Vitodens 100W Power Pump Module Failure

Beaver33

Vitodens 100 WB1B stopped working a few hours after start-up for the new heating season. Turns out the power pump module has 2 broken resistors (R7&R8) that connect in series with the room thermostat. Other components on the board appear good. What would cause the resistors to blow?

I'd like to replace the resistors. Does anyone know the ohms value of these resistors?

Thanks.

Alan (California Radiant) Forbes

Call Viessmann tech. support (844)649-5886 and have your boiler's model and serial number ready.

I don't know if they will have the information you want, but they can give you a part number for a new board.

Beaver33

I prefer replacing the broken resistors rather than purchasing the Viessmann, 7831 396 Power/Pump Control Module for US$476 at Boston Heating.

Can anyone with visual access to this module tell me the color bands on the 2 resistors labelled R7 & R8 on the board, located near Connector X4 in series with the RT connection (labelled 3 & 4)? Thanks.

GGross

@Beaver33

just to make sure does your boiler serial start with either of these 2 sequences?
7374861 or 7374862


or does it start with one of these?
7457960 or 7457961


I have one of these boilers installed at my house I can try to get into that module to get you a picture, but I'm not sure if we have the same power pump module. there are 2 different ones depending on which series boiler you have, they kept the models the same but one is capable of hooking to the old combi kit and the other is not. I'm not certain that the 2 modules are exactly the same other than the combi connection

GGross

Here is a photo of the power pump module for the combi capable boilers. As far as I know the boards are the same, but the connections to it are different. I had this one new on my shelf it would be for serials starting with
7457960 or 7457961

Could be worth a try if you have the capabilities

Beaver33

Thanks very much for your reply and the photo. My boiler number starts with 7374862.

From the section of the board in your photo, the board appears the same. To help confirm this, could you send a photo showing the entire board? Thanks again.

GGross

Beaver33

Thanks. Your board is populated with additional components. Most of the original components appear identical. (However C6 appears to be a larger value capacitor, 3.3uF compared with 1.0uF on my board.)

Can you confirm the stripe colors on R7 and R8 to be Brown, Blue, Black, Gold, Brown? I appreciate your assistance.

Beaver33

My board. Note the burnt resistors (R7 & R8) next to the X4 connector.

109A_5

Hello @Beaver33,

Looks like 16 Ohm, 1%, 1/4 Watt resistor to me.

If you do not know what caused the problem yet, I would put in 1/2 or 1 Watt resistor, may give you some margin for error. Be advised the higher wattage resistors may have thicker leads.

Clean the carbon off the circuit board and mount the new resistors up away from the circuit board.

It could be a defective bridge rectifier or a severely electrically leaky capacitor, bad relay coil. Or an external or improper voltage was applied somewhere it should not have been.

Looks like either input may energize the relay and/or just energizes an opto-isolator.

109A_5

Hello @Beaver33,
Your R5 and R6 look a bit beat up too (getting dark). Since they look like they are on the LED side of the Phototriac coupler, I'd say you were applying an additional voltage where maybe just a contact closure was needed. Or there is two transformers out of phase somewhere in the system. Not sure if a secondary of that transformer on that board supplies power to that part of the circuit.

The Phototriac coupler ( looks like PC3SF21YVZ Series ) may have been damaged by the situation and / or its life shortened, running the LED too bright ( too much current ).

Beaver33

I agree with your identification of the resistor value. Thanks for your additional comments related to R5 & R6. I will replace them.

To be sure the Phototriac is OK, I could swap them since the second one is related to the Domestic Hot Water circuit which is not in use.

I've attached a portion of the circuit diagram around the broken resistors. The secondary of the transformer does supply power to this portion of the circuit. The capacitors (C1 & C3) test OK.

Is the transformer a likely cause? I measured 27V AC output voltage (no load). The windings measure 46 ohms (primary) and 2.5 ohms (secondary). Could the transformer create a fault only when under load? Are there any other tests I should perform on the transformer to help confirm its condition? Should I replace the transformer in addition to the resistors?

Throughout the module most resistors are in pairs in parallel. Why would this be done, rather than using a single resistor of equivalent resistance?

Thanks for your help.

GGross

Beaver33 said:

Thanks. Your board is populated with additional components. Most of the original components appear identical. (However C6 appears to be a larger value capacitor, 3.3uF compared with 1.0uF on my board.)

Can you confirm the stripe colors on R7 and R8 to be Brown, Blue, Black, Gold, Brown? I appreciate your assistance.

That looks correct to me

109A_5

Hello @Beaver33,

Did your unit have a jumper X4 - 3 to 4 (RT) like the @GGross board ?
I suspect the 22uF Cap. is for power factor correction and maybe some transient filtration.
Two like value resistors in parallel doubles the Wattage, may have simply been less expensive if they used those values elsewhere instead of a single higher Wattage resistor.

Interesting your board's issues and the @GGross board appears to have no ill effects. Maybe your R7 and R8 were populated incorrectly.

Is the relay coil related to that circuit or just in the vicinity and totally unrelated ? Looks like its coil is 24VDC.

Where is X4 - 3 to 4 (RT) in the circuit ? Between the transformer and the two burned up resistors ?

109A_5

Hello @Beaver33,

Beaver33 said:

I've attached a portion of the circuit diagram around the broken resistors. The secondary of the transformer does supply power to this portion of the circuit. The capacitors (C1 & C3) test OK.

Is the transformer a likely cause? I measured 27V AC output voltage (no load). The windings measure 46 ohms (primary) and 2.5 ohms (secondary). Could the transformer create a fault only when under load? Are there any other tests I should perform on the transformer to help confirm its condition? Should I replace the transformer in addition to the resistors?

The transformer voltage seems normal, probably much closer to 24VAC at a 10 VA load. Fault under load ? That would raise the output Voltage of the transformer, I suppose a partial primary short could do that. However the transformer does not look like it was running hot, so I think that is an unlikely issue, but not impossible. You could run it under load, either a separate resistor about 56 Ohms, 15Watt or greater or with the repaired circuit board monitored closely.

Since I don't see a jumper X4 - 3 to 4 (RT) like the @GGross board and both sets of resistors were running hot, to the point of one set failing, I would think the most likely cause of that type of failure was an in phase voltage was applied to X4 - 3 to 4 (RT) so the circuit that was designed for nominal 24 VAC had a significantly higher Voltage applied to it.

Beaver33

Thanks very much for your comments.

X4 - 3 to 4 (RT) on my unit connects to a room thermostat (powered by 2 AA batteries). You are correct that RT is between the transformer secondary and the burnt resistors.

The board contains 2 relay coils. Neither appears related to the portion of the circuit in question. One is related to the pump and the other to domestic hot water. (The DHW terminals on X4 have never been connected to anything.) The schematic below (from inside the cover of the unit) shows the complete Power Pump Module. (Unfortunately it is not a true representation of the circuit because it does not show some capacitors that are present on the board.)

Given your comments regarding the transformer, I'm wondering now if the likely cause is a failure of the rectifier bridge. I will remove both bridges for testing.

As an aside, the resistance I measure between terminal X4-4 and the 22uF capacitor is 17 ohms. It seems that one of the burnt resistors is still close to its nominal value of 16 ohms.

109A_5

Hello @Beaver33,

Well lets say the bridge rectifier had a short like defect on the AC side how does that explain the overheating of R5 and R6 that limits the current to the LED inside the Phototriac coupler? It would explain R7 and R8 burning up only.

Let's say the LED inside the Phototriac coupler failed shorted, its normal forward voltage drop is so low (probably less than 2.5 VDC) I would not expect the seriously overheated resistors.

I think at this point I would replace R7 and R8 with the 16 Ohm resistors and mount them up above of the circuit board and away from C3 and power it up on a workbench and see what it does. As previously mentioned I would go with 1/2 Watt resistors. As a test you could try 1/4 Watt resistors which is what I believe the originals were.

Considering your battery powered thermostat, it almost seems like the transformer's 120 VAC primary was connected to 240 VAC (and the transformer survived) or the building has a poor neutral issue, but I think you would have noticed those things by now.

So far to me the only thing that explains all the symptoms is excessive Voltage applied to the circuit.

How old is this unit ? Or how long has it been in service ?

Beaver33

Thank-you for your comments.

The unit was installed in 2011 and running without any problems or significant changes. Power to the unit was off during the summer. When the power was turned on it started normally without any error. I did however think I heard an unusual soft buzzing sound coming from the unit. I made a point of checking the unit a few hours later and noticed it off - there was no display on the led screen even though the power button was still depressed (turned on).

I will replace the burnt resistors. I can try to power it up slowly on the bench using a Variac.

According to the specification sheet for the phototriac coupler an average forward current for the LED is about 20ma. R5 and R6 in parallel have an equivalent resistance of about 4.75k/2 ohms. The voltage drop across the resistors would be V=20ma * 2.38k ohms= 48 V. Isn't this voltage larger than the voltage output from the bridge rectifier?

Thanks for your assistance.

109A_5

Hello @Beaver33,
It seems the current limiting resistors for the LED is going to run a bit warm. So that they are a bit dark for many years of service, this may be 'normal'. The resistors that burned up should run warm too but a bit cooler, but not much.

As for the failure, C3, C6 or the bridge rectifier could have excessive electrical leakage. Looking at the numbers more closely I would replace all four of the resistors actually in use in this circuit with 1/2 Watt devices.

Beaver33 said:

According to the specification sheet for the phototriac coupler an average forward current for the LED is about 20ma. R5 and R6 in parallel have an equivalent resistance of about 4.75k/2 ohms. The voltage drop across the resistors would be V=20ma * 2.38k ohms= 48 V. Isn't this voltage larger than the voltage output from the bridge rectifier?

In this case (48 Volts) R5 and R6 would burn up, at almost 1 Watt being dissipated between the two.

Also just for my fun... The transformer, resistor, capacitor part of circuit I actually built matched the data below closely.
https://www.redcrab-software.com/en/Calculator/Electrics/RC-Series-Circuit

Beaver33

Thank-you for all your assistance and identifying the calculator used to determine the current and voltage.

I replaced the burnt resistors with 10 ohm and 41 ohm resistors in parallel, both 1/2 watt, having an equivalent 8 ohm total resistance. (These resistors were the closest ones to the burnt 16 ohm resistors that I have on hand.) I shorted the X4-3 and 4 terminals (RT). Using a Variac, I brought the voltage up to 120VAC. I measured about 27VAC across the transformer secondary and 26 VAC across the 22uf capacitor. These figures closely match the values you calculated.

However, after the bridge I was expecting about 36VDC (Vpeak=27*sqrt(2) less voltage drop of 2 diodes=2*0.6) but measured 25VDC. As a comparison, I replaced the RT short with a short on X4-1 and 2 terminals (DHW). Once again I measured 25VDC after the bridge. I do not understand the reason for the difference between the calculated and measured values. I would appreciate your comments on this issue.

I also noted at the time of the unit's failure that the led screen was not lit. I opened the Burner Control Unit and found that the 2.5 amp slow blow fuse was open. The on/off switch is part of the Burner Control Unit. I have attached the schematic of the Burner Control Unit.



Line and neutral flow through directly from the Power Pump Module to the Burner Control Unit. Would the finding of the blown fuse add to your belief that a power surge caused the problem? I don't believe an independent failure of any component of the Power Pump Module would cause the Burner Control Unit fuse to blow. Please let me know if this belief is correct.

I have replaced the 2.5 amp fuse. What do you recommend as next steps?

As an aside, I noticed that a phase angle of 86 degrees yields a power factor of about .07 (=cos 86 degrees). Why is the power factor so low?

Thank-you for all your help to date. I look forward to your further assistance.

GGross

Is it working now? I probably can't be much help on the remaining trouble but I am really curious if replacing those resistors fixed the problem.

Beaver33

I have not yet put the board back into the unit so I'm not sure yet whether the boiler will work.

I will let you know if this effort is successful. In the meantime it's getting cold, so I need check soon.

If I do need a new board (US$476) (and perhaps a Burner Unit Control board (US$650) as well), then I'm wondering if it may be more prudent to buy an entire new boiler?

GGross

new boiler new warranty. Though you will have a lot of money wrapped up into a new install as well. Fingers crossed that you fixed it

109A_5

Hello @Beaver33,

Beaver33 said:

I replaced the burnt resistors with 10 ohm and 41 ohm resistors in parallel, both 1/2 watt, having an equivalent 8 ohm total resistance. (These resistors were the closest ones to the burnt 16 ohm resistors that I have on hand.)

That may work just the dissipation will not be equal between the two resistors.

However, after the bridge I was expecting about 36VDC (Vpeak=27*sqrt(2) less voltage drop of 2 diodes=2*0.6) but measured 25VDC. As a comparison, I replaced the RT short with a short on X4-1 and 2 terminals (DHW). Once again I measured 25VDC after the bridge. I do not understand the reason for the difference between the calculated and measured values. I would appreciate your comments on this issue.

True, but 1uF is not enough capacitance to filter out the ripple Voltage (Pulsating DC output from the Bridge Rectifier), even with just an LED and the current limiting resistors as the only load. So the ripple Voltage is excessive as compared to a circuit with a much larger filter Capacitor. And the meter may average or (RMS) the ripple component of the DC voltage. See scope view below. Ripple Voltage is between about 9 VDC and 36VDC. A benefit of the high ripple voltage is lower resistor dissipation.

Would the finding of the blown fuse add to your belief that a power surge caused the problem?

Yes, Sounds like something electrically violent happened at some point. Power utility antics, nearby Lightning strike, intermittently open neutral at the electrical service. I would think the thermostat had to be closed at the time for it to damage the connected (RT) circuit an not the DHW circuitry too. Or the surge jumped the switch in the thermostat.

What do you recommend as next steps?

Do a through visual inspection of all the electronics. Replace the obviously damaged components. Maybe check what you can with an Ohmmeter for hidden damage and power it up and see what it does.

As an aside, I noticed that a phase angle of 86 degrees yields a power factor of about .07 (=cos 86 degrees). Why is the power factor so low?

It is low but leading (Capacitive). Probably to compensate for other internal lagging (Inductive) loads. And the non-sinusoidal current waveform of the Phototriac coupler and other control circuits. ( probably why they kept C6 as small as possible, yet not have the LED shut off). Probably to to satisfy the European Power Factor specification.

Beaver33

Thanks for your explanation.

I shorted RT, attached the transformer primary to a Variac and measured the voltage across the triac. With RT closed the AC voltage measured was the same as that input. With RT open, the voltage never exceeded a few volts AC. So, it appears to me that the RT circuit is behaving normally. (I wanted to do the same test on the DHW circuit but couldn't understand why the output lead of the triac fed into the X1 connector but did not lead out to the Boiler Control Board (or anywhere else)?)

I'll power up the unit and let you know what happens. Thanks.

109A_5

Hello @Beaver33,

Not sure I exactly follow some of what your last post contained. However it appears the Triac output for the DHW input is not used in your version of that boiler, nothing connected to it at X1 (as you stated). Although depending on which diagram you look at it seems to change or maybe the diagram is not complete. Maybe there is different versions of the same basic control system, or an Engineer had an idea and then abandoned it. It appears the DHW control output from that PPM board for your boiler is the relay K2, not sure where those contacts go to. And there is a different K2 on the other board (Burner/ Boiler Control Unit).

Beaver33

I removed the Burner Control Unit (photo below).




No obviously damaged parts. Two round fuses test OK. Capacitors on the higher voltage side seem OK. Transformer was not shorted. Resistance measurements did not stabilize on one side of switch (when closed) prior to the fuse. After opening and cleaning the contacts it seems OK.

I placed the Burner Control Unit and Power Pump Module into the boiler, attached all the connectors and turned on the breaker. I lowered the thermostat set temperature making the RT contact open. When I pressed the On/Off button, nothing happened - nothing on the led screen and no sound from the unit. After removing the unit I found the 2.5A slow blow fuse had blown. (I had used the replacement fuse provided by the manufacturer.) The fuse on the Power Pump Module was OK.

Even with DHW and RT open, power would flow through to the Burner Control Unit. I suppose a short on the unit could blow the fuse, but so far I can't locate a short. The manufacturer does not provide a detailed schematic of the board.

I would like to check the secondary voltages of the transformer on the Burner Control Unit on the bench, but don't have any 2.5A slow blow fuses.

Once again, I would appreciate any comments you may have on the best way to proceed from here.

Beaver33

After inserting a new 2.5A slow blow fuse into the Burner Control Unit on the bench top I applied VAC to L & N on the unit. Output voltages (18VAC & 150VAC) from the transformer secondary appear to be OK. There is no excessive current.

Inside the boiler I did not find shorts at the Gas Valve terminals (that enter via connector X16), the Burner Fan (X18) or the Pump (X19).

So, I'm not sure what would cause the 2.5A fuse (F4) to blow when the printed circuit boards are installed in the boiler and the on/off button turned on?

109A_5

Hello @Beaver33,

The MOV may have clamped the surge and blown the fuse... Do you think the fuse just blew or was blown at the time of the surge ? If the fuse continues to blow but only when the board is installed into the boiler, possibly a device not on the board is shorted to ground but otherwise OK. If the fuse continues to blow put an incandescent lamp like 100 Watt in series with the Hot wire to limit the current so you don't keep blowing fuses during troubleshooting. A 100 Watt lamp quickly limits the current to about 0.83 Amps. Normally the incandescent lamp will get bright when the power supply caps charge and then dim down. It there is a short it will remain bright. Removing the short the lamp would dim. With a normal system, you may not be able to run all the loads with the lamp connected in series, like pumps, etc.

If the short is on the board, I circled in red some of the points I would be interested in. I don't see an Fuse 'F2'. MOV, safety caps, bridge rectifiers, voltage regulators. If after all the fuses are known good and don't blow and it still appears dead, the unregulated and regulated Voltages would be next to verify. Once powered up correctly do any of the chips get hot ?

There should be 5 or 3.3 VDC to power the logic stuff. 24 VDC regulated (or maybe unregulated) for the relay coil power. The high voltage secondary from the transformer may be for the spark coil circuit.

Beaver33

Thank-you.

Results so far. I am using the 100W bulb in series after the Variac. I am not sure when the 2.5A fuse blew.

On the bench with 120VAC input the board uses about 5 watts. Seems OK since no connectors are attached.

The 2 circled fuses, the 2.5A fuse, 4 circled diodes and the switch measure good.

The MOV next to the fuse has 70 ohms across the legs. It is not shorted so may be OK.

At 120VAC line in, the 7805 shows 25V input and 5V output. This appears correct.

C73 (smaller blue capacitor) measures 77nF (Markings on top X7- HIM305V ?)
C82 (larger blue capacitor) measures 3.9uF. (Markings on top u33K100 X7 ?) I will check further to better identify the markings.

Rectifier DI1510 shows no AC input at any input voltage. I will try to follow the tracks to see if it is switched on from a relay.

I will check the transistor tomorrow.

Thanks for your help.

109A_5

Hello @Beaver33,
Hopefully with the MOV measurement you were seeing something else in the circuit. I don't think a MOV should read 70 Ohms with the low voltage of an Ohmmeter.
C73, sometimes the marking or numbers are hard to deal with. C82 seems reasonable.
Rectifier DI1510 is a 1000Volt unit and may be associated with the high voltage secondary of the power transformer and may only be energized when spark is needed.
can you connect the user interface only to this board and see if there is any sign of life ?

Beaver33

Today's results:

MOV is connected across primary transformer leads. Seems to explain the low (70 ohms) resistance across the leads.

C73 marking probably reads as X7-H1M305V. May mean 1U at 305V. Couldn't get in-circuit reading - perhaps because it is across the transformer secondary? Resistance is 169 ohms.

DI1510 does indeed provide DC output to the gas valve. AC input is turned on via a relay.

The component you circled in the bottom left of the control board is a MCR 22-6 Silicon Controlled Rectifier (SCR). I wasn't able to fully confirm its status in-circuit, but what I could check was OK.

When I applied 120VAC to the control board on the bench the LED screen did not display anything. The screen is not connected with ribbon cable but rather makes a direct contact with pads on the board. I confirmed that a good contact between the two was attained.

At this point I decided to try placing both boards in the boiler, but only attach the lower current connectors X21 and X20 to the control board. On the Power Pump Module I disconnected the RT leads and pump power leads. Finally, after depressing the on/off button the LED screen came to life showing an F1 error.

I continued sequentially adding connectors (everything except the burner fan and RT connectors). Now the display shows an FC error indicating faulty electrical fan control. Before the error message appeared the display showed the water temperature and the circulating pump turned on for several seconds then off.

I have not be using the current limiting lamp during these tests.

What do you suggest I try next? So far, it appears both boards are functioning, but the RT circuit on the power board and the fan, igniter and combination gas valve connected to the control board have not been tested.

Could a fan malfunction draw excessive current, enough to blow the control board 2.5A fuse and the power board R7/R8 resistors? If so, is it worthwhile to remove and open the fan to check for any issues? To remove the fan I would need to remove the burner unit (not a problem if you think I should check the fan). Or should I just continue to add the RT connector and then the burner fan? I have one back-up 2.5A fuse if the one in use fails during testing.

As always, looking forward to your comments. Thanks for your help.

109A_5

Hello @Beaver33,
I would expect running it with things disconnected may make alarms. Which is good, since the logic stuff appears to work. So far I don't recall you saying that F1 (6.3 Amps) on the Power Pump Module has ever blown. That seems like a main fuse. I would parallel 2 or 3 100 Watt lamps (1.66 Amps or 2.49 Amps) and substitute the lamps for the 2.5 Amp fuse and give it a normal try, everything hooked up. You may want to jumper RT manually so you can monitor the lamps.

Sure a defective fan could draw excessive current, enough to blow the control board 2.5A fuse. I don't think it would bother resistors R7 & R8. I would only go after the fan if I had a high degree of confidence that it was the issue.

You may be just mopping up from what a power surge blew up (the 2.5A fuse and R7, R8). I suppose it is possible R7 & R8 acted like a fuse during a power surge event.

Beaver33

The saga continues.

You are correct, the 6.3A fuse on the power board has never blown.

I set up 2-100 watt bulbs in series and attached the leads to the 2.4A fuse on the control board. The bulbs burned brightly as soon as I depressed the on/off switch with the fan connector attached. The fan made no noise and did not spin. I am assuming now the fan is the root of the boiler's problem. The fan is an EBMPAPST NRG118/0800-3612-020210 120V 60Hz model (55667 30470). It does not appear to be readily for sale based on a quick internet search.

I removed the burner unit to access the fan. After removing the fan cover and the fan from the burner unit I took photos of the top & bottom.



Visually the board appears fine with no obvious burn marks or broken components. The AC power input has surge suppression, smoothing capacitors and choke, rectified using a bridge, smoothed DC by the large electrolytic capacitor and then input into an H-Bridge (I believe) comprising 4 MOSFETs. The PWM signal from the control board connects to the MOSFETs to control the speed of the fan. The fan has a Hall Effect sensor that outputs the speed to the control board. Some quick checking indicates the electrolytic capacitors, bridge, large copper coil forming a circle in the center and fuse are OK. (The board is hard to test with its conformal coating.)

My concern is with the 4 power MOSFETs. The identification number appears to be IRF064 Y9H AC. Some tested using diode mode show shorts from drain to source and source to drain. Could a short on 1 or more of these components cause the 2.5A control board fuse to blow. (There is a 2A microfuse on the fan board.)

Should I replace the 4 MOSFETs? If so, what more common ones would be suitable to use? Is there a way to test the fan after replacing the MOSFETs to ensure it is in working condition prior to reinstalling it in the boiler? I would appreciate any other suggestions you may have.

Thanks for your help.

109A_5

Hello @Beaver33,

Beaver33 said:

I set up 2-100 watt bulbs in series and attached the leads to the 2.4A fuse on the control board. The bulbs burned brightly as soon as I depressed the on/off switch with the fan connector attached.

Is this a typo ? "I set up 2-100 watt bulbs in series"

I hope you put the Lamps in 'Parallel' with each other so the current capability of each lamp add. Otherwise the test may have been meaningless and may lead you in the wrong direction. Putting the lamps in series would basically give you a 240 VAC, 0.83 Amp incandescent current limiter.

MOSFETs often (or always have a diode junction Source to Drain). If with the Ohmmeter function they read shorted or the same very low resistance with the Ohmmeter leads either way and you are not seeing a motor coil or the like (metering in circuit) I would replace them with the same part number. If the fan motor was bad (and/or its control circuitry) I would expect that small 2A fuse to blow first. Assuming that the fuse is for the whole motor assembly. Looks like the 2A fuse is right after the line comes onto the board.

You would have to have an appropriate function generator and know how to connect it up correctly to generate the appropriate PWM signal to run the motor independent of the boiler.



Not sure about an H bridge configuration. Yes 4 transistors, and there appears to be 4 diodes too. Not sure if the coil orientation supports a motor driven with only one H bridge. Maybe something like a Four-phase permanent magnet synchronous motor. With the PWM drive signal maybe each pulse is the next phase, so pulse width and frequency easily controls the motor.

Beaver33

I removed two IRF634 N-Channel 250V, 8A MOSFETs from the fan - both tested short. I don't have exact duplicates. I do have K3569 N-Channel 650V, 12A that I used to replace the 2 faulty ones. Do you think they could function correctly in the circuit? Could I test the fan with this set-up with the idea to replace all 4 if the fan is functional. Or must I replace all 4 to test the fan?

The circuit includes a pair of IRS2106 Half Bridge Drivers that feed into the gates of the MOSFETs.

Should the fan turn on if 120VAC is applied to the line terminals, or are the PWM signals required for operation? I have attached some specifications for the 240ACV version of the fan. Under normal operation the fan uses 70 watts or about 1/2 A.

I have a signal generator that I may be able to get operational if needed to test the fan on the bench.

Thanks for your assistance.

109A_5

Hello @Beaver33,

Do you have a scope? If so can you maybe monitor the PWM signal from the control board if the boiler startup process gets that far. Be careful of the grounds (circuit common may not be the same as the scope's ground) when using the scope, don't want to make things worse.

If you have plenty of the 2SK3569 I would try them. I think I would replace all four. You don't know how beat up the other two IRF634 are and it may keep the circuit more balanced. The 2SK3569 may run a bit warmer than the IRF634 (on resistance is a bit higher). Transistor dissipation may not be an issue anyways since there was no heatsink. The gate capacitance is higher but since the motor RPM is not too fast they may be fine.

The IRF634PBF (Lead free) is still available. The 2SK3569 may provide better over-Voltage tolerance to future surges or transient Voltages.

Do you think the IRS2106 is still good ? Maybe the IRF634 gate resistor (?) helped save it.

Other than S1, I don't recall seeing the AC power to the fan being switched anywhere else.

Depending on what all is on that motor board it may not turn without a PWM signal. Many micro-controllers have a built in hardware PWM generator so it is easy to generate the PWM on the boilers control board. The motor board may just follow the PWM signal to it. What drives the IRS2106 ?

So i'm thinking maybe the

kept the 2 Amp fuse on the motor board from blowing and maybe saved the IRS2106 too.

Beaver33

I found a document from the manufacturer that describes the 3-lead communication connector input to the fan. Under PWM it states that "not connected = blower off". I am hopeful that the PWM is required for operation because when I fed AC into the power connector, nothing happened and no current was drawn.

The two IRS2106 appear good. They both have similar resistance & diode readings. Resistors and other small nearby components also appear OK.

The 4 copper coils are all connected and terminate in 2 leads near the Hall sensor.

It appears that the PWM signals go through the TLP759F and the Hall signals through the SFH610A optoisolators. The board also contains a LINK304GN oscillator. Not sure what it does.

I can't figure out how to input the PWM signal, so tomorrow I'll place the fan, burner and boards back in the boiler and see what happens. For now I have only replaced 2 MOSFETs. If the boiler works, then I may replace the other 2 MOSFETs and add a small heat sink if there is enough space.

Thanks for your advice.

109A_5

Hello @Beaver33,

Beaver33 said:

I found a document from the manufacturer that describes the 3-lead communication connector input to the fan. Under PWM it states that "not connected = blower off". I am hopeful that the PWM is required for operation because when I fed AC into the power connector, nothing happened and no current was drawn.

Good information if it gives good detail of the PWM signal. The only current I would expect is the large electrolytic cap charging, then very minimal draw.

It appears that the PWM signals go through the TLP759F and the Hall signals through the SFH610A optoisolators. The board also contains a LINK304GN oscillator. Not sure what it does.

Optoisolators are good practice and may be necessary so there are no 'Ground Loops' between the two DC power supplies (motor and boiler control). I believe the LNK304GN is a switch mode power supply for DC power to the rest of the circuitry in the motor assembly. BTW it looks like there is a fuse (green with a 3 on it) connected to pin 5 of the LNK304GN, 300mA I would guess, 3 Amp makes no sense.

I can't figure out how to input the PWM signal,

Your signal generator would just drive the LED side of the Opto, through a current limiting resistor.
Any heatsinks probably need to be individual on each transistor, unless insulation is provided.

Beaver33

I reinstalled the fan and burner, keeping the lamps across the control board fuse. (I erred before in stating the lights were in series, they were in parallel.) The boiler started it initialization and the fan turned on for a few seconds (but not as long as it would normally). The display showed a fan error.

(The fan turns smoothly and the bearings seem fine.)

I thought the problem may relate to mismatched MOSFETs, so I replaced the remaining 2 originals with the same type I previously used to replace the shorted ones. I then applied AC to the fan and briefly applied 5VDC through a 1K resistor to the PWM (and ground) contact. The fan started up momentarily and then stopped. I found the micro 2A round fuse on the fan blown. If I can find a replacement I will reinstall the burner unit in the boiler to see if the fan's behaviour changes compared with me applying 5V on the bench.

I'm not sure what the problem is now. I can't detect any bad component on the board, but I haven't fully checked several of the various ICs.

Perhaps it's time to buy a new fan? (May not be easy to find.)

109A_5

Hello @Beaver33,
If I had to guess, the gate capacitance of the 2SK3569 being higher than the IRF634 may diminish the switching dead band time of the H bridge switching enough that one transistor is not completely shut off before the other is turned on, causing a brief short circuit, thus blowing the fuse. As the motor speeds up that would get worse. The IRF634PBF (Lead free) is still available, Digikey.com and Mouser.com. Not too expensive, you probably can get a replacement fuse or two there too.

https://www.digikey.com/en/products/detail/vishay-siliconix/IRF634PBF/811830
https://www.mouser.com/ProductDetail/Vishay-Semiconductors/IRF634PBF?qs=cvaI6ThkwxtQmbf/PykvWg==

One example of many (MRT, MSF, MST, etc.) of that basic type of fuse. Not sure if yours is fast or slow blow. You may have to check the marking on the fuse and compare to some of the data sheets to figure it out.
https://www.digikey.com/en/products/detail/schurter-inc/0034-7118/2644586