Metcal MX-500P cold

[jclampett]

Hello! This is my first post on the forum, so apologies in advance if I buck any conventions. Last week I bought a used and ostensibly working MX-500P from eBay, along with a brand-new cartridge and -AV hand-piece from Amazon. I plugged it in and it heated up, but I almost immediately realized I forgot tweezers so I turned it off, set it down, and about 40 s or so I came back and found the tip glowing red so I unplugged it. A minute or so later, I plugged it back in, heard a pop and then something rattling around the case.

Here's what I've found so far:
- It still had an OKI International warranty sticker, so it hasn't been opened before.
- My unit is a MX-500P (not MX-500P-11 or MX-500P-21). As far as I can tell, this just means that C53 is soldered on the back of the PCB, just like C7 on the schematic.
- The top and bottom of C6 (50V 2200uf) were expanded (the rubber on the latter was ruptured) and a lead had come free of its solder joint. I imagine the expansion was pre-existing and the shipping pulled the lead free.
- U4 (LM2576HVT-ADJ voltage regulator) had exploded - part of the front of its package was what I heard rattling around in the case.
- Nothing else showed any signs of damage.



I replaced C6 and U4 and now the unit is stable and the LEDs are behaved but the iron does not heat up. Following the repair notes, I checked the FB pin on U4 and it's about 0.3 V when it should be 1.3 V. On a whim, I checked the two generator MOSFETs and, if I'm testing it correctly, it seems the gate and source Q4 (IRF530 N-channel MOSFET) are shorted (I removed it first, since I believe the gate should always appear shorted to ground due to L10 right next to it).

In typing this, I also just realized that I made a mistake in reconnecting C53, which should have been between pins 2 and 4 (OUT and FB), to pins 2 and 5 (OUT and OFF).

So, I have a few questions for how I proceed:
1. The original International Rectifier IRF530 (https://www.infineon.com/dgdl/irf530npbf.pdf?fileId=5546d462533600a4015355e386b1199a) is discontinued. Is Vishay's https://www.mouser.com/datasheet/2/427/irf530-1768291.pdf a valid substitute?
2. I presume Q4 was damaged by U4. Is there anything else I should look at when I order a replacement for Q4?
3. Does it seem like my mistake with C53 should cause any significant damage? Nothing seems wrong that I can see - perhaps I can just rewire it and hope for the best?

[artag]

That regulator circuit is pretty wacky.

[quadtech]

I have the following repair notes gathered from various forums - (Pl note - I have not needed to use this info myself) -

Q4 - IRF640N has been suggested as an alternative

If Q4 failed, check and replace C27 (130 pf 500V silver mica) if needed

Check and replace if needed -  100nf C0 (C5) and C24  (one option is Kemet PHE850ED6100MD18R06L2 - 0.1 µF Film Capacitor 300V AC),
as well as 330nf C21 and C25

Likewise, check and replace the 100 ohm resistors (R27, R28 , 1W) which run hot

These are the potential failure points and some suggested options for replacement (make sure
to check transistor pinout for any substitutes) -

R2 - 180R 2W - 1
R27/28 - 100R 1W - 2
C0/24 - 100nF,63V - 5
C21/22/25/4/1/7 - 330nf,63V - 6
D2 - 6.8V 1W Zener (1N4736A)
Q1 - TIP32C (or BD140) - 1
Q2 - BD139 - 1
Q3 - IRF510N - 1
Q4 - IRF640N - 1

[jclampett]

Thanks! I think the main problem involves the regulator feedback circuit. The voltage on the FB pin is 0.3 V which causes the regulator to throttle itself and I'm trying to figure out what to test to determine why it's too low. Is this a valid reconstruction of the feedback-determining circuit aside from D8 and D9, which I believe can only raise the voltage?

[artag]

No. You've left out the zener and the ground at the bottom of RV2 (not RV1). These are critical to the function.

You have a DC supply of about 50V, a dropper resistor and a zener. Because of the zener, these will give close to 5v6 at the pot. If RV2 is set to 90% (if I understand your notes correctly) and is followed by a 13k/3k3 divider, you will have about 1V at FB. The 2576 expects about 1.23V there when in regulation so it will switch the output in such a way that it will be pumped higher (not sure why you think it will be throttled when the FB is lower than expected).

The FB is rather odd, as I noted above : usually it's just a simple fraction of the regulator's output it's driven from a fixed bias (as just explained) + (as you say) a correction from D8/D9. I think what they're after is regulating the signal out of Q4 but what's derived is way less than 1.23V so they have to add an adjustable bias. After changing U4 this might require adjustment but  you can't get 0.3 up to 1.23 with that : there must be some other reason why it's so low.

Check you actually have 5v6 at the top of RV2, and about 5V at the RV2 end of R6. If they're lower, find out why.
Check with a scope that you've got about 50V without much ripple at U4 input. A DVM may mislead you.

Q1 and Q2 also do something odd. I don't know what that's about.

[jclampett]

Ah, thanks. The missing ground was just a mistake when I undid once too many times before taking the screenshot. I didn't realize the Zener was critical, but I get how they work now - thanks! With that said, voltage at the diode is about 5.4 so that seems basically correct. I also simulated it based on a fairly precise calculation of the potentiometer, and it seems to suggest that my reading of 0.3 V is exactly correct. I'll go ahead and order an IRF640 for Q4. That should restart the dynamic feedback at D8 & D9 - maybe there are no more issues, actually. I'm just worried that the (current spike ?) that shorted the gate and source of Q4 may have damaged something else as well...

[quadtech]

Q1 and Q2 also do something odd. I don't know what that's about.

Trying to reason out Q1 / Q2 - disclaimer - my electronics "expertise" is just at hobbyist level - so my reasoning below may be incorrect
 
See this post for the voltage on the center tap with respect to ground -

https://www.eevblog.com/forum/repair/metcal-mx500/msg599673/#msg599673

The pdf with the schematic says you should see 26V on C2 at power up.

The LM2576 may be turned off by multiple conditions - autosleep state (Q11), temp sensor is triggered (TS1/Q9), power switch is off (Q6/Q7), very high power at output (T3/Q10).

Q2 is turned on only when the LM2576 is on (not in off state for any of those reasons) and via Q1 , enables power to
the RF driver transistor Q3.
D1/L1/C3 look like a buck converter - perhaps the switching freq of the LM2576 of 52Khz
drives Q1 ,  Q2/D1/L1/C3 to act as a buck converter from 26V at C2 to about 15V at C3?

[jclampett]

Ah, that makes perfect sense. I get about 25V at C3, but the datasheet for U4 notes that "[w]hen the voltage on the FB pin (ADJ) falls below about 0.58 V the switching frequency is
dropped to about 11 kHz." so perhaps that explains it? Otherwise, perhaps one or both of those transistors have been damaged.