What does this have to do with photography? Nearly nothing. The only connection I can make is that sometimes a thing you hear about seems like a really good idea and you run out pursue it. I’ve done that with lenses that I thought would be fun to shoot with but I rarely use, or a big custom, external battery I built that I’ve needed exactly once, and lots of little accessories that sit in a drawer. The latest episode of this occurred when I “upgraded” my old Singer 403A sewing machine to add some useful features that I saw in a YouTube video; except that they do not work at all. I’m putting the blog here because it can be found and other owners of these old, but still common, machines might find this useful. As for photography, think about that new gadget, you so dearly want, for a while - you may realize you don't need it at all. 

I’m not a big-time sewer. I’m a sailor. That means that occasionally I need to do some sewing and typically I sew things like Dacron sails and Sunbrella canvas. If I had to do a lot of this work, I’d invest in a proper machine like one of the Sailrite machines. But these cost $800 to $1,300 and that’s a lot of money to do occasional work. So, I’ve settled on an old Singer 403A. For the most part, this machine gets the job done even if a little more power could be used at times; it can’t punch through more than about six layers of Sunbrella or four layers of Dacron.

I’ve often wondered if there was a motor upgrade for this machine that would give it a little boost. Despite looking around for a solution, I never saw any options until I watched the YouTube video How to Upgrade a Singer 401a, 403, or 404 Motor Using a Transplant from a Singer 620. In this video, the presenter tears apart a Singer 620 and puts the motor and high/low speed selector into an earlier Singer 401. The 401 and the 403A are essentially the same machine, so I knew this modification could be done to my machine just as easily as to the 401.

In the video, the presenter claims that the 620 motor is more powerful than the 40x motor. Further, she claims that this motor has a “third phase” that allows the motor to operate at half speed but with full power. In a post-upgrade demonstration, she shows how this modified machine sews through twelve layers of denim and thick leather with ease. I was impressed – gotta find a sacrificial 620 and make these modifications!

In the process of finding a sacrificial 620, I started finding more information about this “hack”, which made me doubtful of how valid this upgrade really was. On the question of the motor upgrade, the stock 40x motor is a model PA9-8, which is rated at 0.70 amps. The motor from a 600 series of machines is a PAJ23-9 and it is rated a 1.00 amps. (A note about these motor numbers: They are all over the map and there does not seem to be any resource that definitively states what the power of the motors are or which machines they were used in. I’m discussing the two I got involved with and they seem pretty standard.) An increase from 0.70 amps to 1.00 amps would be a 42% boost in power. That should result in a significant increase in machine power.

The high/low feature, which the presenter claims is from a special circuit in the 620 motor, is even more puzzling. It is highly unlikely that any of these Singer motors have a “third phase”. All these motors are brush-type, universal electric motors capable of running on AC or DC power. They all have only two brushes – one positive and one negative. In the comments section of the video, a couple of posters commented on the presence of a diode soldered to the back of the high/low switch of a 620 machine. A diode constricts electricity in one direction. Placed on an AC circuit, this has the effect of turning the alternating current into a pulsed DC current. This would result in the machine running more slowly, though not necessarily more powerfully.

Armed with this information and a degree of skepticism, I continued to try and find a sacrificial 620. It quickly became apparent that a 620 was not that easy to find but PAJ23-9 motors, available from many of the 6xx models, are plentiful. I gathered some information on an electronics forum that steered me to a diode, Kyocera FSF05A40, I could use to create the high/low feature of a 620. I also purchased a copy of the 620 service manual, which includes some (very rough) electrical schematics for the 620 machine. 

I studied the 620 service manual and came to some conclusions. There were eleven variants of the 620 machine: 620E1, 620E31, 625E1,E6,E7, 626E1,E6,E7 and 628E1,E6,E7. Only the 620E31 has the high/low switch. You can determine this by the fact that the switch for that machine has four leads: neutral, positive to light, positive to motor and second positive to motor with the diode in line. This explains how the speed reduction is accomplished. The part numbers for the motor of the 620E31 are also different from the other models. There is no top-level motor part number described for any of the machines, so it is impossible to tell which, if any, are the PAJ23-9 motor that I ultimately installed. That the part numbers for these motors is different yields credence to the idea that the 620E31 is special but it is unclear what is different about the 620E31 motor other than their component part numbers. It is readily apparent from the parts diagrams that all these motors are simple two-brush affairs and none of them have a special “third phase” winding. The foot pedal of the 620E31 is also unique. It has its own part number and it has a three-wire power cord. All other models have a two-wire power cord. The electrical schematics are too rough to make any further determination as to what purpose this difference serves.

Before making any changes to my machine, I ran some tests. I don’t have any fancy toque measuring equipment, so I had to resort to gut-call measurements based on performance. Using some new, scrap Sunbrella I increased the number of layers of fabric until the machine could no longer punch through the material. This is subjective as the position of the needle on start makes a big difference as to how much fabric the machine can sew. Also, once the machine has some inertia, it can keep sewing even if it could not initially break through the fabric without assistance on the hand wheel. I use the term “stalling” to mean that the machine won’t turn – just hums - without giving it a nudge to make it go.

Without making any electrical modifications, I installed the PAJ23-9 motor. It fit right in and the only remotely complicated part was swapping the worm gear from the old motor to the new one. (As far as I could tell, they are the same gear but it is prudent to keep the old gears, which have worn together, as a pair.)

Here are the results of the two motors:

Layers                                   PA9-8                                                    PAJ23-9

1                                              perfect                                                 perfect

2                                              perfect                                                 perfect

3                                              perfect                                                 perfect

4                                              perfect                                                 perfect

5                                              perfect                          stall, then punched through without assist

6                                              stall, required assist       stall, then punched through without assist

8                                              stall, struggled after assist      stall, OK after assist        

This imperfect measurement seems to indicate that both motors can easily work with up to four layers of Sunbrella fabric. At five layers they both start to have difficulty, though the PAJ23-9 seemed to do a bit better at eight layers once you assisted it in getting going. If the PAJ23-9 motor is stronger than the PA9-8, it is certainly not 42% stronger. I don’t think the motor swap was a worthwhile upgrade.

The next step was to add a diode and see if it would create the half-speed mode demonstrated in the video. The advantage of this mode is that is allows you work very slowly without having to stomp on the control pedal, which results In the machine “running away” once the needle begins sewing. This would be very useful when sewing heavy material that needs two hands to move around the table.

In my research of this topic I began to realize that the type of foot pedal involved might make a difference in whether this modification would work in my case. The original foot pedals for these old machines is based on a stack of carbon disks. The harder you press the pedal, the tighter the stack becomes and the more electricity the stack can pass – this is a very crude form of rheostat. Because this is a crude rheostat, the difference between off and full power is poorly regulated. This is where the half speed feature becomes useful. However, since the 1950s and 1960s, when these machines were produced, low cost, electronic foot pedals have become available, and these produce a nice, progressive control. I have one of these and discarded the old foot pedal years ago. I realized that it is doubtful that the electronic foot pedal would even work with a diode placed in “front” of it, and if it did work, the effect may not be as dramatic as with the old style pedal. Nevertheless, for about ten bucks I was able secure a few diodes that would perform the speed control. I went ahead and tested this.

The diodes I ordered were a model FSF05A40. These can handle up to 5 amps at 400 volts and should effectively half the power on a 120VAC circuit. When these arrived, I sacrificed an old extension cord and soldered the diode into the power side of the cord. In actuality, it does not matter which side of the cord the diode is on, but it must be orientated correctly to the source of the current. I checked this with an incandescent lamp, and yep, when the light is plugged into the extension cord, it only lights up to about half its normal brightness. The next step was to plug in the sewing machine and test it.

To my surprise, the electronic foot pedal is not bothered by the presence of the diode at all. The machine performs perfectly normally except that it only runs about half as fast. But here’s the issue, it also only has about half of its normal maximum power. With the diode in place, the machine could not start on a piece of test fabric I have been using. Once started with the hand wheel, the machine could barely continue sewing. With the diode removed the machine had enough power to sew the test piece and control with the electronic pedal was good enough that a “run away” was not a problem. The presence of the diode is not a significant benefit when using an electronic foot pedal.

In conclusion, my sewing machine upgrade was a fail. The machine is not more powerful and I am unable to get any significant benefit out of limiting its power by half. It’s kind of like the 200mm Takumar lens I have in a drawer. I thought it would be cool to use on my M4/3 camera where it effectively becomes a 400mm lens, but it really does not work very well - its hard to focus and way too slow. Time to put this stuff on Ebay and unload it.