I work for a toy company and to date all of our electronics have used masked chips. This incurs a chip masking charge and a lead time of 45-60 days to mask and manufacture/program the chips.

Recently one of our vendors offered us an OTP (one-time programmable chip option). No chip masking charge, shorter lead time to get chips programmed (~30 days). The one downside is the cost of the chip is higher.

Besides a cost analysis of when using an OTP chip will be better than a masked chip (based on volume produced) to help me decide which option to go to, what are other pros/cons of using a masked chip vs an OTP chip?

  • \$\begingroup\$ OTP relies on the fact that fused links can't be unburned. That doesn't necessarily rule out the possibility of burning more links later. But in some cases, the last step in programming is to burn the programming links, protecting the chip against further modification. I would find out from your contact which applies here. \$\endgroup\$ – Ben Voigt Mar 31 '14 at 19:02
  • \$\begingroup\$ I think you have it (but I've never used either type of part in the real world). You might double-check with the vendor if there is any difference in yield or reliability between the two types. \$\endgroup\$ – The Photon Mar 31 '14 at 19:03
  • \$\begingroup\$ @BenVoigt - for production ready chips, don't we want things to be permanent? \$\endgroup\$ – milesmeow Mar 31 '14 at 19:03
  • \$\begingroup\$ @milesmeow: In your case, I would expect that "protection against further modification" is more desirable than "continued limited ability to program", yes. \$\endgroup\$ – Ben Voigt Mar 31 '14 at 19:04
  • \$\begingroup\$ @milesmeow, probably. If they weren't, that would introduce a new failure mode not found in the masked part. Which is what I'm pretty sure Ben was trying to suggest. \$\endgroup\$ – The Photon Mar 31 '14 at 19:04

OTP also usually allows you to buy a programmer and program them yourselves, reducing that 30 day lead time further - to nil, if you keep unprogrammed parts in stock.

This allows you to trade the extra device cost against:

  • reduced inventory costs
  • faster time to market - delayed sales can be lost sales.
  • flexibility to deal with unsteady sales

And for steady sales of bulk product you can still reap the benefit of cheaper mask product.

| improve this answer | |

The good thing is that OTP chips generally require an external high voltage supply for programming, so they have less tendency to lose their memory than flash and EEPROM-based chips, so it's much less likely you'll get mysterious bricking of the product if there is no high voltage generator on the chip, and it should not be much different from mask chips.

Mask programmed chips might be a bit more easy to reverse engineer since the code is visible photographically. But I might guess that it's easier for a small scale copier to source OTP chips and there are still ways to break OTP code locking.

You might want to figure in the effect of such things on barrier to entry for copiers, depending on what compatible products are available.

Another difference is that OTP chips often don't have as many options available as mask-programmed chips (EPROM cells are less flexible than metal links), so you may incur additional costs if (say) an on-chip boost regulator is not available.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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