BK Precision 1900 power supply broken knob

[JackM]

I am attempting to repair a broken rotary encoder on a BK Precision 1900 bench power supply. On this unit the current adjustment knob has sheared straight off due to what I can only guess was some excessive perpendicular force.

The trouble I am facing now is finding a part number for the rotary encoders used in this power supply. I cannot find a parts list for this BK Precision model, and the only other thing I have to search is a marking on the encoder base which says "2N4". This is a 7-pin rotary encoder part with a combined push button. The rectangular base dimensions are 13.75mm by 11.7mm by 6mm (LxWxH) and the pole is 20mm in height from the top of the base. It has about 20 detent positions on it as well.

Where or what should I be looking for? Would any rotary encoder with the same number of pins, dimensions and number of detent positions work just fine, or is there some other variation with rotary encoders that I don't know about?

[cowana]

Where or what should I be looking for? Would any rotary encoder with the same number of pins, dimensions and number of detent positions work just fine, or is there some other variation with rotary encoders that I don't know about?

That looks very standard - Bourns or Alps both have a fairly good range of incremental encoders that will drop right in. There are quite a few parameters used to specify the exact type - although getting most of these different to the original will still just work, and just 'feel' a little different.

Note that the ratio between pulses per rotations and detents per rotation can be 1:1 or 2:1. Getting that wrong will mean it counts two per 'click' or one every two 'clicks'. Without being able to turn the original though, you might have to just use trial and error to determine which you need.

Relevant rotary encoder parameters for that footprint of incremental encoder:

• Detents per revolution
• Pulses per rotation
• Rotational torque required
• Push switch 'click' depth
• Push switch force required

[JackM]

I think I may have found the exact part, TT Electronics encoder model EN11-HSB1AF20 and Mouser has some in stock. EN11-HSB1AF20 Datasheet

I checked the dimensions and it all matches. The only thing I haven't verified is the number of pulses per detent. Looking at this series of encoders from TT Electronics it appears as though they only have one type with 20 detents, which also has 20 pulses per revolution, so a 1:1 ratio of detents to pulses. I'll have to check the original encoders in the BK Precision 1900 to see if they also have a 1:1 ratio of detents to pulses to be absolutely sure.

[Gertjan]

Both encoders in your power supply have the same "2N4" makings.
So you can test / measure the other one....

regards, Gertjan.

[JackM]

After a quick test probing the output pin on the encoder while the unit is powered on and looking at the waveform produced from turning the knob, it would appear as though the encoders do indeed have a 1:1 ratio of detents to pulses.

[Ordinaryman1971]

I think they are actually 24 pulses, that's what the 2n4 stands for, does it have two connectors on the back side of the encoder or just three on the side you show in the picture?

[JackM]

I think they are actually 24 pulses, that's what the 2n4 stands for, does it have two connectors on the back side of the encoder or just three on the side you show in the picture?

The encoders used on the front panel have a total of 7 pins. A pair of pins on the top facing side are for the push button, two larger pins on either side are NC and for physical support, and then three pins on the bottom facing side are the encoder outputs. Center pin is ground and the other two on either side of the center pin are encoder output A and B. Visible traces on the PCB seem to show that one of the encoders makes use of both outputs, but the other encoder seems to only use one of the encoder outputs.

[cowana]

Visible traces on the PCB seem to show that one of the encoders makes use of both outputs, but the other encoder seems to only use one of the encoder outputs.

Are you sure there aren't traces on the other side of the PCB? Only using one of the two encoders would mean the software would be unable to determine the turn direction - only that the knob had turned a number of steps...

[JackM]

Are you sure there aren't traces on the other side of the PCB?

I didn't do a very thorough check, so yes it's likely that there is another trace somewhere on the board that I didn't notice.

I have the new encoder part in hand now, and I plan on replacing the broken one soon once I have the time. I'll test out the functionality and see if the new encoder is acting in the same manner as the old one would have.

[JackM]

Repair complete!

I ended up using Digikey part 102-1768-ND instead of the one from Mouser just because I was already making an order from Digikey at the time.

I desoldered the broken encoder, added in the new one and confirmed that it works just like the old one. I knew for certain that the original encoders had 20 detents per revolution, and I even spent a few minutes searching for encoders that might have 20 detents with 24 ppr, but it seemed to me that with encoders such as these the number of detents matches the number of pulses that you get.

20 detents with 20 pulses per revolution was indeed correct; it never skipped a value on either the course or fine adjustment settings. Something that I did notice that was slightly different was that the encoder that I chose has a more "muddy" feel to the detents where as the original encoder has sharper detents that are more defined. One other difference is that the body/base of the new encoder is very slightly narrower than the original, so when I placed the new encoder onto the board it sits maybe a millimeter higher above the board than before. This doesn't cause any problems in operation though.