Hi,

I am working on a device using four DG408 analog multiplexers and am looking for the most convenient way of converting 3.3 V or 5 V single supply voltage into +/- 15 V dual voltage needed by multiplexers.

LT1026 would be the perfect solution but it can only double the input voltage. I was thinking about rising the voltage with LT3433 to about 20 V, and then using 7815&7915 but LT3433 needs too many (and too complicated) external parts.

The next I was thinking was to use MAX232's RS232 transmitting outputs as +/- 15 V sources. Of course, the input voltage would be 5 V and I would set the appropriate input TTL levels. Has anyone tried something similar? I am not sure if it would work. However, I've found this article and it says: "On the IC, use the 'negative output supply voltage Vs-' pin (pin 6 shown below) for up to negative 15V and up to 10mA of current.". I am not sure how is that possible :-/ According to the datasheet, Vs- is -10 and not -15 V. RS232 transmitting outputs are +/- 15.

I hope someone could recommend a solution that won't take too much space on a board. And if it is some low cost solution - even better

 

Use a dual output flyback converter. National Semi has their simple switcher line that will give you a design from their software.

FYI: whenever you try to design a power supply, you have to specify the load current required.

 

Use a dual output flyback converter. National Semi has their simple switcher line that will give you a design from their software.

Do you mean something as

http://www.national.com/rd/RDhtml/RD-136.html

?

FYI: whenever you try to design a power supply, you have to specify the load current required.

Yes, you are right. Sorry, I was thinking more in the terms of just finding a voltage converter and not designing a power supply (that I've already done ) ). Here are the partial data, the currents are very low, a few µA in exception of one of the ratings which is 0.075 mA = 75 µA and I am not sure when this condition occurs and if it is only for the positive voltage supply pin or for the both ones - positive and negative :-/ The complete datasheet of DG408 can be found here.

I thought there is some IC similar to LT1026 (which is a voltage doubler + inverter and doesn't need any external parts) but to be a voltage tripler or maybe a quadrupler. Maybe two LT1026 in a row could be used as a voltage quadrupler?

Or, what do you thing about taking the +/- 15 V from the RS232 outputs of MAX232? With 0 and 5 V at TTL inputs and 5 V supply, + and - 15 V should appear at RS232 outputs but I am not sure if these could be used to power DG408 :-/

Thank you for your reply
Regards

 

Your switch can operate with both single and dual supply. And also lower supply voltages than +/-15 volt. Have you looked at that fact. The input voltage can swing up to the supply rails voltage

 

Your switch can operate with both single and dual supply. And also lower supply voltages than +/-15 volt. Have you looked at that fact. The input voltage can swing up to the supply rails voltage

Yes, I know that. However, I am going to use the mux to switch RS232 signals, that's why I need +/- 15 V.

 

I would have looked at something like this
https://www1.elfa.se/data1/wwwroot/assets/datasheets/asRB_RA.pdf

 

If you're actually dealing with RS232 signals, then I'd suggest converting to logic level and handling the data in that form. If your system is RS232 in and out, you can always convert back to RS232. If you use analog multiplexers, you'd be putting some impedance on the lines, and that wouldn't be good.

How about RS422? It's really better all around.

 

I would have looked at something like this
https://www1.elfa.se/data1/wwwroot/assets/datasheets/asRB_RA.pdf

Thank you very much!! That's it!! ) RB-3.315D seems as a perfect match. And the price is awesome, just about $5. An hour ago I was noticed about TEM2-0522 but it's too expensive (I am going to need more of the selected converters).

Besides, there are analog multiplexers (which I wasn't aware of until a few minutes ago) that can switch up to +/- 25 V signals operating from just 3.3 V single supply!! So I might even decide to use Beyond-the-Rails muxes and avoid voltage converters.

Thank you again for your information!
Best Regards

 

If you're actually dealing with RS232 signals, then I'd suggest converting to logic level and handling the data in that form. If your system is RS232 in and out, you can always convert back to RS232. If you use analog multiplexers, you'd be putting some impedance on the lines, and that wouldn't be good.

How about RS422? It's really better all around.

The device I am working on neither receives nor transmits the data, it just changes the route of the RS232 signals. It could in fact be done with the relays but I wanted to avoid the mechanical parts.

 

Yes, you are right. Sorry, I was thinking more in the terms of just finding a voltage converter and not designing a power supply (that I've already done ) ). Here are the partial data, the currents are very low, a few µA in exception of one of the ratings which is 0.075 mA = 75 µA and I am not sure when this condition occurs and if it is only for the positive voltage supply pin or for the both ones - positive and negative :-/ The complete datasheet of DG408 can be found here.

Power that low gives you options: depends whether you want the cheapest or the easiest. You can use a free running 555 timer to boost up the 3.3-5V input to a higher voltage (like 20 - 30V) and feed it to a 78L15 reg. You can use another 555 to do an invert and feed it to a 79L15 for the -15V. Those parts are dirt cheap.

Or, there are IC switcher solutions that do it easier with lower total part count at higher cost.

 

Do you mean something as

http://www.national.com/rd/RDhtml/RD-136.html

?

Here is the simple Switcher design tool

http://www.ti.com/ww/en/simple_switcher_dc_dc_converters/index.html

 

Power that low gives you options: depends whether you want the cheapest or the easiest. You can use a free running 555 timer to boost up the 3.3-5V input to a higher voltage (like 20 - 30V) and feed it to a 78L15 reg. You can use another 555 to do an invert and feed it to a 79L15 for the -15V. Those parts are dirt cheap.

Or, there are IC switcher solutions that do it easier with lower total part count at higher cost.

It seems RB-3.315D wouldn't cost much more than 2xNE555 + 78/79L15 + accompanying passive parts. For the comparison, ATMega8 is here $4.75 and 16f628 is $6.6

Here is the simple Switcher design tool

http://www.ti.com/ww/en/simple_switcher_dc_dc_converters/index.html

Thank you very much, it sure will be a very useful information for some of my future projects. As for the current project, I have decided to, instead of DG408, use MAX14778 - it operates from 3.3 V single supply and can switch up to +/- 25 V signals!!

Here is the Beyond-the-Rails product line overview. MAX14778 datasheet can be found here. And here is a video of MAX14778 demo.

If they only, besides in the TQFN, produced it in the PDIP package too :-/

 

It seems RB-3.315D wouldn't cost much more than 2xNE555 + 78/79L15 + accompanying passive parts.

I don't know about that. Digi Key lists it at $6.75

I could do the 555 cheapos for a lot less than that but a lot more parts. With current that low, you could use a simple 555 free run at about 50 kHz driving a 2N2222 with a tiny inductor and a 1N914 diode and cap for a boost and just clamp the output with a 15V zener.

 

I've changed my mind. I think am going to use this step-up converter to rise the voltage up to about 20 V, and then 7815&7915 to get +/- 15. I think something similar to this configuration (you can click on the picture to zoom it in) should allow 7915 to work as expected.

 

I've changed my mind. I think am going to use this step-up converter to rise the voltage up to about 20 V, and then 7815&7915 to get +/- 15. I think something similar to this configuration (you can click on the picture to zoom it in) should allow 7915 to work as expected.

That circuit has two different grounds. I don't think it will work as drawn. The "-15V" output is with respect to the top of the positive voltage which is 24V with respect to the shown ground. Ergo, it will be 9V WRTG.