Best way to isolate the Propeller from 1200 volts - use an optoisolator?
ElectricAye
Posts: 4,561
Hi all,
I've got my Propeller taking pulses from a circuit that consists of amps, comparators, etc. whose signal sources are some photomultipliers, which involve very little current but their high voltage supplies tend to hover around 1000 volts. My concern is that something could go wrong one day and send a 1000 volt pulse through the circuits, through the Propeller, and ultimately zap my laptop. I'm not sure how likely that is, but I can't afford to smoke laptops, so I was wondering if there is a cheap way to isolate the photomultiplier side of the circuit from the Propeller, perhaps using optoisolators??? My pulse reading rate would probably not exceed 10Mhz, and would most likely be around 1 Mhz, I think. So does anybody have any suggestions for what might work nicely with the Propeller's pins configured to operate as pulse counters? What's coming out of the AND gates and heading toward the Propeller right now are 4 volt pulses.
thanks,
Mark
I've got my Propeller taking pulses from a circuit that consists of amps, comparators, etc. whose signal sources are some photomultipliers, which involve very little current but their high voltage supplies tend to hover around 1000 volts. My concern is that something could go wrong one day and send a 1000 volt pulse through the circuits, through the Propeller, and ultimately zap my laptop. I'm not sure how likely that is, but I can't afford to smoke laptops, so I was wondering if there is a cheap way to isolate the photomultiplier side of the circuit from the Propeller, perhaps using optoisolators??? My pulse reading rate would probably not exceed 10Mhz, and would most likely be around 1 Mhz, I think. So does anybody have any suggestions for what might work nicely with the Propeller's pins configured to operate as pulse counters? What's coming out of the AND gates and heading toward the Propeller right now are 4 volt pulses.
thanks,
Mark
Comments
I've started making my own PCB's because alot of parts that do a great job aren't available in DIP. TI Makes some GREAT isolators. I ordered a few samples from them a while back and they are amazing little chips. A run of the mill optoisolator is pretty slow(I've found), and you get really bad signal skewing and distortion with them. The ISO72 series uses a different method of isolating which is amazingly fast, and provides something like 5kv isolation. Anyway, attached is a picture of one of my isolator PCBs I made. It's specifically for converting up to three of these ISO7240/ISO7242 chips to a DIP mount for a breadboard. Also, here are a couple links to the chips if you want to try them. Oh, these are ---150mps--- chips. Like I said, fast. One more thing, if you haven't read the datasheet, one side of the isolator can be 3.3V, and the other 5V and anything in between. These are very versitile chips which I love(if you can't tell).
ISO7240 - 4 inputs, all same direction, 4 channels total
http://focus.ti.com/docs/prod/folders/print/iso7240m.html
ISO7242 - 2 inputs, 2 outputs, 4 channels total(can be used for 2 transmit, and 2 recieve channels)
http://focus.ti.com/docs/prod/folders/print/iso7242m.html
Post Edited (Philldapill) : 11/30/2008 9:30:56 PM GMT
*Peter*
these are certainly cool devices you both have here. My first impression is that they might be overkill for my application. I only need about 10Mhz, maybe even as slow as 1 Mhz. Also, I don't have the ability to work with anything other than Dippy thru-hole type devices, so your fancy soic solderwork leaves me pouting in the dust. Unless, that is, you are marketing these things???
Mark
I use the 6N138 & 6N139 optoisolators in DIP. I don't know if you will achieve 1MHz with these, but you could give it a try. Without looking at the specs of the 6N138/9, I don't know what their isolation voltage is.
So, in summary, you need track isolation and air isolation as well as the opto's internal isolation to achieve decent voltage isolation. You will need a fast opto to achieve the speed you are looking for. I used to design modems and had to meet these isolation requirements.
Hope this helps
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
We used to marvel at Maxwell Smart·having a phone in his shoe. Now we just say what a stupid place to put a phone!
BTW, don't expect to get Mhz out of a darlington opto, they may be sensitive but they are sloooowww.
Yeah, nice and fast so I'm sure you will have really good decoupling and ground planes to combat all that ringing
DING - next round
*Peter*
Electric Aye, no I'm not trying to maket my boards, but I did make a few extra. All the board is, is just the PCB with holes drilled, no pins. I could send you one through the mail if you're in the US. I've also made a bunch of SMD to DIP adapters since that's what I use for prototyping. These are also homemade, but work well. [noparse]:)[/noparse]
From memory I got 38.4kbaud with no problems and the resistors could be changed to improve that performance. 1MHz is still 30 times greater than this. However, if all one is looking for is a pulse, then it would be possible to get a lot better than 38K. As I said, I did not check the specs, and thought it could be worth a look. I know the 4N28 will not do this. You will need a high speed opto. There are specialised digital optos that may also do the trick.
Otherwise, you may have to use a transformer to get the isolation.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
We used to marvel at Maxwell Smart·having a phone in his shoe. Now we just say what a stupid place to put a phone!
that's a very good point about duty cycle. I had overlooked that. Cluso, too, makes a good point about layout, though I'm protoboarding everything right now, so the separations are pretty good. I'm not sure how to approach using the parts Peter suggests since I do not have the ability to solder surface mount anything. Unless somebody is marketing something I can just plug into a protoboard and solder, I'll have to keep looking.
thanks to all for keeping me from making another stupid mistake,
Mark
You'll never know how much I've agonized over your super-generous offer of a freebie isolator, but as much as I'd love to snap it up, I'm afraid it goes against my policy of forcing myself to learn everything I need to do to make my project work. I've thrown myself into this project with a sort of "sink or swim" mentality, so if I need isolators and if the ideal isolators require a whole new way of flailing in the electronic waters, then flail I must. I shan't allow myself to get greedy. Your advice and sharing of knowledge is more than I could ever ask for. I am now jonesing one of those fancy soldering stations as seen on Sparkfun, especially after spending time last night oogling their tutorials. They have some SOIC to DIP adapters that look possible.
I greatly appreciate the offer but for me you've done enough already by encouraging me to learn to deal with those surface mount teenie-weenie leads.
thanks again,
Mark
PS. If I screw up really bad, though, I might come back begging with my tail between my legs.
*Peter*
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
We used to marvel at Maxwell Smart·having a phone in his shoe. Now we just say what a stupid place to put a phone!