Copyright 2021-23 by H. Gragger. All Rights Reserved. All information provided herein is destined for educational and D.I.Y. purposes only. Commercial re-sale, distribution or usage of artwork without explicit written permission of the author is strictly prohibited. The original units  with their associated  trade-names are subject to the copyright of the individual copyright owner. The Author is by no means affiliated with any of those companies. References to trade names are made for educational purposes only. By reading the information provided here you agree to the Terms of Use.

Although I will soon go back to my roots by means of language (i.e. German) there is so incredibly much helpful information to be found from mostly the English speaking welding community. So this is my homage to all you guys out there.

A decade ago I felt like learning to weld, and I had bought a huge used professional MIG / MAG welder from ELIN. Surely a brute.
I did (and still do) not have a suitable indoor room that would allow for welding without setting something on fire, so my adventures are restricted to outdoor activities.

It was too early, time was not right yet. I gave it away to my brother.

The very same brother recently found a Güde Easy 170 gas / nogas MIG/MAG unit standing at the scrapyard, looking like new and waiting to be salvaged. He took it home and found that it worked somehow, but appeared to have no ooomph.

Since he has welders galore anyway, I decided to try to repair it for myself. A new epoch opened up for me - the iron age.

There is many a thing I discovered through many an hour I spent on the internet, and the outcomings may be of interest for y´all and may give you a steeper learning curve.

Back To Index

---

The Welding Rectifier

I was totally unknowledgeable as far as the interior of welders goes.
However, being a trained engineer of electronics, a first look inside with a "technical" eye revealed a huge transformer with taps going to a beefy front panel switch, a choke, a feeble control panel, a fan, and a rectifier with a heat sink.

Since there was voltage on the secondary, I looked closer at the rectifier. Some diodes had their connective leads blown, one was fallen off, so clearly - the rectifier needed some attention.
First idea was to find those press-fit diodes, however the crystal ball was ignorant about their type.
On welding sites some guys were retrofitting rectifiers into their welders, mainly simple old-school AC stick welders.

The routes taken were mainly:

1. super heavy (Chinese) block rectifier bridges or single rectifier diodes.
   Those have incredible power capabilities on paper, but people distrust them. Rightfully so, since the Chinese are prone to exaggeration and component faking.
2. Stock rectifier bridges in parallel to boost their current capability. People claim they worked, but we do not know for how long. See the discussion below.
   
3. One thing nobody seemed to mention, was dedicated welder rectifiers. It took me some days to find them, but once knowing the key words, they seemed to pop up. Their price tag cannot be beaten.
   

The built in rectifier has a label on it that says "SCOMES  PMS 30". SCOMES turned out to be one of the key words.
Indeed there are lots of scribblings in Italian inside the machine, so this was definitely assembled in Italy.

SCOMES is an Italian company^[1] (and indeed, very many companies to do with welding are located in Italy) that mainly produces welding rectifiers.

Be prepared that they don´t speak to you. They refused to communicate or were not very helpful (practically all of them), either because of the language (I wrote in English to them) or just because of sheer arrogance towards foreigners. Hard to tell. Maybe they only talk to you if you promise to take hundreds of units.

However, you find the numbers there and then you can go hunting the web.

Indeed you will find the diodes^[2] (they are allegedly used in car alternators), but

1. removing the old diodes would probably damage their bore and thus create a bad  mechanical fit for the new ones
   
2. I believed the rectifier was not fit for the task anyways (my brothers very similar Güde MIG 170 has a much bigger one in it with 12 diodes...), so an upgrade was advisable
3. I turns out, buying a new, much more powerful one costs ridiculously little and comes ready-to-use. Having more diodes that conduct less current means less voltage loss overall, meaning more beef to weld
   

The one I got finally is labeled PMS-180, and is double the length of the old one and has two plates like the old one.

By the way, the number of plates indicates the number of phases:

2... single phase (what they call monophase, like in PMS xx)
3... dual phase: I don´t see it on their list, but a dealer here has one and he says, that there are half-bread machines that are not quite three-phase, but use two phases for more power.
4... three phase.

This can be helpful, because their catalog is not very explanatory.

The PMS-180 can deliver 180 A at a 60% duty cycle (see explanation below) and has 20 diodes, the defective one had 8 diodes. The latter is no  longer listed, but logic dictates that it was on the lower side power-wise and was certainly pushed to the limit. The bill was paid...

The price was ridiculously low at about 50 Euros, and the 120 Ampere type (PMS-120) would cost hardly any less. Sometimes the things you buy bear no relation to the price tag - both ways...

Duty cycle is always measured in 10 minute intervals. A 60% duty cycle thus means, you can weld at the rated current for 6 minutes and must maintain a 4 minute cooling period, or the machine will cut out by thermal overload. This is further subject to several other variables^[3].

MIG 170 schematic (section) (click to download .pdf )

PMS-30 disassembled (click for larger image)

Rectifier diodes are wired as full wave bridge rectifier, see schematic (the parallel ones are not shown). Note that this is the schematic from MIG 170, which is the successor. It has slight differences, such as a dedicated transformer tap for the motor supply. To the right side you see the disassembled PMS-30. The bottom left diode is the only one left on this panel that is outwardly intact, the top left has fallen off entirely. It seems the pills are (soft) soldered to the tabs, they are not spot welded. Obviously the rectifier has become too hot, being way too small for the task. The successor MIG 170 has a rectifier with 12 diodes in it. Those who have a center-tapped secondary may short the AC terminals together and have double the power available into the bargain.

Paralleling diodes might create some hazard: do they really share the current? What if one has a lower threshold, does it not tend to take the major current load and burn out? Let´s look at this subject:

thermal runaway: due to their negative temperature coefficient, when diodes heat up, their resistance decreases, and they heat up even more thermal coupling: sharing the heat by mounting them on a common heatsink will alleviate the problem. using bridge rectifiers with shorted AC terminals enforces thermal coupling the diodes are run far below their limit, so an alleged imbalance will not have dramatic consequences connective leads and wiring acts as a small series resistance that helps balancing. - discussion at https://electronics.stackexchange.com/questions/6151/is-paralleling-diodes-a-bad-idea

practically all ST common cathode diodes in the same package can be connected in parallel without any precaution. Since these diodes are on the same die, the variation in forward voltage is very low. Low voltage Schottky diodes  can also be connected in parallel without precaution. A good value of maximum allowed variation in forward voltage can be 40 mV. With this value, the current will be equitably shared between each diode and the thermal effects will become negligible. - summary from ST AN4381 application note (abbreviated)

Parallel diodes can be forced to share current by connecting a very small resistor in series with each diode. Although currentsharing is very effective, the power loss in the resistor is very high. Furthermore, it causes an increase in voltage across the combination. Unless using a parallel arrangement is absolutely necessary, it is better to use one device with an adequate current rating. - summary from https://www.daenotes.com/electronics/basic-electronics/diode-in-parallel

Original defective PMS-30 (click for larger image )

New PMS-180 (click for larger image)

So I wondered about the current distribution in the SCOMES rectifiers. They are to be found in virtually all transformer-based welders, so this must be a proven and reliable method. Then it dawned on me: the relatively thin wires to the diodes are the resistors! Just a few milliohms are enough. This turns out to be a very well thought about and effective concept. You will probably never consider warming up your soldering iron again for knitting your own  bridge rectifier stack for that price tag (letting alone their unproven longevity...) There was space enough inside the case, but having two mounting studs, I needed to screw an aluminium bracket down to the floor to accommodate the second one. Trivial. So this way I got the machine working again.

Wire Feed Rollers

Since I planned to use flux (cored) wire without gas, a.k.a. self shielded flux wire  (Deutsch: selbstschützender Fülldraht),  I needed the appropriate rollers.

Wire Feed Rollers (click for larger image)

Flux cored wire is very soft compared to a standard MIG (steel) wire, so it cannot be transported with a standard V-groove roller. It needs a knurled groove, V or U. The counter pressure roller (normally a simple ball bearing roller) should be tensioned enough to prevent slippage, but not as much as to damage the wire and certainly far less than a standard wire asks for. The Easy MIG (and MIG 170 at least) use a two roll wire feeder unit made entirely out of plastic, motor bearings an all. They rollers are screw mounted to the axle. I found that spare rollers made by DECA[4] fit, namely #010647 Fe ø 0,6 / 0,8 #010628 Fe ø 1,0 / 1,2 #010629 Al ø 0,8 / 1,0 #010627 Flux ø 0,9 For reference: they are 30mm OD, 10mm ID and 18mm depth.

Note that you probably need a different liner for #010629 (aluminium welding), plus the power will probably be too low for aluminium.
Same applies for 1.0 to 1.2 mm wire. There is a reason the machine comes stock with 06/08 steel.

Note that you probably need a different liner for #010629 (aluminium welding), plus the power will probably be too low for aluminium. Same applies for 1.0 and 1.2 mm wire. There is a reason the machine comes stock with 06/08 steel.

DECA refused to answer my questions in a way that was helpful and were not very friendly overall. I had to jump in at the deep end for the rollers and bought them from a local Italian shop who had them on the ´bay. They indeed did not answer my questions on the dimensions too, despite several attempts. Hmm. Well they were cheap enough for 10 Euros each. They sent me #010627 which has two knurled grooves stamped 0.6 and 0.8, so it looks that both 0.8 and 0.9 are meant to run on the 0.8 groove. And they do.

I then found the same rollers to be used in another machine made by ELMAG for their Euromig 160 series (the pictures don´t fit the descriptions btw.) for four times  price; that´s how to make money out of nuthin...

• #54700 Fe ø 0,6 / 0,8
• #54701 Fe ø 1,0 / 1,2
• #54699 Flux ø 0,9

Those welder units probably stem from the same time domain, like many, many others destined for the hobbyist segment, and are 230V machines that have been equipped with, erm, more economic parts. I have not seen the wire feed assembly per se being offered by anybody any more, but there must be thousands of them installed in those units.

Beware: Clarke has similar design rollers (screw mounted type), but those have a different bore (9mm vs. 10mm).

Unfortunately, if this wire feed unit decides to fail on day, a different replacement unit with a 24 Vdc drive will have to be installed. These days they are made of metal rather than plastic for the relevant parts. I recommend inquiring about the rollers used beforehand, to have a supply at hand  if needed.

The fact the MIG 170 and Easy MIG units  were equipped with DECA rollers (and presumably a wire feed drive made by them) plus the fact that the machine was internally full of Italian scribblings made me think that DECA built it originally, but a welder salesman with experience remembered that Güde welders were built by CEM (likely coincident with CEMONT), but  later outsourced to Slovakia.
Indeed CEM style welding accessories are sold by Chicago Electric / Harbor Freight and others.

Then there is the Trafimet company (maker of the work clamp I used), which is Italian too. Funny there is so much welding stuff built in Italy. All those entry-level units use peripherals that look darn similar, they all seem to use the same components.

Back To Index

The Work (-Lead) Clamp Is Not A Ground (Earth) Clamp

... and the workpiece lead (wire) is not a ground lead...

"The Workpiece Connection is not a Ground Clamp Ground clamp and ground lead are common terms used by many welders. The workpiece is connected to a welding cable typically by means of a spring loaded clamp or screw clamp. Unfortunately, a workpiece connection is often incorrectly called a ground clamp by many welders and the workpiece lead is incorrectly called ground lead. The welding cable does not bring a ground connection to the workpiece. The ground connection is separate from the workpiece connection." - Lincoln Electric: Grounding And Arc Welding Safety https://www.lincolnelectric.com/en-us/support/process-and-theory/pages/grounding-safety-detail.aspx

They are however somewhat clumsy in their description, because they  show a picture of such a clamp right beside the paragraph, where they advise against calling it grounding clamp. This is as useless as a double negation to a child.

The apparatus´ case itself is certainly earthed, but there is NO earth connection to the transformer secondary anywhere. Kind of an insulation transformer. I am not sure why this is not earthed automatically.  I am sure now.

Any electrical device that is not double insulated, i.e. any device that has a metal case has to be grounded (earthed). This is according to standard safety regulations and is valid for all electric appliances. However, the welding voltage itself is not normally earthed, because it is subject to different regulations:

"According to ANSI Z49.1, "Safety in Welding, Cutting and Allied Processes," the workpiece or the metal table that the workpiece rests upon must be grounded. We must connect the workpiece or work table to a suitable ground, such as a metal building frame. The ground connection should be independent or separate from the welding circuit connection." (highlighted by the author) - Lincoln Electric: Grounding And Arc Welding Safety https://www.lincolnelectric.com/en-us/support/process-and-theory/pages/grounding-safety-detail.aspx

But note: this is valid only for those states, that are subject to ANSI Z49.1! In Europe (Germany) they strictly state that a welding table must not be earthed! (BGI 553, Abs. 3.5)

The term "ground" is wrong for other reasons too:

• in an AC system (like a stick welder) both leads are live
• in a DC system the electrode might be negative or positive
• any reference potential can be called "ground", usually having no earth connection. Earth is Earth.

It is a sad fact that many manufacturers or dedicated welding sites keep mixing up that nomenclature and thus add to the confusion, but Lincoln has it right! You are well advised to adhere to the official safety regulations and not to those sites, and indeed my writings.

As a short aside: for earthing, from what I deduct, most hazards associated with often quoted stray currents (for either the welder or the line grid installation) come from

• either welding processes that have permanently live electrodes (stick welders)
  
• working on big steel structures that your are standing on during welding or
• welding processes that require AC/DC voltages higher than those considered safe for touching.

Neither is the case for MIG/MAG machines (except maybe the second one, if you repair your tractor), so we are relatively safe.
Also, with MIG/MAG there is little high frequency components involved that may be the cause of RFI pollution.

But back to right nomenclature for the clamp: it is more aptly be called "work lead clamp" or "work clamp" and "work lead". It is the return current line. This eliminates all potential for confusion.

For the Güde, since the existing work clamp was creating a problem^[5] I upgraded it:

Original useless work lead clamp (click for larger image )

SACIT Nevada 300 work lead clamp (click for larger image)

The work clamp mounted on the Güde was utterly ridiculous, a toy and a disgrace. The prongs were bending under the pressure, the contact claws were just iron with copper plating. There was no mesh wire connection between the claws. OK for a battery charger. I bought one from Trafimet, SACIT[6] NEVADA 200 (their distributor sent me the 300A version which looks equal), which is a really respectable unit. Massive copper contact blocks interconnected with substantial copper mesh wire and a massive spring. This thing is indestructible. Again, the price of less than 9 Euros is ridiculously low for that quality. Why do Güde not ask for, say 50 Euros more and build something respectable? Penny pinching for what? The mind boggles over that.

Back To Index

This welder´s design tries to be efficient while being effective. In so far those designs are so minimalistic that they are fail-safe and brilliant. But it turns out that most of those simple units suffer from certain peculiarities. Some are harmless, some annoying.

Originally, the wire feed circuitry is connected directly to the power rectifiers output. Pulling the trigger would activate the welding supply by means of  the primary relay besides simultaneously activating the wire feed.

Looking closely, this takes a fraction of a second to get going, but this is not usually a problem.

Fellows have noticed (like me) that the feed motor continues to run an indefinite time after releasing the torch trigger, spouting out wire. Now this is a problem, because you have to aim for a constant CTWD^[7] (Contact Tip to Work Distance, a.k.a. known as stick-out; both terms bear the potential of misconception by the way) throughout.
I found myself more trimming wire than welding wire, particularly when tacking...

The problem arises, because the motor is wired directly to the DC supply and has no low resistance path to get rid of the energy stored in its inductance, once the trigger is released. Every DC Motor needs to be braked in a controlled way or it will overrun.

This lead to a modification known as the wire feed mod^[8]. This simple mod utilizes a small auxiliary relay, slaved  after the primary relay (230V type). It has to have a SPDT contact at least.

Motor brake relay (click for larger image)

Brake Relay Wiring (click to download .pdf)

The positive supply to the motor turns on, when the aux relay is energized, and the contact shorts the motor to ground upon release. Note that I added a flyback diode, that should always be there. When DC power gets removed from an inductor (relay, motor etc.) abruptly, the inductance will kick back and produces a reverse voltage, which can result in  huge transits, albeit of very short duration (hence the term "kickback diode"). This was not part of the original mods I have seen, but it is part of commercial designs that utilize this technique. The time duration from removing the motor power to shortening the motor will nevertheless be very small. There is no contact sparking noticeable, so protective measures beyond that seem overkill. I used a relay from Finder (40-61 series), which indeed is beefier than the relay used in the primary (40-52 series) together with a socket, because this eliminated the need for a PCB and can be mounted on a piece of rail. The numbers correspond to the Finder relay datasheet.

Needless to say, this mod stops the motor dead in its tracks. An effective mod by all means and really worth while.

Back To Index

---

Erratic Wire Feed And The Cure

The transformer secondary on the Güde that provides the welding current, after the rectifier, produces a voltage between 19 Vdc and 28 Vdc (open loop) - basically an ideal range for the small DC motor that drives the wire feed unit. Unfortunately that voltage varies with the tap chosen on the primary, which is de facto a power setting. A higher power setting will invariably increase the wire feed speed (hereafter: WFS). This comes out as a faux synergistic effect, since higher voltage needs higher current settings anyways and thus is not necessarily problematic, although real synergistic welder units will actively fiddle with that due  to tried and tested parameters from program presets.

Note: those voltage values were measured by a cheap digital volt-meter. #1 this device uses a certain crest factor for calculating the values, meaning it relies on a sinusoidal wave form of 50 Hz, while the rectified waveform has 100 Hz. #2 the values are measured under no-load conditions, meaning the voltage sag upon application of a load is not considered. Therefore, those values are only rough estimations and will invariably be incorrect.

Some guys have noticed that peculiarity and thought this was something that needs to be remedied and installed a separate supply of sorts for the motor. Arc geometry (linked to voltage) and current (linked to WFS) are strongly intertwined, particularly in the short-circuit welding or short-arc welding mode we are concerned here with. This is a kind of self-regulating process. The faux synergistic effect works not bad on the Güde, but some machines seem to suffer from the voltage sag under load so badly, that the WFS becomes erratic to a point of being unusable (SIP and COSMO machines). That said, it was noticeable on the Güde too, and particularly annoying when starting with laying a bead.

With MIG 170, which apparently is a direct successor to the EASY 170 device being dealt with here, Güde have made a change to the welding transformer secondary by spending a small additional winding for the wire feed motor. They have by the way also made changes to the speed control electronics on the front panel (like using an encoder instead of a potentiometer).

Note that the speed control circuitry has no galvanic connection to the output transformer leads, like it did on the EASY 170 model (common minus potential, look at the schematic above). It is therefore floating.  This is different from the mods shown in a welding forum (post #176).

This decision does not interfere with the primary relay control circuitry on the front panel as this has (to have) its own supply, fed by the small transformer / rectifier assembly you typically find on those boards. This circuitry also produces the low (presumably 12 or 24 V) control signal that goes to the torch trigger. They are all isolated from each other.

So it becomes very easy to provide a stable voltage to the speed control assembly just by cutting the leads to the original (welding-) transformer / rectifier and inserting a dedicated transformer / rectifier for the motor circuitry.

I do not like the idea of using a switchmode power supply left over from a defunct laptop. Although workable, a welder is a blunt instrument and everything inside has to be rugged, and a SMPSU belongs onto a desktop and not inside a welder. A transformer and a rectifier play in the same league as the welder.

The DC Motor present in the Güde has a comparably small current demand, it only uses 500 mA @ 20 Vdc at full throttle (equivalent to 10 W, which suggests it genuinely is a 15 W unit as some claimed)  which is small compared to bigger machine´s 90W brutes (that look deceptively like a windshield wiper motor). By the looks of it, those little machines all use the same motors. Bigger machines use bigger motors, but those very likely have a properly designed driving circuit that needs no upgrade, resulting in a whole different price-segment.

I had a custom wound 80VA transformer lying around that are just fine power-wise. It has two secondaries rated at 18V full-load, but for a system that just uses a FWCT rectifier and no capacitor, and a light load,  it was kind of hard to anticipate, if the resulting motor speed was in the right range or not. It turns out, the range is perfect.

In many welder forum pages that perform this mod people bought 24 Vac transformers, and while this does not hurt technically, it shifts the drive circuit to excessive speeds. They then had to fumble with the drive circuit. Keep in mind that the transformer ratings specify 24 Vac at full load, so at no or light load this may rise another volt. Those values are referring to a crest-factor of a sinusoidal wave, meaning the peak values are by a factor of 1.4 higher or even more, since we have 100 (resp. 120) Hz after the rectifier. So the motor control board might see some 32 Volts.

Auxiliary Transformer Assembly (click for larger image)

Auxiliary Transformer Wiring (click to download .pdf) (w/ brake relay shown)

Having two secondaries, you can either parallel them or use just one of the secondaries together with a FWBR (full wave bridge rectifier). Paralleling is generally advised against, although the maker thought it possible. I did not feel well with that. Ignoring the second winding was feasible due to the low power needed, but appeared a waste. I then decided to join both secondaries in series and use a FWCT (full wave center tapped) rectifier that needs only two diodes. This wastes some power capability (due to the fact that there is always just one winding / diode energized), but has the advantage of losing just one diode drop upon rectification. The transformer´s specs are 2*18 Vrms. I measured 2*20 Vdc open and 19.5 Vdc after the rectifier. The primary is connected via a fuse to a point right after the mains switch. It is permanently live.

The rectifier I had, a KPBC2506 (25A, 600V) is an overkill too, but it does not harm. The (-) terminal is left open.

If I had to buy a transformer, I may strive for a single 20 Vrms secondary (being a stock value) and a bridge like the above to make for a full wave bridge rectified arrangement. Since you can safely assume that all drive units installed in those little machines you ever are liable to meet (in case you  have to swap one out) use motors in that power range, there is little need to prepare for dramatically bigger motors. In the lights of that, 100 VA, as many modders have suggested, appear like a total overkill and a waste.
For Güde Easy 170 a recommended mod, for MIG 170 not necessary.

This modification was another 100% success. Very steady wire speed. The estimated "center" speed value (as it works out from the total available voltage) appears no different than before, so I believe the 18Vrms secondary is a good choice. There is no need to change any trim pots.

Back To Index

---

Burnback

Despite fixing the motor overrun problem, the welding current persists a fraction of a second after releasing the trigger. I made a request in a welding forum and a friendly mate turned me onto the right key-word: burnback. Unfortunately this can mean two things: #1 the wire burning back into the contact tip due to false settings or false usage or #2 a controlled amount of burn-back after the trigger is released.

It is the latter we are concerned with and the amount of control thereof.

I noticed that both after pushing the trigger knob and releasing it, there is a delay of a fraction of a second until a reaction appears.
This delay is unconsequential upon starting and probably helpful by establishing proper gas-flow before igniting the arc, however after releasing the welding voltage remains active for the same period of time. The way it is wired means that there will be a certain amount of wire spouted out too.

This has a very logical and comprehensible reason:

"Burn back control isn’t available on all machines but that doesn’t mean it’s not there. It’s just not adjustable. (...) If it just stopped dead your wire would be stuck to the job. And too much you’ll be changing the tip every run. " -Richard, (Mig Welding Forum Member), post #11 https://www.mig-welding.co.uk/forum/threads/burnback-control.46/

The way this is ensured is by putting an electrolytic capacitor in parallel to the relay coil. Now that could be fixed by separating the mains relay from the motor relay and adding different time constants to the motor relay. That said, once a certain setting has been found, any change in any other parameter will render this setting futile again. So for the most, you can find  a burnback control setting that works only for a given setting of the other parameters, and that will give you freedom for exactly that one job. In the end, you probably have one control more to worry about...

"As far as burnback control goes, i dont have it set too high on the set i use most of the time at work, because of how as the wire burns back, you end up with a blob on the end of the wire (...). That blob unless you snip it off, can cause problems with popping and farting as you start the next weld." -Liquid Metal, (Mig Welding Forum Member), post #14 https://www.mig-welding.co.uk/forum/threads/burnback-control.46/

You can twist this every way you wish, it probably boils down to just that:

" (...) there was always an inch of wire poking out after every weld. Then I discovered that if I left the torch in place briefly after I released the trigger it would actually burn back to exactly the length it needed to be. I assume that is built in burn back." -Morrisman, (Mig Welding Forum Member), post #17 https://www.mig-welding.co.uk/forum/threads/burnback-control.46/

That said, I looked what can be done on the Güde. After a day of reverse engineering the control PCB, I know their schematic inside out. It is not as basic as some SIP machines appear to have it, but still crude enough that a burnback feature cannot be incorporated without unreasonable effort (meaning: a completely reworked control board). What is needed is a true off delay timer relay for the welding transformer primary similar to the way shown here^[9] .

Back To Index

---

Back To Index

Pro vs. Hobbyist Grade And Upgrade

---

Verdict

I learned a tremendous amount on how welders work in general and how those transformer-based units work.
Note I highlight transformer-based, because contemporary units are inverters, meaning they are mostly all-silicon.

Such stuff usually is full of bells and whistles, programs that let even a beginner weld good (so they promise...), but let some decades go by, like on the unit I speak about above, what will be with those electronic gew-gaws?

---

In 20 years from now, a simple transformer based  unit
will either be working, or repairable.
A cheap all-electronic device will  be on the scrap-yard.
Guaranteed.

---

Should I ever feel the need for something bigger or more flexible, I would know what key elements to look at.