Broken Metcal RFG-30 soldering base unit fixed

[wkb]

Today I finally managed to fix my beloved Metcal RFG-30 soldering base unit   

Metcal uses a more than ingenious heating system that uses a 13.5 MHz HF power feed into the soldering iron itself. In the actual soldering tip is a coil that inductively heats a copper 'slug' which is itself covered in a metal alloy with a high magnetic permeability.

It looses this high permeability once it's temperature reaches the Curie point.  This in turn stops the heating of the tip by the HF feed.
Anyway, long story short: go and read US patent 4,626,767 for much more information and some sample circuit diagrams.

Over all it turned out that I:

• had 2 broken heating elements/tips.  One died on me while I was using it, the other (used one) proved dead as well
• had a broken coax cable.  The Metcal uses a very nice silicon-rubber, double shielded coax which is next to impossible to refit to the handle / connector 
• forgot to put the series cap in the handle on reassembling the cable    This made the whole thing no longer a tuned LC circuit  which did not warm up, not even after the generator was fixed
• had a short due to a misaligned(?) isolation shim on the heatsink

The heatsink story is a nasty one, there are 2 power FETs and regulator IC mounted on it, using a collection of nuts and bolts, some isolated from the heatsink, some not.  For more fun the mains inlet mounts into it, as does the F-connector for the hand piece.
See the picture for details.  Design for serviceability: NOT!

The short proved to have blown a transistor, which manifested itself in a 100 Ohm resistor getting hot enough to 'smell'.  What got me confused initially was that I still had 13.5MHz HF on the output connector.  I first replaced the resistor, later I got the correct transistor type and now things are OK.  The transistor is apparently part of some feedback circuit (?) that throttles the HF amplitude.  The transistor in question is Q1, with a little heatsink, as shown in the detail picture.

For the benefit of others Googling the web for info on fixing their Metcal I have also attached schematics I found on the WWW for the Metcal MX500 base station.  That one is different from the RFG-30 but there are enough similarities to make the schematic useful.  I never found a RFG-30 schematic unfortunately.

Included are 3 scope screenshots, the first one showing the HF while the generator was broken.
The 2nd one shows the HF amplitude during warm up.
The 3rd one shows the HF amplitude with the iron idling at it's Curie temperature.

[whonline]

I appreciate the post.  I have an RFG-30 that needs repair but I can't figure out how to remove the top cover without breaking it.  Can you explain the procedure?

Thanks in advance.

Walt

[wkb]

I appreciate the post.  I have an RFG-30 that needs repair but I can't figure out how to remove the top cover without breaking it.  Can you explain the procedure?

Thanks in advance.

Walt

There are 4 Phillips head screws in mine that hold down the plastic cover with the cutout for the mains switchs.  Remove these screws.  Underneat you will find an aluminium plate that also has the mains switch mounted on it.  Pull that plate out, the wiring of the mains switch will hold it down.  Those wires have Faston connectors that allow you to separate the switch & plate from the rest of the unit.

Now remove the bottom cover, again 4 Phillips head screws.  No wires this time.

Now remove the Allen head screws from the heat sink.  And take the power inlet out.  Carefully shift the heatsink/PCB assy out of the aluminium enclosure.  I had to cut a (riveted) earth wire, don't forget to reconnect that on reassembly!

That should do it.  Not too tricky, just a bit awkward.

[whonline]

Your description convinced me the top screws on my unit were hidden by a plastic overlay.  I found the edge, peeled it up and exposed the screws. 

My repair was trivial.  Hardly a repair at all.  I was annoyed by the burned out light bulb.  I replaced it with an LED and am good to go. 

Thanks again for your help.

[longpole001]

i also have a problem with my metcal

Checking  the voltage at the 74hct04 inv is only 2.5v , and the rf output is at 3.5mhz , done the norm change of electro caps which are over 15 years old etc , doing diode check on transistors , and diode showed nothing , and slowly working the schmatic of mx-500  as a guide that  you have here and comparing to the rfg-30 unit , so far with out any success ,

i have found that the resistance on the board over the q4 is close to dead short , even when the q4 is removed , which surprised me , will try replaceing all the transistors if i can with equiv if can as a test , but i am bit lost at this point

if anyone can can verify what the sig / voltage  at the test points on the board should be would be very helpfull

What i see at the moment :

1.  voltage over R1 (100 ohm )  is 20V -  DOES THIS LOOK CORRECT
2.  Q1 collector is 6.8v
3.  R2 via L2 voltage is 6.8v , but on other side of R2 is 2.5v , which is appears wrong
4. 74htc04 pin14 vdd = 2.5v , - looks wrong

 Some input on this would be good guys

Cheers

Sheldon

[vk6zgo]

i also have a problem with my metcal

Checking  the voltage at the 74hct04 inv is only 2.5v , and the rf output is at 3.5mhz , done the norm change of electro caps which are over 15 years old etc , doing diode check on transistors , and diode showed nothing , and slowly working the schematic of mx-500  as a guide that  you have here and comparing to the rfg-30 unit , so far with out any success ,

i have found that the resistance on the board over the q4 is close to dead short , even when the q4 is removed , which surprised me , will try replacing all the transistors if i can with equiv if can as a test , but i am bit lost at this point

if anyone can can verify what the sig / voltage  at the test points on the board should be would be very helpfull

What i see at the moment :

1.  voltage over R1 (100 ohm )  is 20V -  DOES THIS LOOK CORRECT
Don't know,what is voltage on Q1 emitter?
2.  Q1 collector is 6.8v
The Q1 collector is connected to L2 as in #3
3.  R2 via L2 voltage is 6.8v , but on other side of R2 is 2.5v , which is appears wrong
According to the MXP500-11 circuit,D2 ,which is the only 6.8 Volt zener on the  board is on the opposite side of R2 from L2,so the voltage on L2 should be higher.---different circuit?
Check across D2 to see if it is working.
4. 74htc04 pin14 vdd = 2.5v , - looks wrong
Definitely wrong,with 2.5 volts on the 74HTC04,you will have no oscillator or buffer.

 Some input on this would be good guys

Cheers

Sheldon

[longpole001]

continue repairs 

 the 74hct04 is pulling the voltage down . 2.5v , when removed its goes to 8v on pin 14 , after replacing the 74hct04 , it remains at 8v now
output of pin 2 7404 is now 13.5mhz as expected , but the buffered combined outputs of the 7404 which feed Q2 [ VN016n5] ( Q3 on mx500cct) drop to 3.5mhz. and this is reflected at the final stage of output

removing the gate Q2 from the board restores the 13.5hz sig , clearly Q2 has issues

finding that VN0106N5 ( N ch power Mosfet) hard to find - is now been dropped from the   supertex product list , and VN0109N5 is also dropped from what i can see, some suppliers on the net show some stock and i am looking to get one.


VN0106N5 - To220 - 60v @2.5A  , VN0109N5 - 90v @2.5  , there are many equiv for voltage and current but not many with a @ 5ns turn on / turn off  delay time , most mosfets have higher delays typicality  about 40ns  or more but have greater power output that i have found so far.

if you know of an VN0106N5 equivalent me know ASAP

Also note that if your looking for Q4 - IRF130 - T03 package they are on the net  ,  Q4 on the MX500 shows a IRF 530 on the cct  , and its in T0220 package , with same spec's but not fun to retro mount by the looks on RFG-30 heatsink

the repair continues

cheers
 

[longpole001]

looking at the possible replacement for VN0106N5 and the IRF510-512 series came up in searches

given the input to gate of  13.5mhz  gives  a 74ns cycle , half a cycle is 37ns

rise half cycle is  = turn on delay  time + rise time
fall half cycle is turn off delay + fall time


thus spec for IRF510 -513 series  says
max rise turn on delay =20ns
max rise time =25ns ,
same for for fall times

so half cycle rise time is max 45ns , fall half cycle is 45ns  , i need a max of 37ns at a freq at 13.5mhz

thus this equivilant  mosfet wont fit this application

am i correct in my calculations ??


cheers

[longpole001]

also i find that the irf130  ( Q4) spec sheet has over 115ns combined delay + rise time , and 105ns delay + fall time  , clearly the gate input is not required to recieve a 13.5mhz signal   

 more like a 4.8mhz given those time of rise and fall specs ( assuming my other calculations are correct )

Q4 is driven via the T1 and what looks to be RC filter arrangement look at Mx500 cct 

so what is the freq i should expect on the SMA connector for the hand piece ??

what your thoughts

cheers

sheldon

[vk6zgo]

I was going to say that the turn on/turnoff specs for use as a switch would be worse than the capabilities of the device in linear mode,but looking at the circuit,I can't see any biasing,so it certainly looks like it is being used in a non linear class of operation.
If that is so,the filtering (LPF after Q4) is probably just to get rid of any harmonics of 13.5MHz.

C24 is just a coupling capacitor,R27 & R28 constitute a 50 Ohm termination for the secondary of T1.
L10 doesn't have any resonating caps--perhaps it is self-resonant at 13.5MHz?

There are no circuit components included that can divide down in frequency ,so I suggest that you will still see 13.5MHz  at the handpiece in an operational unit.
I have seen PowerFets used in "homebrew" Ham Radio transmitters,so that may be somewhere worth looking into.
P.S.:Re-read the OPs  posting ,it will help more now you have reached this point.

[vk6zgo]

Had a look around on Google--doesn't look very hopeful.
Quite a few designs using  switching type Power Fets as RF amps.
Almost universally,comments are "low gain","hard to match,due to high input capacitance","not much good above 7 MHz",etc.
The devices used by Metcal seem to be fairly unusual in having much more useful characteristics at 13.5MHz.
Perhaps you need to look at devices originally designed for RF use?

[wkb]

I was going to say that the turn on/turnoff specs for use as a switch would be worse than the capabilities of the device in linear mode,but looking at the circuit,I can't see any biasing,so it certainly looks like it is being used in a non linear class of operation.
If that is so,the filtering (LPF after Q4) is probably just to get rid of any harmonics of 13.5MHz.

C24 is just a coupling capacitor,R27 & R28 constitute a 50 Ohm termination for the secondary of T1.
L10 doesn't have any resonating caps--perhaps it is self-resonant at 13.5MHz?

There are no circuit components included that can divide down in frequency ,so I suggest that you will still see 13.5MHz  at the handpiece in an operational unit.
I have seen PowerFets used in "homebrew" Ham Radio transmitters,so that may be somewhere worth looking into.
P.S.:Re-read the OPs  posting ,it will help more now you have reached this point.

The power amp section is class-C, followed by a filter to reduce harmonics.

The frequency is 13.5MHz at the handpiece, this is an ISM frequency. 

[Joshua Bretz]

Looking at the MX-500 shematic, a short on the gate of Q4 is normal because there is an inductor from gate to source.  I found that C24 was blown in my unit which explains why Q1 fried: with C24 blown, this inductor shorts the supply.  Replaced C24 with Kemet PHE850ED6100MD18R06L2 and unit works fine.

i also have a problem with my metcal

Checking  the voltage at the 74hct04 inv is only 2.5v , and the rf output is at 3.5mhz , done the norm change of electro caps which are over 15 years old etc , doing diode check on transistors , and diode showed nothing , and slowly working the schmatic of mx-500  as a guide that  you have here and comparing to the rfg-30 unit , so far with out any success ,

i have found that the resistance on the board over the q4 is close to dead short , even when the q4 is removed , which surprised me , will try replaceing all the transistors if i can with equiv if can as a test , but i am bit lost at this point

if anyone can can verify what the sig / voltage  at the test points on the board should be would be very helpfull

What i see at the moment :

1.  voltage over R1 (100 ohm )  is 20V -  DOES THIS LOOK CORRECT
2.  Q1 collector is 6.8v
3.  R2 via L2 voltage is 6.8v , but on other side of R2 is 2.5v , which is appears wrong
4. 74htc04 pin14 vdd = 2.5v , - looks wrong

 Some input on this would be good guys

Cheers

Sheldon

[mamalala]

And just in case:

https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply

(sorry for the shameless plug)

Greetings,

Chris

[mamalala]

• forgot to put the series cap in the handle on reassembling the cable    This made the whole thing no longer a tuned LC circuit  which did not warm up, not even after the generator was fixed

I'm curious about that. In the original patent for the handle, there is a series cap mentioned. However, in my actual RM3-E handles, there is no such cap. It is a straight connection to the tip cartridge. A Talon tweezer i just got also has no series cap.

What handle do you have/use? A friend of mine got a newer handle (those shiny metal/aluminium thing with replaceable grips), and no cap in there as well. So unless the RFG30 is "missing" something in it's RF output section, i'd say that there is no cap required at all.

Greetings,

Chris

[AndersAnd]

finding that VN0106N5 ( N ch power Mosfet) hard to find - is now been dropped from the   supertex product list , and VN0109N5 is also dropped from what i can see, some suppliers on the net show some stock and i am looking to get one.


VN0106N5 - To220 - 60v @2.5A  , VN0109N5 - 90v @2.5  , there are many equiv for voltage and current but not many with a @ 5ns turn on / turn off  delay time , most mosfets have higher delays typicality  about 40ns  or more but have greater power output that i have found so far.

if you know of an VN0106N5 equivalent me know ASAP

Also note that if your looking for Q4 - IRF130 - T03 package they are on the net  ,  Q4 on the MX500 shows a IRF 530 on the cct  , and its in T0220 package , with same spec's but not fun to retro mount by the looks on RFG-30 heatsink

Old topic I know, but I recently came across this topic looking for repair tips for a couple of defect Metcal MX-500 power supplies.
So the below might be useful to others searching for replacements for obsolete MOSFETs in Metcal power supplies.
The text is a copy of a reply I made in the DIY Metcal 13.56 MHz RF Supply topic: https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply/msg434497/#msg434497

I've just repaired two MX500 power supplies (won at an eBay auction), with the help of the schematic and documentation here: https://www.mikrocontroller.net/attachment/193474/MX-500P-11.pdf
Mirror: http://scopetechniques.com/Metcal/MX-500P-11.pdf

The documentation says:
Q3 = VN0109N5 http://scopetechniques.com/Metcal/465-1342-0-VN0109.pdf / http://scopetechniques.com/Metcal/VN0109N5.pdf
Q4 = IRF530(N) http://www.irf.com/product-info/datasheets/data/irf530npbf.pdf

But in the two defect MX500's I bought they were different:
Q3 = IRF510 http://www.irf.com/product-info/datasheets/data/irf510pbf.pdf
Q4 = IRF640N http://www.irf.com/product-info/datasheets/data/irf640npbf.pdf
Both branded International Rectifier.

The amber/orange led lit up all the time in my two defect devices while the green led was always off.
I found out Q4 was shorted in both devices and after replacing it with a new IRF640N both devices work again.
It looks like Q3 (IRF510) had been replaced before in at least one of the devices, so not sure if they came with Q3 = IRF510 and Q4 = IRF640N from the factory. But it seems to work, at least so far.
Has anyone else seen an MX500 with Q3 = IRF510 and Q4 = IRF640N?
Looks like VN0109N5 from Supertex is obsolete, so maybe Metcal started using IRF510 instead?
Now when you search Supertex.com VN0109 seems to only be available in wafer / dies and TO-92 housing called VN0109N3, but no TO-220 option.
And for Q4, IRF640N is a 200 V MOSFET, while IRF530(N) from the schematic is only rated at 100 V, so maybe they have replaced it because there was problems with too high voltage spikes, toasting the IRF530's? Doesn't look like IRF530N is obsolete.
Haven't done any measurements after repairing them, so not sure about the voltages Q4 handles? But since the IRF640N was toast in both of them, maybe that's not the issue?


Much to my surprise I also received a Metcal solder stand with the defect power supplies, even though this wasn't advertised. A very nice surprise as Metcal solder stands are actually very expensive and I didn't have any. Later I received a new and very nice MX-H1-AV handpiece as a birthday gift.
This came from RS Components where MX-H1-AV seems to be quite cheap compared to many others for some reason, even though RS usually isn't know for being cheap. So if anyone needs a new MX-H1-AV alloy handpiece for your DIY, Metcal or Thermaltronics/Easy Braid power supply, try to check out the price at you national RS Components site.

[wkb]

• forgot to put the series cap in the handle on reassembling the cable    This made the whole thing no longer a tuned LC circuit  which did not warm up, not even after the generator was fixed

I'm curious about that. In the original patent for the handle, there is a series cap mentioned. However, in my actual RM3-E handles, there is no such cap. It is a straight connection to the tip cartridge. A Talon tweezer i just got also has no series cap.

What handle do you have/use? A friend of mine got a newer handle (those shiny metal/aluminium thing with replaceable grips), and no cap in there as well. So unless the RFG30 is "missing" something in it's RF output section, i'd say that there is no cap required at all.

Greetings,

Chris

Talking about a delayed answer...

Not sure which handle type I have. It sure is an old one, I am pretty sure it dates back from the 1980-1990s.

[mamalala]

Talking about a delayed answer...

Not sure which handle type I have. It sure is an old one, I am pretty sure it dates back from the 1980-1990s.

No worries. As it turns out, there is no cap in there. A friend of mine checked with a network analyzer, and found no hint of a cap being there. Also, the tip detection works by biasing the tip, which in turn requires a DC path, any series cap would make that impossible. This applies to the RM3E handpiece, as well as the newer ones, and also the Talon tweezers, from what i have been able to find out.

Greetings,

Chris

[AndersAnd]

Talking about a delayed answer...

Not sure which handle type I have. It sure is an old one, I am pretty sure it dates back from the 1980-1990s.

No worries. As it turns out, there is no cap in there. A friend of mine checked with a network analyzer, and found no hint of a cap being there. Also, the tip detection works by biasing the tip, which in turn requires a DC path, any series cap would make that impossible. This applies to the RM3E handpiece, as well as the newer ones, and also the Talon tweezers, from what i have been able to find out.

Isn't there a feed-through capacitor instead of a series capacitor as we talked about earlier from here onwards: https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply/msg310381/#msg310381

[mamalala]

Isn't there a feed-through capacitor instead of a series capacitor as we talked about earlier from here onwards: https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply/msg310381/#msg310381https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply/msg310381/#msg310381

From what my friend told me, there _seems_ to be none, according to the analyzer. Of course something could have gone wrong, and there is one and he just missed it, i don't know. After all, there is also the capacitance from the cable, so that any internal cap was simply "hidden" due to that.

Greetings,

Chris

[AndersAnd]

Isn't there a feed-through capacitor instead of a series capacitor as we talked about earlier from here onwards: https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply/msg310381/#msg310381https://www.eevblog.com/forum/projects/diy-metcal-13-56-mhz-rf-supply/msg310381/#msg310381

From what my friend told me, there _seems_ to be none, according to the analyzer. Of course something could have gone wrong, and there is one and he just missed it, i don't know. After all, there is also the capacitance from the cable, so that any internal cap was simply "hidden" due to that.

Greetings,

Chris

Found these old pics of a disassembled Metcal MX-RM3E https://imgur.com/a/RvDVu#0
They were posted in this topic in Oct. 2012: https://www.eevblog.com/forum/reviews/for-jbc-fans-out-there-jbc-tips-anatomy/msg151939/#msg151939

The last pictures shows there actually is a small capacitor in parallel with a 22µH inductor.

[brunos67]

Looking at the MX-500 shematic, a short on the gate of Q4 is normal because there is an inductor from gate to source.  I found that C24 was blown in my unit which explains why Q1 fried: with C24 blown, this inductor shorts the supply.  Replaced C24 with Kemet PHE850ED6100MD18R06L2 and unit works fine.

i also have a problem with my metcal

Checking  the voltage at the 74hct04 inv is only 2.5v , and the rf output is at 3.5mhz , done the norm change of electro caps which are over 15 years old etc , doing diode check on transistors , and diode showed nothing , and slowly working the schmatic of mx-500  as a guide that  you have here and comparing to the rfg-30 unit , so far with out any success ,

i have found that the resistance on the board over the q4 is close to dead short , even when the q4 is removed , which surprised me , will try replaceing all the transistors if i can with equiv if can as a test , but i am bit lost at this point

if anyone can can verify what the sig / voltage  at the test points on the board should be would be very helpfull

What i see at the moment :

1.  voltage over R1 (100 ohm )  is 20V -  DOES THIS LOOK CORRECT
2.  Q1 collector is 6.8v
3.  R2 via L2 voltage is 6.8v , but on other side of R2 is 2.5v , which is appears wrong
4. 74htc04 pin14 vdd = 2.5v , - looks wrong

 Some input on this would be good guys

Cheers

Sheldon

[brunos67]

Looking at the MX-500 shematic, a short on the gate of Q4 is normal because there is an inductor from gate to source.  I found that C24 was blown in my unit which explains why Q1 fried: with C24 blown, this inductor shorts the supply.  Replaced C24 with Kemet PHE850ED6100MD18R06L2 and unit works fine.

i also have a problem with my metcal

Checking  the voltage at the 74hct04 inv is only 2.5v , and the rf output is at 3.5mhz , done the norm change of electro caps which are over 15 years old etc , doing diode check on transistors , and diode showed nothing , and slowly working the schmatic of mx-500  as a guide that  you have here and comparing to the rfg-30 unit , so far with out any success ,

i have found that the resistance on the board over the q4 is close to dead short , even when the q4 is removed , which surprised me , will try replaceing all the transistors if i can with equiv if can as a test , but i am bit lost at this point

if anyone can can verify what the sig / voltage  at the test points on the board should be would be very helpfull

What i see at the moment :

1.  voltage over R1 (100 ohm )  is 20V -  DOES THIS LOOK CORRECT
2.  Q1 collector is 6.8v
3.  R2 via L2 voltage is 6.8v , but on other side of R2 is 2.5v , which is appears wrong
4. 74htc04 pin14 vdd = 2.5v , - looks wrong

 Some input on this would be good guys

Cheers

Sheldon

Joshua what blew your C24 cap? curious!

[hurtmanissimo]

is possibly to replace VN016n5  Q2 (q3 on mx500) by irf510, but need to short c20 (mx500 schematic) to work properly.

without that it start to work on 7mhz   and very low heating.  may be different characteristics of gate capacity irf510 vs VN016n5 make lc-filter like.

hope it will help someone

[clauberty]

Hello

Thank you for the useful post i found here and solved/repaired my mx500

but i still have one problem, the station does not go off. Meaning green led after 30 min measured by me on other station is off, on my station will remain on has long has the switch is ON

yes there is a little torx / hex screw on the side of the station and it is screwed in so switch is activated, has is on the station that this standby of is working

i made the testing with the same  Hand Piece and Work-stand from the link

https://gokimco.com/oki-mfr-uk1-cartridge-hand-piece-and-workstand.html

Thank you