Is it worth making something for this?

[paul23]

For years I have trying various different manufacturers for a suitable pump for a particular application.  Each one I tried had a problem (reliability, size, noise, etc).  Eventually I found the prefect pump, it´s silent, reliable (not had a single breakdown yet) and small.  Then the bastards went and changed the design.  The new pumps look identical, but the electronics are different.   Before there were two versions, one that detected a temperature differential and one that had an extra wire that detects mains.  They´ve ditched the mains detection one (which is the one I was using) and replaced it with a plug that shorts out the socket for the temperature sensor (so it runs all the time it has power).  This is where the problem is.  When they detected the mains, they would run on for a few minutes after the signal was lost.  Obviously, this is not possible when the only way to control the pump is actually cut the power to it.  I need the run-on, its the most important part.

So, I though I would just redesign the pcb to include a timer circuit and ditch their temperature circuit.  Pretty easy to do and had it working on the breadboard.  When it came to laying out the PCB I couldn´t physically fit all the components on the PCB, and there are not hardly any parts.  I re-used some of the original parts (transformer, bridge rectifier, triac, opto-triac and caps) and added a couple of transistors, a HCF4060 and a voltage divider for the mains detection.  All the parts I tried are in the smallest SMD packages I could find.  So I either need to think of a better way to do it with less components, or make my original design and put it in a box.  The thing is, if I do that, then I need to buy all the parts(I wont be able to re-use the parts on the PCB because I´ll have to leave that in place), and a box.  This is now getting expensive and the pumps are already expensive to start with (about 150% more than what I used last).

So is it worth spending any more time on this trying to figure out a solution, or do I just cut my losses and buy delay timer relays?  What would you do?  It must be possible somehow as they used to do it in the same case in the old pumps.

Some specs;
Pumps consumes 0.07amps 230Vac
The PCB is 45 x 45mm, and available space after the transformer, connector for the motor and cap for the motor is 10x33mm (double sided).  In that space I would need, the timer circuit and mains voltage detection.  In that space now is an op-amp and some resistors.
The transformer output is 9V 0.35VA and has a bridge rectifier(MB2S) immediately after it.

[WBB]

Would it be possible to retrofit the circuitry from the old pump to the new one? I'm aware of how obvious that seems!

[paul23]

If you mean the old design, then that would be ideal.  Except, I don´t have any of the old pumps and the manufacturer (apparently) don´t either, so I don´t know how they used to do it and they wont tell me, except for saying its a similar design to a bathroom extract fan. 

[daedalus]

I guess it depends if this is a one-off, or if you need lots of these pumps, either for your own use or to sell. Personally I would avoid a custom pcb job for a one-off, as you don't see a return on your time.

Can you get your hands on an old version pump, even if its broken? tracing out the old circuit might give valuable insight into how to solve the problem.

without knowing more about the application its hard to give you advice, but I assume there is a good reason you cant dump a bigger pcb in an enclosure nearby?

[paul23]

I guess it depends if this is a one-off, or if you need lots of these pumps, either for your own use or to sell. Personally I would avoid a custom pcb job for a one-off, as you don't see a return on your time.

They are condensate pumps for air conditioning units.  I have 10 of the new ones in stock and would like to use them on future installations.  We are talking maybe 10-20 a year.

Can you get your hands on an old version pump, even if its broken? tracing out the old circuit might give valuable insight into how to solve the problem.

No, you can only buy the new ones now.  The ones I already have installed, I can´t really remove as you have to break the cases to get in to them.  The new ones, although they look identical, you can get in to them without breaking the cases.

without knowing more about the application its hard to give you advice, but I assume there is a good reason you cant dump a bigger pcb in an enclosure nearby?

Size and cost being the main reason.  I´ll get hit with shipping and customs charges on everything I import,  so one charge for the PCBs, one charge for the box, one charge for components, then charged again for the actual pumps.  At the minute I pay around 48€ for the pumps and break even at selling for 160€.  If I have a separate enclosure and buy all the components then it´s going to be nearer 250€ per pump.  Clients aren´t going to pay that for 48€ pump.

[daedalus]

As an importer into europe you are responsible for checking that there is CE certification on the pumps, and although the manufacturer you are using probably has CE for their design, as soon as you swap the pcb out that becomes your problem. I cant see the profits of sale of 20 pumps a year ever recovering the testing costs involved.

If it were me I would see if the pump manufacturer is open to doing you an OEM model with the old pcb, even if they are 2x as expensive, that would be a better option at your volumes.

[paul23]

As an importer into europe you are responsible for checking that there is CE certification on the pumps, and although the manufacturer you are using probably has CE for their design, as soon as you swap the pcb out that becomes your problem. I cant see the profits of sale of 20 pumps a year ever recovering the testing costs involved.

If it were me I would see if the pump manufacturer is open to doing you an OEM model with the old pcb, even if they are 2x as expensive, that would be a better option at your volumes.

I mean this as a genuine question, not as to offend or be a smart arse or anything, but how much of that reflects real word situations rather than "what the rules say".  For example I find it very hard to believe that more than 50% of the stuff on ebay is CE certified. 

If the pump remains the same (it is CE marked, I just checked) and I add additional circuitry, does that matter?  Or if I somehow mange to find an old design and copy it exactly?

I already tried them on ordering the old boards and they won´t do it.  The costs would be too high anyway.

[PA0PBZ]

I already tried them on ordering the old boards and they won´t do it.  The costs would be too high anyway.

If they don't want to do anything with the old design, maybe they are willing to send you the details?

[paul23]

No, they wont do that either.  That is the first thing I asked.  Which in retrospect was probably a bad first question to ask, I should of perhaps asked if they would manufacturer the old boards as OEM first, and then ask for the design.  I totally understand though that they don´t want to send me any of their design files.  I also totally understand why they wont send me the old boards, because I asked about the design, it´s clear that I probably would reverse engineer their boards, then I would have the design files (or near enough) anyway.

[PA0PBZ]

Re-reading the thread I think your best option would be to use something external, either a timer relay or making something like what's inside these bathroom fans yourself. Or using the bathroom fan circuit on a pcb that replaces the one inside the pump, 45 x 45 mm should be enough space to do that.
What is the existing circuit inside the new design doing? If I understand correctly the pump just runs with mains applied?

[kripton2035]

what is the price of ONE old board delivered to you ?
you seems to tell us it is too expansive, but for one ?
then reverse engeneer it and make your owns.
if they don't produce it, they can't blame you if you make them for your "personnal" use ?

[Fraser]

I am not an air conditioning engineer but I have seen plenty of AC condensate uplift pumps in my job.

My personal opinion is that you are getting too focused on the current model you have been using and has now changed. In such a situation, I would dump the idea of modification and spend some money finding another supplier of a pump that meets your needs and buy a bulk load of those instead of messing with a pump that no longer meets your needs.

I know you said that this is the only pump on the market that is acceptable.... I advise you to recheck the marketplace as such pumps are so common that I find it hard to believe that they are all not fit for purpose.

Just my view but modification of commercial parts for commercial use is always a last resort due to development costs and personal liability issues..... I have seen a few failed condensate pumps that have lead to damage due to water overflow..... you don't want to be held responsible if it is discovered that you have DIY modified a failed pump do you ?

Fraser

[paul23]

I agree with you, It would be better just to cut my losses, put delay timers on the ones I have and find another supplier.  The problem with that is that I pretty much think I have tried every single one. 

Generally there are two methods used.  Peristaltic pumps work quite well and are quiet and trouble free at first.  With the climate we have here, the tube used for the peristalic action in the pump dries out quite quickly if the pump isn´t used and then pump doesn´t work and you get water leaks from air conditioning unit.  The pumps are also a bit bulky.

The other method used (which is the most common) are mini pumps, which are a piston type pump (I don´t know what the proper name for them is).  These are generally rubbish.  If they run dry, then they burn out quite quickly.  Again the climate here seems to effect the sensor and the pump runs dry.  It get about (rough estimate) 90% failures within 2 years with these styles of pumps.  Normally the sensors fail and the pumps run dry and burn out.  I have tried 9 different manufacturers over the years and always had the same problems with all of them.  They are really noisy as well.

There are tank pumps as well, which are great.  But no suitable for this application.

The pumps I am using now, use a different kind of pump.  Its some kind of patented black magic, which will run dry all day long without trouble, they are fine with dirty water, silent and small.  perfect for me really.  If I could find another manufacturer which had a similar design, then I would switch over in a second.  However, Google says no. 

[Fraser]

I understand your problem.

The pumps that I have worked with are not black magic and work on the diaphragm pump principle. These use a push-pull action on a neoprene diaphragm and are both reliable and quiet. The valve flaps are the only weakness as these can get blocked open by debris but they generally are self clearing in such a situation.

It is worth noting that ships bilge pumps often use this type of self priming diaphragm pump design.

I will see if I can find the manufacturer of such AC condensate pumps for you

UPDATE:

Here you go.... Just search for diaphragm pumps for AC condensate. These are much better than any piston design and will run dry without damage.

http://www.directindustry.com/prod/homa-pumpenfabrik/diaphragm-condensate-pumps-for-air-conditioning-31233-575704.html


http://www.bluediamondpump.co.uk/


Fraser

[paul23]

Thanks,  the first pump you linked to is a tank pump and not suitable for my usual application (too big and needs to be lower than the air conditioning unit).

The second one is the one I am talking about.  I didn´t want to mention them by name, but oh well.  They are brilliant pumps the best I have ever used.