Author Topic: PSA: do not use the TPS61099 boost reg in your designs  (Read 11325 times)

0 Members and 1 Guest are viewing this topic.

Offline fireworks

  • Contributor
  • Posts: 38
  • Country: hk
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #25 on: January 23, 2019, 09:34:22 am »
Their own engineers couldn't decipher their own documentation...  It took 5 days and a few friends who specialized in math and physics to reverse engineer the notation in their .pdf data sheet and to get the IC to dynamically be programmed to do exactly what I want. 

  Hi Brian,

  Which TI datasheet was that ? I want to take a look at the weird formulas ☺
 

Offline TheDane

  • Regular Contributor
  • *
  • Posts: 209
  • Country: dk
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #26 on: January 23, 2019, 09:49:22 am »

If anyone has any ideas about how I could work around something like this, I'm all ears  :-//

Thanks,
Jeremy

Not much of a TI device work around; more a protection in general: ESD Suppressors / TVS Diodes UNIDIRECTIONAL 3.3V + a polyfuse on the input.

Getting the correct part for your device can be tricky, as operating voltage vs. clamping voltage is quite close so not to blow up everything.
What is your device supply absolute max rating, and the device current draw?

https://www.mouser.com/datasheet/2/80/CPDH3V3UP-HF_RevB034941-1481125.pdf
Clamping voltage - 5V5, I_PP = 1 A
Punch-through voltage - 3V5, I_PT = 2uA
Snap-back voltage - 2V8, I_SB = 50mA
Fig.3 - Clamping voltage vs. peak pulse current - 8V@5A down and stops at 1A@5V5, extrapolating it down could result in a working protection solution for you?
 

Offline JPortici

  • Super Contributor
  • ***
  • Posts: 3573
  • Country: it
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #27 on: January 23, 2019, 10:02:06 am »
These things are IC design equivalents of spaghetti code from a total beginner. If this was a software, a fix would be another overcomplicated "UVLO filter unit" which would reduce the number of issues, but increase complexity again. But ICs can't be updated.

And yet newer TI chips ARE updated. I've used some of their newer LDOs in a design and between batches (like 6 to 12 months later) the exact partnumber didn't even exist anymore, but a partnumber followed with "A1" was available.  then "A2" and so on...
« Last Edit: January 23, 2019, 10:24:17 am by JPortici »
 

Offline forrestc

  • Supporter
  • ****
  • Posts: 720
  • Country: us
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #28 on: January 23, 2019, 10:50:25 am »
Sounds like typical TI engineering. I use their power conversion/management ICs less and less - they are overcomplicated and have excess state space, and therefore tend to hit unwanted states - and a typical "wrong state" tends to mean: blow up everything.

I miss National Semiconductor.   Back before TI gobbled them, I found the national power conversion IC's to be top notch and just work.   I never had the same luck with TI.

Right now I'm sort of looking for a replacement for a buck converter which was an original National design.   It's old enough that it's got a way high I(q) but it sure makes me nervous to try something else especially when you hear stories like this on a fairly regular basis.
 

Offline jeremyTopic starter

  • Super Contributor
  • ***
  • Posts: 1079
  • Country: au
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #29 on: January 23, 2019, 11:24:44 am »

If anyone has any ideas about how I could work around something like this, I'm all ears  :-//

Thanks,
Jeremy

Not much of a TI device work around; more a protection in general: ESD Suppressors / TVS Diodes UNIDIRECTIONAL 3.3V + a polyfuse on the input.

Getting the correct part for your device can be tricky, as operating voltage vs. clamping voltage is quite close so not to blow up everything.
What is your device supply absolute max rating, and the device current draw?

https://www.mouser.com/datasheet/2/80/CPDH3V3UP-HF_RevB034941-1481125.pdf
Clamping voltage - 5V5, I_PP = 1 A
Punch-through voltage - 3V5, I_PT = 2uA
Snap-back voltage - 2V8, I_SB = 50mA
Fig.3 - Clamping voltage vs. peak pulse current - 8V@5A down and stops at 1A@5V5, extrapolating it down could result in a working protection solution for you?

Thanks for the pointers. I’ve been measuring some zener diodes I have around but as I suspected even a good 5.1V zener had around 100uA of leakage at 3.6V. Unfortunately my entire power budget during sleep is 100uA @ ~3V at Vin (incl regulator efficiency loss and Iq), but of course I’m trying to beat that. I’ll have to do some reading on these esd devices, I didn’t realise the reverse leakage was so good.

System can withstand up to 4.2V, and needs about 250mA peak at 3.6V when radio activity is heavy, but it normally stays around 20-30mA otherwise.

My other concern with clamping is that it doesn’t actually reset the converter and in my experience with this problem it needs a reset to be fixed. I’ve managed to get my double voltage supervisor down to 10uA, which I think I can live with. But testing it with this intermittent problem is tricky...

Edit: sorry, I see you mentioned polyfuse as well. That could cause a UVLO reset perhaps
« Last Edit: January 23, 2019, 11:28:01 am by jeremy »
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9327
  • Country: fi
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #30 on: January 23, 2019, 12:36:16 pm »
Not saying the IC is perfect but I see some fault on your part for not assuring a stable Vin before enabling the converter.

Sorry to use such strong word, but: bullshit.

Ramping and intermittent voltages are completely normal on any system. This is the exact task for an UVLO circuit, this is why it exists, and because the need is so common, this is why it's integrated.

If they claim they have UVLO, it must accommodate exactly the ramping and bouncing Vin case, of course within maximum ratings.

It's ridiculous to ask the user (only after sales, of course!) to provide an external UVLO on the enable pin when they claim to have one internally.

Outputting too high of an output voltage due to input brown out cannot be justified even without an UVLO, let alone with one, unless the input bounces outside the allowed range, which probably isn't the case here.
« Last Edit: January 23, 2019, 12:41:42 pm by Siwastaja »
 
The following users thanked this post: OwO

Offline OwO

  • Super Contributor
  • ***
  • Posts: 1250
  • Country: cn
  • RF Engineer.
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #31 on: January 23, 2019, 01:19:53 pm »
Is there a reason for using the TPS610995 in the first place? If the requirements are just <1V startup voltage and low shutdown current there are many cheap boost converter ICs that can do the job, for example the SGM6603.
Email: OwOwOwOwO123@outlook.com
 

Offline splin

  • Frequent Contributor
  • **
  • Posts: 999
  • Country: gb
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #32 on: January 23, 2019, 01:25:05 pm »
Damn, that's good to know. This IC looked interesting on paper.

Reading the datasheet, we can see it's got quite a few different modes of operation. Between the soft-start feature, the "down mode", "pass-through", bust mode and synchronous mode...

Ah, that must be the one the OP was using.
 
The following users thanked this post: jeremy

Offline jeremyTopic starter

  • Super Contributor
  • ***
  • Posts: 1079
  • Country: au
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #33 on: January 23, 2019, 01:35:24 pm »
Is there a reason for using the TPS610995 in the first place? If the requirements are just <1V startup voltage and low shutdown current there are many cheap boost converter ICs that can do the job, for example the SGM6603.

Not low shutdown, low Iq and high efficiency at light load. Also relatively high current at 250mA (peak inductor pulse is up to 0.8A)
 

Offline OwO

  • Super Contributor
  • ***
  • Posts: 1250
  • Country: cn
  • RF Engineer.
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #34 on: January 23, 2019, 02:09:21 pm »
Not low shutdown, low Iq and high efficiency at light load. Also relatively high current at 250mA (peak inductor pulse is up to 0.8A)
some more parts to investigate:
PT1301
ME2108
XC6367
Email: OwOwOwOwO123@outlook.com
 
The following users thanked this post: jeremy

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15794
  • Country: fr
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #35 on: January 23, 2019, 04:35:12 pm »
Reading the datasheet, we can see it's got quite a few different modes of operation. Between the soft-start feature, the "down mode", "pass-through", bust mode and synchronous mode...
Ah, that must be the one the OP was using.

Ahah, this typo was unintentional but that's a good one. ;D
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15794
  • Country: fr
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #36 on: January 23, 2019, 04:44:19 pm »
I suggest the LTC3528 as an alternative. Linear has never failed me.

 

Offline Bassman59

  • Super Contributor
  • ***
  • Posts: 2501
  • Country: us
  • Yes, I do this for a living
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #37 on: January 23, 2019, 08:38:36 pm »
With TI, it's not just their power management ICs.  Their new norm on documentation SUCKS shit.  On their clocking ICs, I tried to get information as their formulae in the data sheet made no sense and had non-standard characters.  I was told from TI support, just use their PC configuration app to get the IC2 configuration settings.  NO... The IC said I can program any frequency I wanted, both setting the reference and feedback dividers and their config software didn't always make the most stable oscillator configuration unless I manually tweaked it.  Their own engineers couldn't decipher their own documentation...  It took 5 days and a few friends who specialized in math and physics to reverse engineer the notation in their .pdf data sheet and to get the IC to dynamically be programmed to do exactly what I want.

I've been using the Silicon Labs clock ICs (synthesizers, etc etc) and while you have to use the Clock Builder software, the parts behave and the support has been responsive (don't use the forum, just send email to their support staff).


Quote
I also found bugs in a few of their other ICs. 

I have an issue with their not-new-anymore audio SRC chip, its receiver simply won't lock on S/PDIF signals whose sample rate is a multiple of 44.1 kHz. The 48 kHz family works fine. At least the support guy is honest: "none of the people on the design team work here any more." Still, there's no help and no support. This is why I haven't considered their microcontrollers for anything, even now that they've made the development tools free.
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #38 on: January 23, 2019, 08:47:13 pm »
Their own engineers couldn't decipher their own documentation...  It took 5 days and a few friends who specialized in math and physics to reverse engineer the notation in their .pdf data sheet and to get the IC to dynamically be programmed to do exactly what I want. 

  Hi Brian,

  Which TI datasheet was that ? I want to take a look at the weird formulas ☺

I've attached the original cdce925_2007.pdf I had to work with at the time.  See page 21, part 2 and the line after.

I wish I had my notes, this ended up being simplified down to easy.
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #39 on: January 23, 2019, 08:49:00 pm »
With TI, it's not just their power management ICs.  Their new norm on documentation SUCKS shit.  On their clocking ICs, I tried to get information as their formulae in the data sheet made no sense and had non-standard characters.  I was told from TI support, just use their PC configuration app to get the IC2 configuration settings.  NO... The IC said I can program any frequency I wanted, both setting the reference and feedback dividers and their config software didn't always make the most stable oscillator configuration unless I manually tweaked it.  Their own engineers couldn't decipher their own documentation...  It took 5 days and a few friends who specialized in math and physics to reverse engineer the notation in their .pdf data sheet and to get the IC to dynamically be programmed to do exactly what I want.

I've been using the Silicon Labs clock ICs (synthesizers, etc etc) and while you have to use the Clock Builder software, the parts behave and the support has been responsive (don't use the forum, just send email to their support staff).


Quote
I also found bugs in a few of their other ICs. 

I have an issue with their not-new-anymore audio SRC chip, its receiver simply won't lock on S/PDIF signals whose sample rate is a multiple of 44.1 kHz. The 48 kHz family works fine. At least the support guy is honest: "none of the people on the design team work here any more." Still, there's no help and no support. This is why I haven't considered their microcontrollers for anything, even now that they've made the development tools free.
I'm sorry but TI's forum didn't yet exist in 2007, or didn't have anyone useful there when I was working on the stuff.
 

Offline Daixiwen

  • Frequent Contributor
  • **
  • Posts: 367
  • Country: no
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #40 on: January 24, 2019, 08:41:38 am »
I can relate to that. I had some problems several times with some components from Benchmarq and was told by TI that after their merger with Benchmarq the engineers that worked on the BQ components weren't there any more and TI couldn't offer any support on them.
So I can't say I've had a good experience with TI's support either ;)
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #41 on: January 24, 2019, 02:46:27 pm »
I would say something similar happened to Unitrode when it was absorbed by TI.

Unitrode had *many* outstanding and legendary Engineers. Their app notes were superlative. The products performed, and any pitfalls were clearly spelled out

Back in the 90s I went to a power seminar. The speaker was Laslo Balogh (sorry if the spelling is incorrect) and he was pure genius.

 

Offline kony

  • Regular Contributor
  • *
  • Posts: 242
  • Country: cz
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #42 on: January 24, 2019, 03:18:22 pm »
I wouldn't speak so highly about Unitrode. Literally the only switcher chip I had problem with from TI portfolio was inherited acient unitrode universal switchmode controller (don't remember exact PN anymore). Noise sensitive as hell.
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #43 on: January 24, 2019, 03:51:11 pm »
No one is perfect............
As I mentioned in my earlier post, Unitrode parts had their shortcomings clearly spelled out. Which were several, because of the "ancient" technology.
Their seminars were full of design tips, based on an excellent analysis, to circumvent those pitfalls.

That was one of their greatest strengths, which unfortunately TI has not followed through

 

Online Hydron

  • Super Contributor
  • ***
  • Posts: 1085
  • Country: gb
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #44 on: January 24, 2019, 06:07:32 pm »
When we started having large numbers of TI DSPs go bad in service (having used them in 10s of thousands quantities for years, so we knew when something was up) they refused to do anything useful to help track down the issue.

Got someone to analyse them and sure enough, very poor wirebond quality compared to earlier samples, bad enough to cause reliability issues - the parts should have never made it out of TI's door. Now in the process of designing them out of everything...
 

Offline IDEngineer

  • Super Contributor
  • ***
  • Posts: 1950
  • Country: us
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #45 on: January 24, 2019, 07:02:32 pm »
sure enough, very poor wirebond quality compared to earlier samples, bad enough to cause reliability issues - the parts should have never made it out of TI's door. Now in the process of designing them out of everything...
Sheesh, what happened to TI? Used to be they were like Linear... pricey but you could rely on them for great products with great support. Sounds like someone is now optimizing for short term profits instead of long term design wins to maximize the area (profit) under the curve.

If I had to guess, maybe their semiconductor division used to be run by IC people but is now run by MBA's?
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #46 on: January 24, 2019, 08:02:46 pm »
Your guess may be accurate.

Also, they have grown to be very large through acquisitions. When that happens, the culture shock in the acquired companies cause the best talent to migrate somewhere else.
 

Offline jeremyTopic starter

  • Super Contributor
  • ***
  • Posts: 1079
  • Country: au
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #47 on: January 25, 2019, 12:44:57 am »
A neat little tidbit from the LTC3528 datasheet which might give a clue to what the issue could be with the TI converter:

Quote
The converter is shut down by pulling the SHDN pin below 0.25V, and activated by pulling SHDN above 0.88V. Although SHDN can be driven above VIN or VOUT (up to the absolute maximum rating) without damage, the LTC3528/LTC3528B have a proprietary test mode that may be engaged if SHDN is held in the range of 0.5V to 1V higher than the greater of VIN or VOUT. If the test mode is engaged, normal PWM switching action is interrupted, which can cause undesirable operation in some applications. Therefore, in applications where SHDN may be driven above VIN, a resistor divider or other means must be employed to keep the SHDN voltage below (VIN + 0.4V) to prevent the possibility of the test mode being engaged.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8525
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #48 on: January 25, 2019, 02:29:48 am »
That's better than nothing, but I think the best way to design something like that is either fully document what the test mode does and how to exit it, or fuse it off permanently so it would never be entered again. Weird undocumented functionality in an IC is overall bad for a variety of reasons.
 

Offline LukeW

  • Frequent Contributor
  • **
  • Posts: 686
Re: PSA: do not use the TPS61099 boost reg in your designs
« Reply #49 on: January 25, 2019, 02:45:40 am »
Who makes decent affordable DC-DC silicon other than TI?

Don’t say high-price LT.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf