We seek to understand and document all radio transmissions, legal and otherwise, as part of the radio listening hobby. We do not encourage any radio operations contrary to regulations. Always consult with the appropriate authorities if you have questions concerning what is permissible in your locale.

Author Topic: Class D audio amps - protection circuits  (Read 1267 times)

Offline W9RAN

  • Jr. Member
  • **
  • Posts: 20
    • View Profile
    • Email
Class D audio amps - protection circuits
« on: April 27, 2023, 0339 UTC »
I've learned a lot from this forum, so thanks, first of all!

One thing is the use of class D amplifiers as modulators by connecting a class E PA directly to one of the H bridge (speaker) outputs.   Beautiful hi-fi quality and couldn't be simpler or less expensive!   I've been using the TPA3116 amplifiers to modulate 5-8 watt transmitters but have one of the TPA3118 jobs enroute.    My question/problem is about the built-in protection circuitry which is common to both.

I've been getting frequent shutdowns that seem to be due to a DC Detect fault, which causes the outputs to go Hi-Z.   The datasheet describes this fault:  "A DC Detect Fault is issued when the output differential duty-cycle of either channel exceeds 60% for more than 420 msec at the same polarity."      It could be triggered since we are applying a load to one output but not both - or it could be for some other reason as I'm just guessing, and know the amp hasn't run enough to be thermally related.  I've had it happen at start-up and can often produce it by connecting an input device when the amplifier is powered-up and idle, thus both outputs are a 1/2 Vcc.  There is no current draw until the transmitter is keyed but sometimes I see a brief current spike on power-up then the outputs fault.  Other times it just stops randomly.   Very annoying to say the least.

I'm curious to know if others have observed this and any suggestions as to how to deal with it would be appreciated, thanks.
Bob

Offline Charlie_Dont_Surf

  • Hero Member
  • *****
  • Posts: 787
  • Charlie's Point, California
    • View Profile
Re: Class D audio amps - protection circuits
« Reply #1 on: April 28, 2023, 0318 UTC »
Yes, familiar with it. In fact, I mentioned it over here: https://www.hfunderground.com/board/index.php/topic,89881.msg348346.html#msg348346

How do you know that it is "DC Detect Fault" and not simply the overcurrent protection turning on? Overcurrent is simply drawing too much current from the output(s). This could be that you have an inappropriate load (likely too low impedance) or you are just asking for too much output based upon the load that is attached. In the latter case, you just have to turn the gain down or reduce the input.

If we are talking about the same thing, I generally I observe it as momentary drop in audio; basically just asking for too much from the amp. This is most likely to happen when there is heavy bass content to the program material. Not many alternatives for immediate relief other than reduce the bass content, turn down the gain or reduce the input. Also, it is important to note that these transmitters do not look like a pure resistor to the Class-D amp output; they will look like a complex impedance, especially below some arbitrary audio frequency. The Class-D amp may be "happy" with the load that the transmitter looks like at 1 KHz but very "unhappy" with the (probably) different impedance at 50 Hertz, for example.

The reason why bass tones are more likely to cause a problem, is for one thing, bass tones are low frequency, which means longer period and thus the Class-D output is on for a comparatively longer period of time. The overcurrent protection is only going to be triggered when the output is on, so, if there is an impedance problem that would trigger an overcurrent fault, it's more likely to occur on bass than treble just by sheer probability.

Beyond that, I have noticed some interesting turn on characteristics. If I have an audio input present at turn on that is too strong, it can create some sort of fault condition that is difficult to clear without turning off the amp, lowering the input and trying again.
I don't STRETCH the truth.

"Every minute I spend in this room, my signal gets weaker.
Every minute Charlie squats in the bush, his signal gets stronger."

Offline W9RAN

  • Jr. Member
  • **
  • Posts: 20
    • View Profile
    • Email
Re: Class D audio amps - protection circuits
« Reply #2 on: April 28, 2023, 1858 UTC »
Thanks for the reply Charlie.  I was just spitballing - it could be overcurrent, overtemperature, or DC output offset as those three conditions result in latched protection (vs. over/under voltage which is self-resetting).   I will do some additional testing but one thing I found is - there are differences between boards, as usual when buying "eBay specials" they don't come with configuration control, or any assurance they're built to the same schematic.    I'll review your other thread as well and post back here if I can identify anything definite. 

Since I have had this happen most consistently at power-on when bass notes (or any kind of input signal) is not present, I wonder if a turn-on transient may be having the same effect.  I'm also going to look at the coupling caps and if necessary, switching the load with PTT as in my previous transmitters I've done that (no load at all on the output until PTT is keyed) and never had this  issue.   On this board the modulated drain voltage is applied from power-on and the PA is held off with gate bias, so there could be a race condition.

Counterfeit ICs were one of my fears, but hopefully not the case.  I'd certainly be willing to pay more for a guaranteed TI part, after all these are dirt-cheap and virtually ideal modulators even at a premium.

Bob

Offline W9RAN

  • Jr. Member
  • **
  • Posts: 20
    • View Profile
    • Email
Re: Class D audio amps - protection circuits
« Reply #3 on: April 30, 2023, 0152 UTC »
Well I've learned some interesting things and while I suspect what I discovered would have been instantly apparent to someone using a Lulu, I'll post it here in case it's useful to someone down the road.

I've been buying generic "TPA3116" and/or "TPA3118" (the only difference is whether the thermal pad is on the top or bottom side) from the lowest priced eBay sellers, assuming (and we know what that means) that they were more-less the same.    Which they may be, but that begs the real question as all of them I've got do not in fact have either 3116 or 3118 ICs on them, even though that's how they are marked.   They cannot, because the real TI parts are in 32 pin packages and the ones I've got are in 28 pin packages!    However, they may be genuine TI parts, just different ones - namely TPA3111d1, which is a 10 watt amplifier.    TI offers a family of parts built on the same platform in various power levels, and for those like me who are interested in low power QRP transmitters, the good news is they all can take up to 26 volt supplies, so the same modulator trick can be done.    I should note that one board appears to be DOA and one bridge output on another is shorted to ground, so maybe the protection circuitry isn't quite as good as claimed.

I was planning to implement a change to cause faults to automatically reset by connecting the FAULT and SD (shutdown) pins together, but that's when I realized the pinout was wrong.   Both parts have the same functionality but the pins are different and on the TPA3111 it's pin 1 and 2, and they're already connected on the  PCB, and the datasheet clearly says:  "If automatic recovery from the short-circuit protection latch is desired, connect the FAULT pin directly to the SD pin. This allows the FAULT pin function to automatically drive the SD pin low that clears the short-circuit protection latch."  So why then am I getting latched faults?      No idea as of yet, but it is definitely related to connecting input devices with power applied.

The input impedance relates to frequency response and gain as described in the datasheet, which also states:  "The impedance at the inputs must be limited to an RC time constant of 1 ms or less if possible. This is to allow  the input DC-blocking capacitors to become completely charged during the 14-ms power-up time."  Also:  "The input capacitors used should be a type with low leakage, like quality electrolytic, tantalum or ceramic  polarized type is used the positive connection should face the input pins which are biased to 3 Vdc."     

While I don't understand why it is not automatically resetting, the fault that occurs when connecting an input is probably as I guessed, a DC Offset fault:  "To avoid nuisance faults due to the DC detect circuit, hold the SD pin low at power-up until the signals at the inputs are stable. Also, take care to match the impedance seen at the positive and
negative inputs to avoid nuisance DC detect faults."      The real 3116/8 parts have a separate MUTE input but that pin is omitted on these misbranded TPA3111 parts, and the MUTE terminal is actually connected to pins 1 and 2, which are SD/FAULT.    So it may be possible to connect a capacitor that will act like a power-on reset circuit to hold the SD input low at power-up.   

I have one of the Wondom 100 watt amplifiers on the way, which I assume will be a genuine TPA3116 for use with a Lulu-type transmitter and it will be interesting so see how it compares.

Bob

Offline W9RAN

  • Jr. Member
  • **
  • Posts: 20
    • View Profile
    • Email
Re: Class D audio amps - protection circuits
« Reply #4 on: May 01, 2023, 0400 UTC »
Additional findings:
1.  I did learn from the TPA3111 datasheet that DC offset faults cannot be self-cleared:  "To clear the DC Detect it is necessary to cycle the PVCC supply. Cycling SD does not clear  a DC Detect Fault."    This may vary with other parts in the family.

2.  I found that some of the mis-branded amplifiers would not produce 100% modulation and guessed correctly that they'd been configured to limit the output voltage swing.  This is done by selecting how the PLIMIT pin is connected, and I found a 100K resistor to ground.   Inspection of the PCB showed that by repositioning a 0 ohm resistor the limit could be removed, and that did work.

A nasty trick for those using them as modulators, but the feature is there to allow designers of TVs, bluetooth speakers, etc to reduce the voltage swing if needed to prevent speaker damage.

Still quite frustrated by unwanted protection shutdowns, but I'm going to have to wait for more amplifiers to show up now.

Offline Charlie_Dont_Surf

  • Hero Member
  • *****
  • Posts: 787
  • Charlie's Point, California
    • View Profile
Re: Class D audio amps - protection circuits
« Reply #5 on: May 01, 2023, 0433 UTC »
I've been buying generic "TPA3116" and/or "TPA3118" (the only difference is whether the thermal pad is on the top or bottom side) from the lowest priced eBay sellers, assuming (and we know what that means) that they were more-less the same.    Which they may be, but that begs the real question as all of them I've got do not in fact have either 3116 or 3118 ICs on them, even though that's how they are marked.   They cannot, because the real TI parts are in 32 pin packages and the ones I've got are in 28 pin packages!    However, they may be genuine TI parts, just different ones - namely TPA3111d1, which is a 10 watt amplifier. 

Here's the deal: occasionally the offshore assembly house (the subcon that puts the die in plastic packages) will goof up the top mark (what used to be called "brand" in the old days) and I can guarantee you that TI does not want mis-branded production material getting out to a customer. They tell the subcon to scrap the material. (You can't re-brand anymore.) Someone we shall call "Grey Market Joe" will take it off their hands and pay a little money for them - which is more attractive than the dumpster - and magically Joe has some "TPA3116" for his friend in Shenzhen who puts it on Ebay. Equally likely is that these were old engineering material that they tried stuffing into a different package for experimentation purposes and they somehow made their way onto Ebay by more or less similar means. I'm not making this up.

TI offers a family of parts built on the same platform in various power levels, and for those like me who are interested in low power QRP transmitters, the good news is they all can take up to 26 volt supplies, so the same modulator trick can be done.

OT note: they may be all based upon the same die with different wirebonds or different redistribution layers and substrates for the different part numbers. Or they may be what is called a "metal option" - the part numbers are decided through higher level metal masks in the fab process. Either way, doing this shaves development costs and time to market, which means everything in consumer electronics.

I was planning to implement a change to cause faults to automatically reset by connecting the FAULT and SD (shutdown) pins together, but that's when I realized the pinout was wrong.   

While I don't understand why it is not automatically resetting, the fault that occurs when connecting an input is probably as I guessed, a DC Offset fault:  "To avoid nuisance faults due to the DC detect circuit, hold the SD pin low at power-up until the signals at the inputs are stable. Also, take care to match the impedance seen at the positive and
negative inputs to avoid nuisance DC detect faults."  The real 3116/8 parts have a separate MUTE input but that pin is omitted on these misbranded TPA3111 parts, and the MUTE terminal is actually connected to pins 1 and 2, which are SD/FAULT. So it may be possible to connect a capacitor that will act like a power-on reset circuit to hold the SD input low at power-up.   

I think that you are going in the right direction. I would have done the same thing. I'm a big believer in making the extra effort for a controlled power up and power down with time delays and sequencing, etc. Pins labeled ENBL, STBY, MUTE, SD and FAULT are your good friends.

I have one of the Wondom 100 watt amplifiers on the way, which I assume will be a genuine TPA3116 for use with a Lulu-type transmitter and it will be interesting so see how it compares.

Mine does not allow viewing the IC without pulling the big heatsink off and I have not attempted that. (The DROK modules are similar in this regard.) Also the backside of the PCB gives no indication as the chip - whatever it is - is SMD. I can say that it does seem to be able to drive ~2-ish Ohms, though I have not gone into great rigor on that and I don't have high-wattage 2 Ohm resistors.
« Last Edit: May 01, 2023, 0436 UTC by Charlie_Dont_Surf »
I don't STRETCH the truth.

"Every minute I spend in this room, my signal gets weaker.
Every minute Charlie squats in the bush, his signal gets stronger."

Offline W9RAN

  • Jr. Member
  • **
  • Posts: 20
    • View Profile
    • Email
Re: Class D audio amps - protection circuits
« Reply #6 on: May 04, 2023, 1827 UTC »
Yep, agree these ICs are based on a common die with metal and bond-out/package options.  Defeatured parts like the 3111 go into the 28 pin package, actual 3116/8 parts are 32 pins.  So while it's true that if you have a part branded 3118 in a 28 pin package, it's a fake.   Having a 32 pint part doesn't guarantee that it's genuine but the one I've had in a transmitter for a couple of years looks genuine and I've never had any of these problems with it.

In fact I wonder if metalization options may even determine the power rating, based on the number of output FETs that are actually connected.  Also how many of these rebranded 3111 parts are final test failures?   I say that because I've now broken at least four of them.   Three were the kind with screw terminals which came with PLIMIT configured to limit the output voltage to 1/2 of PVCC.   After changing them to allow full voltage swing, all have died with the outputs at full PVCC rather than half.  That indicated a shorted high side FET, even though the 3111 is rated to 26V just as the 31118 is (and they did work, for a while at least).   Speculate they may have been test failures that were sold with the stipulation that the output had to be derated (as a former test engineering manager I well know the manufacturing mantra that there's a buyer for everything..at the right price).

The biggest culprit here is me - as I didn't follow my elementary school teachers advice about paying attention.  If I had been I'd have known the 28 pin amplifiers could not be true  3116/8 parts despite what the friendly eBay seller claims.   A bag of 10 more showed up yesterday, so I've got plenty of LM-386 replacements now :-)

I'd started laying out a PCB to make my own, but it seems there may be some reliable suppliers of genuine TI 3189 parts, which is all I am after (although the ability to control power-up sequencing and integrate the modulator with the RF section would be pluses).     I also have one of what I think are referred to a the "Drok amp" which has the onboard boost converter to make it's own 24V, which has worked well, but seems to be hard to find.   The actual board is marked with the name "Shareelec" and that comes up in search as CIRMECH:   https://shareelec.aliexpress.com/store/1101230787

One part on this page is a $4.50 board that claims to have a TPA3116D2 although it's hidden by the heatsink -  has anyone used any of the CIRMECH amps?

A company with a range of products like this seems to be a cut above the bargain-basement eBay sellers of products they don't understand or have any control over.  I'll be buying and trying more of these amps as modulators but lesson learned - only 32 pin parts!

Any sources of the Drok-type class D BTL amplfiers with the built-in boost converter?  I think they were sold as "bass amplifiers".

Bob