Hi all I need some advice, i am not a electrial guy so hope some one can help me on this
Current i need to design some sort of electrical control switch to distrubute a signal of varible voltage from 0 to 80V of current less than 50mA.
It will be a one to four configuartion where the path will be control electrically by PIC.
I was thinking of using relays for that but i am not very sure if this is advisable.
> Current i need to design some sort of electrical control switch to
> distrubute a signal of varible voltage from 0 to 80V of current less
> than 50mA.
It will be a one to four configuartion where the path will be
control electrically by PIC.
I was thinking of using relays for that but i am not very sure if
this is advisable.
_____________
Tell us more.
This job can be done easily by a number of means. Could use relays or
FETs or bipolar transistors or ...
What is the variable voltage usef for?
Can the output devices all have common grounds. or common positive
supplies, or do they need to be separated from each other. ?
What accuracy of switching do you need?
How is the variable voltage made? - is it varied by some other control
or do you require the switch/distributor to be the voltage control as
well.
> I was thinking of using relays for that but i am not very sure if this
> is advisable.
Sometimes relays are a viable option. I've a couple of products that
use them rather than FETs because of the crossover switching needed.
I get them for a good price, less than it would cost to do it with
transistors
My project is to increase the number of possible channel but at any one time
the same number of channel support by the device
I just need to relay the signal form one electrodes to the other so i
thought of using a relay (contact switch )so i will not physically
affect the signal if i were to use transistors i may indirectly change the
source signal which is what i dont want to.
where as for switchng accuracy or frequency maybe one or two change per sec
so i think relays should be able to do it fine
as the signal is control by the TENS and i am not making any changes what i
need is to change the path of flow to different electrodes. that all sound
simple :)
> Thanks for the reply
>
>
>
> The siganl source is from the follow device.
>
>
>
> http://www.paintechnology.com/046.htm
>
> TENS, Transcutaneous Electrical Neural Stimulation
>
>
>
> My project is to increase the number of possible channel but at any one time
> the same number of channel support by the device
>
> I just need to relay the signal form one electrodes to the other so i
> thought of using a relay (contact switch )so i will not physically
>
> affect the signal if i were to use transistors i may indirectly change the
> source signal which is what i dont want to.
>
>
>
> where as for switchng accuracy or frequency maybe one or two change per sec
> so i think relays should be able to do it fine
>
>
>
> as the signal is control by the TENS and i am not making any changes what i
> need is to change the path of flow to different electrodes. that all sound
> simple :)
>
>
>
> I am confuse with what type of relay I can use
>
>
>
> can you kindly please advice me
>
>
>
> Regards
>
> Soon Lee
>
>
>
>
> WebSite : http://soonlee.agreatserver.com
> The siganl source is from the follow device.
>
> http://www.paintechnology.com/046.htm
>
> TENS, Transcutaneous Electrical Neural Stimulation
>
80V 50ma presents a problem for relays especially if the signal has a
DC current associated with it.
>
>
> My project is to increase the number of possible channel but at any
> one time the same number of channel support by the device
>
So you are basically switching what path the signal takes at any given
moment?
> I just need to relay the signal form one electrodes to the other so i
> thought of using a relay (contact switch )so i will not physically
> affect the signal if i were to use transistors i may indirectly
> change the source signal which is what i dont want to.
>
A relay will affect the signal, in fact no matter what you use it will
affect the signal somehow. You could make a high voltage CMOS
demultiplexer thay you drive with a digital devices. This would
require 2 pins from the pic for selection and one for deselecting the
device (turning it off completely so no signal is passed).
> where as for switchng accuracy or frequency maybe one or two change
> per sec so i think relays should be able to do it fine
>
Relays are not reliable enough. Let's take there life expectancy
first.
Typical mechanical is 1000000. Once a second lets say. These operate 8
10 hours a day maybe? That's 28800 to 36000 actuations per day. You
will be replacing relays maybe once a month. Not reliable enough. Yes
you can get higher actuation cycle units 5 10 million. This just shifts
the time from 1 to 4 and 8 months respectively.
Your next problem is the voltage, it's too high. Especially if the
signal has ANY DC component in it, you are at too high a voltage for
the majority of relays. You could use MDR's for this (mercury
displacement relay), but the DC component is something you need to
avoid with all relays. This creates arcing and even if you 'say' it's
50 ma arcing causes current surges. Relays are just to well suited for
this at all. 32VDC is the common highest voltage for DC switching on
relays. It's that for a good reason. MDR's also contain mercury, so it
has a hazardous material problem.
>
> as the signal is control by the TENS and i am not making any changes
> what i need is to change the path of flow to different electrodes.
> that all sound simple :)
>
> I am confuse with what type of relay I can use
>
> can you kindly please advice me
>
If you are using this for medical purposes you have to go for ultra
high reliability. The TENS units are designed for single signal path
and are made for only that type of application. Using an NMOS
transistor as an analogue switch is perfectly fine for this
application. Since 50ma is your max current you can use relatively
inexpensive 100V transistors for the application. Switching the
transistors might be a bit more challenging, but it is rather unlikely
the signal will be attenuated by a measurable amount (as in if you are
off 1/10 of a mv on an 80V signal I don't think you will be in
trouble).
For any medical application you must take reliability and safety
precautions up 3 orders of magnitude. These are the most important
features, and supercede the fact if the device works or not.
__________________________________
Do you Yahoo!?
Yahoo! Mail - Easier than ever with enhanced search. Learn more. http://info.mail.yahoo.com/mail_250
> The siganl source is from the follow device.
>
> http://www.paintechnology.com/046.htm
>
> TENS, Transcutaneous Electrical Neural Stimulation
>
> My project is to increase the number of possible channel but at any
> one time the same number of channel support by the device
Rather than switch the signal, why not use amplifiers to copy it ? Then
switch signal paths on or off
Another option could be to use a PIC to examine the original signal
(2-150Hz, 60-250us pulse width). It could then drive a FET or FETs.
An 80V supply is not too difficult. Simple oscillator/inductor for example
>> The siganl source is from the follow device.
>>
>> http://www.paintechnology.com/046.htm
>>
>> TENS, Transcutaneous Electrical Neural Stimulation
>>
>> My project is to increase the number of possible channel but at any
>> one time the same number of channel support by the device
>
> Rather than switch the signal, why not use amplifiers to copy it ?
> Then
> switch signal paths on or off
>
> Another option could be to use a PIC to examine the original signal
> (2-150Hz, 60-250us pulse width). It could then drive a FET or FETs.
> An 80V supply is not too difficult. Simple oscillator/inductor for
> example
IF the signal is AC relative to ground (as something in the text on
that page suggests it MAY be) then solid sate solutions are not as
simple as it the pulses are always say positive relative to ground.
Despite the good stuff that was written about the reliability of
relays, I consider that they would be by far the easiest way of trying
the idea out and may be an excellent long term solution. Cost of
trying this out is relatively low and implementation is simple. IF you
can arrange the relays to switch when the TENS pulse is not present
then you could get a very long contact lifetime. Also, I expect skin
current would in fact be rather lower than the peak currents
mentioned. If you switch the relays with low voltage battery based
controllers then safety aspects would be easily enough addressed.
Mains powered devices are not at all recommended unless you are
exceedingly competent.
Solid state switching provides a superior solution but requires a far
better knowledge of the signals involved.
The TENS units can be expandable by 2 according to the website, instead of
using the connector given i was think of using a push button to on and off
the path but pusbutton is not controllable that is the main intention of
usng a relay or some sort. Relat is just a contact (just like a push button)
it will just connect and disconnect the signal.(my intention)
but do you have any circuit for the method you have suggust,(sorry i am not
a electrical guy)
> If we copy the signal am i changing the strength (current of the
> signal)? The TENS works on current and it use current to stimulate
> the human
You might even improve the signal. I've got a TENS unit here that's
not very good quality (left here for me to have a look at years ago
and never picked up). I forget the details but do remember it uses
output transformers which put out up to 300V unloaded in very
short pulses that are hardly square. The build quality is not good
and the output would be much better with a couple of dollars spent
on it.
You could find that the units you're looking at are average performance
for an average price (for the lowest production cost), which you should
be able to enhance. You know how it is - many products off-the-shelf
are optimised for as wide a range of customers as possible and can be
upgraded
It would be quite easy to make a PIC-based TENS unit. I'd be surprised
if there isn't one out there already. I got into one that a flatmate had in
1990, costing over NZ$500 (US$350), which was based on a lowly 6805
and micros have come a long way in 15 years
singnet.com.sg> wrote: