Help with Water Pump

[alec_t]

I have 10 FSM-Mk1 modules on a board.

You only need two (one for the tidal pumps and one for the wave pumps), since each Mk1 handles up to four 'channels' (pumps). The filter pump system is odd man out and the FSM Mk1 needs modifying to suit. Here's the mod (re-named FSM-F):


I tried drawing the power/gnd wiring for the whole of your mammoth system (wave, tide, filter) but found that it wouldn't fit on an A4 sheet as the text was too small to read. So I created a hierarchical drawing with separate wave, tidal and filter sections each on an A4; each section including pump 'sub-systems'. The wave and tidal systems each have 'Type A' sub-systems; the filter system has 'Type B' sub-systems. A separate A4 drawing shows the make-up of the two types of sub-system.
Here you go:


Are you ok with adding extra 'channels' to the Alarm module in the TTAM? Only 4 are shown but the extra 6 are analogous.

And here's the revised filter pump power back-up arrangement:


I hope I've covered all bases, but may have forgotten one or two features

 

[salty joe]

Wow! Those are nice looking drawings. Thanks Alec. I'll get it printed after work and get after it.

I should have said I have 10 channels of the FSM-Mk1 built on one board. (4 wave pumps, 4 tide pumps and 2 filter pumps) So I used three LM324 ICs with six unused pins on the third IC. I used a common terminal point for 12V and another common terminal point for signal ground.

 

[alec_t]

I used a common terminal point for 12V

...but not, I hope, for the FSM-F that will be used with the filter pumps? That will use the 12.6V line.
4 input pins of the 6 unused pins can be grounded (as stated on the FSM-F drawing); but would probably be ok left unconnected, as KISS indicated.
FSM-F has two added 470k resistors. If you don't have that value then 390k will do (which I think you have).

 

[salty joe]

Thank you Alec for the super wiring diagrams. I have them printed, but before I put everything together, I need to re-test the PDMs that I might have damaged by mixing 12V with 13.8V. I’ll do that tomorrow.
I want to make sure I have this right-The Filter Pumps Timer (0-15 min) will get 13.8V from FPPMk4 (what I like to call battery backup). The two channels of the FSM-F for the filter pumps will also get 13.8 from battery backup. The two PDMs for the filter pumps will get 12V from the Vreg.
The battery backup I have is version FPPMk4. There are a couple extra diodes and R5 is 1.8K.
I installed terminal points for and connected fault and Isense to the PDMs.
Thanks for the upgrade on the FSM-F, it’s done.

 

[alec_t]

-

The Filter Pumps Timer (0-15 min) will get 13.8V from FPPMk4 (what I like to call battery backup). The two channels of the FSM-F for the filter pumps will also get 13.8 from battery backup.

Yes (though the 13.8 will be ~12.6 with the mod).

The two PDMs for the filter pumps will get 12V from the Vreg.

No. They'll get the 12.6V (see FPPB-Mk2). Sorry, I forgot to re-label the '12V' input to the PDM on the PSSB sub-system drawing

The battery backup I have is version FPPMk4. There are a couple extra diodes and R5 is 1.8K.

Superseded by FPPB-Mk2 (which still shows R5 as 1.5k).

Edit: IIRC you had the 'Fault' inputs of at least some of the PDMs temporarily grounded. If so, un-ground them.

 

[salty joe]

Thanks for clearing up the voltage feed for the battery backup. So everything associated with battery backup gets 12.6V, except fot the alarm module. That makes it easy to remember.
I already have 10 alarm modules lined up and ready to go. I also already ungrounded the 'fault' pins and connected them to individual terminal points, Isense is also now connected to individual terminal points for all PDMs.

I am going to re- test run all PDMs and then see if I can test all components individally before putting it all together.

I've been waiting forever for 0.33uf caps for C10 in the Tidal and Alarm Module. I have 0.22uf caps on hand. Do you think the 0.22 cap might work to get a wee flick?

 

[KeepItSimpleStupid]

For testing you can put a 0.22 in parallel with the series combination of two 0.22 uf caps.

That's 0.22 + (0.22/2) or 0.11; Total capacitance being 0.22 + 0.11 or 0.33 uf.
Note: formula used is for like values.

------------||---------------
*...........................*------
----||-------------||--------

 

[salty joe]

That's very cool KISS-thank you! If the 0.33 caps don't show up tomorrow, I'll give it a try.

 

[alec_t]

So everything associated with battery backup gets 12.6V, except fot the alarm module. That makes it easy to remember.

Spot on!

I also already ungrounded the 'fault' pins and connected them to individual terminal points, Isense is also now connected to individual terminal points for all PDMs.

Sounds good.

Do you think the 0.22 cap might work to get a wee flick?

I expect it will. The design was based on your original estimate that the old pump took about 1 sec to get up to max speed. I'm now guessing that the new pump, being more powerful, has greater acceleration so should need a shorter pulse than originally for the kick. This is where design gives way to experiment, as the new pump's start-up behaviour is an unknown. Even a 0.1uF might work. The trimmer and its series resistor are other candidates for experimentation if the flick setting proves too sensitive.

I can test all components individally before putting it all together.

Recommended. When it comes to assembling it all, are the various 'Vt' connections clear for you? In the various schematics some have been generically labelled 'Vt'; others have been labelled 'Vt1' etc. I've tried to be consistent (but may have failed ). The principle is that where pumps are paired they share the same Vt; where they are not they have respective different Vt signals.

 

[salty joe]

The new wave timer works and so that's the good news.

Somehow, there is 23V coming out of the 2 TTAM lines that connect to Vt in the PDMs. I did not think to check the V before hooking it up. Am I correct to assume that I probably fried the ICs in four of my PDMs? I temporarily installed a 0.1uf cap at C10.

 

[KeepItSimpleStupid]

Very likely.

@alec
Thoughts whether OVP would be useful?

@Joe
OVP is an acronym for Overvoltage Protection. It was used when logic was 5 V and had a tolerance of 10%. The circuit was also called a crowbar because what it did was to short out the power supply causing other protective devices to trip.

There are IC's that implement the funtion using an SCR (Silicon controlled rectifier)

Since nearly every chip was 5V at the time, a lot of damage could occur with an overvoltage. It has the same properties as a divorce. e.g. Why is a divorce so expensive? Because it's worth it.

PS: I got the news that my hard drive is unrecoverable.

 

[salty joe]

OK, thanks KISS. Tomorrow I'll go over the TTAM and find out what I did wrong. Then determine if the ICs on the four PDMs are shot. I fully expect them to be, but want to be certain before tearing anything apart. Would hooking those four PDMs up to the wave timer be a good way to test the ICs?
This is just my opinion and what do I know, but I think once I get the TTAM right, that's the end of the over voltage problem.

The divorce joke is funny. Luckily, my wife still puts up with me

And I am sorry to hear that about your hard drive. That stinks.

 

[KeepItSimpleStupid]

Here is te datasheet of the CD4093: https://www.google.com/url?sa=t&rct...yCbJkXuDPSwls2w&bvm=bv.48705608,d.dmg&cad=rja If I remember, the inputs don;t change very fast.

The first time around, measure the voltages of all of the pins to ground and report.

wen you look at says pins 1-3, 1 and 2 are inputs and 3 is an output. The gate is a NAND gate.
The truth table is:
00 1
10 0
01 0
00 0

A "1" is very near the power supply voltage and a "zero" is very near ground.

For now, just report the values when the inputs are not changing. We should be able to spot something obvious.

 

[alec_t]

Somehow, there is 23V coming out of the 2 TTAM lines that connect to Vt in the PDMs.

Oh dear! That suggests either a) the input to the 78L12 reg has a wrong connection directly to the '12V' output line or b) the 78L12 has failed short-circuit. I suggest you disconnect the 12V input to the PDMs, disconnect the 78L12 output from everything else in the TTAM then power up the TTAM and see what voltage you get at that output. If it's 12V then reconnect that output within the TTAM and see what voltages you get at the '12V' output and 'Vt' terminals of the TTAM. Only when you get 12V at '12V' and Vt should you reconnect the PDMs (one at a time, testing as you go). With a bit of luck the ICs in the TTAM and PDFs might have survived 23V.
KISS has suggested adding over-voltage protection to the system. That's an option, but personally I think the extra complication could in itself be another source of bugs. Your call.

Sorry to hear about your HDD, KISS.

 

[salty joe]

Thanks KISS, thanks Alec.

Can't go to work until 9:30 today, so I started with the TTAM and will test PDMs tonight.

There is 12V coming out of the Vreg, so I re-connected everything. The tidal timer seems to work just right! My DMM showed it toggling right along at 11.9V. On the flick terminal, after I slow counted to 35 or so the DMM momentarily showed 0.5V or maybe it was 5V but anyway it looks like the flick is active.

I must have had something hooked up wrong, IDK. Fingers crossed for the PDMs to be OK.

 

[alec_t]

I must have had something hooked up wrong, IDK

Looks like it. We've all been there, done that . If the TTAM is now working ok its ICs must have survived, so there's a good chance the PDM ICs did too.

 

[salty joe]

I dodged a bullet. The new tidal timer works like a champ and the four PDMs are all working. I don’t know what I did last night but I aint doing it again!

I was going to test all the FSM modules/channels but realized the need to make the two channels for the filter pumps separate from the other eight. It was a long day today-not a good time to fire up the soldering iron.

I am tickled pink that that new tidal timer and new wave timer both work and especially that the four PDMs that were hooked up incorrectly did not bite the dust.

A friend has a 12V motorcycle battery and if it will hold a charge, I can borrow it to make working on the filter pump backup way more convenient.

On the TTAM, with the 500K flick pot adjusted all the way to one side and a 0.1uf at C10 the flick turns the blade ½ turn. With the pot adjusted to the other side, the flick makes the blade shudder. I am going to add another 0.1uf in parallel to see if it gives a little more oomph.
Thanks guys!

 

[alec_t]

I am tickled pink that that new tidal timer and new wave timer both work and especially that the four PDMs that were hooked up incorrectly did not bite the dust.

I heard your sigh of relief from this side of the pond

I am going to add another 0.1uf in parallel to see if it gives a little more oomph.

That should do the trick.

 

[salty joe]

I was about to test the FSM and assumed Vt on the FSM gets connected to Vt on the respective PDM. Trouble is, Vt is shared between pairs of PDMs since the wave and tide run in pairs.
In other words, we have four pairs of PDMs and each pair has a common Vt. (really 5 pairs of PDMs, but I'm not working with the filter pumps now)

Or maybe I’m going about this all wrong.
Your thoughts would be much appreciated.

 

[alec_t]

Trouble is, Vt is shared between pairs of PDMs since the wave and tide run in pairs

Not a problem. Let's say pumps A and B are a pair driven by PDMa, PDMb repectively and monitored by channels FSMa, FSMb respectively. Connect the single Vt of pair A/B to PDMa, PDMb, FSMa and FSMb. There will, however, be two different Isense signals going into FSMa, FSMb respectively.