Announcement

Collapse
No announcement yet.

EM6910 Control Board Testing

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

  • EM6910 Control Board Testing

    Safely Testing EM6910 Control Board

    For anyone wanting to work on the control board out of an EM6910 the biggest dis-incentive is the very real hazards of the 240V mains power required to power it up.
    After studying the circuit diagrams in the service manual I concluded that it should be possible to get it working with a small 6V - 7V AC plugpack after a few simple mods.
    This was OK in theory but what I didn't allow for was that the circuit in the manual was for the EM6900 board and that it seems to have been modified for the EM6910. My board is marked as EM6900 V1.3.
    Still, after much trial and error and tracing out the circuit I was able to get a beating heart working out of the body and, in the process, was able to learn a lot about how it functions. I was also able to track down a shorted triac in the steam pump driver.

    First of all, I would like to thank TommyV for his donation of the patient in the "Pay It Forward" section of the CS Forum. The clapped out machine he was ready to toss was just what I wanted. Since it was pretty-well beyond repair I was able to do a full teardown with a clear conscience and learn a lot in the process. Unfortunately he was at work when I picked it up so - "A big thanks Tom!"

    The main differences between the 6900 circuit diagram in the manual and the board from the 6910 are-
    Relays instead of Triacs on a sub-board to control the heaters in the two thermoblocks.
    Extra optocouplers to monitor the heater circuits and mains power.

    Basically the board has two sections, the microcontroller with its associated inputs and outputs running on a 5V regulated suppy from a small on-board 240V to 7VAC transformer and the driver circuits controlled by the micro and switching the mains supply to the pumps, heaters and solenoid valve.

    The first thing I did once I had the board out of the machine was to remove the transformer. This came off fairly easily once I had sucked up the solder. (solderwick would probably work just as well as a good solder sucker). I then bridged the 2 pairs of pads with 2 insulated wire links - soldered from each prim to sec pad.

    I rather optimistically figured that at this stage it should work so I attached a little 6VAC (230mA) plugpack to the "Line" & "Neutral" terminals and all of the other peripherals to their respective white sockets down the side of the board - 2 NTC sensors, 2 microswitches, the flowmeter, control panel, piezo buzzer and lastly a little magnet on the reed switch under the white silicon. I did not bother with the low water LED.
    To simulate the 240V water pumps and solenoid valve I used 3 little 12V torch globes wired between each respective terminal and the "Line" terminal.

    OK - time to power it up. All good - no smoke, nothing getting hot but the globe attached to the steam pump terminal was lit (Tom had told me that the steam pump ran as soon as the machine was plugged in so the triac was probably shot). A quick check with the multimeter showed there was 5VDC from the voltage regulator.

    A press of the "Power" button caused its LED to light - but only for a second then it went out and nothing! The only way to get a repeat of this was to power the whole lot down briefly (I wonder where I learned this trick!). Time to start looking more closely at the circuit for something I had missed. But first I tested the driver circuits by jumpering between pin 2 of each opto to ground with a 390 ohm resistor causing the 12v globes to light - so the the triacs seemed quite happy to work with 7V AC!.

    A couple of hours poking around the board and sketching circuits revealed that this later version had some extra optos to monitor the mains voltages through the two thermoblock heaters and also that the mains is present. Two of these optos are on the relay board and another just below it on the main board.
    These had high value (100k/120k) 1w dropping resistors on their inputs. Shunting them with resistors around 4k provided enough current to trigger the sensors at the lower supply voltage. It was also necessary to simulate the heater circuits which normally connect to the relay board. A jumper between the neutrals on the two boards and a couple of 220 ohm load resistors resistors to the "Line" terminal finished the job. Now a press of the "Power" button got the board into warmup mode and jumpering the thermistors with 390 ohm resistors made it think it was up to temperature and light up the relevant LEDs.
    A press of one of the brew buttons had the water pump globe flash a couple of times with pre-infusion then come on solid. Sucking air through the flowmeter hose to spin its impeller caused it to go out after a few seconds. Pressing the steam microswitch caused the steam pump globe to emit a string of double flashes - all good!

    My next project, now that I have a working test rig, is to see if I can find the most suitable value resistor to use in parallel with the steam thermistor to get the steam thermoblock to hold a steady temperature below boiling point. I figure that if I can switch this in and out of the circuit I will have a convenient way to pump descaler through the steam circuit without the problems of it boiling. This would make regular descaling of this section simple and safe as it would not be necessary to take the covers off the machine.

    As soon as time permits I will try to post some photos.

    Trev

  • #2
    Hi Trev

    I have a dead 6910 I scored as a "pay it forward" a while back. Let me know if you need a few stray parts, as it is getting near the bin. Missing steam pump and most accessories, otherwise complete. Too expensive to freight the whole thing to you. Yet another Perth tap water victim (corroded and leaked onto the main board, which is nominally shot).


    Good work!

    TampIt

    Comment


    • #3
      Trev,
      Wonderful work,would be wonderful to comfirm the viability of descaling solution flow modification,would save so many machines.
      Once you have nailed the problem components,many more machines will live again instead of wasteful landfill.
      My one wish is to somehow incorporate wear pads into the casting of the group head,such an unfortunate design!
      In my opinion the design of the group head was "speedy and planned obsolescence!
      These units are ,I believe,the best models made in the Sunbeam range......just have to work on the noise reduction problem...they would be perfect then!
      I'm slowly replacing copper pipe for flexable tubing,that is going a long way to improve things with my collection.
      Keep up the good work and please keep us informed.
      Cheers,
      Mick.

      Comment


      • #4
        Tamplt-
        Thanks for the offer of the bits - will keep it in mind if I need anything.
        Incidentally, I have a planned use for the case from the old machine I pulled apart so I have put it away for safekeeping - should be good for a laugh!
        Also, I managed to completely clean out the steam thermoblock which was almost totally blocked. I was able to warm it up nicely out of the machine by running its heating element from a 24V garden light transformer and forcing straight vinegar through it using a syringe with a plastic tube attached - the tubing consisted of the insulation stripped from a piece of cable and I found I would fit into the tube through the thermoblock so that the descaler could be forced right up to the blockage. Once I got a flow started I coupled the syringe to the water inlet/outlet using a piece of teflon tubing from the peristaltic ink pump from an old inkjet printer (perfect size and nice and flexible) - lots of bubbling and fizzing from the heated thermoblock as I forced the solution through but it came up very clean. Now I have a good spare.


        Mick-
        Yes, I can't wait to get some spare time to work on this but I have been held up with trying to get the bathroom renovations finished. Also the surf has been quite good here for a couple of weeks now - not sure which is worse, caffeine addiction or surf addiction!

        Re the mods, the first problem to solve will be finding the optimum resistor value to parallel with the NTC thermistor which reads about 150k at room temp (it's only about 15deg here at the moment) so would be specified as a 100k thermistor (25deg C). The temp curve graphs for a 100k thermistor would probably be a good match and could be a good starting point for some parallel resistor calculations to save a lot of trial and error testing (does anyone know the normal operating temp of the steam thermoblock?).
        The shunt resistor would need to be carefully chosen so that the controller cycles on & off as I think it may otherwise signal a fault and cause a shut-down - another function which I hope to look into with the 'simulator'.

        Once I have the shunt sorted I will need to work out how to conveniently switch it.
        I plan to try to source some suitable insulation displacement cable joiners to tap into the wires to the thermistor rather than strip insulation and solder.
        For switching I thought I could either use a slide switch fitted to the back plastic panel or a magnetic reed mounted on a little bracket close to the back panel and activated by a magnet on the outside. Unlike the slide switch option this would not need disconnection should the back panel need removal and would also enable the cables to be tied down safely instead of flapping around inside the machine and getting into trouble.

        (I am putting my ideas up here in-case someone wants to experiment before I get around to it).


        I agree with you that its good to keep these old Sunbeams going. The good thing about them is that they can be picked up relatively cheaply now so you can keep a spare on hand while fixing the other one. It reduces the anxiety of caffeine deprivation should one break down.

        Trev

        Comment


        • #5
          Originally posted by Xanthine View Post
          Safely Testing EM6910 Control Board

          After studying the circuit diagrams in the service manual I concluded that it should be possible to get it working with a small 6V - 7V AC plugpack after a few simple mods.
          This was OK in theory but what I didn't allow for was that the circuit in the manual was for the EM6900 board and that it seems to have been modified for the EM6910. My board is marked as EM6900 V1.3.
          I posted my copy of the 6910 manual on this site a while back. Got (mildly) abused by Java"I love his posts anyway"phile because it is already here somewhere. My copy may be different - I have not tried to find the one he mentioned. Mine is EM6910 Service manual.pdf. It may help. Murphy's Law - I have never used it as I had already sorted my 6910 out by then (with invaluable help from Coffee Classic).

          Originally posted by Xanthine View Post
          Tamplt-
          Thanks for the offer of the bits - will keep it in mind if I need anything.
          Incidentally, I have a planned use for the case from the old machine I pulled apart so I have put it away for safekeeping - should be good for a laugh!
          You are welcome.
          Planned use - care to share? I have had a couple of thoughts for the dead one as the casing is unmarked.

          Originally posted by Xanthine View Post
          Also, I managed to completely clean out the steam thermoblock which was almost totally blocked. I was able to warm it up nicely out of the machine by running its heating element from a 24V garden light transformer and forcing straight vinegar through it using a syringe with a plastic tube attached - the tubing consisted of the insulation stripped from a piece of cable and I found I would fit into the tube through the thermoblock so that the descaler could be forced right up to the blockage. Once I got a flow started I coupled the syringe to the water inlet/outlet using a piece of teflon tubing from the peristaltic ink pump from an old inkjet printer (perfect size and nice and flexible) - lots of bubbling and fizzing from the heated thermoblock as I forced the solution through but it came up very clean. Now I have a good spare.
          More good work. I think you have clarified something for me. My WA friends have blocked the steam wand (from the nozzle end) to the point that nothing flows through. They are all switched on enough to use filtered / rain / anything but tap water (most WA tap water should have "severely corrosive" warnings). The other eastern states guys I know of via this site encounter blocked steam thermoblocks instead - I wonder if that is their local tap water impurities fouling it up from the water tank end?

          I have explained to so many people that the high temp & pressure of an espresso machine means that any water impurity will create havoc... to my mind that is common sense, however it clearly isn't!

          Originally posted by Xanthine View Post
          (does anyone know the normal operating temp of the steam thermoblock?).
          No idea, however a quick reading thermometer should be able to work that out if you have a working 6910 handy. If not, let me know and I can measure my older (Aug 2008) one for you next time I play with it (about 4 weeks - it is "just visiting" family at present). Page 35 of the 6910 user manual just tells you that you can vary the steam temp by +/- 5 or 10 degrees without telling you the default temp. Very helpful...not.

          Originally posted by Xanthine View Post
          (via Mick) I agree with you that its good to keep these old Sunbeams going. The good thing about them is that they can be picked up relatively cheaply now so you can keep a spare on hand while fixing the other one. It reduces the anxiety of caffeine deprivation should one break down.

          Trev
          +1. I mainly use a 7000 these days, having partially retired my older 6910 and my 2 group La Pavoni (I sold my newer 6910 to a friend in need cheaply, still under new pump warranty).

          BTW - the 7000 has a very thorough inbuilt cleaning cycle (both steam and espresso thermoblock systems). It may be worth your while to see how it works. Ironically, the 7000 also tends to build up a lot less gunk than any other espresso machine I know (way, way less than a 6910) so to a large extent the 7000 cleaning cycles would be a lot more useful on the 6910.

          Good luck with the 6910 & the surf - our extra swell here has been magnificent.

          TampIt
          PS: the 6910 DVD is quite useful - I presume you have it. If not PM me.

          Comment


          • #6
            Note:

            My copy may be different - I have not tried to find the one he mentioned.
            This copy seems to be the same - if it is a later edition they don't seem to have updated the circuit diagrams as they are still titled EM6900!

            Planned use - care to share?
            The project with the case is a work in progress - will post details when finished if it works out.

            My WA friends have blocked the steam wand (from the nozzle end) to the point that nothing flows through.
            Could the blockage in the stem wands be due to sucking milk up at the end of steaming rather than contaminants in the water?

            BTW - the 7000 has a very thorough inbuilt cleaning cycle (both steam and espresso thermoblock systems). It may be worth your while to see how it works.
            It would be interesting to look at a copy of the 7000 service manual to see what has changed but not sure if one is available. It is probably very similar to the 6910 with most of the changes being to the programming (probably still got the circuit diagrams labelled as EM6900 )

            our extra swell here has been magnificent.
            The surf got a mention on our news here. Looks like you got the full serve from that storm in the Indian ocean which had 50' - 60' waves. We only got the leftovers but it's still been good.

            PS: the 6910 DVD is quite useful - I presume you have it. If not PM me.
            Have a copy thanks.

            Mick

            The de-scaling mod is still underway when time permits. Once I have had a play with the 'simulator' board to be sure it is all working OK I will put a 100k trimpot in place of the steam thermistor to try to find the on/off resistance values. I have swapped some of the globes for LEDs as they seem to load the plugpack down too much. I have also ordered some T-Tap Quick Splices online as they may be a good way of connecting to the thermistor wires - am just hoping that the red ones suit the gauge of wire in the leads.

            My one wish is to somehow incorporate wear pads into the casting of the group head,
            I wonder if you have considered soldering some brass strips over the wear surfaces. I believe that many aluminium alloys are solderable with the right technique which basically involves mechanically breaking down the aluminium oxide film while applying the solder. Once tinned it may be possible to sweat some brass strips in place.

            Photos-
            As promised, I have uploaded a couple of photos showing the added resistors, jumpers in place of the transformer and the removed steam pump triac (it had cracks and charring so I guess its u/s) It looks a bit rough with the long resistor leads but makes them obvious. I have ordered some replacement 2N6075AGs but they are probably 'on a slow boat from China'!


            Click image for larger version

Name:	Bottom.jpg
Views:	1
Size:	98.9 KB
ID:	739051Click image for larger version

Name:	Top.jpg
Views:	1
Size:	109.6 KB
ID:	739052

            NOTE :This board is only operating from a 6V ac plugpack. In it's present state it would be highly dangerous to operate it from 240v and would probably produce some spectacular fireworks!

            Comment


            • #7
              I have just carried out an initial test with the control board by substituting a 50k ohm variable resistor for the steam thermoblock and obtained the following results -
              Pressing the power button caused the steam and water thermoblock lamps to light solid as in a normal startup (the water thermistor was connected as it was originally but not heated and the 50k pot was at max resistance)
              The the resistance of the pot was gradually reduced while watching the steam lamp. When it got to 7.5k the lamp began to cycle on and off at about 1 -2 second intervals.
              As the resistance was further reduced the cycles became longer with the off times proportionally longer than the on times.
              At around 50ohms the lamp switched on permanently and the Power button led turned red and flashed (Fault).

              So it looks as if the micro controls the steam heater by altering the mark-space ratio and period of the control relay in response to the NTC resistance during warm-up.
              The heater locking on at the low resistance (hot) came as a surprise and will possibly affect my plans for descaling - I will have to give this some thought to see what affects it will have.

              In the meantime I would strongly advise against attempting to fit a shunt resistor to a working machine as it could cause overheating of the steam themoblock. This would result in the thermal cut-out being activated and may also cause the thermal fuse to 'blow'.

              For my next test I hope to also replace the water thermistor with a variable resistor and see how it responds and ultimately simulate a range of machine faults and record the results.
              Hopefully the data obtained will help with fault finding and I am still keen to find a simple and safe method for descaling the steam circuit. Any suggestions would be appreciated.

              Trev

              Comment


              • #8
                Hi Trev

                https://www.sunbeam.com.au/Root/Spec...M7000_ib_1.pdf page 23 has the 7000 descaling cycle. The 7000 also has a different cleaning cycle, plus a poor one for the wand (same as the 6910). The descaling one is far more comprehensive than the 6910's cleaning cycle.

                Yes, the WA folk were clogging the steam wand up by sucking up milk. If a newbie follows the 6910 manual "to the letter" the end result is dried milk in the wand. OH & S gone mad. Like most people with prior espresso use, I didn't even read that section of the manual until the third friend dropped in with a clogged steam wand and I got curious. No one I know that has any experience in frothing milk has had a clogging problem... or steaming injury for that matter.

                TampIt

                Comment


                • #9
                  https://www.sunbeam.com.au/Root/Spec...M7000_ib_1.pdf page 23 has the 7000 descaling cycle. The 7000 also has a different cleaning cycle, plus a poor one for the wand (same as the 6910). The descaling one is far more comprehensive than the 6910's cleaning cycle.
                  Thanks for that - the descaling process certainly looks straightforward and efficient - pity about that as it will probably greatly reduce the number of cheap s/h 7000s coming on to the market with the "steam not working" fault .

                  For those sad technical types bothering to read my posts, here's latest -

                  The 6910 control board continues to run me around in circles in my attempts to get it working as a simulator. I fired it up with both thermistors replaced by variable resistors hoping to get it past the warmup stage. I started with both at max resistance and gradually decreased first the steam until its load lamp started flashing then the water one. It proved to be difficult to get it to the 'ready' stage and required repeated alternate tweaking of both resistors to get it there, obviously it is looking for correct temperatures before it is happy. Still, a good sign as it may mean my planned descaling mods may just work after all. Once in ready mode it proved all too easy to get it to signal a fault by pressing the steam switch or one of the brew buttons. Looks like I will have to make it think it has proper thermistors and heaters connected.
                  Not being ready to give up yet I plan to make working model thermoblock simulators using bead thermistors attached to 12v auto globes (thermo-couplers?).
                  This will require the addition of a bigger 12v transformer to power the 'mains' section.
                  I have already ordered some 100k NTCs online - they are cheap and readily available as it appears that they are used to control the extruder in 3D printers.
                  Now all I have to do is wait for them to arrive - at least I will be able to get on with the bathroom renovations with one less distraction for a while.

                  Trev

                  Comment


                  • #10
                    Keen to continue testing and growing impatient waiting for components to arrive, I decided to try to simulate the thermoblock/thermistor combination using parts I already had.
                    Scratching around in the junk box I found a couple of 12V 10W festoon lamps and decided to use these in conjunction with the thermistors I had from the junked machine.
                    Because the existing 6V ac plug pack I was using to power the board was only good for 230mA I decided to add another 12V 1.7A ac transformer (ex garden lights) to power the lamps as heaters. This necessitated removing the jumper links I had previously put on the board across the transformer pads.
                    The 6V plugpack leads were then soldered to the output pads to run the micro-controller and the 12V ones to the Line/Neutral terminals. The lamps were connected between each of the relay outputs and 'Line' with the thermistors taped to them - ready to go!
                    When I switched on both plugpacks and pressed the Power button on the control panel both heater lamps lit and then began on/off cycling after a few minutes, but I waited and no 'Ready' lights. Foiled again! It seemed that the thermal mass of my simulated thermoblocks was too low and the processor was not happy with the faster temperature cycling.
                    Not to be defeated I tried jumpering the thermistor terminals with a couple of resistors and eventually found a combination which forced it into 'Ready' mode. Once there it happily clicked away cycling both 'heaters'.
                    I found that putting a 1k resistor across each thermistor in turn changed the duty cycle of the heater(lamp) from about 6 seconds 'on' to 2 seconds - a good sign. But when I activated the steam microswitch the heater locked on continuously until I released the switch and it went back to cycling. The same thing happened to the water heater when either the hot water switch or the brew button(s) was pressed.
                    So it looks like the 6910 only regulates the temperatures of the thermoblocks in 'idle' mode and turns them on full whenever they are required to produce steam or hot water. Programming for different steam characteristics or extraction temperatures simply changes the duty cycle of the relevant pumps and possibly the idle temperatures as well. Something to investigate later.
                    The question now arises - will the steam thermoblock temp stay below boiling point when I attempt a descale. I guess it is nearly time to try it 'for real' on my spare machine but first I would like to be able to measure the steam thermoblock temp so I have ordered a cheap IR thermometer online - another thing to wait for!

                    Trev

                    Comment


                    • #11
                      Tamplt wrote-

                      Planned use - care to share?

                      Click image for larger version

Name:	Letterbox.jpg
Views:	1
Size:	64.9 KB
ID:	740900

                      Comment


                      • #12
                        Absolutely love it, well done!!!

                        I love the old "cheapie" p/f addition, just completes the art.

                        Hmm, I wonder if I can "enhance" the rust on my current letterbox surreptitiously? Plotting, plotting, plotting...

                        TampIt

                        Comment


                        • #13
                          Tamplt-

                          Glad you liked my letterbox. There's not too much of the original 6910 left as I didn't want to waste any of the 'good bits' on it.


                          Click image for larger version

Name:	Controls.jpg
Views:	1
Size:	65.3 KB
ID:	740906

                          The control panel buttons are ex computer keyboard.

                          I bent the wands from a straight length of chromed steel tubing from a lamp fitting and stuck them in the original holes with epoxy putty.

                          The water tank lid was sealed in place with silicone left over from the bathroom reno, as was the cup warmer tray on the top and the steam and hotwater knobs on the sides.

                          The group handle cost 40cents at the local op-shop.

                          I drilled drain holes in the waste tray at the bottom and the coffee cup so they wouldn't hold water and breed mozzies in summer.

                          The coffee cup is bolted and siliconed to the perforated cup stand.

                          The letter slot was cut with a thin abrasive disk in the angle grinder. This was also used to remove all of the internal mounting spigots under the lid so the mail could slide in.

                          Click image for larger version

Name:	Inside.jpg
Views:	1
Size:	39.1 KB
ID:	740907

                          Aluminium panels were fitted over the water thermoblock hole top and bottom and another half-height one behind the door at the back to stop the mail falling out when it is opened. I siliconed a magnet to this to hold the door shut as the original ones came out with the plastic water tank liner.

                          So, as you can see there was quite a bit of modding to do, but all good fun.

                          Now to get on with the steam thermoblock descaling experiments - will post more on this soon.

                          Trev

                          Comment


                          • #14
                            G'day all.
                            I'm new here, been looking over your shoulders for a while now.
                            The reason I'm posting is that my EM6910 has started to play up, and I have been looking through thesse posts for some clues.

                            My machine now goes into warming mode, with relays clicking and thermoblock heating, but it never gets to the ready stage. The power light just keeps flashing, and therefore no way to make coffee.

                            My first thought was the thermistor, so I measured its cold resistance, about 150kOhm, then fired it up and let it sit for about half an hour so it was nice and hot, then switched off and unplugged and measured again as quickly as I could. The resistance was now around the 10kOhm range (from memory). According to the spec I found (see ATC Semitec GT-2 NTC Thermistor) this would equate to the 80-90 DegC range.

                            So I think my problem is not the thermistor.

                            I did a system reset, hoping to maybe fix something, but no change.

                            So does anyone have any suggestions on where to look next?

                            Originally posted by Xanthine View Post
                            Re the mods, the first problem to solve will be finding the optimum resistor value to parallel with the NTC thermistor which reads about 150k at room temp (it's only about 15deg here at the moment) so would be specified as a 100k thermistor (25deg C). The temp curve graphs for a 100k thermistor would probably be a good match and could be a good starting point for some parallel resistor calculations to save a lot of trial and error testing (does anyone know the normal operating temp of the steam thermoblock?).
                            The shunt resistor would need to be carefully chosen so that the controller cycles on & off as I think it may otherwise signal a fault and cause a shut-down - another function which I hope to look into with the 'simulator'.
                            Xanthine, I was told that the optimum temperature for the water into the coffee was 92DegC. The spec for the NTC above would equate this to about 7.2KOhm. So your 390 Ohm would equate to 200DegC. I would hope it thinks this is hot enough!

                            I might try jumping a variable resistor across the NTC to see if i can get mine to settle.

                            Oh, just found another thread that says the steam block and its NTC also as to be operating. Hadn't checked that, as I just assumed that group heater was all thats need to start. I'll have a look tomorrow and see if that's my problem.

                            Steve

                            Comment


                            • #15
                              Hi Steve
                              It sounds like your problem could be with the steam side of things.
                              If you listen carefully during warmup you can hear the two relays (steam and water) clicking on & off independently.
                              I would suggest that most likely the steam thermoblock mains circuit is at fault rather than its thermistor.
                              If the steam thermoblock is not heating you would need to check the following components in its circuit -
                              Heating element - two terminals at the top of the thermoblock. 45R to 50R.
                              Thermal cutout - lower left of thermoblock. close to 0R.
                              Thermal fuse - lower right of thermoblock in a white silicone sleeve clamped under a small metal plate. 0R resistance.
                              Thermistor - upper left of thermoblock. resistance depends on temp.

                              The last possibility is the relay on the control board which could have burnt contacts or an o/c coil (or, least likely, the control board itself)

                              Good luck with your search and let us know how you get on.

                              Trev

                              Comment

                              Working...
                              X