Multiplexing solenoids-defying the laws of physics

[GTechno13]

Ok so this is hard to explain but I am multiplexing solenoids so figure about 7ohm loads running at 24VDC.

I have 20 solenoids to run split into two groups. Each group has 10 wires going to each solenoid and a common wire going each. My circuit has 10 darlington NPN (TIP120) for the low side switches and two TIP125 (darPNP) for the high side. The PNP use 3904 NPNs to pull the base low with a resistor pulling it up to the emitter voltage (24-30V). I then have 2 flyback diodes going from each NPN (the 120s) to each power rail switched by the PNPs. (see pic)

My problem is that no matter what, I get two solenoids firing together. More specifically if group one is divided into solenoids 1-10 and group two is 11-20 then 1,11 always fire together and 2,12 and so on. I have checked everything 10x over and the PNPs are not both on. I have followed every possible way current can flow and can not for the life of me figure out whats going on. It seems current goes through the correct solenoid and then follows backwards through the other but then I can't find the path to ground. I cut the connection to the flyback diodes and tried adding diodes in every place imaginable to make sure current is flowing in the right direction and nothing helps.

If I disconnect one of the PNP outputs which is the common wire for one of the groups...IT STILL FIRES BOTH!

I don't have a schematic but I have the business end of the PCB. It's very simple through hole stuff. The two groups of connections on the right are the positive rails switched by the PNPs (1-10 and 11-20) and the horizontal row is the NPNs. Each of those big holes paired together is a (1,11) and (2,12) and so on. Hope this makes sense.

View attachment 63766

 

[ronsimpson]

You are not going to like this one!

See attached schematic.
Top transistors, TIP125, right is on, left is off.
Bottom right TIP120 is on. The coils are 7 ohms.
Bottom left is 10 parts in parallel. These 10 transistors are all off. The coils are 0.7 ohms because there are 10 in parallel.
Current, in red, flows from 24V through, TIP125 through the coil and via a TIP120 to ground. 24V on right coil.
Current, in red dots, flows down through 10 coils to the left transistor(s)-C, then up through 10 coils to the off TIP125-C.= 20 volts. You have 20 volts across the left coil.

To say it another way. The left coil has a current path that starts out with 0.7 ohm coil, 0.7 ohms coil, then a 7 ohm coil. The coil will be on at 20 volts.

 

[GTechno13]

Wow, thank you. This really had me confused but your explanation is clear. I am replacing the manufacturers boards and I wondered why they used all low side switches. So to stay with the current design I would need to put a diode in series with every solenoid. Not something I want to do but maybe just for this board. I have an additional set of 20, a set of 12, and a set of 32 solenoids to drive and luckily I didn't make all four boards at once. What I am still a bit confused on is then why I haven't encountered this problem before. But now that I think about it...I just rebuilt the electronics from scratch for a vending machine with 50 or so motors and I believe the motors had a series diode built in. Learn from your mistakes huh?

Again, I want to thank you for taking your time to draw and scan me a picture (or google it, either way). People like you provide help drive the engineers of the future. Now I got to share your pic with my buddy at school who couldn't figure it out either...it seems so simple now.

 

[ronsimpson]

Where in Nebraska? I spent 4 years in Lincoln and have a house there.

 

[GTechno13]

Born and raised in Lincoln! I currently live in Omaha because UNL's electronics engineering program is at PKI at UNO.