Monday, 25 November 2013

Garter carriage problems all solved! Thank you The Answer Lady and Jack

Now then, why did I never find this wonderful resource before?  The Answer Lady and Jack have a series of youtube videos explaining exactly how to fix problems like the ones I was having with my garter carriages.

In short, my 'thorough' cleaning hadn't gone far enough.  The gunge that was used as lubricant when the machines were first made had become thick and was clogging the mechanism and preventing it from operating smoothly.   Once this was all cleared out of both machines they were re-lubricated and are both now running like new.

Details are in The Answer Lady and Jack's highly recommended videos.

Friday, 25 October 2013

Simple knitting machine cover

Just a tablecloth with an eyelet for the mast. 

Tuesday, 20 August 2013

Replacing the fuse in a garter carriage motor

In my experience, these fuses can and do fail, but never (so far) when the carriage is actually in use.

The first situation is where the power is on and the tip of the DC connector is allowed to touch metal (for example, various parts of a knitting machine!)  This is why instruction books give a very specific sequence for connecting and then powering the equipment.  The reason is that the tip is not recessed, as it probably should be, creating the potential for a short circuit.  So, if you don't like playing with fiddly power supplies, be careful.

The second situation is where the machine is left plugged in for a period while out of use.  This turns out to be because of a process called 'electromigration' which you are welcome to look up here if you wish!

In both cases, the fuse has to be replaced and this is thankfully fairly easy. 

 The power supply box is held together by three screws on the top.  

The holes may be sealed with small rubber plugs which can be rooted out with a small crochet hook or similar. 

Then the screws can be taken out with a long-shanked crosshead screwdriver.  

Replace the fuse (on the circuit board near the output cable).

Use a 2 amp miniature fuse.

 Reassemble.  Job done.

Monday, 8 April 2013

Fixing the timing on an overlocker (serger)

Like Lorna at I did something unspeakable to my overlocker.

While happily sewing away the other day, there was a crunch which resulted in the right needle breaking.  The needle had come down right on top of the metal hook that sits in front of the lower looper in my machine.  For the record, it's a Frister+Rossman Euro-Lock 4DS MkIII, but I believe many other models are very similar.

I have no idea why this happened, and like Lorna I found precious little help online.  There was lots of general information about buying and using sergers, but as far as servicing guidance was concerned, very little.

As I have said before, marrying an engineer has its benefits.  Not one to be beaten by a mere lump of metal, mine eagerly took up the challenge and identified four potential problem areas which we have indicated in the diagram below:

1. The needle position in relation to the loopers might have shifted.  Maybe the wheel which drives the needle up/down had slipped on its shaft.

2. The depth of the lower looper arm on its shaft seemed to be much less than might have been expected.  Additionally, there is the angle (theta) by which the arm might have rotated from its correct position.

3. There is the angle (phi) by which the upper looper might have rotated from its correct position.  This would determine the separation between the upper looper and the needles as they pass each other.

4. The wheel which drives the upper looper might have slipped on its shaft, changing the position of the upper looper in relation to the other elements of the overlocker.

The innards of the machine appeared to be pretty robust, so options 1 and 4 were set aside, at least for the moment.

Options 2 and 3 each had possible sources of mis-alignment.  In the case of 2 the angle (theta) is critical, as is the depth of the arm on the shaft.  Either of these could explain why the right needle hit the hook.  So, we slackened off the nut that secures the arm to the shaft and tried it in various positions without success.

While the lower looper arm was so loose that it was completely 'out-of-play' we observed the interaction between the upper looper and the needles and here we found what appeared to be the entire cause of the fault.  The drawings below show what happened as the upper looper moved across the needles.  So close was it to the needles that as it continued its journey the back of the looper bent the needles so that the needle on the right hit the hook as it went through the face-plate.

Slacking off the bolt on the top of the upper looper shaft and drawing the looper slightly away from the needles completely solved the original problem, but the timing was now completely out.

We now got some help from but what was even more useful was an excellent little video at

The creator of this video takes two scenarios.  In the first instance, she considers the positions of the loopers when the needles are at top dead centre.  By luck the inside of her machine looked just like ours and it was immediately clear that the upper looper was in its correct position.  As you turn the handle towards you the needles start to move down and the upper looper starts to cross them.  As recommended in the eHow reference the distance between the needles and the looper was set at about 1mm.  This is shown in the photograph below.

The second step in the video considers where the lower looper should be when the needles are at bottom dead centre.  Initially we set it to the right of the position that was recommended.  This worked fine for three thread operation (left needle removed).  However, it did not work at all when all four threads were used. Success came only when we rotated the arm so that the back of the looper was in the position relative to the cutter depth control shown in the next photograph.

At the same time we were attempting to optimise the depth of the lower looper arm on its shaft.  There are various factors here.  One concerns the way in which the hook on the looper arm restrains the needles when they are fully down.  The apparent purpose of the 'hook' on the lower looper arm can be seen in the photograph below.   The needles are temporarily trapped so that they don't move when the dogs move the cloth.  If the arm is too far out along the shaft then it will foul the feed-dog support.  If it is too far in then the loopers will foul each other.  At one point we thought that we had got everything positioned correctly, but once the overlocker was running we could hear the loopers clicking against each other, so the depth of the arm on the shaft had to be eased off slightly before the locking bolt was tightened.

At last we had the machine running perfectly, probably better than it ever had before.  Long may it remain so!

Thursday, 4 April 2013

Fish hook weight for machine knitting

This is just a large fish hook with the barb rubbed off, tied to a 1 oz. weight with fishing line. It's brilliant for slipping very accurately between stitches, even on the row just knitted, and it never gets caught in the carriage.

Thursday, 17 January 2013

Garter carriage blues

My usually reliable workhorse garter carriages have been playing me up.  After taking a very deep breath and abandoning any thought of actually knitting something, I tried to narrow down the problems in a systematic way.

Internet searching yielded a service manual here and an excellent set of hints and tips here, but nothing seemed to be relevant to  the health issues of my particular GCs.  I checked the needles, magnets and brushes, all seemed fine on both the GCs and the main bed.  I tried different yarns and tensions.  I tried the same card using fairisle.  All of this led to the solution of a few minor niggles (and beautifully cleaned machines!), but I was left with three intractable problems so I called in reinforcements.

Problem one (KG88) solved: the wrong needles were being selected

Himself is never truly happy without a screwdriver in hand, so he was quick to suggest a strip down of one of the knitting machines, which is a KH 881.  We set up a row of yoghourt pots to hold the  bits and removed the plastic cover for the card reader and the built-in knitleader, so it was possible to see the whole length of the needles.  The problem was immediately obvious.  Although they looked fine from in front one, just one, was bent at the back end.  I must have replaced this sometime in the past and not noticed it was bent.  While the selection mechanism for the main bed was apparently able to cope, the GC could not.  Problem solved.  A new needle was an easy repair, but the thing had to be reassembled.  Too many little screws!

Problem two KG 88II not solved: the GC would not turn at the left-hand end of the row

One garter carriage was now working properly.  Unfortunately, this was the KG88, which does not cast on.  The KG 88II was still sick.  We partially dismantled it and found a botched previous repair, but tidying this up had no effect.  It had to be the pawl, or 'feed change plate' as it is called.  The service manual simply said 'replace the feed change plate' but that isn't an option these days, so we had to try a repair.  Although it was (so far!) unsuccessful, I have documented what we did in case anyone finds it useful.

Figure 1
We had the idea that the springs inside the reversing pawls might be 'tired'.  The way the garter carriage chunters along impulsively, we hypothesised that the pawl might jump the first out-of-work needle, if the spring were not strong enough.  Figure 1 shows our attempts to determine the weight needed to push down the pawl against the spring.  An old baby bottle with millilitre (ml) graduations along the side was suspended from the pawl.  The bottle weighed 30gms and as each ml of water was injected into the bottle, it added 1gm to the weight acting against the spring.  In the 'good' garter carriage the weights were 50gms on the left-turn pawl (the one on the right) and 80gms on the right-turn pawl (the one on the left).  The weights in the case of KG88II were significantly less: 40gms on the right-turn pawl and less than 30gms on the left-turn pawl!

Figure 2
Now what are the chances of being able to replace the entire pawl?  It just so happened that we had a spare pawl and so any experiment to improve the performance of the spring could be piloted on this one.  So, we decided to remove the plastic part to expose the spring.  This is held in place with a pin that is very like the pins that hold a bicycle chain together, but only a lot smaller.  In the case of a bicycle chain it is possible to buy a pusher tool, but in this case, we were on our own.  The tools that we used were: a small centre-punch, a small nut and a small hammer.  The nut shown in figure 2 is perhaps a little larger than it should be.  The objective is to punch through the pin without distorting the steel frame beneath it.
Figure 3

The component parts are shown in figure 3.  The spring can just be seen on the extreme right

Figure 4
Figure 4 shows that the spring has two bends.  A small-nose pliers was placed below the lower fold and that part of the spring was held firmly while the spring above the fold was pushed gently upward.  The process was repeated at the second bend, so that the final position of the spring was much higher than before the exercise was started.

Figure 5
Reassembly required putting the plastic in place over the spring, lining up the holes and inserting the pin.  This is made easier by using the centre-punch as a 'drift'.  As the pin was hammered in, it displaced the centre-punch (Figure 5)

Once the pawl was reassembled and inserted into KG88II the force required to displace the spring was measured and found to be 100gms.  The garter carriage was then set in operation and the left-turn pawl worked perfectly for a while before failing once again.  So our next step will be to re-test the springs.

Someone also pointed out to us that there was originally a spiral spring connecting the two pawls.  The 'good' GC has never had one in my ownership, so I'm not sure if this is important, but it is probably worth fitting one anyway.  Who knows?

An oddity:  Since I wrote the above I've noticed that the problem never happens on 'ordinary' rib, just while knitting patterns ...

Problem solved!  See post of 25th Nov 2013

Problem three (KG 95) not attempted: the row count-down and stop button were non-functional

Never one to learn from experience, I bought a third GC to work with my my lovely new KH 965i.  It went back to the seller because it would not count down properly and the red stop button was non-functional. We could have attempted a repair, but chickened out and bought a KG93 which, I'm relieved to say, works perfectly.