Intelia HD8751 Pump Only Receiving Low Voltage (70/90V) 230V UK

[Inominata]

Mechanikkenntnisse vorhanden

Amateur but competent

Messgerät vorhanden

Multimeters

Genaue Bezeichnung der Maschine

Saeco Intelia HD8751

My problem is that my Intelia HD8751 stopped descaling mid cycle. Pump did not run. Could not get it out of descale. Went into Test Mode. Most functions worked except the pump. Nothing from the pump. So, I removed pump. Bought a new pump. As I was stuck in Descale, I tried Test Mode to check the new pump. Still nothing. Put multimeter test leads on the ends of thd spades to pump. Only got 70V and then up to 90V when hit OK to continue descaling, which was just a second or so, but no sound. I removed the control board. Checked main Triac BTB12-800W on heat sink in circuit, seems ok. Have checked triac BTA204 marked T1 and suspect it may be bad, if so, is that the likely cause of low power to pump? I have not checked the U6, (ie L7805CV voltage regulator) as it only priduces low voltage output by design.

Any suggestions for a likely cause and fix would be much appreciated. Machine is about 5 years old. UZK, so 230V.

 

[Grisu]

the pump runs with full 230V, but depending on DMM you may see different values because it will run in some interupt mode - not full time on - to control water flow (more or less in different modes).
BTB12 is a 12A Triac (maybe for heater), pump only has a 4A Triac like BTA204.
You may first repair pump to get out of descaling mode, otherwise it wont leave it.
Testmode will work all the time. Pump in testmode should have full 230V all the time I think. To check triac and voltage on pump you could use an old edison lamp instead lightning up when pump is active.
And there may be a clixon on the pump, small white overheat protection fuse.

Maybe others know this type despite me (never had one).

 

[Inominata]

Thanks. That's very helpful. I checked the Klixon fuse before I replaced the pump and concluded it was ok as the multimeter showed continuity. I was/am really puzzled by the 70V and 90V but that was in normal descale mode and I didn't measure voltage in Test Mode, perhaps I should? However, there, there was just no response from the menu on either of the pump options in Test Mode (ie EV or not), Not even a red screen, but I didn't use the multimeter to see what voltage was being delivered, however briefly to the new pump in Test Mode.

The failure in descale mode happened after some months of pump weakness. I was not looking at the machine when it failed.

As you will appreciate, the absence of a schematic for the PCB limits what non experts can logically analyse. I don't know if the BT204 is definitely bad yet. Is it the likely culprit if it is bad, and why would it blow? Does it control the switching of anything else? If it has blown, what may have blown with it? I don't want to take it out until I have a replacement. So far, my readings of M1, M2 and G of it in circuit make me think it may have partly failed.

It was the 90V which threw me, and then I saw something posted about lower voltage programmed on your forum......how is that done?

I'll test the BT204 as you suggest.

 

[Grisu]

It cant be done with a triac, they're open (for a halfwave sinus) or closed, so 230 or nothing. But they can be opened intermittent. Maybe thats been meant with lower voltage.
Usually those pump triacs wont go bad as there is minor current on them, but never say never. Change it to be sure, only <1€. Most times for each function there is one triac (maybe 2 for grinder).
You have to follow (or measure) the print from pump connector to the proper triac to be sure, there are more triacs of same type on the board.

Dont know if there is an opto-triac which drives the pump triac, they may be named ISOxx, often those go bad (LED on input side goes week) and then output side triac wont open to open main triac.
Have a look at this ISO (if there are used in this machine)!!!

 

[Inominata]

Thanks. Looks like the BTA 204 800B is obsolete. I can find Es but see these have a more sensitive gate. If looks like a T435-800T from STMicroelectronics is the cited replacement, so would that be best? I can't see another Triac on the board other than the high amp BTB12 800W, and the other device which looks like a Triac is in fact a voltage regulator. I've seen images of older machines with a row of Triacs but not on the Intelia boards. There are indeed ISOs. I'll follow your advice on testing. In the meantime I'll order the T435-800T if you agree it is compatible but I'll order something else if you advise.

 

[Grisu]

should be ok, but really I dont know this layout and schematics.

Better to wait for others to help!

Found another thread: Saeco Intelia HD 8751/11 Pumpe läuft nicht, Elektronik defekt?
And it says there is not triac on newer models, only an opto-triac like this: https://media.digikey.com/pdf/Data Sheets/Sharp PDFs/PR39MF51NSZ.pdf
There it has been this optotriac, but 4-8 weeks for replacement from China.

A new board costs only €80: Saeco Vollautomaten Platine für HD8751 / HD8752* 421941229373

 

[Inominata]

Thanks, that's a pretty close description of my problem, except mine happened when in Decale. Unfortunately, the link to the optic coupling triac chip and pump jpg is broken, and only one of the other two links works for me. Any chance of fixing them or emailing me the unavailable image and a working link?

My board is the same as this except mine is 00 not 01.

 

[Grisu]

you will find the part R39MFS on your board (I can see it!).

Sign In | AliExpress Mobile

Mobile Online Shopping for Apparel, Phones, Computers, Electronics, Fashion and more from China; Shopping on AliExpress from mobile site, the world's online marketplace!

de.aliexpress.com

Maybe it is easier to find a TLP3616 and this one is even better (more temp-reliable and more current): Frage - Optokoppler

https://www.ebay.de/c/876252278?mke...-53477-19255-0&campid=5338903651&toolid=10001

https://www.ebay.de/itm/TLP3616-Opt...-53477-19255-0&campid=5338903651&toolid=10001

Please do a bit more search on your own, I really dont know any more about it than you and need to google not different to you.

 

[Inominata]

Many thanks for your help (and patience).

I was away from the board when I posted the last message. Indeed there is a R39MFS on my board, but my board has a number of differences compared to the 0 I uploaded before. Some are just repositionings of components, but there are different variants of components too. Furthermore, the R39MF5 is a Surface Mount (Gull Wing) chip, not a Through Hole DIL, and that too is important when looking for a replacement. I can only find a Through Hole TLP3616 sadly, and given what I say and show next, (which I probably would not have noticed were it not for your help), I can see why it would be wise to find a Surface Mount version of the more tolerant TLP3616 or something like it. Any suggestions?

I attach two images of my PCB which I hope is in focus enough to show the odd state of the R39MF5 on my board.

There should be 7 pins on a R39MF5. It appears that one of mine has actually melted. That is rather worrying. Whilst this makes it the likely culprit, it's rather bad that this could happen surely? Surely some fuse should have blown elsewhere to protect the circuit?

 

[Inominata]

I see someone else had the same problem back in 2016:

Kein Strom von der Leistungsplatine pin jp24(Pumpe)

Um dir schneller bei deinen Problem zu helfen, bitten wir dich folgende Angaben zu machen: - Mechanik- und Elektronikkenntnisse vorhanden: - Messgerät vorhanden: - Genaue Bezeichnung der Maschine: reperatur jura f50 und x7 messgerät vorhanden intelia hd 8750 neues problem,wahrscheinlich durch...

saeco-support-forum.de

No follow-up sadly.

 

[Grisu]

you wont get any SMD type, couldnt find one anywhere.
So you have to play a little bit and solder the DIL version on it, you can bend the ends of the pins 90° to make it easier for soldering.
And have a look where the one melt pin should go to and repair the print in this area.

First desolder the old one, then make a good picture only of the small part of print, then I may have a look what to do.

Very cheap 5pcs for €1,55 incl. shipping!!!

1.92€ |5 Teile/los, TLP3616 P3616 3616 DIP 8, Neue Original Produkt, Neue original freies verschiffen schnelle lieferung|Integrated Circuits| - AliExpress

Smarter Shopping, Better Living! Aliexpress.com

de.aliexpress.com

 

[Inominata]

Just something to think about....

Attached are two close-ups of the damaged optotriac. According to the Sharp data sheet, the pins on the input side run 1,2,3,4 with 2 being the anode and the other three being cathodes. It appears that only 2 and 3 are connected on this PCB. Presumably these tracks controls the light source on that input side of the chip. As we go round to the other, output, side, pin 5 is the Gate of the triac, and (the melted) pin 6 is output T1. There is no pin 7, and pin 8 is output T2. You can see where T2 goes but I don't see any tracks for pins 5 or 6, and remember, this is a Surface Mount chip, not Through Hole. Is anyone able to explain the circuit switching logic by pins? It would seem that the straining/struggling pump has drawing too much current through pin 6? But it's only a 48 Watt pump, which at just 200mA is well below the 900mA this chip is supposed to be able to handle. So, a question for engineers, why does this happen, and why not put a PCB fuse near JP24 to protect the components? The manufacturer of the pump put a Klixon on the pump, and Saeco/Philips know that these pumps start to struggle after a few years. The board is much more expensive than a pump. Furthermore, how do later boards differ? Finally, I've been looking for even 6 pin alternatives to the very hard to source in UK R39MF5 or Toshiba TLP3616. There must be other optotriacs which would do the same job, and SMT (Gull Wing) Surface Mount ones too?

 

[Grisu]

There is no need for any fuse, never seen one on those triacs and there are many on other boards - and the triac acted as fuse!
Whats the problem to solder a DIL only on one side like I wrote before?
Even much better as triac will be bit larger and better cooled with some space to the board and last forever

Klixon is only for overheating not for short circuit like a fuse!

I dont know any others for <=1A.

Ever tried to resolder this pin? Maybe triac still ok ...
Seems there is a track between pin 5 and 6, make a large solder point over both and furthermore to the conductor whole in front of pin 6.
If not working desolder the triac that we can have a look at the board and tracks.

 

[Inominata]

There's nothing wrong with your helpful advice regarding soldering a DIL type instead of a SMT, my problem is getting a chip in the first place from a supplier in the UK other than eBay (not everyone is happy to use it). On the other points, well, here is a better close up of the absent leg location at pin 6 and two others showing both sides of the PCB in that area with a light source behind the PCB to show why i am so puzzled about that (nominally 3 legged) side of the optotriac. Only one leg on the triac side seems to go anywhere, which is very odd. If one looks at other variants of the board, that side of the circuit varies considerably. I'm trying to work out the logic of how it is supposed to work.

Incidentally, my original pictures are quite high resolution at 2mb. I then downsize them to under 200k, which look OK before I upload, but online they don't look so good. Does the site do its own additional conversion? I'm trying to make sure you and others can clearly see the chip and its tracks, thinking this may be helpful to others. What size should I upload?

 

[Grisu]

pics are well done, for me all is very clear.

This device has 2 triacs inside, same as it was done before with 2 separate triacs.
First ist the opto-triac on pin 5, from this it internally goes to gate of triac 2 and output of triac2 is pin 6.
Pin 5 is only for special cases where someone wants to trigger output triac not with input-LED but (additional) with this pin 5 or use output of optotriac for other purposes.

You could use a standard opto triac with 4 pins and externally you would have to rebuild circuit for 2nd triac (and a resistor to its gate).

There is no pain with ebay for a 5€ device, the loss would be minimal.
Sorry, dont understand whats your matter, if you want to repair it then take it, so much time and then fail on ebay, too much for my brain.

You cant do any more, you have to get a chip and repair it - NOW!
Print is ok, only one part to replace, nothing easier than that.

 

[uhrmacher]

Why not a fuse?
I have never had a faulty pump spool and have already repaired a lot of machines. A fuse could break for no reason more often than the pumps do, increasing the machine's failure risk.

 

[Inominata]

All sound advice for which I am grateful, but....How about a Panasonic AQH3223 as an alternative to the Toshiba TLP3616?

Is the failed Sharp R39MF5 device not just an optically driven Solid State Relay under another name i.e. photocoupler/triac?

Fehler E05. Kleine Pumpe läuft nicht

Um dir schneller bei deinen Problem zu helfen, bitten wir dich folgende Angaben zu machen: - Mechanik- und Elektronikkenntnisse vorhanden: Ja - Messgerät vorhanden:Ja - Genaue Bezeichnung der Maschine: HD 8856 Hallo Leute ich habe folgendes Problem. Die Maschine sagt mir nach dem Einschalten...

saeco-support-forum.de

If so, the advantage to me if suitable would be that I can get one from a reliable local source, and even get a Surface Mount version. If so, looking at the specs, which one would be best?

AQH3223 Panasonic Solid State Relays - PCB Mount | Mouser United Kingdom

Panasonic AQH3223 Solid State Relays - PCB Mount are available at Mouser Electronics. Mouser offers inventory, pricing, & datasheets for Panasonic AQH3223 Solid State Relays - PCB Mount.

www.mouser.co.uk

 

[uhrmacher]

AQH 2223 could work

 

[Inominata]

Thanks. This company Is a probably better source for small quantities in the UK given lower handling/postage:-

https://uk.farnell.com/panasonic-electric-works/aqh3223aj/ssr-600v-1-2a/dp/2095718?ost=Aqh3223a&ddkey=https%3Aen-GB%2FElement14_United_Kingdom%2Fsearch

The AQH3223A is the Random Turn On (aka Non Zero Cross) version, and A is for SMT. How does that Random Turn On functionally differ from the AQH3213 Zero Cross version? I can see the difference in the datasheet but I am not clear how it relates to the Intelia PCB circuit. I note you have referred to the lower, 0.9A version, the AQH2223 rather than the 1.2A AQH3223. I thought you were advising to go for the higher 1.2A (so that's what I did)?

I'd still like to know how this circuit is supposed to work, but can see why that is difficult given the boards vary, and schematics are unavailable. I'm a bit puzzled by your saying that there are two triacs in the original component rather than two outputs T1 and T2 and G.

 

[Grisu]

take 3223, no zero crossing as old one had non too! 3223 with 1,2A is better than 2223 only 0,9A.
with zero cross it will activate only with next zero crossing of mains sinus wave, means later as old one which opens suddenly when activated.

Its the same from Panasonic!

Thanks for finding this identical replacement.

In the schematics you can see that there are 2 triacs inside, only an optotriac wont give enough current, so amplified with a second one controlled from first via output1 to gate2 (45° line is not output of T2 but its gate).

https://www.panasonic-electric-works.com/pew/eu/downloads/ds_aqh_en.pdf

Also available at Conrad: Conrad AQH3223A