2
\$\begingroup\$

I am currently designing a system for a satellite and I am looking into the components to generate the required low-voltages, and I was wondering if maybe the collective knowledge of stackexchange could be of some help.

Some specs:

  • input voltage: exact value still TBD, but min 5V and up to 20V if needed
  • output voltage: range 1V up to 5V
  • ideally, the value should be adjustable during operations (controlled by an FPGA)
  • rad hard: TID > 100 kRad, SEE > 60 MeV.cm2/mg

At the moment, I selected a couple of LDOs from TI and the plan would be to control them following this scheme (a DAC in the feedback loop of the LDO): link to control scheme

I was wondering if you knew either of some other control scheme or maybe of components which include such capability directly in one chip package, instead of having a separate DAC. Thanks for any help you can give me!

EDIT: some additional information:

  1. this is the first space-project of my organization (research institute, not a company), so I don't really have tried and tested product recommendations. We got the rad hard requirements for TID and SEE that I wrote in the post, and that's basically it (we got them from the project office).

  2. Here are the datasheets of the components that I selected. These are a basic starting point for what I need, but I am not married with these choices

\$\endgroup\$
7
  • \$\begingroup\$ If you are designing stuff for space then what does your organization provide you in terms of tried and tested rad hard product recommendations? What acceptance criteria do they impose any any new component - these are the things you should be investigating before asking this question. Also what were the rad hard parts you selected - links to data sheets please. \$\endgroup\$
    – Andy aka
    Jun 15, 2020 at 10:26
  • \$\begingroup\$ Hi Andy, thanks for your comment. 1) this is the first space-project of my organization (research institute, not a company), so I don't really have tried and tested product recommendations. We got the rad hard requirements for TID and SEE that I wrote in the post, and that's basically it (we got them from the project office). 2) for the datasheets of the selected components, here they are: - ti.com/lit/pdf/slvsdw6 - ti.com/lit/gpn/tps7a4501-sp EDIT: sorry for the first empty reply, it was an accident. \$\endgroup\$
    – Cesco
    Jun 15, 2020 at 10:28
  • \$\begingroup\$ Please don't thank me - fill-in the blanks! \$\endgroup\$
    – Andy aka
    Jun 15, 2020 at 10:30
  • \$\begingroup\$ Add this information to your question please and not as comments. And, properly explain why you are undertaking this project despite having limited experience in rad hard products for space. \$\endgroup\$
    – Andy aka
    Jun 15, 2020 at 10:57
  • \$\begingroup\$ A lot of the regulators from Linear Technology have a rad tolerant/rad hard version. We've used the RH1185 and RH1086 in designs, \$\endgroup\$
    – SteveSh
    Jun 15, 2020 at 19:51

2 Answers 2

0
\$\begingroup\$

How about an interesting idea. Not sure if it's perfect solution or even correct one, I hope someone adds to my comment confirming/busting it. But it won't hurt if you just try it on a breadboard.

Your LDO must have an external voltage divider to set its output voltage. How about you replace one of them with digital potentiometer? The one that works with I2C. You can change one resistor, thus you can change output voltage with reasonable steps via I2C. Of course, you have to see for yourself what exactly its resistance swing should be. You can even have two of them for finer tuning (then you probably need two different ones so that their slave addresses don't match or they should have I2C address select pin; or address shifter, which is too much extra circuit imo).

A little unorthodox solution, but I would try it on breadboard just for fun first, there's a good chance it will work. Especially with two digital pots. But be careful when changing voltage with 2 variable resistors, they will change values not at the same time, which may lead to problems.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ This is not as simple as it seems - digital potentiometers are not exactly good resistors at all - they have inductance, capacitance, limiting frequencies... everything that can mess with the regulation loop of the LDO. What Cesco described in his question is one of the recommended solutions. \$\endgroup\$
    – asdfex
    Jun 15, 2020 at 16:37
  • 1
    \$\begingroup\$ Thanks for the idea, I will give it a try. As asdfex said, it probably will not work: I was cautioned against using digital potentiometers. \$\endgroup\$
    – Cesco
    Jun 15, 2020 at 20:24
  • \$\begingroup\$ depending on your requirements, abilities and skills (and they need to be better than mine, I'll be honest, I'm not a pro), maybe an MCU with DAC? you could then connect the lower side of the voltage divider not to ground, but to analog output? Of course I'm not sure about this either, I'm just throwing in as many ideas as come up in my head to find the better one from the idea list \$\endgroup\$
    – Ilya
    Jun 16, 2020 at 8:50
0
\$\begingroup\$

I can't help with radiation hardness, but there is at least one more option to adjust voltages digitally: TI has a series of LDOs, e.g. TPS7A7200 with an internal voltage divider that can be switched using several digital inputs to adjust voltage in steps of 50mV. You might want to try to resemble this by building an external circuit using some FETs and Open Drain Switches around your favorite rad hard LDO.

\$\endgroup\$
1
  • \$\begingroup\$ thank you, I will have a look! This is exactly the kind of suggestion I was looking for. \$\endgroup\$
    – Cesco
    Jun 15, 2020 at 20:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.