Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Trouble shooting dosing pumps
#1
I just finished built a temporary holder for my dosing pumps.
I thought hooking them up could not be simpler. But it turns out I must have made some boo boo during the construction.

I'm measuring roughly 2.5v on the pump outputs. But that does not change when the pumps should run and drops right away when loaded.

When I measure the pins on the Ardiuono (Board rev 1 I think/hope) I get a short impulse of a few volts, but they drop away almost instantly. Looks like something is loading those pins too high.
My suspicion is that I did something wrong when soldering the drivers.
Which chip is that on the board? Once I know which, I can probably trouble shoot it with the datasheet.
Reply to top
#2
Hi Tilman, sorry to hear about the trouble. The way its setup is the 12v from your power supply goes directly to four of the RJ45 pins for the pump, one 12v line for each pump. The grounds for each pump are switched using 4 of the 5 small transistors that are hiding under the mega. If you check the voltage using the ground and hot coming from the dosing pump RJ45 you shouldn't have any voltage unless the controller has the pump running. You can run the pump by pressing manual dose on the dose settings screen.

Check the voltage on the hot lines coming from the RJ45 but use the 6 pin ground header as a ground. If you do have 12v then I think everything is ok, I can't imagine all 4 transistors being bad and you can only mount them one way so I really doubt it was you. Do you have 12v on the connector board marked 12v?

What kind of cable are you using for the pumps? If its cat5 try a different cable, some don't let any voltage through them. I have a 4ft patch cable that won't power up my display, the same thing happens, as soon as there's a load it drops. I was quite surprised by that as I'm using 25ft of nothing special with no problems but that 4ft just won't work.

EDIT: Oh and I just remembered, in the power bar sketch did you change the board version to 1? I have it set to 2 but I believe you have the original boards so it needs to be changed. That's on line 20 of the sketch. This changes a couple of the pumps pin outs.
Reply to top
#3
I won't be able to track this down much more before the weekend. I will have to dissemble everything and take it out of the case, so that I can probe the pins.

The 12v supply is working fine. Unless I needed to hook up a jumper or other power supply line for the pumps separately.
I don't remember soldering those transistors, but I guess they are there then. :)

Actually, that reminds me. I was missing the fet/transistor that drives the feeder. I got four out of five. Perhaps I happen to measure the one that was not installed by pure bad luck.

PS: Yes. I have board version 1 code running. The right pump pins on the arduino wiggle when I dose manually. But as I said, the pins don't stay high but droop quite fast.
Reply to top
#4
I was starring at your build pictures and it just hit me, you actually have v2 power board so only 2 of the 4 pumps would work right now. Change the sketch to v2 and I bet it works.
Reply to top
#5
Oh, wow. That would explain a few things. :)
Reply to top
#6
(01-19-2016, 10:09 AM)TilmanBaumann Wrote: Actually, that reminds me. I was missing the fet/transistor that drives the feeder. I got four out of five. Perhaps I happen to measure the one that was not installed by pure bad luck.

Sorry Tilman, I forgot about those missing parts, I'll get them out right away.
Reply to top
#7
No worries. Everything I need working works. I had only put the pullups at the usb ports which I use. And I think the caps that where missing where not important for me as well.

And, no surprise really, the pumps work now. Unfortunately I cheaped out a bit on them.
And I bought some really slow ones, thinking that would improve accuracy. But they are maddeningly slow.
And quitet wonky too. They are designed to run 24v but are speced for 12v too.
Reply to top
#8
Great, glad they work kinda. On the power supply there's a small adjustment for voltage, turn it as high as it can go, it should then put out around 13.8v which 13.6v or so will get to the pumps. If the power supply is 12v exactly only 11.8v or less is making it to the pumps, maybe that little bit will help. Another thing you could try is a large power supply, the onboard regulators are good for 18v input so you can use an 18v power supply instead.
Reply to top
#9
I got a relatively ok Dosing schedule now. As a baseline I dose 1ml per day. (I decided that 1ml accurate dosing will be enough. So the next step may well be 2ml for some components)

However I had to bend the system pretty far to make this possible. The maximum dosing time for 1ml is around 2000ms.
I would need 4000ms to dose 1ml.

I'm running it in 400ms and define my 1ml dose as 10ml. That works quite ok. Not too confusing.

But what is pretty useless is the container content. I can set it to max 990ml. That is not very big for regular users I suppose. But for me that means only 99ml by my factor of 10.
Even more funny is that the pump to container size wraps over at 390ml. I can't figure out why that is the case.


I suppose I need to find faster pumps. That would probably fix all issues.
Reply to top
#10
I will be making some changes to the dosing, I'll see what I can do there. Currently the way its setup is the container size holds a max of 99 doses or 2 liters max, that's why you get max of 990ml. Not sure about the 390ml, that would be set at 3-4ml dose. Either way it'll work better soon.
Reply to top
#11
I thought you would enjoy a picture of my horrendous bodge. :)

[Image: 2016%2B-%2B1]
Reply to top
#12
   
Hot-linking images is not the best idea
Reply to top
#13
lol, I love it, a little bit of Red Green in there.
Reply to top


Possibly Related Threads...
Thread Author Replies Views Last Post
  Trouble understanding port number notorious.jatt 13 553 07-15-2019, 03:24 AM
Last Post: ZeeSocial
  dosing pumps Neongt 35 8,248 04-06-2019, 11:50 AM
Last Post: Rob F

Forum Jump:

Current time: 10-20-2019, 10:42 PM