VCCS questions

[NeoMoses]

I've been doing LOTS of reading and all I can say is Wow! I love the new LCD anodizing instructions. Excellent documentation! Those are the kind of answers I've been looking for. Did you write that documentation, FiberGeek? The instructions sound a lot like your writing.

Anyway, I've been planning an upgrade to my existing system, and the new LCD system looks like a good way to go. The only problem is this. I bought all of the components necessary to build 2 large, unregulated, 12V power supplies (Transformers, bridge rectifiers, LARGE capacitors...) I have everything needed for the "old" method. I plan on simply adding another anodizing kit, and keeping the old one in service while I ensure that I can get quality results with the new kit. (I've seen how touchy anodizing can be to changes... Just covering my butt. )

I see in the LCD instructions that the author built a VCCS (Voltage Controlled Current Source). How difficult is this? From the picture, it appears pretty simple, using a rheostat, an IC, some resistors, and possibly other stuff. Does anyone have any good instructions or good links for making a current source? I just don't really want to pay the $300+ for a rectifier when I have all the materials for 2 large PSes. If it's possible to add some control circuits to a PS to make it a current source, I'm all ears.

I noticed that with higher current densities, peak voltage can top 20V, so I'm assuming I should wire these 12V transformers in series to make a 24V PS. Any other tips?

[Fibergeek]

Thanks for the kind words. Mike wrote the LCD instructions, I wrote the paper that the LCD instructions are based on.

I make my living as a circuit designer; very advanced and expensive telecom circuits and systems, I've been doing this for nearly 30 years. If you have the necessary electronics backround, a current source like a VCCS is child's play. BTW, a VCCS is only 1 of many available current source designs. Since I don't know your backround, I can't say if it would be easy or hard for you.

You need to start by first determining two things; the maximum surface area you want to do, and the maximum current density for want to operate at. Based on these decisions, you will be able to see if you have the right components. If what you have isn't right, you can adjust max surface area and/or max current density so that it is right.

[NeoMoses]

Let's say I want to start out small, with a maximum surface area of 1 square foot and a maximum current density of 6 A/ft^2.

I'm a mechanical engineer, so I'm somewhat limited in my circuits knowledge, but I'm always willing to learn! I'm a quick study, so feel free to drown me in knowledge, I'll learn to swim.

[Fibergeek]

Under your conditions of 1 sq.ft. max surface area; and 6 A/sq.ft. max current density, and also assuming PAR = 2.5 Ohms/sq,ft.,
you will need:

V = I x R, V = 6A/ sq.ft. x 2.5 ohms/sq.ft. = 15 Volts.

This doesn't allow for any voltage loss due to less than ideal electrical connections. So add say 15% for this:

15V x 1.15 = 17.25 Volts.

If you were tightly bolting the anode connection to the work with aluminum hardware, like I did in my paper, you would not really need to add this 15% to the required voltage. If on the other hand, you were just hanging the work from aluminum or titanium hooks, as shown in the instructions, the connection resistance could be quite high, requiring much more voltage to overcome. doubling the voltage requirement to 30 volts would not be unreasonable.

Let's assume you will tightly bolt the connection, so you only need 15 Volts. Now for the required current :

6 A/sq.ft. current density x 1 sq.ft. area = 6 Amps anodizing current.

So you need to have your VCCS be capable of 0 to 15 Volts, and 0 to 6A minimum.

If you go to the online manual on the caswellplating.com site, at the bottom of the page, click on the title "Building a Power Supply". The circuit described by Don Foreman is a VCCS. I would replace the Darlington power transistors he used (TIP 120) with power MOSFETs, which are now cheaper than Darlingtons, easier to design circuits with, and are easier to get these days. For 6 Amps, you would only need one power MOSFET, two if your heatsinking wasn't that good. Start there.

Next question; your 12 V transformers, what is their output current rating (in Amps) or their VA (Volt x Amp) rating?

Next question; your rectifiers, are they diode bridges (4 diodes in one package) or full wave rectifiers (2 diodes in one package) ?

Next question; what is the voltage and capacitance (microfarads) ratings of the big capacitors you have?

[NeoMoses]

I don't think I currently have access to the online manual, but I may purchase it if necessary.

My transformers are 10A each, 120 W I have 2 of these.
My rectifiers are diode bridges.
The caps are 27,000 uF, 63V.

I think I've sufficiently sized the components to make a large power supply. If at all possible, I'd love to be able to use the 2 transformers in series to make a 24V/10A VCCS.

[Fibergeek]

Unfortunately, you do have a sizing problem.

If you wired your two 12V 10A secondaries in series, and then used one of your diode bridge rectifiers, you will get about 31 VDC at 5A, not 10 A. If you allowed the voltage to sag you might get 10 A, but the voltage would be around 16 Volts or so. The transformers will also get hot.

If you wired the same transformers in series, and then used a full wave rectifier you would get about 14.8 VDC at 10A.

Like everything else in life, there is no free lunch in electronics. If you could return the 12V 10A transformers and get one of the proper size, you would be better off. Either that or reduce your max voltage and/or current specs.

[NeoMoses]

OK, I now see the error in my thinking.

How about this transformer available from MPJA? It's a 24V, 10A transformer.

http://www.mpja.com/productview.asp?product=7846+TR

I just paid the $15 for the online manual, and I'm looking through it now. I'll let you know if I have any questions.

[Fibergeek]

If you used that 24V 10A transformer with a full wave rectifier (not the bridge) you would get 14.8V - 15V at 10A, depending on the actual diode losses. You won't make it to 20V, you would need a 30V secondary to do that.

I'll show you how to select some power MOSFETs to replace the Darlingtons when you are ready for it.

[jnbsystems]

Hi guys,

I'm going to jump in here, since I am in the same boat - trying to come up with an "economical" constant-current power source.

I too have looked at the "Forman Paper" in the online manual, and I think I am going to use this a basis for my supply. My education in electronics is a little more than basic, but not nearly as high as Fibergeek and others. Any help here will be greatly appreciated, and my results will be duly reported here

I found a transformer from Stancor, part number RT-206. It is listed as a "Rectifier Transformer", and has several primary taps, and a single secondary with a center tap.

MPJA has the transformer, listed here: http://www.mpja.com/directview.asp?product=14466+TR
The spec-sheet is here (PDF): http://www.mpja.com/download/14466tr.pdf
It looks like this unit will supply around 29 volts at 6 amps when used in Full Wave Bridge mode. Am I looking at this correctly?

Stancor lists several other models with much more power in the PDF sheet, but MPJA does not list them.

Fibergeek, I will also need some help with the power MOSFETs. I know what they are, but I have never used them. How are they used in this application? I am just about ready to purchase some parts.

Thanks!

[Fibergeek]

29V at 6A DC is what it says. You would be able to reduce the DC voltage to something in the low 20s or so by selecting the appropriate primary taps, you will still get 6 A with a FWB rectifier. You really don't want too much more voltage than you need, because it will make the power dissipation go up, and thus heatsinking gets harder. If you are "racking" parts, or hanging them from hooks, you will want 10-20% more voltage available to compensate for the connection loss. If you tightly bolt the connection to the work, you won't need this extra voltage.

I notice that MPJA also lists a 13.5 V 20 A transformer, that would provide 17-18 VDC 10 A, including diode losses, with a FWB.

I havent heard anyone talk about an enclosure yet, only a fool would leave any of the 120 VAC wireing exposed. You also need a fuse or circuit breaker in series with the AC line input.

You guys have a ways to go before your're ready to select MOSFETs. I will make some minor refinements to Don's design besides MOSFETs, to make it better suitable for anodizing. Give me a few days.

I would suggest www.digikey.com or www.jameco.com for the MOSFETs. opamps, resistors, etc. that you will need. You will pay through the nose at RadioShack, and their selection sucks.

[jnbsystems]

I'm concerned about the connections to some of my pieces, which is why I was looking at the higher voltage transformer. Although they are very small, some of the parts I have may only have the wire wrapped around them in one place. I am still trying to work out the total surface area of these little buggers. I have them drawn up in Autocad, and I'm trying to cheat by using the software, instead of actually measuring them (without much luck, I must admit ). They have several complex shapes, external threads, holes, o-ring grooves, etc. I will probably just end up using estimating and trial-by-error.

I will look at addressing the enclosure issue, since I've done some panel assembly and industial electrical work.

I'm curious about the heat sinks. I would think there is some way of calculating surface area for the power devices, assuming things like max power handling, ambient air temps, convection vs forced air cooling, etc. I've looked at quite a few various sizes, ranging from very small to very big. What is the best way to do this? It will aid in figuring enclosure configurations and such.

[Fibergeek]

Yes the exact shape/size of the heatsink can be calculated, but none of you will like the math.

In Jnbsystems case; his worst case power dissipation would be 29V x 6A = 174 Watts. In the case of a VCCS used for anodizing, the worst case only occurs at maximum current (6A) and at startup, where the output voltage will be very low, and the VCCS has to absorb most of the 29 V. Fortunately this condition only exists for a matter of minutes, after some anodize forms, the required voltage goes up, taking some of the load off of the VCCS.

I noticed at MPJA a heatsink; 9206 HS, that is already drilled to mount three TO-220 packages, it costs about $5. TO-220 is the package that we will select for the MOSFETs. If you used this heatsink with a small 120 VAC fan; or even better, a small blower, moving air along the fins, you should be fine. MPJA has suitable fans and blowers.