Broken Metcal RFG-30 soldering base unit fixed

[memed]

Hello
I've got Metcal MX-5200 not heating up the tip at all.
Main MOSFET Q14 IXFH12N50F is getting very hot only when iron connected.
Checked all diodes and caps around.
U4 LM2576HVT gives on output 45V all time when iron is connected.
U3 LM2574HVM gives on output 5V all time when iron is connected.
Vcc 5.69V on 74hc04

Any ideas what can be wrong?
Regards

[fireworks]

  Can you post a picture of the board ?
  Do you have a schematic ?

[memed]

I don't have schematic for MX-5200 or even MX-5000.
I found online only schematic for MX-500p-11 but looks to different for me
I will post pictures of the board front and back on evening.
LM2576HVT gives Uout-45V with feedback-0.8V is it correct?

[fireworks]

  I'll help you repair it.
  Do you have an oscilloscope ?
1. First please post pictures of the front and back of the board.
2. If you have an oscilloscope, please post the waveforms at the 3 terminals of the output MOSFET right after powering up the station - the handle needs to be connected to the station and a known good tip must be in the handle. 

[memed]

Thanks, really appreciate Fireworks.
Unfortunatelly I don't have an oscilloscope but still hoping can manage some how
When I put a probe on output MOSFET then DMM geting crazy and switching off.
All my tips and handle are 100% working.

74HC04N
PIN 8- IN - 2.09V
PIN 9- OUT - 3.80V
PIN 14 - 5.69V

U4 - LM2576HVT
IN  -  48V
OUT - 45V

Q13 IRF510
DRAIN - 26V

Q14 IXFH12N50F
DRAIN - UNREADABLE BY DMM

Board pictures

[fireworks]

  Without oscilloscope is more difficult to fix it, but still doable.

1. The board is not too complex. Can you reverse engineer it and draw the schematics ? Shouldn't take more than a few hours.

2. Without oscilloscope, you need to build a RF voltmeter. It's very simple, 3-4 diodes 1N4148 in series, a 1M resistor and a ceramic capacitor of tens of nF.

3. First step: unsolder the output MOSFET and check it with a multimeter.
  Also, measure the voltages at its 3 pins while unsoldered.
  If your multimeter has a frequency meter, measure the frequency (and duty cycle if available) at the gate of the unsoldered MOSFET.

[memed]

I'm looking to buy some good oscilloscope as it will help me to fix more devices and get experience
Thanks to you now and will be back later to get it sorted

[lanuser]

Hello Fireworks,
Can I ask what the voltage output of U4 on pin2?
Thank you.

[fireworks]

  Hi,

  I don't have the voltage. Are you trying to do a repair?

[lanuser]

Yes, I am trying to repair my Metcal station.

[fireworks]

  I can help you with your repair but I need more details.
1. Which Metcal station do you have ?
2. What are the symptoms ?
3. What did you measure so far ?
4. What equipment do you have ? An oscilloscope will be very useful but you can do without.

[lanuser]

1. Which Metcal station do you have ? ( Metcal Model# mx-500p-11 )
2. What are the symptoms ? ( LM2576 U4  burned out )
3. What did you measure so far ? ( U4 showing sign of burned marks )
4. What equipment do you have ? An oscilloscope will be very useful but you can do without. ( Oscilloscopes, reworks stations in one word everything to start repairing the Metcal )

Until I receive U4 replacement part  I was trying to use external power supply replicating U4 voltage output, however, nobody knows correct voltage output as I was aking you before.
Thank you.

[fireworks]

  Since you have a Mx-500p-11, it should be easy to repair since the schematics is available.
  Can you post some pictures of the board ?
  Regarding the voltage: the real value is not important. You can try using 10-20V with current limiting to 1A for example.
  Since U4 burned, there might be shorts in the rest of the circuit. Even before connecting a power supply in place of U4, you should check the components "downstream" from U4.

[lanuser]

I have checked for short of the circuit prior to replacement U4 IC.
After replacing U4 everything went back to normal, however, I would like to properly adjust station to full 100% functionality.
I will appreciate all the help I can get.
Thank you.

[Wolfram]

As this is the currently active thread on Metcal repair, I'll add a story here, even if it doesn't directly concern the RFG-30 but the MX-500p-21.

The unit in question was completely dead. Measurements indicated that the output transistor Q4 was shorted. After replacing the transistor, the station worked but died after a short time of use. After again replacing Q4, I measured the output stage supply voltage (across C8), which turned out to be way too high at around 40 V. The LM2576 regulator for the output stage was tested and confirmed to be operating correctly. The feedback signal to the LM2576 was at 1.23 V, hinting that the RF loop was in regulation, but that the output stage was delivering much less power than expected, causing the buck converter to compensate. The gate waveform of Q4 looked reasonable, if a little high (the driver stage voltage is derived from the buck converter output, so this was a symptom of the high buck output voltage) but drain waveform was a mess, with a significant oscillation at four times the 13.56 MHz operating frequency. I measured all the passives in the matching network, and all components measured within tolerance, except for C27 (130 pf silver mica) which measured 60 pF. I borrowed C27 from a second broken MX-500p-21, and measuring it before installing it showed that it was also out of tolerance at 45 pF. C28 (which is identical to C27 and connected directly in parallel) measured spot on in both units. Replacing C27 in the first unit with C28 from the second unit restored it to normal operation, with the output stage supply voltage within the normal range. It appears that the second unit had the same fault, so this might be a common problem. It is probably a good idea to measure C27 and C28 in MX-500s with blown output or driver stage transistors, and to replace both if any of them are out of tolerance.

Next up is a pair of MX-PS5000s with the same symptom; dead buck regulator. After replacing it, they work for a short time before the regulator fails again.

[lanuser]

Do you know by any chance, how much difference in the schematic between MX-500p-21 and MX-500p-11?

[Wolfram]

As far as I know, the only difference is the default configuration of the supply transformer primary wiring, the -11 is configured for 115 V and -21 is configured for 230 V. Either can be changed to the other by moving a few solder jumpers on the board.

[miked]

I made a successful repair of a STSS-002 model RFG-30 Serial 6011xxx   This version has two diffused T 1 3/4 LEDs installed, one Green (power) one Red (RF lockout)
There appears to be *many* versions of this excellent soldering system, all with minor differences and component changes.

The hard part:  take the screws out of the top and disconnect the insulated power connectors.  The two 1/4" quick connects to the mains transformer are a pain. (a lot of wiggling needed, these were very tight !)
Now, you can drop the bottom and remove the power transformer after desoldering the ground lead from the case lug.

Q1, the ZTX749 PNP with small round heatsink, had a collector to emitter short.  I suspect the heatsink does not make good thermal contact in this application, and might be something to watch for.  As I didn't have anything like this in the junk box, I used a On Semi NSS60600MZ4T1G from a sample kit (price was right, and was a lower VCE(sat) than the original part)  The kit included small surf boards with copper heatsink islands.  I wired the eval board in vertically with respect to the case with clippings from some capacitor thru hole leads.  I had to be careful not to let the small board extend up above the main board edge,  so it fit back in the case.  I ended up trimming the ON semi supplied eval board a bit.  If I did this again, I would have clipped that Q1 out, and removed the leads one by one, because the main board was thermal damaged either by me or by the failed transistor,  and I lifted a pad. Oh well.

Unfortunately, that was not all that was wrong.  Old corroded solder flux on a tiny glued in sub board kept the unit in the two LED illuminated RF lockout state.  I think this board just puts a small DC bias on the tip, and locks out RF if there is an intermittent connection to the wand (bad wand detector)  It simply needed the connections re heated and the old flux removed with 91% isopropyl alcohol.  After verifying, I used GE silicone II to attach the two sub boards back to the main board.  I held them in place with tape overnight for a full cure.

I do wonder how long these electrolytic capacitors will last.  My unit was made around 1990.  I additionally soldered a 0.1uF 100V disc across the 1000uF Capacitor on the top of the board (this feeds the RF lockout board) and put a 1000pF disc between the collector of Q1 to ground to bleed off any stray RF created with that wing board.  Seems to work fine now.