Honeywell ST799 keeps turning off

[Bob2016]

Hi

The ST799 controller on my heating/hotwater has become very intermittent. It will run for around ten minutes (either HW or CH) and then the display goes blank and the boiler turns off.

If I press the reset button it will run again for a few minutes then turn off again.

I've Googled around and the usual advice is to replace the battery but took the front panel off, checked the backplate connections are ok, checked the battery voltage which is at 3.2V and have noted even with the display off overnight when reset next morning the time is correct so it looks like the battery is doing its job?

Has anyone come across similar symptoms? as usual its picked a great time of the year to fail

thanks

Bob

 

[Echo the husky]

Sometimes there are caps in the power supply that fail, they can be replaced if they are defective. Otherwise a new programmer should do the trick.

 

[cjcoffin]

hI Bob, did you ever find a cure for this, I have exactly the same symptoms. Batt checks out at 3.1V so not that. Keen to find a fix as I see s/h one is now £189.99 on ebay(!)

 

[Deleted member 267285]

I see s/h one is now £189.99 on ebay(!)

S/h? A new programmer doesn’t have to be a straight swap, just change the wiring over.

 

[cjcoffin]

Or new £400 on Amazon(!). I would like to replace like-for-like as there is a lot of complicated wiring including an internet-base control box and a Honeywell evohome setup all connected to the backplate and there is not room on the panel for a wider controller.

But seeing the prices I have taken the current one to pieces. It is fairly easy to work on being mostly through-hole discrete components. The big 680nF cap in the power supply tests OK but there are scorch marks on the pcb round the two zener diodes as they have obviously got very hot over the years.

So I am trying to debug the circuit ATM (am electronics engineer). It looks as though I could replace all the components hereabouts for less than a tenner, but if anyone already has a circuit diagram or knows which components usually fail that would be a great help!

 

[RandomGrinch]

if anyone already has a circuit diagram or knows which components usually fail that would be a great help!

Sorry, not helpful, but this was the first hit I had from YouTube!

 

[Deleted member 267285]

Sorry, not helpful, but this was the first hit I had from YouTube!

A nice easy fix

 

[ianmcd]

got to love this post a "De-bug" engineer hi-jacks a 7 year old post wanting to change components on a 20 year old thermostat that can be replaced for buttons

 

[cjcoffin]

A pity your reading isn't as good as your sarcasm <there is a lot of complicated wiring including an internet-base control box and a Honeywell evohome setup all connected to the backplate and there is not room on the panel for a wider controller>.

 

[cjcoffin]

So it turns out the Drayton Digistat has an identical 0.68uF dropper capacitor in the power supply section which is similarly prone to failure as described here. But the original Iskra parts seem to have a limited life so I am going to fit one with a slightly higher voltage rating (310V) from Wurth Elektronik, the lead pitch is 3mm different but that will not matter.

While I have got it apart I plan to replace the two BZT03C39 zener diodes which get very hot with higher rated 1N5366 5 watt zeners, the two ordinary diodes with 1N4004s and lastly fit a new 100 ohm series resistor, for a total parts cost of £2. Then hopefully the power supply rail will be restored to its proper 39 volt level. ATM it is working again (just) but when both relays are energised the voltage droops to 30V which is only marginally above the relay "must operate" voltage of 28.8V.

39V seems rather an odd choice for a circuit which uses 48 volt relays, no doubt Honeywell had their reasons, maybe the lower current drain meant they could get away with a smaller, cheaper capacitor. Also the 5-way jumper lead joining the two pcb's is very stiff, and prone to fracture at the terminations. The backlight is pretty much useless now but I can do without that so will not attempt to replace the surface mount LEDs.

BTW if anyone reading this needs a replacement clock battery as well, they are also available from Farnell or from the Small Battery Company.

 

[HSingh]

So it turns out the Drayton Digistat has an identical 0.68uF dropper capacitor in the power supply section which is similarly prone to failure as described here. But the original Iskra parts seem to have a limited life so I am going to fit one with a slightly higher voltage rating (310V) from Wurth Elektronik, the lead pitch is 3mm different but that will not matter.

While I have got it apart I plan to replace the two BZT03C39 zener diodes which get very hot with higher rated 1N5366 5 watt zeners, the two ordinary diodes with 1N4004s and lastly fit a new 100 ohm series resistor, for a total parts cost of £2. Then hopefully the power supply rail will be restored to its proper 39 volt level. ATM it is working again (just) but when both relays are energised the voltage droops to 30V which is only marginally above the relay "must operate" voltage of 28.8V.

39V seems rather an odd choice for a circuit which uses 48 volt relays, no doubt Honeywell had their reasons, maybe the lower current drain meant they could get away with a smaller, cheaper capacitor. Also the 5-way jumper lead joining the two pcb's is very stiff, and prone to fracture at the terminations. The backlight is pretty much useless now but I can do without that so will not attempt to replace the surface mount LEDs.

BTW if anyone reading this needs a replacement clock battery as well, they are also available from Farnell or from the Small Battery Company.

Hi, Thanks for providing this info. Please could you offer an opinion on the following. I have an ST699 programmer which is probably about 25 years old now and is still working well. It has a brown backplate so may well have the original varta type block cell which can be easily removed for replacement. I bought a second hand unit from ebay for £40 a couple of years ago advertised as the battery no longer working so hopefully if there is a sudden breakdown I can quickly switch it over and keep the system running. The alternative appears to be a new programmer where the wiring will probably need looking at and likely redoing and likely the heating will be lost for a period of time until this is done. The second hand one has the soldered in battery and I check it now and again to make sure it isn't leaking although my understanding is these type of batteries are less likely to leak. I also noticed the BZT diodes you refer to are the ones that appear to heat up more and the browning on the white board around them. thanks for the tip on the jumper lead (will avoid reopening unless neccessary) .
I would prefer to store the spare with the battery out and I have a few questions. I was going to remove the battery by cutting it at the clips. Presumably the unit will still work but it will not hold the time setting in the event of a power cut? I was also thinking about soldering some wires onto the clips running to a battery which will be kept in a separate space. The other end of the wires will be secured to the new battery with conductive tape and wrapped in an insulation sleeve. This will reduce the risk of the battery leaking into the pcb which I have heard can happen (maybe more with the previous block type). Also this will avoid soldering onto a cell button battery which has clear risks I understand and do nt consider myself competent to do this safely. The battery will then also be easily accessible. Also if I use the block type Varta battery rather than the cell button one will this matter? Would be grateful for any further suggestions/ improvement. If this is feasible I will probably look at snapping up a new old stock one if it is offered at a good price while they are still available Thank you.

 

[cjcoffin]

I would prefer to store the spare with the battery out and I have a few questions. I was going to remove the battery by cutting it at the clips. Presumably the unit will still work but it will not hold the time setting in the event of a power cut?

I imagine it will work OK without any battery at all but have not tried it.

I was also thinking about soldering some wires onto the clips running to a battery which will be kept in a separate space. The other end of the wires will be secured to the new battery with conductive tape and wrapped in an insulation sleeve. This will reduce the risk of the battery leaking into the pcb which I have heard can happen (maybe more with the previous block type). Also this will avoid soldering onto a cell button battery which has clear risks I understand and do nt consider myself competent to do this safely. The battery will then also be easily accessible. Also if I use the block type Varta battery rather than the cell button one will this matter? Would be grateful for any further suggestions/ improvement. If this is feasible I will probably look at snapping up a new old stock one if it is offered at a good price while they are still available Thank you.

I don't think you should do this. THERE IS A DANGER OF ELECTRIC SHOCK. The use of a voltage dropper capacitor for the power supply means the circuit inside is not isolated from the mains so ALL PARTS OF IT ARE LIVE INCLUDING THE BATTERY CONNECTIONS.

The risk of these small rechargeable batteries leaking is pretty small so my inclination would be to keep it as it is. If you are worried about it then fit a new battery now then you will have a spare unit already to go (or lend or sell on). It is not particularly difficult to solder in the coin cell type. If I were doing it again I would replace the 0.68uF cap but probably not bother with the other parts as they did not make any difference. The main issue was having to replace the jumper between the two boards. Replacing that with individual insulated stranded wires was quite easy with a desoldering gun but entirely feasible without one as there is plenty of room to work in once you have unclipped the two boards.

 

[HSingh]

Thanks for your quick reply. My understanding was these types of batteries have a very low risk of leaking also and will not touch anything. Will probably buy a new old stock unit and keep it as a spare as it is. I was enquiring after watching this video.

 

[HSingh]

BTW I wasn't thinking of keeping the battery outside the unit if that is what you thought but still inside the unit but in a location where it was not able to leak over the pcb as it is now. I won't be doing this at all in any case and it was just an enquiry.

 

[cjcoffin]

It wasn't entirely clear and so I wanted to make sure in case someone re-reading this thread in the future (as you have done!) thought having the battery externally was a good thing to do.

Video quite impressive but he very nearly shorts out the battery with the long bit of copper tape on the -ve side.. Also I would have doubled the wires back over the battery to provide some element of strain relief.

If you have found a source of new old stock for less than the £189.99 in #3 upthread please post it here! Bear in mind if buying one the battery may be end-of-life.