Repair Idromar SH-5000

Hi everybody,

I have repaired a UV water treatment unit, it´s Idromar SH-5000. It had PFC circuit damaged (TD4862, 2 x 1 ohm resistors, bridge rectifier, 1N4148, IRF840, 22uF 450V capacitor, 1,6A fuse). To be sure it works properly, can I connect any 220V fluorescent lamp with wattage less than 38W? It uses two 38W lamps. It seems it works properly but I know usually these systems detect when the lamp is not connected so I would like to test the unit before giving it to my customer.

I have substituted "ordinary" fluorescent bulbs (two pins each end) into germicidal UV light fixtures with similar bulbs, except for spectrum (UV bulbs were low pressure mercury), with no problems. However, to answer your specific question, we need to know more about the UV bulbs used in the circuit you have. A picture of the proper UV bulb might help.

John

Thanks Jpanhalt, I found this:

https://www.novacquashop.com/produc...H0110+PER+STERLIZZATORE+UV+SH_2500_5000~.html

That lamp is distinctly different from common fluorescent lamps in the US. Is that base common in Spain/Mallorca for ordinary fluorescent lamps? It does look like a low-pressure mercury lamp, rather than the other options.

Assuming the lamps are low-pressure mercury and the base is the same with the same apparent connections, I don't see a problem in testing it with ordinary fluorescent bulbs. In fact, ordinary fluorescent bulbs have a small amount of mercury in them. The color you see is due to a phosphor coating on the inside of the tube.

The UV lamps we are discussing emit almost all of their energy at 2537 Angstroms (254 nm). At the wattage you are describing, I would strongly advise against looking at the UV lamps without adequate eye protection. When we used them in the lab to decontaminate processing areas, the labs were locked to prevent entry, and the lamps were on timers.

John

OK John, I´ll test it and I give you a feedback. This is the typical fluorescent here in Spain:

https://media.djmania.net/images/extra/5AKt3bg_u1.jpg

Thanks, Iván.

That design is the same typical fluorescent bulb in the US too. The plug you showed in post #3 is not like that. Do you know what the filament connections are for that type of socket?

John

It'd probabyl work, however you run the risk of flashing something over, fluo's on startup require a high voltage spike, with the incorrect tube this might get too high in voltage.

OK, finally I had to give it to my customer, they will install it and we´ll see if everything works, I couldn´t test it with load. I send you some pictures and I made a quick schematic (sorry about the quality, Monday morning and fast) about lamp connections looking at board:

http://www.restoretronic.com/descargas/Idromar/1.jpg
http://www.restoretronic.com/descargas/Idromar/2.jpg
http://www.restoretronic.com/descargas/Idromar/3.jpg
http://www.restoretronic.com/descargas/Idromar/4.jpg

Thats a complex looking board, I'm sure you could run fluo's with something simpler, my point being you were right not to try a diffrent tube as the board could have got upset, or said 'bang'.

It didn´t work properly, it says LAMP1 OFF and LAMP2 OFF. They give me back the unit now with original lamps but when I was checking it again I don´t know what happened, I had very low value between gate and source, I removed the mosfet and when I measured it was OK, but when I measured at IR2156 pin 13 (low side) I didn´t get the proper signal as in pin 11 (High side). I decided to open gate resistors to see if something was overloading the driver IR2156 and it seems after that both mosfet IRF840 were damaged. Now, without the mosfets, when I measure at pin 11 I have a proper PWM signal but when I measure at pin 13 I don´t have a good one, I measure against ground. I think I should get PWM signal at both but opposite phase, isn´t it?
I ask this because I still have IRF840 to replace them but I should order the driver in case the low side is damaged...

I agree with you. That is what the datasheet shows:


There should also be a dead time between phases. Do you have a scope? If not, I think your voltage measurements are pretty clear.

John

I saw that image at datasheet too but this is not what I get at LO pin.

Yes I have a scope, I will post a picture of the type of signal I get. Anyway tomorrow I´ll order the driver and I´ll test with the original bulb. I´ve seen the original bulb also has 4 pin but all of them on one side.

I'd agree with that too, you could have a pair of fets in parallel but unlikely with that driver, push pull is the most likely.
Most common for push pull is that there are 2 n channel fets, the source on both being grounded, however double check the layout, one might be high side and the other low.

Put a bulb in series with the supply when testing, you'll need something of reasonable wattage not to loose to many volts.

Hi,

dr pepper said:

I'd agree with that too, you could have a pair of fets in parallel but unlikely with that driver, push pull is the most likely.
Most common for push pull is that there are 2 n channel fets, the source on both being grounded, however double check the layout, one might be high side and the other low.

Put a bulb in series with the supply when testing, you'll need something of reasonable wattage not to loose to many volts.

Click to expand...

Yes, they are in push pull connection, drain from one to source from the other.

There are 2 ways (I know of) for push pull, one common + to the centre tap of the primary and a fet switching each end of the primary to ground, or as I suspect in your case 2 primaries, one with an end grounded and a fet to the supply, the other with an end at + supply and a fet to ground.
Maybe thats been done to balance flux in the transformer core or some emi reason, kinda makes it tricky as the drive voltage for the high side fet needs to be higher than the rail, requiring a gdt or other tactics.

Hi guys,

I still have problems with this unit. After replacing mosfets and driver I found a 330 ohm resistor at 500 ohm value and a 18V Zener damaged. The problem I have is that I don´t get voltage at TDA4862 for PFC. I only see one way to get voltage at its pin 8 (see at drawing), no other power source as for example in datasheet example a startup resistor directly from rectifier or from secondary coil used in PFC.

I gave direct DC power at D6 zener diode without 220Vac input just to see what I got at TDA4862 and at IR2156. IR2156 pin 11 and pin 13 shows a perfect PWM signal. However TDA4862 pin 7 shows nothing and when I cross threshold voltage startup it shows DC voltage, not PWM... I tried with another new one and is doing the same so maybe it needs high voltage input.

When I give 220Vac input I see how U5 blocks the unit with no voltage at IR2156 and "Press Start" message on screen, then I press start unit release but I only get 10,6V at pin 2 and 3 at IR2156 (not enough to start it up as I saw it needs 12,2V to start up) and of course 0V at pin 8 of TDA4862.

I attach a schematic I did:

http://www.restoretronic.com/descargas/Idromar/Esquema Idromar.jpg

Thanks for your help.