full-load voltage (at both the point of supply and the pony-board) would be an easy check that should identify supply voltage aberrations or voltage drop between the point of supply and the pony board, if any exists.

              Comment


              • #8
                Originally posted by skydragondave View Post
                Thank you, this is a very comprehensive response. I am in Canada. Not sure of any differences between countries other than electrical code.
                I was not aware that voltage could spike after the load switched off. Do I have that right?
                Would an auto-reset type be available?
                A slow turn-on device sounds more practical but I wonder if either increasing the gauge of the pony panel supply wire or swapping in a lower power heating element might be a better solution.
                Canada / US - I do not know either.

                Yes, you have it right. Virtually any power delivery system / grid set to automatically compensate for power loadings will create some sort of surge when the load eases off. The faster the load drops, the higher the voltage surge usually is. There are are whole pile of new electronic control systems that are supposed to minimise the surge. I am now out of that loop directly (now semi retired) but other current industry colleagues remain unconvinced they are making the kind of differences that the makers are claiming.
                The auto resets are exactly what blacked out 1/3 of the US in 2003 / 2004 for a few weeks. Generally they are not a good idea, however in your case you may be able to hunt one down.
                Hate to be the bearer of bad news, but increasing the size (i.e. overall conducting capacity) of the cable actually speeds up the power delivery, so surges impact more severely if it makes any difference at all. At least a bigger cable may not heat up as much under sustained loading. My two power amps were warming the wires all the way to the mains board when I let them have their head for half an hour. Crap spaghetti wire...
                A lower power element is one of the many mods La Marz did to make it work. It just took longer to recover and, being used to Aust power, I beat the machines over there far too often. Now I think about it, maybe it was the US Strada I read about via coffee geeks. Having tried both US and Aust versions of both (plus a few others), they are literally a different world. Mind you, having to put up with crap amplifiers for 18 months probably makes me less than objective on this subject - that really hurt.
                FWIW, now I think about it a little more, I doubt the grid could feed a big UPS without special wiring, in which case a direct connection would also be viable with a slow turn on anyway.
                Conclusion: A slow turn on is really your best bet, or you can start nobbling the performance of the machines - which I personally find unacceptable.

                Hope I have clarified rather than confused.

                TampIt

                Comment


                • #9
                  Hi, Thanks again for the response. I will start calling around and asking if anyone has ever heard of a slow turn-on. A quick google search reveals nothing for a 220V circuit - only sound system related stuff.
                  I wonder why this is an issue at this customer and not at any of my others. There's thousands of volumetric espresso machines with high powered elements around here on 220, they can't all have complicated voltage regulation systems installed at the panel and they're not all frying cards left and right. Something just doesn't make sense here.
                  Thanks again for your help.

                  Comment


                  • #10
                    I'd be looking around the shop and nearby neighbors for equipment that is causing nasty voltage spikes and/or RF line interference. The most common culprits I've run into are typically older or cheaply made and inadequately filtered motors powering air compressors, air conditioners (Especially roof mounted ones.), and the compressors on fridge/freezers (Especially walk-in models). You'd be surprised how many PCB's I've had to replace in equipment over the years that were taken out by a 'dirty' motor someplace else in the building or even in rare cases in a neighboring one.


                    Java "RFI sucks!" phile
                    Toys! I must have new toys!!!

                    Comment


                    • #11
                      Yikes, how do you test for this? The client has a bar/small restaurant with live music and audio equipment being brought in by acts all the time, sometimes several per night. Do I just concentrate on the 220V equipment or also the conventional 110V stuff?

                      Comment


                      • #12
                        High draw audio amps could account for line voltage/cleanliness issues.

                        The easiest way to test for line issues is to get a UPS or line conditioner system that connects to a computer and reports instances of bad line supply issues used in conjunction with software (Frequently supplied with the UPS/Line conditioner.) that records these reports. Hook it and the recording computer up on a circuit as close as possible to the troubled one, set it to maximum sensitivity, and let it run for a few days/week and see what it reports.

                        If looking just for RFI a handheld AM/FM radio can make for a cheap and simple detector. Walk around the site with it turned on and hold the antenna/radio close to any suspected sources and see if you get static when the suspected source turns on.


                        Java "Line supply problems suck!" phile
                        Toys! I must have new toys!!!

                        Comment


                        • #13
                          Originally posted by Javaphile View Post
                          I'd be looking around the shop and nearby neighbors for equipment that is causing nasty voltage spikes and/or RF line interference. The most common culprits I've run into are typically older or cheaply made and inadequately filtered motors powering air compressors, air conditioners (Especially roof mounted ones.), and the compressors on fridge/freezers (Especially walk-in models). You'd be surprised how many PCB's I've had to replace in equipment over the years that were taken out by a 'dirty' motor someplace else in the building or even in rare cases in a neighboring one.


                          Java "RFI sucks!" phile
                          Too true, I had suppressed the memory of faulty air conditioner brushes in the next door warehouse some 50 metres away that kept futzing up some servers in Sydney. Cost me three plane flights before I nailed the SOB.

                          FWIW, I doubt domestic US supplies could generate the kind of power to do that, however a nearby factory....

                          skydragondave: good luck with sorting it out.

                          Comment


                          • #14
                            Hi all,
                            I have done several searches for UPS/Line conditioner devices and have found no commercially available units that could supply more than 2000W of protected power. Most of the units available are meant to protect audio/home theater devices and plug into a standard 110V wall socket. For 220/230V heavy duty appliances, the choices are even more limited, but nothing close to either the old machine's 3000W or my VFA's 3500W element.
                            The 220V outlet for the espresso machine is the only one, all others are 110V. Even if I did spend the $500+ these systems and an appropriate laptop would cost, plugged it into a grinder or water tower, and found a faulty ice machine to blame for frying a card on a machine I just rebuilt, how on earth could I expect to convince an angry customer that a) he needs to get a new ice machine and b) some other machine or traveling band's guitar amplifier isn't going to do the exact same thing all over again in two weeks because I can't suggest any kind of device that will protect the power coming in to his espresso machine?
                            Unless there is some cost-effective solution I can install in the pony panel straight-off?
                            Not liking my chances of solving this one.
                            Thanks again for the comments and concern.

                            Comment


                            • #15
                              Originally posted by Dragunov21 View Post
                              Have you measured supply voltage under load yourself?
                              To which I'll elaborate and include "under start, under load and under stop".

                              Anyone with a suitable oscilloscope should be able to check the quality/cleanliness of your power for you.

                              Comment

                              Working...
                              X