Unmarked component

[Mark56]

Hi, I am in need of some expert guidance to help identify a replacement device with a TO-220 encapsulation. The device has a heatsink fitted about 1" x 1 1/2". This device is in circuit with a 24V DC motor, there is no wattage on the motor and I have been unable to measure the current demand, it is heavily geared down to about 120 rpm so there is little stress on it. The motor is only about 4" x 2". There are 3 motors on the chair that perform massage and two rams that move the back and foot rest, it has a 1.6A slow blow fuse on the 240V side of the transformer, the supply goes through a full wave bridge rectifier on the circuit board.

With my limited electronics knowledge I have confirmed the device has gone short circuit but my problem stems from the fact that all the numbers on it have been ground off. The only mark I can see is 3F moulded into the encapsulation just above the centre pin.

The device has three pins, looking at the face of it with the metal plate at the back pin 1 is on the left. Pin 1 goes back through another circuit board presumably to the remote control. Pin 2 is connected to the motor and pin 3 has a reading of -35V. There is an intensity control on the remote but on this model it has no effect, the motor always runs at a constant speed. There is a 1/4 Watt 4.7K resistor across pins 1 and 3.

I have guessed this to be a fixed voltage regulator, possibly an LM7924 but may be an LM7824.

I would most appreciate some help in confirming what this device is likely to be and where I can order just one for postage to Spain.

I do have another identical fully operational massage chair that I can take further readings from if need be, or I can use the faulty board (failed device removed) to take open circuit readings.

Thank you in advance for any help that can be offered.

 

[shrtrnd]

If possible, you want to see if you have the markings on the TO-220 Pkg in the working chair.
If not, you can use a volt meter, to read the pins of the device that is NOT damaged,
(Download the data sheet on-line, if you're pretty sure you've got a 7824 (positive 24V regulator), or 7924 (negative 24V regulator)).
Since you believe you've got a damaged voltage regulator, read the voltage on the good
chair device, determine if that's what it is, and what the value is (+/-24VDC).
That's where I'd start, if you can't get an actual schematic from the manufacturer,on
your chair.
Good luck. If you run into a problem, somebody here can probably give you advice.

 

[Mark56]

Thank you for such a fast response.

Unfortunately there are no markings on any of the voltage regulators on the faulty or the working massage chair, not even a makers name plate anywhere to be seen. Lord knows why all the numbers were removed, it is the lack of identification that has caused me such a headache, compounded with only very basic electronics knowledge.

On the working chair I get a reading of -24V on the connection that goes to the motor, pin 2, on the spec sheet for the pinouts of an LM7924 it show this as the input and pin 3 as the output which reads -35V, this is where I get confused and need some guidance.

With my lack of knowledge I am confused with the -35V, which comes directly from the supply, being attached to pin 3 which is the output on the LM7924. My logiic tells me that should be the input, but is it reversed with a negative device? Or may be I am looking at the wrong device and should be after a different regulator with different pin outs.

Should the pin connected to the negative side of the motor be the input or the output?

 

[davenn]

some equip manufacturers scrub off the component ID marks to stop others copying their design. Its a total pain in the butt..

maybe time for a couple of 640x480 max and sharp/clear photos of the board and device... BOTh sides of the board.

Dave

 

[Mark56]

OK, here are the photos. I have simply rolled the board over from top to bottom to get the underside view.

On the top view the central two pin connector, between the diodes, is the AC input and the next connector to the right (M2) is the one that goes to the motor, I connected the AC supply to this by mistake which is the reason the voltage regulator went short circuit and the track on the board burned through. Everything else is still working. The relay to the right has already been replaced which was the original fault. I am still kicking myself for making such a daft mistake.

Anything you would like me to take readings from then just tell me where to put the red and black leads of my AVO and I'll post them back.

It looks like the motor can be reversed through the relay but it only operates in one direction.

If it helps, the band on the four diodes nearest to the bottom of the board have the band uppermost, the next two are at the bottom and the upper two are at the top.

 

[shrtrnd]

The -35V you read is the UNREGULATED voltage that comes from your supply.
It is the reason why you HAVE the voltage regulator to give you the voltage you need.
(-24VDC).
The four diodes are there in a 'bridge' configuration. They take the AC voltage from
your power supply, and convert it to unregulated DC voltage (which is where your
voltage regulator comes in).

 

[Mark56]

Thanks for that information Shrtrnd but I had already worked that part out, what I need help with is selecting the correct voltage regulator to replace the unmarked one.

As I said earlier in the thread I had guessed at an LM 7924. What is confusing me is the pin configuration, this is where I need help in confirming that this device is correct. And if wrong, what device I need to use.

EDIT
Just to explain further. If the LM7924 is correct and I used that device in this board the input would be connected to the motor and the output back to the -35V DC supply. Due to my lack of electronics knowledge I am thinking that the output should connect to the motor, or would that just be the case with a positive regulator.

 

[davenn]

The more I look at the cct The more I suspect that it isnt a 3 term V-reg. There are resistors between what would be the GND and Output pins if it was a reg.
I suspect its more likely a power FET.
Also, a LM7924 in that case is only going to handle 1A max without cooking even with that heatsink. I dont really think it could handle driving any sort of heavy motor.
Whereas there are power FETs that handle many 10's of amps.

It would be so easy for me to test the device and find out what is was .... i have one of the nice small transistor testers that are available on the market, maybe you need to buy one and confirm the type of device. Else we are all just stabbing in the dark.

http://www.peakelec.co.uk/acatalog/jz_dca55.html

these units are just brilliant, couldnt live without it in the workshop

for a modest cost outlay, its going to answer you Q's virtually immediately and you have it for future testing on other jobs

cheers
Dave

 

[(*steve*)]

Mark56, your reasoning makes sense (in the edit). And it points to the device not being a 3 terminal regulator.

 

[Mark56]

OK, thanks guys, I have a digital AVO with a transistor tester. Obviously testing the burned out device will tell us nothing but I can remove the device on the other end of the board to test. Do you think that would be an identical device or should I take the one out of the other massage chair?

On my AVO there is a row of four holes marked E,B,C,E with hFE written below it, I am guessing this is what I need. On the selector of the AVO there are two sections which I believe relate to this, one has hFE with positions for NPN and PNP, the other has Hz(Trig) with Hi and Lo positions.

If you can tell me what the pin configuration is going to be so I know which way round to put it in the E,B,C,E holes and what setting to put the AVO on I will get the results posted.

I doubt the current running through this device is much more than 2 to 3 Amps as the motor is quite small and heavily geared down to about 120RPM and the relay is only rated at 5A. The 240V transformer supplying the board has a slow blow 1.6A fuse so from that I'd say the maximum current demand is about 11A split between all the motors.

Thank you for trying to get me through this.

I hope I can do this with my transistor tester, if I buy this other tester which looks like a very usefull device, I think it unlikely I will ever need it again. I work on computers and household wiring so my AVO is all I need.

 

[Mark56]

Davenn, if I splash out and buy the analyser and identify what the component is how easy is it going to be from there to calculate exactly what component number I would have to use.

 

[Resqueline]

I'm willing to bet there are 6 MOSFET's and two voltage regulators on that board.
On the picture; G = Gate, D = Drain, S = Source.
Patch the burned track with some wire.

Measure on any of the other transistors (in circuit):
There should be no continuity (other than the 4.7k resistor - remove this if you wish(*)) between Gate & Source, independent of polarity.
With red probe on Source & black probe on Drain there should be a reading of some 700 - 900 mV (with the DMM in the diode test range). Open circuit the other way around.
If you can't confirm this expectation; remove one of the other transistors and take diode measurements every which way on it (6 measurements in all).

With power on the board (no need to attach motors), measure the Gate to Source voltage on the removed transistor pads - with that channel activated on the remote.
Confirm the G-S voltage, measured on one of the other channels (with intact transistor in place).

Measure the output voltages of the two regulators. What voltage are the relay coils made for?

(*): With the red probe on Gate & the black probe on Source, the diode test voltage may be enough to make a MOSFET conduct.
Subsequent Drain-Source tests may show as a short until the Gate voltage is shorted to Source, or reversed.

 

[Mark56]

Thank you for offering additional assistance, got your post just before I was going to order the semiconductor analyser.

I repaired the burnt out track before taking any of these measurements.

The two voltage regulators you labelled on the photo are correct in part. These are the only two transistors on the board that do have the numbers on. The one you labelled as 5V is correct LM7805. The other you labelled as 12V is actually 24V L7824CV.

The relay coils are rated at 24V with a 5A/30VDC rating on the contacts.

With red probe on Source and black on Drain all transistors measure around 510mV (in circuit). The only exception is the transistor, top left of board in your picture and second one down. That is open circuit across Source and Drain in both directions but with red on Gate and black on Drain it shows 574mV, with red to Gate black to Source 599mV,open circuit with all other pin combinations.

If this helps, for all the other transistors in circuit red to Gate black to Drain, red to Gate black to Source & red to Source black to Gate all read 1.8V to 1.95V.

I removed the other transistor with the identical heatsink to the faulty one and confirmed open circuit on all pin combinations except for red to Source and black to Drain giving 530mV.

I then connected the board to the supply and activated the remote for the faulty component, the voltage across the Gate and Source is 7.9V. This supplies a 24v motor confirmed on the motors label.

The reading on the other transistor with the matching heatsink shows an output to its associated motor of 35V with 6V across Gate and Source. Up in the top left hand corner of the board (component side) that device reads 12V across Gate and Source and has 37V at the output to its associated motor.

I will be happy to supply any other readings you require.

 

[Resqueline]

Ok, those readings confirms they're MOSFETs, and not neccessarily logic-level controlled. Strange that there are different gate voltages btw..
Now the only question is how high a voltage rating they need. I'd guess that 200V would be enough to keep them relatively safe.
So in essence look for 200V transistors with the highest Drain current available (= lowest Rds on) at a reasonable cost. At least 5A of course.
Higher voltage ratings leads to higher on-state resistance (Rds on) making them run hotter. It's not critical here, you just want to keep it at a reasonable level.

@ Dave: would the analyser be able to tell the breakdown voltage, &/or Rds-on?

 

[Mark56]

Resqueline, thanks again for helping me. As my knowledge of electronic components is very limited and non existent when it comes to MOSFETS would you be able to suggest what ID number I should be looking for?

 

[Resqueline]

There are a multitude of part numbers that will fit the bill. Newark for example comes up with 18 suitable transistors. Find a suitable supplier and use their search engine.
Find first: single MOSFET's
Apply filters: Polarity = N-channel, Drain current = 5-80A, Drain-Source voltage = 200-300V, On resistance = 0.01-0.06 ohm, Case style = TO-220 & TO-220AB

 

[Mark56]

Thanks again.

I am a bit confused with the selection process as there are several parameters that I don't have any knowledge of, On resistance RDS, RDS voltage, Threshold voltage, to name but a few. I selected this component on another site using a similar process, would this do the job. I just want to make sure before ordering.

http://uk.farnell.com/stmicroelectronics/stp50nf25/mosfet-n-ch-250v-45a-to-220/dp/1752160

Just to save me the time, and if it's not too much trouble, could you quote me a few part numbers, then I can confidently hunt around for a supplier that will ship to Spain.

Would any of these do, I'd rather avoid the ones in packs of five as they will be wasted.

http://uk.rs-online.com/web/search/...610&binCount=2540&multiselectParam=4294612489 4294611819 4294610907 4294604845 4294607873&selectAttribute=0.041Ω#breadCrumb

 

[Resqueline]

I think they'll all work, but the first one is best suited, due to its lower Treshold Voltage. Of the others I believe the Ixys is the least suited, for the "same" reason.
Look at the Output - and Transfer Characteristics curves in the datasheets to get a feeling for the Treshold Voltage and the On Resistance.
A couple of other good choices:
STP80N20M5
STP75NF20
Shipping usually costs, so why not buy a couple of spares.

 

[Mark56]

OK, I'll pick one of them and go for it. I'll report back with the results. Thank you very much for all your help, I could never have figured this out without your assistance.

 

[Mark56]

I was expecting to be able to come back here and give you all a big thank you but sadly, after waiting four weeks to get the component delivered via the UK to Spain, it all went wrong.

As soon as I put the power on, the Mosfet burned out and took the small diode (an IN4004 just infront of the relay to the left of the missing Mosfet in the photo) with it, which now has a crack down the middle. I had nothing else connected apart from the power supply and the link to the other circuit board that connects to the remote control.

Before you ask, yes I checked the PCB after soldering the component back in and there are definately no short circuits. I used a very strong eye glass just to be sure.

So, what went wrong guys. If you can make any more suggestions then please fire away. I must emphasize that I am extremely gratefull for the help you have already offered even though the outcome was not to good. I understand it must be very difficult to give advise at a distance.

The Mosfet I put in the Board was from Radio Spares, FDP61N20 N-CH 200V 61A T220.