I received the two Mabuchi SF-266SA motors this afternoon, not bad 11 days from China.
The motor looks like it is very close to the Canon EN22 motor. I haven’t figured out how to check the torque yet but comparing it in a Rivarossi Y6B frame it appears to be very close to the EN22. The RPM is 7000 but has plenty of power at 5000 RPM. The published specs are right on the money, 1.1amp locked rotor at 12 volts. The motor draws .041 amps no load at 12 volts and under similar load in the Y6B draws 390ma, same as the Canon EN22. Both the EN22 and the SF-266SA pulling 11 heavy passenger cars up my 3½% grade at a scale 50MPH draw the same current.
I’m going to remove two Canon EN22 motors from a AC-9 and install the two SF0266 motors. I can’t run it on DCC on my layout until I finish my control panel overhaul. I presently have two Canon EN22 motors in many of my Rivarossi articulateds so the real test will be running my 120’ mainline loop with it’s 3½% grades for several hours.
Thanks and good to hear your report on the SF-266SA. I first began looking for EN22 motors back in 2017 because of your postings on the MR forums and your web site. But could never find them. Finally when I found the SF-266SA in the repower/regear Groupio I figured it was a winner based on what those guys had done with it. It looks like you have proved that. By the way, I bought 6 in two orders and each of those came in 12 days - fast! Most electronic stuff from eBay China for me has been over a month of shipping. Good luck with your next ”266” adventure. Also, I don’t know if I had said this - I removed and discarded the “266” plastic end cap and cut the shafts down to 10mm. Thanks again.
I didn’t look closely at the plastic end cap until I read your post. I guess I was too evolved with connecting the power to those weird connectors.
I haven’t removed the plastic end as of yet but that would knock off about an ⅛” off the length of the motor frame. Even with the end cap on it’s the same length as the Canon so I don’t see a problem with leaving it on. I kinda like the Mabuchi contacts, just plug them in.
I have noticed that the low RPM torque is less than the Canon EN22, most likely from the difference that the six poles aren’t skewed like the five pole Canon. I does take more voltage than the Canon at low speed, 3 volts for the SF-266 2.4 volts for 2000 RPM. From a little under half throttle to full throttle they seem pretty close.
EDIT:
I’m rounding off the RPM as it is easier to do my tests that way.
Does anyone out there know of a way to measure the torque on one of these tiny motors? Currently I use a Faulhaber 2224RS as a generator with a shunt ammeter and a digital RPM meter. That works for motor comparison but not accurate for torque.
I have put the Mabuchi SF-266SA motor through all my testing procedures and as far as I’m concerned it is close enough to the Canon EN22 specs to be an excellent replacement motor.
I built up a Mel Dynamometer consisting of a pair of CANON EN22-R11N1B 12VDC - 5400 RPM motors as generators. I used my Fluke 179 Multimeter on the 10 amp position measuring the Canon EN22 output current, direct short across the motors. I use 2mm metal universals to connect the motors being tested.
I used a cheapo multimeter on the second Canon EN22, it checks out very close to the Fluke, ±2ma between 200ma and 300ma.
I powered a MABUCHI SF-266SA and a CANON EN22-R11N1B with separate variable power supplies. I ran both motors at the same voltage from 0-12 volts watching the digital ammeters. Both the Canon and Mabuchi at the same voltages and I was amazed at the results. Both ammeters were within ±4ma with the crossover at 8.2 volts. The Mabuchi was on the low side up to 8 volts then drew more current up to 12 volts. The RPM was also along the same curve. Both motors maxed at 620ma at the generator moto
I’d think with your power setup plus an accurate measurement of RPM, yooou could calculate a close enough torque. You have the elements needed, just have to get them all in agreeable units. Even just measuring the stall torque plus the no-load RPM should give you what you need, on a permag motor there is a linear relationship from one endpoint to the next. Torque at any other RPM should then be easily calculated.
A skewed wound motor, all else identical, should theoretically have better loow speed torque. And better starting characteristics. I suppose this can be made up for somewhat by a good drive system, goood BEMF, PWM with torque coompensation, etc. I’m going to predict that you will see a greater difference with these straight wound motors between the MRC decoder and the TCS BEMF ones, as the better drive of the TCS will make up for the lack of skew winding.
Are they really 6 pole. not 5 or 7? There are 6 commutator segments? Even poles is rather odd.
Edit: Can’t find a data sheet, but found an eBay listing with one disassembled - wow, indeed a 6 pole motor. Odd. I may have 2 od them - in my car. Seems they are commonly used for folding the side mirrors. Cheap enough, too.
I replaced a Canon EN22 with the Mabuchi 266 and it seemed to run better than the Canon on DC I can’t run DCC on my layout for a couple more weeks, I have my control panel all screwed up.
Something I haven’t figured out is the RPM of both the Canon and Mabuchi are very close at 12 volts, 7700 RPM no load and 6200 under my full load. The Canon specs say 5400 RPM full load. I’m figuring full load at 400ma at 12 volts.
I wonder which one is in my car - not about to take my mirrors apart to find out though.
Max torque should be when stalled, at rated voltage and current. Between there, 0 RPM, and max RPM with 0 torque load on the shadt, again at rated voltage, is s strictly linear relationship, so the torque at any intermediate shaft RPM should be easily extrapolated. Thoough this will be the torque at the rated voooltage, not torque at whatever RPM it turns when fed say 6 volts.
The sixth pole may make up for the lag of a skewed armature - might be a sweet spot between a 5 pole skew wound armature and a 7 pole. I’ve seen it mentioned in the past when it was complained that the 7 pole motoor in questin wasn’t skew wound that with enough poles, it doesn;t really matter, the skew helps the most on 3 and 5 pole armatures. I sort of remember from old magazines back in the day of open frame motors, there were 5 and 7 pole versions, and I kind of remember seeing someone advertise a 15 motor.
Mel and Randy:
Since you guys might do testing of this motor when using DCC, there are a couple of things I read about the Mabuchi SF-266SA when it comes to their use with DCC. You may want to look into the motor’s back electromotive force (BEMF) and its internal capacitors (it has two tiny SMD components acting like capacitors). I am not a DCC user but I found that this motor is perfect for my DC HO repower projects. The “RepowerRegear Groups-io” site where I first learned about this motor last spring has lots of postings about this motor and those topics. Apparently, some years back members were testing some other motors with capacitors and found that there were problems using those other motor with some DCC decoders but the problems were eliminated when the capacitors were eliminated. The BEMF stuff I ran across was over my head but they get into that quite a bit. I never understood the lingo when reading about these topics. Anyway some members have taken the capacitors out of “266” circuit and it appears easy to do from the photos and explanations that are there. That Groupio site is full of info. If you go to the group and use its search function with “SF-266SA BMF” and “SF-266SA capacitors” you will find lots of postings and that stuff can better say anything I can post here.
As I have my control panel down for modifications I can’t operate my layout on DCC but I did install a SF-266SA motor in a locomotive for testing on DC. It preformed perfectly on DC.
This morning I tried the other motor in my Mel Dynamometer with a MRC 1730 sound decoder and a Digitrax SDH164D, I didn’t detect any problems but . . . . testing on the bench and in a operating locomotive on my layout can be like the difference between apples and oranges.
I swapped the test motor from DCC to DC and I couldn’t tell the difference on my meters.
The locomotive I installed the SF-266SA in is wired for DCC and I tried it on DCC on my bench test track but I can’t really tell anything in 36” except that it runs both directions.
Well, if those capacitors aren’t easily accessible, that is a bummer. Yes, they do interfere, not only with the BEMF, but also with the high frequency PWM drive from the decoder. They would be the equivalent of the external capacitors on many Bachmann locos.
Looking again at the ebay listing that shows one disassembled, I can see the caps on the brush cap. And it looks like the motor was pretty much destroyed to pull it apart like that. If that’s the case, this motor is a non-starter.
Having studied that RepowerRegear Group about the SF-266SA quite a bit and popped off and discarded that plastic end cap on the motors I am using, I have studied the crimping of the metal end cap to the case, I have to agree with what members of the group said even though I have not done it as I’m all DC - it is easy to get to those capacitors and cut the connections and it was easy to reassemble and when installed in locos ran fine. Heck, at just over $2 a motor (currently), you could toss a few if you did mess up compaired to the cost of other motors. Now, there is a NO capcitor version of that Mabuchi, the SF-266SH, which has lower rpm than the SF-266SA. Depending on the loco, a few group members were looking at it.
I removed two Canon EN22 motors from a Rivarossi Cab Forward and installed the pair of SF-266AS motors, the Cab Forward actually performs better with the new motors on DC.
I setup a small loop of track on the floor and hooked it up to my DCC controller. The decoder is a MRC 1731 and it seems to run pretty good to me. What should I look for? I didn’t change any settings in the decoder, it is still configured for the EN22.
When I had a SF-266 and a EN22 running in parallel on DC both motors ran very close to the same RPM at the same voltage. On DCC the SF-266 runs much faster on DC than a Cab Forward with EN22 motors. It’s the same on DCC, about 20% less throttle for the SF-266 on both DC and DCC.
I can’t see any difference in the operation other than the higher speed.
The power connections are reversed from the Canon.
Hmm. I searched some of the posts, and most sound a bit complicated - maybe not in practice, but when written down it doesn;t make it sound too easy to get them apart. One mentions removing the knurlign from the output shaft so it fits through the front bearing - frankly that’s already sounding like a bit too much work just to save some money. My biggest concern is getting those brushed back in place without bending them, even if there is no need to fully remove the armature (which seems odd - I do think the person that posted that wanted to take it completely apart to drill mounting holes without drilling into the motor and also allowing a way to clean out any chips.).
Considering I have very few locos at this point which I would even consider needing to remotor - maybe my brass RS-3, but it already has a decent can motor, and the Mantua 0-6-0, which is a newer version with some sort of can motor, not an open frame motor - I might just spend more money and get the Canon motor. Most of my stuff is less than 20 years old and has decent motors, so unless one absolutely fails, I see no pressing need to swap them out. They all run nice and slow with a good decoder, and quiet. Any issues are possibly in the fore and aft slop in the worm gears causing bndng on grades, which is realtively easy to fix with thrust washers.
The “removing the knurling” only applies if you are pulling the armature out of the case in order to get the shaft thru the output end’s bearing. This is done if you are drilling holes in the forward end of the case for attach points. When cutting the lead on capacitors, you only need the metal end cap off. I read one modeler used chisel end blade(s) in Xaxto and a couple of fine tip screw drivers to separate edge crimps. When reassembling, you need to spread the brushes to fit over the armature commutator.
I’m checking out a replacement for a EN22 motor. The pair of SF-266AS motors in my Rivarossi Cab Forward have been running at 42% throttle (about 40MPH scale) on my MRC Prodigy using a MRC 1731 decoder for about an hour. The only difference I can tell is the SF-266 runs a bit faster than a like Cab Forward with two Canon EN22s.
Without touching the throttle setting I swapped the two Cab Forwards and they both draw 190ma, the FS266 a little faster than the EN22.
The decoder isn’t even running warm, the decoder is in the tender so I’m swapping tenders so it’s the same decoder on both locomotives.
This is the Rivarossi frame with the two Mabuchi SF-266AS motors.
I comment, the Mabuchi motor doesn’t change RPM very much by adding load. It’s much better than the EN22 or as Tim The Toolman says, MORE POWER!!!
SF-266SA source is China via eBay. There appears to be no USA source like Motorman. Assumption is this motor, made for automotive outside mirrors, had a big overrun. Thus far, they are very cheap and are free shipping. Orders have arrived in less than 2 weeks.
Over about the last year I have installed these Mabuchi SF-266SA motors in 25 Hobbytown drives ,I am very happy about performance results . The Hobbytown drives are all flywhell drive unit tha ran very well with the supplied DC 70 motors . With these Mabuchi motors they run even better at slow speed and throughout the speed range.
In the last few months I have used these Mabuchi motors on ten brass diesel drives in Alco models and a Red Ball diesel I am very pleased with the results it has turned some shelf queens into fine running units. I did two Alco Models S2 switchers with one motor they now run very well compared to the old motors.
All the others are setups that had double ended motors so I mounted two motors since these are single ended, These all run much better. All the brass diesels have some form of KMT drives with these new motors there is a great improvement.I run DC only . I did remover the white nylon end caps which shorten the motor a bit . I soldered power wires to the tabs at the end of the motor, The motors were all mounted on a bed of silicone adhesive.
