Reversing polarity for electrolysis project

[BHinote]

The light at the end of the tunnel is "right in front of my face".

Ronv,

I made a couple adjustments, which I will share with you a little later, but everything is looking awesome. (The timer and NOR Gates are switching sides, I added LEDs to indicate which is on, the ma value from the top to the middle of the H-Bridge are only loosing 0.X ma , and I added a switch to turn the timer off so it can be run without switching so the ma can be adjusted without it switching or event to run constant...)

Anyway, I have a question and/or concern about the Resistor being used in the middle of the H-Bridge.

Obviously the current resistor value of 2.5K is producing the milliamp results that we want, however when this resistor is changed the milliamp value also changes. The design must maintain a Constant Current Value throughout the entire process. So when the Resistance Changes the Current Must be adjusted accordingly. This is why I originally used the LM317 for my Constant Current Source. It should adjust the current based on the Load/Resistance change… At least I think this will occur…

What are your thoughts on this and am I correct that changing the resistor should not affect the current when the current is regulated?

Best Regards,
BHinote

 

[KeepItSimpleStupid]

A couple of things. 3700 ohms would be the maximum value of R for 10 mA (37/0.010) given a 37V supply. Actually it will be a bit less. The resisitor should not affect the current unless it's greater than ~3700 ohms.

-----

I uploaded the simplest drive interface for the LCC110. At least this much you can model.
Lots of devices can be used for the 2n3904 and base resistor. The ULN2003 is my favorite.

if you incorporate it together with the H-bridge, you need to connect the NC (Normally closed) and NO (Normally open) MOS transistors so that the polarity matches the H-bridge.

LCC110's are about $4.74 each
a 2N3904 is about 0.42 each
The 4.7K is about 0.45 in metal film 1/2W
the 390 ohm is 0.23 in metal film

OOPS. D1 is backwards!

 

[BHinote]

KISS,

I have a couple of Questions:
1. It is correct of me to assume that the Resistor in the middle of the H-Bridge is only for simulation and will be removed in the actuall circuit? (i.e. Resistor comes out and each end becomes the connections to the Silver Electrods.)
2. The Resistor and Potentiomater used to adjust the Current, will still be correct values to adjust simular results when attached to the Silve Electrods. (i.e. Adjust the Current between 1ma and about 10ma or so.)
3. I am not sure that I understod your drawing as you intended, but I attached a mock-up of what I thought you ment. (Please fogive my Ignorance...)

Best Regards,
BHinote

A couple of things. 3700 ohms would be the maximum value of R for 10 mA (37/0.010) given a 37V supply. Actually it will be a bit less. The resisitor should not affect the current unless it's greater than ~3700 ohms.

-----

I uploaded the simplest drive interface for the LCC110. At least this much you can model.
Lots of devices can be used for the 2n3904 and base resistor. The ULN2003 is my favorite.

if you incorporate it together with the H-bridge, you need to connect the NC (Normally closed) and NO (Normally open) MOS transistors so that the polarity matches the H-bridge.

LCC110's are about $4.74 each
a 2N3904 is about 0.42 each
The 4.7K is about 0.45 in metal film 1/2W
the 390 ohm is 0.23 in metal film

OOPS. D1 is backwards!

 

[BHinote]

Constant Current LM317

When I use one of the LM317 Constant Current Calculators, found online, they are indications that 100 ohms = 12.5mAs and 1106 ohms = 1.3mAs. This appeared to work the first time that I only used the LM317 to product actual Colloidal Silver, but I am not concerned that this is not going to be the case when the complete circuit is employed...

Any thoughts...

BHinote

 

[ronv]

BH,
The resistor represents the resistance of your solution. The current source can only maintain constant current as long as there is enough voltage and low enough resistance for it to maintain the set current. There is a certain amount of overhead in the current source above and below which it can't regulate. The transistors perform the same function as the regulator in your first circuit except the regulator became unstable ( Oscillated) when the current was below it's minimum. I'll run you a simulation of them both.

 

[KeepItSimpleStupid]

1. Yes
2. Yes
3. My fault too. Migraine day.

Remove the right hand transistor, one LED and both resistors.
Remove one LED from the left side of your drawing.

----

LED1 will never turn on. On your other drawings you had an inverter which is needed
To drive LED1.

The idea is to have the LEDs in the LCC110's in series with the bands toward ground.

One LED belongs inside one LCC110 (pins 2 & 3 from memory)
and the other LED in my drawing belongs in the other LCC110.

My schematic is WRONG in the sense D1 needs to be turned
180 degrees.

The 74HCT05 replaced the transistor and 1 resistor.
The added resistor to +5 converted the OPEN drain inverter to. a conventional
Inverter.

We can do the drive and LEDs with a ULN2003 which is a very common part and a resistor.

 

[BHinote]

Sorry to hear about the Migraine. Hopefully this does not intensify it...

I understood most of what you were indicating, but not all. Sorry...

1. You indicated the LED arrangement I setup to indicate Left or Right electrode was Positive, would not work. (OK, the simulation seemed to like it, but who am I to contradict this with an expert. Answer: Nobody...)
2. I believe you indicated that the two LEDs in you LCC110 box actually represented to two sides of the H-Bridge, so I believe the attached drawing addresses this.
3. You indicated the 74HCT05 would replace the transistor and 1 resistor, but I did not see these on your latest upload, so I looked back at your prior upload and believe that I changed it accordingly... I Think...
4. Added resistor to +5 converted the OPEN drain inverter to a conventional inverter. (I am not sure what this means, but again I tried to combine both of your drawings into the one I am uploading now.

Please let me know if I am close to what you were saying or simple missed the boat...

Best Regards,

BHinote

 

[BHinote]

Ronv,

Understood... I assumed that if I changed the Resistance on the center Resistor the numbers would still be the same, which is what should happen in the solution with Constant Current. However, I am not sure why it does not follow this type of pattern, if it is a simulation of what should happen... (i.e. When the resistance of the solution changes the current should remain the same so as not to allow it to runaway when the resistance is much less. More Parts per Million.)

Is this correct or ????

Best Regards,
BHinote

BH,
The resistor represents the resistance of your solution. The current source can only maintain constant current as long as there is enough voltage and low enough resistance for it to maintain the set current. There is a certain amount of overhead in the current source above and below which it can't regulate. The transistors perform the same function as the regulator in your first circuit except the regulator became unstable ( Oscillated) when the current was below it's minimum. I'll run you a simulation of them both.

 

[BHinote]

I have to go help my dad, but when I get back I plan to do the following:
1. Upload the latest schematic.
2. Put together a parts list.
3. Pray that I don't overlook something and have to wait for another order to arrive.

I greatly appreciate all the help that both of you have been providing me.

I know that the assembly could still present a problem, at least with me assembling it, but I do not expect this to be a problem.

Best Regards,

BHinote

 

[ronv]

I haven't been able to show the results as a curve. But the basics are that I = E/R. So If we had the whole 37 volts to work with (we don't, we loose some) the maximum resistance we could have in the solution would be 3.7k for 10 ma and 37k for 1 ma. So anything lower that 3.7k we can do whatever we want up to 10 ma. The current source is not perfectly linear. What are you seeing?

 

[KeepItSimpleStupid]

I'm about an hour away from home now.

Your close.

Pin 5/8 of the LCC110 has to go to ground.

Usually an IC with multiple gates use U1a, U1b, U1c etc. U1 is the IC # and a-f are the 6 parts.

Gates depending on the family have different characteristics especially asymetric source
And sink current. LED's should not see a reverse voltage although 7 V. is usually OK.
Sink current is usually higher. Circut would be virtually the same except for pin numbers.
There are 7 drivers in the ULN2003 and 6 in the 74HCT05.

Can you upload a pic of the LCC110 that you embedded in your drawing?

If you use a 74HCT05 it HAS to be surface mount. If we use a ULN2003 it can be a DIP
Package. Your choice.

So basically the inverters need to only SINK current rather than source it to drive LEDs and relays.
The ULN20003 is a 7 relay driver with an interesting property. If an input is unconnected, the "relay" is off. I've used this part a lot.

Layout can change a design. E.g It might be easier to route and extra gate and resistor because resistors can jump tracks.

The scanned scribble I did had LED indicator drivers (3 gates and a resistor). It had. 2 gates and 2 resistors for the LCC110 which can be simplified to 1 gate and 1 resistor.

Sorry for the confusion. I tried to download LTSpiceIV today and got an error: This is not a WIN32 application.

I may be back at this in an hour, although it's bedtime.

 

[BHinote]

Honestly, I just thought if I changed the center resister it would be like the water changing resistance. So I tested tape water, which will have less resistance than Distilled Water, but substantually more resistance than the solution when it reached more of the consistance that I am aiming for. I decided to bump the value higher and lower and expected to see the ma stay the same. When it didn't I wasn't sure why. (i.e. When I first just had a LM317, Potentiomiter and a resister, the processes started at lower than 1ma. However as the silver started building in the Colloid the Constant Current kept adjusting to maintain 1ma - 1.5 mas.)

Anyway, regardless of what the Simulation does I am hoping that simulare results will be seen with the much improved complete curcuit...

Best Regards,

BHinote

I haven't been able to show the results as a curve. But the basics are that I = E/R. So If we had the whole 37 volts to work with (we don't, we loose some) the maximum resistance we could have in the solution would be 3.7k for 10 ma and 37k for 1 ma. So anything lower that 3.7k we can do whatever we want up to 10 ma. The current source is not perfectly linear. What are you seeing?

 

[BHinote]

KISS,

If I understood you comments, is this what you were sugesting?

I am unable to sumulate this design with Multisim, so it would be untested.

BHinote

I'm about an hour away from home now.

Your close.

Pin 5/8 of the LCC110 has to go to ground.

Usually an IC with multiple gates use U1a, U1b, U1c etc. U1 is the IC # and a-f are the 6 parts.

Gates depending on the family have different characteristics especially asymetric source
And sink current. LED's should not see a reverse voltage although 7 V. is usually OK.
Sink current is usually higher. Circut would be virtually the same except for pin numbers.
There are 7 drivers in the ULN2003 and 6 in the 74HCT05.

Can you upload a pic of the LCC110 that you embedded in your drawing?

If you use a 74HCT05 it HAS to be surface mount. If we use a ULN2003 it can be a DIP
Package. Your choice.

So basically the inverters need to only SINK current rather than source it to drive LEDs and relays.
The ULN20003 is a 7 relay driver with an interesting property. If an input is unconnected, the "relay" is off. I've used this part a lot.

Layout can change a design. E.g It might be easier to route and extra gate and resistor because resistors can jump tracks.

The scanned scribble I did had LED indicator drivers (3 gates and a resistor). It had. 2 gates and 2 resistors for the LCC110 which can be simplified to 1 gate and 1 resistor.

Sorry for the confusion. I tried to download LTSpiceIV today and got an error: This is not a WIN32 application.

I may be back at this in an hour, although it's bedtime.

 

[BHinote]

Ronv,

Please find attached the latest copy of the Schematic.

I probably will at a ON/Off switch for the entire circuit and make adjustments to the Left and Right LEDs. However do you see any other issues.

Best Regards,

BHinote

I haven't been able to show the results as a curve. But the basics are that I = E/R. So If we had the whole 37 volts to work with (we don't, we loose some) the maximum resistance we could have in the solution would be 3.7k for 10 ma and 37k for 1 ma. So anything lower that 3.7k we can do whatever we want up to 10 ma. The current source is not perfectly linear. What are you seeing?

 

[ronv]

Looks like it should work.
Good luck!

 

[BHinote]

Thank You

Ronv,

Thanks again for all that you have done.

I am going to create a parts list and try to nail down where they can be sourced. (It appears that the majority of the parts I previously bought will not be useable , so all new parts it is...)

Best Regards,

BHinote

Looks like it should work.
Good luck!

 

[ronv]

Try Mouser.com
The shipping will be more than the parts.

 

[ronv]

BH,
Before you order parts, can you measure your power supply with nothing attached? The 2N39XX transistors are only rated for 40 volts. If your supply is not well regulated with low load we might smoke them. Since you have to order parts it might be safer to order higher voltage transistors.

 

[BHinote]

Ronv,

It is reading 40V, with the multimeter directly connected to the power supply plug.

Does this mean that we should go up and will this change any of the resister values currently used?

BHinote

BH,
Before you order parts, can you measure your power supply with nothing attached? The 2N39XX transistors are only rated for 40 volts. If your supply is not well regulated with low load we might smoke them. Since you have to order parts it might be safer to order higher voltage transistors.

 

[ronv]

No resistor changes, but we should change the 2N3904 to BC547 and the 2N3906 to BC557. Make sure you look at the picture to get the thru hole part and not surface mount, cause they are really tiny.