Tuesday, 13 August 2019

Textar Brake Pads again.....

I finally got my Textar brake pads sorted! Textar pads are increasingly hard to find and on the point of bring prohibitively expensive.  See my post of April and June 2019 for the story.
Textars......
The plan was to (re?) insert 'pad wear' warning light wires into a set Textar pads that did not have them. Back in the day, not all Ds had the brake wear warning system as part of the loom and it's possible that Textar either sold a variant without wires, or that someone simply cut the wires off.
DS loom with warning wires
DS loom without warning wires......
The casting for the backing plates include a shallow groove where the wire sits and presumably the mould for the brake pad itself similarly makes allowance for the groove and a wire.
The wires sit in a groove behind the brake pad material
I wanted to drill out what I was convinced was a pre-existing hole.....
Tell-tale marks of a hole
......but a standard drill was not long enough to clear the edge of the brake backing plate.
The chuck is blocked by the backing plate
The slow boat from China carrying the extra long drill I'd need failed to arrive so, before I went on holiday, I ordered another one. This time on a fast van from Stourport-On-Severn. I got back from holiday to find it had arrived - so I could crack on!

If you look carefully at the photo below you can just about see that the silver drill on the left is slightly longer than the standard length black drill on the right......
This one should do it!
Taking one of my old Textar pads, I sandwiched the drill bit in the groove between the brake pad and backing to get the drill alignment. It turned out that the drilling line aligned to the corner of the back plate - so was easy to get right. I also took the opportunity to mark the drill for the correct drilling depth.
Confirming the drilling angle
Textar pads are old school asbestos, so some safety equipment was definitely in order. I used goggles (because I was drilling) and a facemask (because of asbestos dust). I also kept my face well away from the drilling area.

I was nervous drilling the first hole but much happier when I saw fragments of copper wire and shredded fabric appear from my drilling hole. These were tell-tell remnants of the end of of a brake wear wire that had still been embedded in the pad.
Drilling the holes out
When all four were drilled, I took the pads out in to the open and gave the holes a quick blast of compressed air. I recovered the four wires from the pads that were worn and broken and cleaned up the brass contact ends with emery paper.
The wires were recovered from my old pads
I had decided to use exhaust sealing paste to glue the wires home. I had reasoned that the wires did not come under any tension and only needed to be lightly secured in place. The paste was non-sticky, happy at high temperatures, and crumbled to nothing when crushed or rubbed.
I used exhaust assembly paste as my glue
One by one the wires were coated in paste and carefully pushed home. It was a simple as that. I resisted the temptation to heat the wires with a hot air gun to help the paste go off. It would do that in time anyway.
Does what it says on the tin: "Problem solved".

Saturday, 29 June 2019

CRC Rebuild - The Riddle of the Seals - Solved!

If you've read my 29 September 2018 post about rebuilding a CRC unit, you will know that my repair kit seem to have an extra 'spare' seal in it. Surely that wasn't right?

I'd done a pretty comprehensive strip of my CRC, but had ducked-out of pressing out the shaft with the cam on. I couldn't really see how it came out or went back together, so thought it best not to disturb it.
Strip Down

Preparing for the rebuild, I started to account for the parts in my kit. I had a spare 'O' ring! 
Where did the mystery seal go?
Was it a mistake in the kit? I checked a few other kits and they too seemed to have one seal more than I could account-for. In fact the kits varied greatly: some included the hook ended springs, others didn't. The CRC on an IE car doesn't have the flow restrictor ball bearing - though at least one of the kits I found included a bearing. It was all a bit of a lottery.
Seal kit bingo.......
Having accounted for everything else, I was pretty sure that the seal was something to do with the rotating shaft, though I couldn't think why an IE car would have the seal but a carb car wouldn't. The shaft seemed to be the kind of thing that should be sealed, but I would have perhaps expected it to have two seals - one either end? However of all the kits I looked at, I could only find one 'O' ring without an obvious home...... 

I made some enquiries about the differences in the kits and had some interesting ideas in response. One person I spoke-to thought the extra one went onto the end of the pipe that the return hose is connected to. Another thought it was an 'anti-rattle' fitting that went between the CRC and the inlet manifold. Another just thought it was included as a handy spare (for what?).

Unconvinced by these ideas, I left the last, spare 'O' ring off. I got the CRC all back together again but that extra 'spare' 'O' ring niggled me. 



It must go somewhere? I carefully put it away in a box. I was keen to get to the bottom of this and shortly after my rebuild I got a response to the thread I had started  over at 'Aussiefrogs'.  Firstly 'Myourd' posted that he'd knocked the shaft out of a CRC and had found a seal. A day or so later Michael ('Michaelr') posted a photo. It solved the riddle of that last seal!

Here's the photo. You can see that the single, black seal fits into a groove on the rotating shaft.
Location of the mystery seal revealed!

So - how to get at that seal? There is a bush on the end of the shaft that does not connect to the carburettor - so the other end. It looks as though the bush (on the right hand side of the picture above) has been pulled off the shaft with mole grips or similar. It looks quite chewed up. More recently I have seen photos of work done on the CRC by 'Timo'. It looks as though he welded a length of threaded rod to the end of the bush and then, by rigging up a nut and spacer around the bush with a piece of tube......
Using a nut to wind off the seal bush (photo by Timo)

.......slowly wound the bush out. Very much like pulling the cork out of a bottle of wine. You can see the bush and welded rod at the bottom of the photo below.
Bush with welded threaded length (photo by Timo)

With a bit of cleaning up, I reckon that bush could be re-used for the rebuild after replacing the seal.

Going back to Michaelr's photo, you can also see that the 'O' ring seal ion the removed shaft looks quite worn and flat. Maybe it leaks.... 

The spare CRC I'd bought at Citromobile was covered in silicon sealant and it was very likely that was because it leaked in some way. Maybe because it lacked the gasket between the two halves, but maybe that difficult-to-get-at last seal was the culprit?
Smeared-on silicon sealent points to a problem of some kind....
The photo from 'Michaelr' was also helpful in another way. Now when people say they need another CRC, they often say the reason why is that the rotating shaft has become stiff or 'knotty'. If you turn them by hand, they do get a bit knotty and 'lock' - reluctant to return. Here's  a link to a video of someone demonstrating this:


Anyway, when the CRC is fitted to the car (on a carburettor car at least), the return springs of the throttle pedal help with that 'return' action. However, it wouldn't hurt to make sure there are no flat spots on the cam.

From my own rebuild, I noticed that the cam was slightly worn and wondered whether it had developed a flat spot and whether it would perhaps benefit from a polish. However as I wasn't confident about getting the cam off, I had to leave that thought there. Seeing Michaelr's photo now, helped to demystify this part of the puzzle for me: I could see that the cam fits to splines on the shaft.Not so sure about that chewed-up part on the right side of the photo though.

As I've got a spare CRC, perhaps I'll buy another rebuild kit and go the whole hog this time - replacing that 'spare' seal and polishing the cam.

Saturday, 15 June 2019

Textar Brake Pads - The Gentleman of Verona

A very exciting parcel has arrived from Italy! I wonder what’s in it?
I love a good parcel - especially when it's from abroad and DS-related
Actually I know what’s in it, I’m just trying to build some mystery and suspense.

Now if you look elsewhere, and can find them, a set of genuine original Textar brake pads is going to cost you about £250. When I was over at the Citromobile thing in Holland back in May, I found a set of Textar brake pads for just 80 Euros !! 
Lovely Textar brake pads.......
That part’s true, but there is a down side (I was just trying to build suspense again). The down side? It was a part-worn set. But they looked as good as new. They’d come from a car that had crashed.

The seller was Italian. Several attempts were made at a joke with the seller along the lines of “I hope the brakes weren’t the cause” but it was completely lost in translation. Anyway, genuine Textar brake pads for 80 Euros - bargain! I was literally counting out the money to buy them when I suddenly spotted that they didn’t have the brake wear warning light wires fitted - did you spot that? They still had the holes for the wires, but they had not been fitted and instead glue or filler was squidging out of the holes. 
Textars: with and without wires for the brake wear indicators......
Prior to the 1969 model year, IDs did not have the wiring to the dash for the brake wear indicator in their loom - it was a DS-only thing. So not all pads needed to have the corresponding wires in them.
Pre-69 ID loom without brake wires
Pre-69 DS loom with brake wires
So no wires. Money hadn’t changed hands. The pads would have been absolutely fine, but I decided to hold on to my money and keep looking for a set with the wires. The deal was off! (more suspense building again). 

I came back to the UK and, as is my way, realised that I could use those Textar pads after all……..  I realised that I could carefully remove the wires from my dead set of pads and re-fit them in the 'tunnels' of the pads I’d seen.
A plan to re-fit the wires.......
The wires are a non-critical component of the pads themselves. They don’t come under any strain or load. They simply sit there, then trigger a 'wear' warning light once the wire inside is earthed against the brake disc. The wire lies in a groove between the backing plate and the pad - a tunnel - and looks as though it is simply held in place by whatever glue is used to bind the brake pad to the backing plate.  From what I remember seeing at Ciromobile, I was gambling on the mould for all Textar pads having the wire groove in the back. I’d just need to carefully clean out the tunnels with a twist drill. I decided I could then use exhaust sealing paste as the glue for the new wires: it’s non-sticky but dries hard and is able to withstand high temperatures. It crumbles to a dust when put under pressure or scraped.

I tested the tunnel theory. I didn’t want to drill into the pad itself - for fear of weakening it. I just wanted to clean out what I was sure would be existing holes. One of my old split pads served to help me align the drill to the correct angle of the tunnel - which was helpful. What I found though, was that the body of the brake pad would block the chuck and would prevent the drill bit going deep enough to clean the holes out.....
The drill isn't long enough
I did a bit of research and found some extra long drills over in China. They are probably not the best quality but only have to do a simple job and are very cheap. Problem solved!
Solution: extra long drills......
Okay: I had a plan. I just needed to put it in to action. The parts seller at Citromobile was Italian and I vaguely remembered where his stall was. Using the programme I tried to work out who it was. 
Citromobile 2019: the stall was somewhere around here. The suspense is killing me......
I remembered an Italian guy who was always at Citromobile and always at the same place - and approximately where I remember seeing those brake pads. Stall 107 - Citro Services! I Googled them and sent off an email.

Wrong guy……I looked more closely at the Citromobile floor plan......I did a bit of digging and came up with a possible alternative name: Perencin - stall 145. He sold DS parts. There were no contact details in the Citromobile programme and a Google search of the given trade name - "Citroen DS Parts Perencin" - drew a blank - other than confirming a probable Italian connection for the name in general.

My research revealed that a "Riccardo Perencin" had also had a stall at the ICCR up at Harrogate a few years ago, but again no contact details. Even though I was narrowing in on a name, I was still struggling for contact details and Riccardo seems to have a low internet profile. I think the breakthrough came from some old forum posts on a German DS site. Riccardo is based in Verona. More digging and I finally came up with a possible email address.

After some heavy use of Google translate and few emails, the deal was back on! The purchase was made and the pads have now arrived. I’m very pleased with them. They have just over 14mm of pad depth on them. I compared that to a brand-new, un-used set of Ferodo pads that I have: they have less than 12mm pad depth from new.
Textar (on the left) still has good pad depth
And, just as I remembered it to be, these Textar pads do indeed have the little holes where the brake-wear indicator wires go - they’ve just been filled in. 

You can see the hole where the wear indicator wire should go
In all other respects, they’re the same as my old Textar pads and carry the same part number.

My drill bits are currently on a slow boat. As soon as they arrive  I will set to work fitting those wires. I’ll keep you posted.

Sunday, 28 April 2019

A Quick Word About.......Brake Pads

Brake Pads
Replacement front disc brake pads are available at a reasonable price but people in the know (and with very long memories) say they are not as good as originals. As people who know about these things, I'll let Citroen Classics tell the story. From their website HERE:

Textar used to make the original brake pads for D models and these were sold via Citroen and later via the aftermarket. They were a bit pricey, but by far the best brake pad available.
Original Equipment - 'Textar' brand brake pads
An inferior alternative came on the market at about half the price and so many people bought the cheap alternative, oblivious to the shortcomings of these pads and only seeing the price, that demand for the good Textar pads dropped to the point where Textar stopped making them. Now people who actually drive their cars and notice if the brakes are working properly or not and who want/need the good pads cannot get them.......

So there you have it.......

From my longish memory, in the 1990s Textars were about £80 and the cheap replacements were about £40. £80 in the late 1990s is equivalent to about £140 now. If you can find a set, Textars will cost you about £250 or more - so you can see how their scarcity has pushed prices up........

So what would you get for your money? In particular, the Textars are felt to be superior as their fin arrangements aided air flow and brake cooling - and so reduced the chances of brake fade. The backs on some cheap copies were, and are still not, finned.
Reproductions: poor fin design (photocopyright Der Franzose)
Now it should be said at this point that if you shop around, you can find reproduction brake pads that are (visually at least) virtually identical to the Textars - though of course the brake lining material will have been changed to meet modern asbestos-free safety requirements.
Copies and Textars (photo Copyright Citroen-Andre)
Anyway, I dug out the pairs of pads that had been removed from my car back in 2014. YES – they were Textars! I would have been happy to put those back into service except….when I dug out the second pair I found that the wearing pad surface of one had broken away from the backing plate – and not cleanly either. It looked quite crumbly.The mating surfaces were both grubby – suggesting the break had happened sometime ago and probably when the car was in use.

I must have discovered this when I first dismantled the brakes several years ago – it had just completely slipped my mind. Even if I could find an adhesive to reset the pad, I felt it was just too crumbly to give a good fix – and I didn’t want to risk a pad breaking up when in use.

Maybe this loose pad goes some way to explain the embarrassing and irritating low speed ‘farting’ my brakes would make when the car was stopping and virtually at a standstill (walking pace down to stopped). I do know there are other possible causes for this though.

This is probably a good time to explain how the wear warning indicators work. With the ignition on, a positive signal is fed from behind the dashboard to wires embedded into, and below the surface of the brake pads. My broken Textar is of some use here: 
Back of brake pad showing wear warning indicator wire embedded
When one or other of the pads wears down sufficiently far to expose the wire, it is earthed through the disc to the chassis completing an electrical circuit. This then illuminates a warning light on the dash. This warning light varies according to what dash your car has, but mine would look something like this:
Dash warning light - in this case showing low hydraulic pressure (Photo: Kelvin Ming)
I had picked up a set of NOS (new old stock) Ferodo brand pads many years ago – but these were the ‘ID19’ version without the electrical wires that connect to the ‘brake pad wear’ warning light. They have lovely fins, though I'm just noticing that while the pads themselves are the same size as the Textars, the backing plates are a different shape and bigger. it remains to be seen whether they will fit. I suspect they are for a pre-1966 car.

I checked my other gearboxes but no spare pairs of pads. They had already been removed. Bugger.


I can either:

- Use my Ferodo pads without ‘wear warning’ wires (assuming they fit)
- Try adding wires – but risk weakening the braking material in the process
- Buy some modern replacements with wires
- Find a part worn set of Textars from elsewhere
- Seek out some very expensive Textars.
In My Dreams: A Lovely new Set of Textars
It’s good to have choices. On the serious side, I’ve got time to explore this: swapping up pads at some point down the line should be straightforward.

Sunday, 21 April 2019

Front Brake Discs - Checking Runout

Checking the runout - checking to see whether the front brake discs are warped.

Measuring the Runout with a dial gauge
My front disc brakes looked quite a mess. Quite rusty. Not so much on the rubbing surface, but everywhere else - where they bolt on and around the rim edges. I guess this is partly down to road dirt when the car was my daily drive and partly down to many years sitting unloved in a barn.
Discs - June 2015
Back then I cleaned them up with a wire wheel on a drill and was fairly happy with the results. I didn’t give them a protective layer of oil for obvious reasons. Returning to them now, they had started to show a patina of rust all over again and generally looked a little tired. Replacement discs are widely available and, thankfully on a DS, are surprisingly easy to fit.  However I had heard that, because of poor steel quality, they tend to wear quite quickly. Especially when in combination with modern asbestos-free disc pads.

Another option is to have my current discs grit-blasted. However there was no point in doing this if the discs were already excessively worn or warped. As a first step therefore, I needed to carry out some checks on the discs. Brake specifications are listed in Operation DX. 451-00 in Manual 518 and Operation D.450-00 in volume 1 of Manual 814.

My disks were well within tolerance for disc thickness and with no obvious cracks, grooves or ‘hot spots’ on the surface. The other key test that you find in the manual is for disc ‘run-out’  - effectively the degree of any warp in the disc. Some degree of variance/ warp should probably be expected, but Citroen specify tolerance for this.

The way to test this is to fit a gauge to a stable surface, with it's needle on the edge of a fitted disc. Rotating the dic reveals the degree of warp or run-out.  
From the manual: fitting and using the run-out gauge

First job though, was to replace the studs in the drive shafts. When I’d removed the shafts, half of the studs had come out with their nuts. These studs needed to be reset in the shaft flanges. I double-nutted them on the opposite ends to the stubborn nuts to give me some purchase. The studs were cleaned up on my wire wheel and degreased. I ran a small wire brush over the female threads and followed up with a nylon brush dipped in isopropanol alcohol.

With the studs and threads de-greased, I applied threadlock to each and double-nutted them back on the flanges.
Studs cleaned and refitted
I made sure there were no contaminants (grit or burrs) on the surfaces where the discs would bolt. Now I was ready.

To test run-out I made up a simple rig to hold my dial gauge. I used the template for tool MR.630-52/21 already included in the Manual for this purpose. You can find it in the ‘Special Tools’ pages at the end of the first section of Volume 2 of Manual 814.
Bracket for measuring run-out with a gauge
I bought a length of 4mm steel bar from my local DIY store and bashed it around a bit in a vice. I hade to use an angle grinder to narrow the end where the gauge fitted so that it was reversible and could be used on both sides of the car.
Homemade bracket for run-out gauge
Gauge Fitted
This test was quite time consuming……..Following the instructions from step 23 onwards of Operation DX.330-4 of manual 518, (or Step 9 of Part II "Removal and Refitting a Brake Disc" at Operation 343-4 in section 1 of Volume 2 of manual 814) the discs were fitted to the differential shafts. The discs don’t bolt directly to the differential shafts. Instead they are sandwiched between the drive shaft and the flange of the differential shaft. Fitting them now, I had to use several washers on each stud/ nut to cover for the absence of the flange of the drive shaft. I tightened each disc down but not to the final torque figure.

Relative to the disc, the gauge needs to be fitted to something that doesn’t move. With a disc in place, the dial gauge bracket was bolted to the top hole where the parking brake calipers would eventually fit. I didn’t have a suitable bolt so used one of the caliper bolts – together with a socket as a spacer. The gauge needs to be set up so that it is contact with the outermost edge of the disc radius on the side of the disc that faces the wheel.
Dial gauge in place on the right hand disc.....
I rotated the other differential. Simply for ease of reading, I found the ‘low’ spot and zero-ed the dial. Rotating the disc again I measured the degree of change in the dial gauge. With the help of my son Tom, I made a video to show this process in action:
VIDEO: Testing Disc Brake Run Out (left hand disc)

I had a fair amount of needle judder (due to the gauges sensitivity and the discs roughness), but (as can be seen) the first test consistently gave a disc run out of about 0.22mm (yes – as much as that!). Joking aside, the Citroen specified tolerance is 0.15mm, so that test was a ‘fail’. The remedy is to remove the disc, rotate it 120 degrees clockwise relative to the studs (and so hub flange), refit and test again. That was the time consuming bit…..Allowing for machining tolerances on the mating surfaces of the disc and hub flange, presumably the aim is to see if there is a particular position (of the three possibles) that gives a reading within tolerance. If you find one - fit the disc in that position.

Before moving the disc, I used chalk to mark the ‘low’ and ‘high’ spots on the disk and hub for ‘test 1’. ‘Test 2’ gave the same results…..As did ‘test 3’…... The chalk marks were in broadly the same areas of the disc circumference – i.e. they moved around relative to a particular stud on the hub - indicating that the warp was on the disc itself and not a high spot on a particular part of the hub. That was good news as I didn’t want to have to pull the differential off again and get it machined.

I did the tests several times - so that’s several circuits - just to confirm my conclusions: run out on the disc and in excess of Citroen-specified tolerance. In this situation, the Citroen manual simply says ‘change the disc”, however the specifications also give tolerances for skimming (grinding) the discs – implying that for some problems, rectification is possible.

I went through the same process for the disc from the other side and found that one to be within tolerance – 0.10mm in all three positions. Just for good measure I then did it all over again trying the discs on the opposite sides! I arrived at a point where one disc was 0.10 and so within tolerance, but the other was 0.22mm and so out of tolerance.

Now 0.07mm (out of) out of tolerance didn’t sound a lot to me! But of course that did represent 0.07mm beyond a given figure – and I guess there has to be some kind of limit. I sought advice on various forums as to how critical these margins were in practice. I got some helpful answers but, as you can probably expect, these fell into both the ‘don’t worry’ and ‘do worry’ camps…..

One option for curing disc problems is to fit special abrasive brake pads. These are available from Der Franzose and so probably the other suppliers too.
"Brake Sharpening" Pads
Franzose say that these are/ were used in Citroen workshops though I’ve not found any reference to their use in in the workshop manuals as one of the remedies for brake disc run-out problems.  To use special pads the car needs to be assembled and running. I guess that they are a crude way of fixing problems and it would be difficult to measure the impact on disc run post-use with everything re-assembled.

In the end I sent my two discs of to my local engineer with the brake units for skimming and blasting. Decisions were needed:

Assuming that my out-of-tolerance disc is of uniform thickness and simply slightly warped, then any high/ low spots really need to be removed from both faces of the disc – to maintain uniformity. It would have been tempting to remove only enough to bring the disc back within tolerance – but that would have had the effect of lopping the top half off the warp ‘hump’ creating a flattened warp peak with a rise and fall both sides. Would that create new brake problems? To prevent this, the whole 0.22mm would have to come off. Same on the other side…..that meant reducing the overall disc thickness by 0.44mm. That was starting to sound like a lot….That was almost half the permissible wear on a disc – meaning that I was reducing the life of the disc and hastening the need for the new discs I was trying to avoid!

More decisions: to give unifom braking performance on both wheels, should the other in-tolerance disc be skimmed – just to give it a clean surface like the other disc?

The answers to these questions will have to wait as my engineer couldn't help me: when skimming discs they need to be cut on both sides at the asme time and he didn't have a big enough adaptor to be able to fit the discs to a lathe.

Saturday, 30 March 2019

Winter Diversions 2 - Learning To Weld

It’s been very quiet around here on the DS front…..

Because of all the building work we are having done, the garage has been emptied and gutted. 
Then......
Now
Still Now
This all started back in November 2018 and there is no way my DS and various parts could have stayed in there while all this work went on. Instead, the car is off site (back at the farm where it languished for so many years). I vowed at the time that the car would be gone no longer than five months. Well that's not working out at all. I think the main building work will take another couple of months and, even then, I want to paint the garage floor and will need to rack out the garage again before I can bring the car back and work on it. At the moment all the parts are stacked in boxes in my workshop……

…..and the wooden shed at the end of the garden. 
Shed
So what have I been up to? Well, I decided to learn to weld! As a DS owner I have a sneaking feeling it might come in handy. I talked myself into it on the basis of being able to save costs by rebuilding my own doors. I was, however fully aware that was a rather skilled art and that I would not be in a position to tackle that job on day one (ed. If at all….). First, I would need to learn some 'craft'. If panel welding proved too tricky, then Plan B was just to use the welder on a number of smaller jobs around the car and for fabricating things. Plan C was that I could re-sell it on if I discovered that me and welding did not get on…..

I won't presume to tell you how to weld because, as a novice myself, I have very little idea. I did think though, that it might be helpful if I explained how I approached all of this. As ever, it started with research.

Research
I  started by eyeing up the different welders that were on the market - new and second hand. I read posts, queries and problems in welding forums, and picked the brains of a couple of people who have restored their own Ds. I reached the following conclusions (Note: other conclusions are also available!)
  • ‘Cebora’ brand welders are well-respected in the hobby/ home, non/ semi-professional market. They have been (still are?) re-branded and badged as ‘Snap-On’ welders with the accompanying hoik in price.
  • The wire spool and feed mechanism is a crucial component for successful MIG welding. Cheaper welders can have poor quality and inconsistent feed mechanisms.
Example of a wire feed mechanism
  • Go for a more powerful welder (more amps). I’d seen dirt cheap, low amp, welders that are only able to cope with the thinnest of metals and wouldn’t be able to cope with some jobs I might want to undertake.
  • Go for a welder that provides for a wide current range and that allows you maximum control over that range. I’d seen cheap welders that had little more than an on/ off switch – so no flexibility to cope with different thicknesses of metal.
  • Similarly, go for a welder that allows you maximum control over wire-feed speed.
Welding current (1 -6 dial) and wire speed (1-11) controls
  • Check the cycle time/ duty time as stated on the spec. plate on the welder. This tells you how much resting ‘down time’ there will need to be between bouts of welding.
  • MIG welding needs gas. Many people see this as an unnecessary inconvenience but the gas is there for a reason - it's not optional. It provides a shield around the weld and ensures that the weld is neat, strong and fit for purpose. The give-away on any prospective welding machine is that the welder will have gas hoses and a pressure regulator valve. You can buy welders that don't use gas (instead they use core-fluxed welding wire) - but that's not really MIG welding. So go for gas! Get a gas welder and you have a choice.
  • There are different regulators/ valves for different gas bottles. You will need the right one for the gas you intend to buy/ use - otherwise you will need to buy an appropriate replacement regulator for your welder. That's pretty straightforward, just more expense.
Older style regulator - but suitable for MIG Argon mix and for standard BOC bottle fittings
  • Buying gas under contract is not cost effective for someone with my planned usage. Research local gas suppliers. Where will you get yours?
  • Different welders use different wire thicknesses. Body panels are only 0.8mm thick (!)  so, for body panel welding, you will typically want to use lower currents and so a welder that can handle the thinner 0.6mm welding wire. However you may also want to do larger welding jobs so need a welder that can also take 0.8mm wire. 
  • Some modern welders have a ‘tack weld’ time function that allows you to have a short controlled burst of weld. Handy for ‘stitching’ thin panels together and for reducing the risk of ‘blow through’ from applying heat for too long.
  • Buying a welder is only the first step. There will be lots of things you then need to source/ buy before you can weld. Most importantly: personal protective equipment (PPE) but also welding gas and other ancillaries and consumables. Leave some budget for these.
MIG Welding Machines
I bought a MIG welder off Fleabay at Christmas.  I ended up with a Cebora ‘Autostar 180’. It's probably 20 years old but in very good condition.
Ebay purchase
The same welder is also marketed as a Snap-On ‘Pro-MIG 160’. The internals are the same but the badging is very individual and only an idiot could fail to tell them apart.
Two brands, one welder.
With a name like 'Autostar 180' it would be tempting to think that you were getting a welder with a max welding current of 180 amps. That's not the case. The actual max welding current is 145 amps. So don’t be misled by a name.

But 145 amps is not bad. I should be able to weld metal up to 6mm thick. Maybe 8mm at a push. There are many small, cheap welders badged as ‘130s’. The actual welding current of these may only be 90 amps, or less in some cases. Added to this, they can have poor duty cycle times, a very narrow operating range and very limited control of how that range is adjusted. If you're shopping for a welder, all the clues are marked on a plate. This tells you about the operating range and other characteristics. Allow me to translate:
Welder Technical Specification Plate
    • The '1' in the top row tells you it has a 'single-phase'  transformer/ rectifier - meaning it can be used on a domestic circuit and has a standard three-pin plug fitted. '3' would mean 'three phase' and require significant re-wiring of your house.
    • The table below this gives you an indication of the welding potential of the particular machine. Related to 'duty time' (see below), the table shows it's maximum capability in one column and in another the power level at which it can comfortably operate without any downtime (that's the '100%' column)
    • The top row of the table  tells you that the voltage and current draw is between 20 amps/ 15v and 145 amps/ 21 volts according to which setting you select (so the maximum welding current is 145 amps).
    • The I2 figures state given welding currents and include the max. value (145 amps)
    • Correspondingly, the 'X' figure and the %s,  tell you what proportion of 10 mins the welder can run for at the corresponding current setting - so mine can operate for 2 mins 30 secs at 145 amps before it overheats. After it's duty time, it needs a 7 min 30 second rest. It also shows you the maximum current it can weld without a break: it can weld all day long (100%) at 75 amps. For welding thin body panels, I expect to be making lots of short tack welds  - so there will be plenty of down time and the duty cycle should not be an issue.
    • The U2 is the corresponding voltage for the approriate welding currents.
    • The I1 figures tell you the corresponding draw on your domestic supply. 
    • The U1 value shows that this model is for the UK market and so a 240V domestic supply.
    • 'IP21' refers to the strength of the case.
    • The big 'S' at the bottom tells you that the welder can take a little bit of dampness but should not be used in the rain.
As you can see, mine has a range of between 145 amps and 20 amps with six settings dividing max. from min. - so a fair degree of flexibility in the settings. The duty time is also more than acceptable for what I think I will need. (If I'm welding body panels, I won't be expecting to weld long beads or the panels will warp). It has a fixed torch (fixed into the body). More modern welders have a ‘euro’ adaptor and the torch unplugs from the welder. It’s designed to handle 0.6mm or 0.8mm wire. The wire feed mechanism seems sturdy and the wire speed control goes all the way up to "11" !! It looks used but not abused and I’m told it was bought secondhand from a body shop – so has probably been used for exactly the same job I plan to use it for. All in all it seems to tick the right boxes. I’m not kidding myself: better welders are available but I think I’ll be happy with mine. 

Further Necessary Expenditure
PPE (personal protective equipment):

  • ‘Auto-darkening’ welding helmet. Very common these days. A bit like polarising sunglasses, these ‘black out’ as soon as the welding arc is struck. When the weld stops, the viewing panel clears. Simples! The difference is that they are either battery or solar powered. Check out the reaction time from light to darkened. Mine will change from light to dark in 1/25000ths of a second. That's a fixed speed - not adjustable. Others can be faster. 
Auto-darkening welding helmet
The helmets have other adjustable features and so tend to have a series of controls inside them. 
Helmet set up controls
Look out how sensitive the mask is to detecting changes in light - i.e whether you can adjust the threshold that triggers the darkening action.  Better helmets have more sensors - which mean, whatever the angle of your head, they are better at sensing changes in light and the need to darken. Mine has four sensors. Also see how long the delay is before the window becomes light again. This can be anything up to a second. Although you want the helmet to go dark quickly, you don't want it going light as quickly as a bit of delay helps protect your eyes as the weld cools. Make sure the helmet can be adjusted to the right level of shade needed for MIG welding – you want something including a range encompassing "shade 10" to "shade 13". (On mine, the 'mode' control is a just a quicker way of adjusting shade). Also, try to get one with as big a viewing window as you can afford. The window in mine is 100mm x 73mm. Doesn't sound a lot, but that's quite generous.
  • Long leather gloves to protect not only your hands, but also your wrists – preferably with a degree of heat insulation. If you’ve got big hands – beware. Even gloves marketed as ‘XL’ size can be a tight fit – making them uncomfortable and impractical.
  • Goggles or glasses. With molten metal flying around, you may want to also wear safety glasses or goggles in addition to your helmet.
  • You might also want to think about a breathing mask. (See further below). 
  • Footwear. Wear some. You don’t want molten metal on your M&S slippers.
Other Equipment
I also bought some tools and some ancillaries:

  • Spare welding tips for 0.6mm wire
  • Spare welding tips for 0.8mm wire
  • A reel of 0.6mm welding wire 
  • A pair of welding pliers. A little extravagant. The main use is for trimming the ends of welding wire, but they are also shaped to help remove hot welding gun tips and to clean spatter from inside the welding shroud.
    Welding pliers
  • Joddlers
Joddlers
I bought a pair with two functions: they can be used for creating a slight offset edge to a panels for overlap seams.....
Joddled Edge
......and they can also punch holes for plug welds:
Punched Holes for Plug Welding
  •       A second cheap angle grinder - again, a bit extravagant, but it means that I can have one fitted with a thin cutting disc and one fitted with a thicker surface cleaning disc. (Thanks for the tip Peter!)
  •       And of course I bought gas.........
The main gas used in MIG welding is Argon. A typical mix especially for MIG welding will be argon (95%) – with a little CO2 mixed in (5%). Some suppliers supply a mix that has 3% CO2 and 2% oxygen.

To reduce cost, some people, on friendly terms with a pub landlord, use pub gas. In other words cellar gas. BUT you have to make sure you use the right one - pure CO2. Some pub gases are a mix of gases to stop the drink (bitter for example) being too fizzy. You don't want to be using that gas....... and bitter shouldn't be fizzy anyway.

BOC (British Oxygen Company) might appear a first choice for a supplier of gas bottle, however their business model doesn’t suit the hobby welder who may only want to weld once in a while – or indeed may have a specific short term project and no requirement thereafter. They require you to set up a contract. Their prices are high and their terms include ongoing costs of bottle rental - whether you are welding or not. 

Instead I went with ‘Adams Gas’ – a hobby supplier. That have outlets all over the place. My nearest one happens to be a farm on the outskirts of town. A deposit needs to be paid on the first bottle (this is refundable if/ when you return the bottle) and of course you pay for contents – but that’s it. No hidden charges. Get a bottle and use it as quickly or slowly as you lilke. No contracts involved. You can buy small 9 litre bottles, but I worked out that a 20 litre bottle offered good value.

I still need to buy one (or two different) pairs of snips and several clamps for holding pieces of steel plate together. I'm sure there will be other things I need - such as hammers and dollies?

First Tentative Steps
I looked into welding courses but there were none near me and they were expensive. With the amount I'd need to pay for a welder, ancillaries, consumables, repair sections and the cost of the course, I could have just got someone to repair my doors for me! So I decided to try to teach myself. Before I even turned on the welder, I watched a good number of videos on Youtube. No matter what video you find, they attract very mixed comments in terms of criticism and the right/ wrong way to do things so, while I picked up some basics, I decided that the best way to learn will be by doing……

Although it’s very unlikely I will be running beads on body panels, I started by practicing running some beads on scraps primarily just to gain confidence in handling the equipment but of course also to see what power and wire feed settings worked on different thicknesses, and to see how the torch position and movement speed affected things. I also tried some deliberate ‘fails’ – just so I could understand how to correct errors (that’s my  story and I’m sticking to it).

Very first efforts…….
Very First Efforts
I got a bit braver.......
One Thing badly welded to Another Thing
Ultimately I will have to aim to become a lot more skilled at this if I am going to be welding thin body panels as this is fraught with welding difficulties:  firstly in terms of the thin metal heating, warping and distorting as you try to join pieces, and secondly from the risk of the weight of the molten weld falling through the panels you are repairing – leaving a bigger hole than you started with…….

To get some practice in, I got hold of a dented car door from a skip (just happened to be a Citroen door!). As I needed to weld to clean, bare metal, this door also gave me the opportunity to experiment with various paint strippers and grinding wheels.
Practicing on a Real Thing
I deliberately ran the welder at too much current to find out what ‘bad’ looked like and how ‘blow through’ appeared. "Blow through" is where the weight of the molten metal falls through the welding surface. The light pressure of the shielding gas may play a part, but I reckon gravity is the main culprit. It looks like this:
"Blow Through"
In hindsight, practicing on this door was not such a good idea! It is lined with sound-deadening material and possibly zinc* and has lots of plastic bits tucked away inside. The heat from the welding soon started to generate some very noxious fumes that were clearly unhealthy.

·      * Do not weld zinc plated metals! Remove the zinc! Exposure to zinc-oxide can cause ‘metal fume fever’ – which has flu-like symptoms (so a.k.a. “Welders flu”). And we all know how bad Man Flu can be eh, men?

I've now bought myself some 0.8mm sheet steel. I'll be using some of it to practice on and will have plenty left over for any patching or fabrication that might be needed on my doors or elsewhere. Though the plan at the moment is to fit replacement door bottom kits and buy complete new door skins - to avoid seams and moisture traps if water runs down the inside of the door.