One of the challenges in an amateur lab is the control of apparati like solenoid valves, pumps, etc, down to the second or less. The T-cell, when operating, requires large and regular dosages of HCl to keep the pH in the optimum range, as close to 6.8 as you can make it.

The first thought is to use a pH controller to do the job. The simplest controllers are those sold to the aquarium market. When the pH reaches a setpoint, the controller turns on a simple outlet. Plugged into that outlet is whatever device you are using to lower the pH. Typically, it'd be a CO₂ system, or some sort of acid pump.

But a perchlorate cell is not an aquarium. The liquors are incredibly harsh, full of potent ionic species that can play hell with even the very best pH probes. By definition, a controller requires continuous immersion of the probe in the solution of interest. A good pH probe is an expensive, precision, and complicated device. To the uninitiated, they appear sealed, but if truly sealed, then they would not function. Inside the probe is a reference cell, and there is a path from that reference cell to the outside world. We know what happens when you have two solutions of differing ionic strengths, and a path between them. You get ion migration, and the precision reference cell in the probe is poisoned, to use an industry term. The liquor in a (per)chlorate cell is especially toxic to pH probes, and continuous immersion spells eventual death to the probe. How long it would take, I'm not sure, but the cheaper the probe (single junction, open reference cell) is, the faster it is going to first drift, and eventually die.

Rather than use a true controller, I settled on timed and controlled dosages of HCl to control pH. The last run of the T-Cell proved to me, at least, that this was more than satisfactory, as the amount of HCl required per mole of product is known by the industry. The HCl pump I've been using is a surplus Hanna dosing pump (eBay) that has all of the wetted head components composed of PVDF, PTFE, and glass; perfect for everything except HF and perhaps a few organic solvents. The problem it gave me was that it did not contain an internal timer. Once on, it remained on, and even on the lowest setting, it would have rapidly overdosed the cell with acid.

I added a cheap consumer timer to it. This timer could not be programmed to deliver less than 1 minute of ON time per day, with a maximum of 24 events per day, meaning I could turn it on once, for 1 minute, every two hours. One minute is too coarse! I needed control down to the second. I also wondered at the reliability of the timer... if it stuck ON, I'd be faced with a major HCl overdose.

I began to browse for effective and inexpensive process (industrial) timers. There are a LOT of them out there, especially from Omron. But one on eBay particularly caught my eye, available from Auber Instruments.

Thirty two bucks! Nice. I place the order, and when it arrived, ran it through its paces. It was going to be perfect, having a range from 0.01s to 9999 days. I gathered enough components to make a complete system a reality.

1. The timer itself
   
2. A suitable box
   
3. A solid state relay to carry the load of the Hanna pump
   
4. A matching outlet
   
5. A fuse
   
6. and an indicator light
   

The first step was to slot and cut the lid of the container, a PVC plastic junction box from Home Depot, for the components. This was done on a vertical mill, but could just as easily be done with a scroll saw or similar. These PVC boxes are awesome. Don't buy a box from Radio Shack, get one of these. They are a vastly better value, and cut and machine beautifully.

Since the edge of this particular cut would showm I wanted to make the cut for the outlet as cleanly as possible, and was pleased with the result:

With project like this which is ultimately placed in an enclosure, it is usually easiest to mount as many components as possible on the lid. It makes wiring much simpler at the end of the project, and it also makes maintenance easier.

The SSR (Solid State Relay) is a 40 amp device that is triggered with an AC signal, so the internal workings of this timer are all at 120V AC; no DC power. Normally, you'd want to mount an SSR on a heat sink, but given the low current of the pump, and the very intermittent duty, I don't think heat will be an issue.

It is always wise to fuse such a unit. In this case, I am using a 20 amp microwave oven fuse in a commercial holder. It can be seen located just above the empty cutout in the picture above. Again, stuff like this can be found at Lowe's or Home Depot; no need to go to a specialty electronics store.

With the lid inverted onto the box, I began the wiring process. The 120V mains hot line (black) goes first to the fuse, and from there, it is daisy-chained to both the switch and directly to one output terminal of the SSR. The other SSR output terminal goes to the outlet hot terminal screw. From the switch, the 120V is also routed to the timer, and additionally branched to the timer's mechanical normally open relay. The other side of the relay transfers 120V to the input side of the SSR; thus, when the timer's relay closes, it delivers 120V to the SSR, turning ON the SSR output and powering the normal electrical outlet on top of the box.

The neutral (return) lines from the outlet, SSR input, and the timer, were tied together for eventual attachment to the power cord. Finally, a GROUND line was secured to the outlet's grounding lug. The finished unit:

Initial tests: With the unit plugged into the wall outlet, I applied a DVM to the unit's outlet, and was a bit shocked to see 120V! Shit. Then I remembered that all solid state relays have a tiny bit of leakage in the OFF state. With nowhere to go, the leakage charges the outlet to a full 120V, but it is like static electricity - all voltage, no current. This was proved by plugging a radio into the device - it did not turn on. Switch on - the timer turned on, no smoke, looked good. I ran a quick timing cycle and watched the neon light turn on at the correct time, and the radio began to play some music. Five seconds later, it dropped out and the cycle began again. Everything worked as designed, and as programmed.

I realize I have been doing a lot of electronics recently, but it is all in support of electrochemistry. This unit will be ideal for HCl pump control. When powered, the Hanna pump is a stroking pump. A piston is retracted, the pump head fills with acid, and the piston is released, creating a fairly loud CHUNK noise. Each pulse delivers exactly 1ml, so it is easy to determine how much has gone into the cell. Further, I am interested to see if the data collection system will detect the addition of HCl when it is injected into the cell. My guess is, I will see a distinct dip in the cell voltage. At a constant current, added HCl will ease flow of electrons, and the voltage should dip for a short time as the HCl gets disbursed.

I metioned the voltage leak thing to the gang at SMDB, and hopefully someone will have a thought. It won't hurt anything, but it bugs me, and I'm thinking a resistance to "ground" will drain away any leakage, but when the SSR is ON, the resistor will see a full 120V, so power dissipation may be an issue.

I promise very soon to get off the electronics bandwagon, and get back to our favorite oxidizers.