Need the advice of the clever ones

Steve2381

Senior Member
Hi

We are trying to fix a bit equipment at work.
Its basically a circuit (SMT so hard to read components and see the PCB tracks)... that can either receive a 230v input or a 24v input and it operates a 12v solenoid.

The 230v input is pretty self explanatory, but we don't understand the point of the LM358 op-amp. Its more than highly likely that we have traced the circuit slightly incorrectly, as access to the PCB is difficult.

I am thinking its for signal cleaning? We are not even sure why its needed? Could the same outcome not be achieved a lot simpler?

For those who use SMT parts, any help identifying the caps. The smoothing cap has 100 & 255 written on it. The other cap has 10 & 165 on it.
In my haste to scribble the circuit down, its quite probable I have got the diodes the wrong way round. My old man eyes had trouble enough reading what I could.

Finally, the large 68R resistor across the output of the 230v - 12v PSU. Is that just a dummy load to stabilise the output?

We are hoping to fix this unit, and then make our own with a Picaxe to control some delay features.

Thanks guys
 

techElder

Well-known member
Steve, I've always approached this kind of project with the idea of defining the functions of the target. Only then would I consider where to go from there.

So, define the inputs. For instance, why is there a 230 input? Why is there a separated 24 volt input? What provides those inputs? Is there a specific kind of "noise" from the inputs?

Is the solenoid the only output? What is the function of the solenoid.

Use your electronic skills rather than just duplicating what another engineer created.
 

Steve2381

Senior Member
The 230v trigger input is a trigger - usually from a simple toggle switch.
However 24v input comes from a digital controller that houses a 2Ah 24v battery. The cable that runs to the 24v input terminals however can sometimes be 20 or 30 metres however.
The reason for both 230v and 24v is purely down to giving the user trigger options depending on the other equipment being used.

The entire system is actually a Pyrotechnics controller. The servo operates a gas valve that is usually pressurised with CO2 at around 7bar. The servo under no load is around 0.9A and under worst case load is around 1.4A.
There is only ever one valve on the output.

Being Pyro equipment, its quite possible that the signal will not be always be particularly clean. This stuff gets a beating during its life.

We have about 50 of these units. Retro fitted into equipment by someone else and they are starting to fail.
It would be nice to understand the thinking behind them.
 

premelec

Senior Member
I don't know if original mfr's specs on unit are available - if so what were it's advertised features... ?? That would give you more of an idea of what it's doing... likely a central safe mode and such... i agree with Tex that it's hard to know what original engineer was thinking - and I've seen many mysteries in this regard over the years... "What WAS he thinking??"
 

Pongo

Senior Member
My best guesses: The LM358 circuit is a delay/filter to avoid false/uncertain triggering. The 68 ohm is to ensure the discharge of the output capacitor of the 230 -> 12 volt supply to speed the response to the 230 volt trigger signal going away.

As others have suggested it would be best to decide what you want and go from there.
 

Steve2381

Senior Member
No spec available. Last guy (who apparently designed these) basically quit and left leaving nothing.

Basically, you plug the 12v solenoid valve in the output. Then you 'trigger' the valve by either sending it 24v DC down a signal line. This can be from a posh digital control panel to a man using the battery off his Dewalt drill.

The 230v input is because some of the old control systems use 230v triggering, or sometimes its basically connected to a lighting effect so that it operates at the same time.

I just wondered what the LM358 brought to the party, and why the signal would need any adjustment or filtering. I do not know enough about op-amps to make a reasoned judgement.

The 230v basically goes through an 'off the shelf' 230v-12v PSU module and directly controls the solenoid via the transistor.
The 24v 'off the shelf' PSU goes through the LM358 and then into the transistor. Why not simply use a multi-voltage PSU (the one he used appears to accept 12-36v) and go directly from its output to the transistor, the same way the 230v PSU does?
 

Steve2381

Senior Member
OK... so maybe that is it.
Would the 'on/off' response of the 230v AC conversion provide a reasonable amount of switching stability due to its internal components (large caps etc).

Whereas, perhaps the 24v input is a little bit more 'twitchy' due to it being DC?

Who knows

I think I need to design my own method. The less complicated the better to be honest.
I will start wandering around RS and Farnell
 

premelec

Senior Member
Seems likely the OP amp was a delay function - and 555 timer would be a better choice which could drive the solenoid with a transistor. perhaps take a still working unit and see what the output is when actuated - i.e. discern a delay. You could do with a PICAXE 08 and appropriate spike protection and fail safe considerations. As a fellow who almost lost my life in a pyrotechnic mishap 66 years ago I'm all for safety! Good luck with it...
 

cravenhaven

Senior Member
The cct seems a bit odd to me.
You have drawn the LM358 pins 6&7 shorted together but not connected to anything else. 6 is the '-' input and 7 is the output so shorting them would imply a voltage follower configuration, but then you would surely have them connected to something else?.
The connectors both have 3 pins, but you have them shown with 2 pins in the cct diag.
There are 2 blanked out areas of the cct board, 1 of which seems to be hiding a component.
What does the label on the bottom connector say?.
Could you label the components on your cct diag with the component numbers. eg R1, D3 etc.
 

tmfkam

Senior Member
GIven the simplicity of the circuit, I might be tempted to repair the original as opposed to trying to reinvent something that presumably has been operating correctly for some time now.
 

srnet

Senior Member
The entire system is actually a Pyrotechnics controller. The servo operates a gas valve that is usually pressurised with CO2 at around 7bar. The servo under no load is around 0.9A and under worst case load is around 1.4A.
There is only ever one valve on the output.

Being Pyro equipment, its quite possible that the signal will not be always be particularly clean. This stuff gets a beating during its life.

We have about 50 of these units. Retro fitted into equipment by someone else and they are starting to fail.
It would be nice to understand the thinking behind them.
Think of the possible consequences of openly soliciting advice on how to fix such a piece of equipment on a public forum.

Is this not a situation where your firm needs to employ someone, a professional engineer, to fix the devices ?
 

Jeremy Harris

Senior Member
These things aren't "pyro" in the normal sense, it looks very much to me like the controller for a system to blow stuff like confetti out of tubes, with a blast of CO2 (CO2 being used because it also adds a "smoke effect). Sometimes these things are just used for stage cannons and the like. It's not particularly hazardous, and the pressure not really enough to be a significant risk (my water supply pump runs at 9 bar, for example).

As above, repairing the existing units looks the best bet. None of the SMT parts are too small to rework, and my inclination would be to just fix them as the break.

Finding a reason for the failure would be nice, and my best guess is that either the DC-DC converter on the low voltage input trigger side (which supplies the 12 V from the 24 V DC input) may be failing due to being operated with an input voltage outside it's limits (I'm assuming that's hidden on the other side of the board) or that the solenoid drive transistor is failing, as there's no short-circuit protection for it not does there seem to be any reverse inductive spike protection (it would be usual to fit a diode across the solenoid coil to prevent reverse voltage spikes, just like you do with a relay).

Once you've found which components fail (and my guess is that it will probably the same one on each failure) then it should be straightforward enough to come up with a fix to make repaired units more reliable.
 

Steve2381

Senior Member
Well as I mentioned, its highly likely I have missed a link or trace off the circuit. Pins 6 and 7 may well link underneath the chip to somewhere else.
No delay. The trigger must be instant. There is never a delay in pyro equipment at the receiving end (well there should not be)... press the button, it should fire straight away.

Yes... its for co2 style effects etc, nothing too horrific.

The reason for the tinkering or redesign is that these modules fail on a constant basis. But now we cannot get replacements, as they are not made any more.

I just wondered why we could not simply get a 230v AC to 12v DC PSU and a 9v-30v DC to 12v DC PSU and simply join the outputs together via diodes to prevent feedback and achieve pretty much the same outcome with far less components.
Obviously a little more refined than that, but you get the idea.

Just sticks in the back of my head as to why they added the LM358

I think there is a diode on the output - hidden from view.
I believe the input from the 24v signal goes through the bridge rectifier to prevent incorrect polarity. The 24v to 12v PSU is on the rear of the board (its actually 9v to 30v input).

The 24v signal is rarely, if ever over the rated 24v. These units start at around 23.8v at the controller but the cable runs to these units usually kills a good few volts before it arrives at this circuit.
 

marks

Senior Member
Hi Steve2381,
looking at your pics you can see the card is designed for
. 24 ac input (longer distantances) will proberly also work from 24dc to 15vdc
. or 12 dc input (probaly local cpu control) 68 ohm is keeps voltage lower on non regulated supply and provides some noise blocking .
. also looks like the output does have a flyback diode not shown on your drawing

the opA could be a pulse generator ie (reducing coil current when continously on and prevent overheating)
so this circuit looks likely made to drive a 500ma 12v solenoid (usually has a 25% duty cylcle)

your circuit doesnt look quite right and for the other opA it is usual to connect this way when not in use.

the smd caps usually marked from top to bottom
0c date or batch
100 rated capcitance (100uF)
25s rated voltage (25v)

edit.
after thinking and looking at your circuit r9 looks to be a 330 k resistor not a cap
so the opA looks to me to actually be configured as a 1 shot (giving an instant short duration high pulse)
 
Last edited:

Steve2381

Senior Member
Thanks for everyone's help on this.

Its not a one-shot. These units power the solenoid for as long as the input is present.

The pulse generator would make sense, but then why have they not bothered on the 230v input?

The trigger signal is always 24v DC. I assumed the bridge rectifier was there to simply provide protection due to incorrect polarity (or an incorrect AC signal).

I have never made an SMT circuit, so the components are foreign to me.
 

hippy

Technical Support
Staff member
The pulse generator would make sense, but then why have they not bothered on the 230v input?
From the circuit diagram it appears the solenoid is only controlled by the op-amp, the internal power rail and op-amp supply is a wired-or combination of 230V and 24V incoming sources. So whatever the op-amp does it seems it does it for either supply.

Because this is a real world problem which needs to be fixed it would be worth spending a lot more time determining how the PCB tracking and what the circuit actually is. Using safe working practices you could take a working board and look at what signals there are using a scope and meter and not have to risk guessing wrongly as to what is happening.

Or perhaps just specify what you want and design a replacement board to deliver that. That would make sense if there is some design issue which is causing them to fail regularly.
 

marks

Senior Member
Hi Steve2381,
that looks correct as you say you can just plug in a 24v input polarity doesn't matter
the 230v trigger is just doing the same as the 24v trigger.
providing the 12v supply and the led will be lit.

the solenoid is switched on the ground side controlled from the opA

in any case just put a meter on the valve connections and see what the signal there is doing.

pin3 of the opA is set at half the supply voltage say 6v
when
pin2 is higher than this value
pin1 will then go low turning off tr1
 

techElder

Well-known member
Ok, Steve, here's the "real world" of engineering in front of you. :D

The original engineer was tasked with designing a solenoid driver that has two inputs; 230 volts and 24 volts. Perhaps some polarity neutral inputs and noise filtering.

What educated designer couldn't do this in a day? On the other hand, what educated designer couldn't reproduce this design in a day? That is what is stumping you now.

The real world says the original engineer added some obfuscation into his/her design just to make it look difficult to design and reproduce thus providing a product that he/she could sell into the future.

Just like I said in the first reply here, determine the needed specs for the application and design your own solution.

PS. Add some obfuscation into it so none of us can build a competing product; just like the original "smart" engineer. :D
 

Jeremy Harris

Senior Member
If you have a spare failed module, then I will happily take a look at it for you, for free, and diagnose the failure and see if I can work back as to the most likely cause. There's no reason why something as intrinsically simple as this should fail so regularly, unless there's a design fault or poorly specified component. Either way, I reckon it won't be a lot of work to find the cause, given a failed unit to examine.

It has to be easier (and cheaper) to just mod all the existing modules to make them reliable than to start again with a new design, I suspect.
 

Steve2381

Senior Member
Thanks guys, and that is a very kind offer Jeremy.
I will try to liberate the module out of its enclosure and we can then have a closer look. Its still currently buried in a piece of kit on a job.

These modules basically need to be 230v on = solenoid on, or 24v DC (or there abouts) on = solenoid on.

The only other reason for designing our own units was to add the option of 110v signal supply. However that is not too important right now.
 
Top