Author Topic: Probing full bridge amplifiers with oscope  (Read 2585 times)

0 Members and 1 Guest are viewing this topic.

Offline Hyper_SpectralTopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Probing full bridge amplifiers with oscope
« on: July 17, 2019, 06:14:55 pm »
Full bridge amplifiers are proliferating rapidly in the car audio market, just 5 years ago the idea of full bridge was off-putting but now they make more power in a smaller package with the same relatively low distortion (remember when everyone hated Class D?)

I'm hoping to get some clarification on how to probe a full bridge multi-channel amplifier (ch = 2/4/6/...) that has only 3 outputs for 2 channels. I.e. for a 2 channel it would be:
Ch1      Ch 2
+    -/+    -

1) how would one probe this amplifier? I've attempted probing it normally, and with the scope on Ch 1's + and Ch 2's -, basically bridged mode, without success. Am I correct in assuming that in this case I would need to use the amplifiers actual input ground (12vDC ground) for the scopes ground instead of the negative rail? Obviously I would then only get half of the waveform, which is fine for detecting clipping, and I'd need a 2ch scope to get the whole waveform?

2) Does this mean the center shared -/+ has the full sine wave on it, with each half 180 degrees out of phase? Could I just probe this while referencing the input ground?

 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Probing full bridge amplifiers with oscope
« Reply #1 on: July 17, 2019, 07:52:20 pm »
Does your scope have an ChannelA  -(minus) ChannelB function?  For digital scopes, this might be in the math functions.
Or, ChannelA + ChannelB with an inversion function for ChannelB?

Set the scope inputs to AC coupled.  GND to the chassis GND.  And, place 1 probe on the speaker + and the other on the speaker -.

The other way would be to purchase a differential probe and just tie that probe's inputs to the speaker's + and - terminals.
https://www.eevblog.com/product/hvp70/

Note that these techniques will image on your scope exactly what the actual speaker is receiving in voltage.
« Last Edit: July 17, 2019, 07:57:53 pm by BrianHG »
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: Probing full bridge amplifiers with oscope
« Reply #2 on: July 17, 2019, 07:53:25 pm »
...
I'm hoping to get some clarification on how to probe a full bridge multi-channel amplifier (ch = 2/4/6/...) that has only 3 outputs for 2 channels. I.e. for a 2 channel it would be:
Ch1      Ch 2
+    -/+    -

Hmm, that looks unusual. How, exactly, do you connect a speaker to the above (specifically, is that +/- terminal used)? Are speakers coming with center-tapped voice coils now?


1) how would one probe this amplifier? I've attempted probing it normally, and with the scope on Ch 1's + and Ch 2's -, basically bridged mode, without success. Am I correct in assuming that in this case I would need to use the amplifiers actual input ground (12vDC ground) for the scopes ground instead of the negative rail? Obviously I would then only get half of the waveform, which is fine for detecting clipping, and I'd need a 2ch scope to get the whole waveform?

For a typical bridged output - which is floating with respect to "ground" - you need to use a floating or differential measurement technique. That could be with an isolated scope (which few people own), a differential probe (much better) or, if your scope has at least two channels and supports basic math functions, you connect the probe tips from two channels to each side of the bridge, respectively, then subtract one channel from the other (the probe grounds can be connected to power ground, if available, or left floating).

There are numerous other tricks that can be done so don't consider the above an exhaustive list.


2) Does this mean the center shared -/+ has the full sine wave on it, with each half 180 degrees out of phase? Could I just probe this while referencing the input ground?

That's the proverbial $64000 question. In a typical bridged amplifier you have two half bridges which are fed signals that are inverted with respect to each other so when the load is connected between them it sees double the voltage, which quadruples the maximum power (assuming there is sufficient current available). This almost seems like a push-pull output stage, but that is really only used in tube amps and power supplies, so very puzzling. Got a make/model number for the amp so we can see how it is supposed to be used?

 

Offline Hyper_SpectralTopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: Probing full bridge amplifiers with oscope
« Reply #3 on: July 17, 2019, 08:23:42 pm »
Does your scope have an ChannelA  -(minus) ChannelB function?  For digital scopes, this might be in the math functions.
Or, ChannelA + ChannelB with an inversion function for ChannelB?

Set the scope inputs to AC coupled.  GND to the chassis GND.  And, place 1 probe on the speaker + and the other on the speaker -.

The other way would be to purchase a differential probe and just tie that probe's inputs to the speaker's + and - terminals.
https://www.eevblog.com/product/hvp70/

Note that these techniques will image on your scope exactly what the actual speaker is receiving in voltage.

The scope I'm using is a dso201 or similar, basically a cheap sub 50khz battery powered mini oscope. That will probably help both of you understand what I'm doing and that it's fine to ground the scope wherever.



Hmm, that looks unusual. How, exactly, do you connect a speaker to the above (specifically, is that +/- terminal used)? Are speakers coming with center-tapped voice coils now?

For a typical bridged output - which is floating with respect to "ground" - you need to use a floating or differential measurement technique. That could be with an isolated scope (which few people own), a differential probe (much better) or, if your scope has at least two channels and supports basic math functions, you connect the probe tips from two channels to each side of the bridge, respectively, then subtract one channel from the other (the probe grounds can be connected to power ground, if available, or left floating).

There are numerous other tricks that can be done so don't consider the above an exhaustive list.

That's the proverbial $64000 question. In a typical bridged amplifier you have two half bridges which are fed signals that are inverted with respect to each other so when the load is connected between them it sees double the voltage, which quadruples the maximum power (assuming there is sufficient current available). This almost seems like a push-pull output stage, but that is really only used in tube amps and power supplies, so very puzzling. Got a make/model number for the amp so we can see how it is supposed to be used?


See above for the type of scope I'm using, it's an isolated ground and single channel (or at least I only have one probe).

I'm unsure of what a center tapped voice coil is, but I'm confident that's not what we're using. I assume it's some sort of voice coil that uses power on the center slug inside the voice coil? lol

Here's a picture of the actual amp in question, although there are many "Brazilian" (i.e. full bridge) designs like this coming onto the market as of lately. Ch 1's negative is shared with Ch 2's positive. Not sure how they do channel separation like this unless something cancels out on the central output they share.
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Probing full bridge amplifiers with oscope
« Reply #4 on: July 17, 2019, 08:33:12 pm »
Since your scope is battery powered, and you will be using 1 channel, just tie the probe GND to the speaker - and the probe tip to the speaker + on the amp.  You will see the correct signal.

If you do this, do not make any contact between any metal frame or connector GND on your scope to any any GND in your 12v system.

Also do not do this if you have 2 channels on your scope as each speaker - output on the amplifier will have a different voltage signal burning out you probe's GND or even the scope itself.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: Probing full bridge amplifiers with oscope
« Reply #5 on: July 17, 2019, 08:36:45 pm »
Alternatively, connect scope ground to car chassis and probe any of the channel connectors to see what's coming out of them. It is particularly curious what happens on the shared, middle tap.
 

Offline Hyper_SpectralTopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: Probing full bridge amplifiers with oscope
« Reply #6 on: July 17, 2019, 08:39:06 pm »
Since your scope is battery powered, and you will be using 1 channel, just tie the probe GND to the speaker - and the probe tip to the speaker + on the amp.  You will see the correct signal.

If you do this, do not make any contact between any metal frame or connector GND on your scope to any any GND in your 12v system.

Also do not do this if you have 2 channels on your scope as each speaker - output on the amplifier will have a different voltage signal burning out you probe's GND or even the scope itself.

Thanks Brian, the scopes case is plastic so I should be alright. The 12v ground of the amp is in parallel with the heatsink so I'll probably just use that.

I'm going to be using sine waves with equal output on the left/right channels so tolerances aside they should be similar. The scope is only one channel so I'm good to go either way. If someone reads this in the future though, that's good knowledge.

I'll report back with my results, which I'm sure will be positive. I too want to probe that central "shared output" to see what's on it.
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: Probing full bridge amplifiers with oscope
« Reply #7 on: July 17, 2019, 09:01:45 pm »
I downloaded the owner's manual for the specific amplifier and near as I can tell - I'm not too familiar with car audio stuff - is that the center +/- terminal is actually power ground and that channel 2's signal is permanently inverted, hence you need to connect the speaker "backwards" to channel 2 if you want to maintain correct phasing when using each channel separately, rather than bridged. So, no big mystery, really.

Given the type of scope you have, I agree with BrianHG's suggestion that you simply connect your scope to channel 1 (+) and channel 2 (-). Note that most amps - especially Class D types - will not produce a reasonable looking output waveform unless a load is connected. If you want to test amps more than just this one time then get some wirewound resistors to use instead of speakers. I suggest (4) 2 Ohm resistors that way you can wire them up in series/parallel combinations to get 0.5, 2, 4 and 8 Ohms total.

 

Offline Hyper_SpectralTopic starter

  • Regular Contributor
  • *
  • Posts: 86
  • Country: us
Re: Probing full bridge amplifiers with oscope
« Reply #8 on: July 17, 2019, 09:14:59 pm »
I downloaded the owner's manual for the specific amplifier and near as I can tell - I'm not too familiar with car audio stuff - is that the center +/- terminal is actually power ground and that channel 2's signal is permanently inverted, hence you need to connect the speaker "backwards" to channel 2 if you want to maintain correct phasing when using each channel separately, rather than bridged. So, no big mystery, really.

Given the type of scope you have, I agree with BrianHG's suggestion that you simply connect your scope to channel 1 (+) and channel 2 (-). Note that most amps - especially Class D types - will not produce a reasonable looking output waveform unless a load is connected. If you want to test amps more than just this one time then get some wirewound resistors to use instead of speakers. I suggest (4) 2 Ohm resistors that way you can wire them up in series/parallel combinations to get 0.5, 2, 4 and 8 Ohms total.

Those would be some large resistors! 200w per channel out of this amp is a lot of power. I've never had a problem getting an output from a Class D amp with the scope until now (but they were all half bridge). I tried connecting the scope to Ch 1 + and ch 2 - and i had no output so either there was an issue with the dsp and the demand of the sine wave or there was no signal.

Generally if I'm concerned that the psu rails will sag under load I'll just back the gain down after reaching the clip point. Most music is recorded at -5 db anyways and I use 0db tones.

I'm hearing two different things now, connect the probe to the + output and then the scope ground to 12vDC ground, and one saying try it in bridge mode. I'll try both?
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: Probing full bridge amplifiers with oscope
« Reply #9 on: July 17, 2019, 09:34:58 pm »
Those would be some large resistors! 200w per channel out of this amp is a lot of power. I've never had a problem getting an output from a Class D amp with the scope until now (but they were all half bridge). I tried connecting the scope to Ch 1 + and ch 2 - and i had no output so either there was an issue with the dsp and the demand of the sine wave or there was no signal.

Class D amps have an LC or LCL output filter that averages the pulse-width-modulated (PWM'ed) square wave into the original audio waveform (well, close to it, anyway), but this filter needs a load present to actually do its job. 50W-100W chassis mount wirewound resistors are maybe $4 each from a reputable source like Mouser, Digikey, etc. Probably even cheaper than that if you get them from a dodgy source like ebay or Amazon. Also, they can be be grossly overloaded for short periods without harm - 10x rated power for 1-2 seconds is easily doable.

I'm hearing two different things now, connect the probe to the + output and then the scope ground to 12vDC ground, and one saying try it in bridge mode. I'll try both?

Well, I don't know about connecting the ground clip of the scope to 12V ground - likely some kind of push-pull step up power supply is inside the amp to boost the nominal 12V from the car battery into +/-30 to +/-50V or so, and the ground reference on the power side might not (shouldn't, really) be connected to chassis ground on the 12V side. So, you could connect scope ground to the +/- terminal and touch the probe tip to Channel 1 (+) or Channel 2 (-) to see if each channel is working, then connect scope ground to, say, Channel 2 (-) and scope tip to Channel 1 (+) to see if bridged mode is working. Again, with some kind of load present to give the best chance of seeing a realistic output from the amp.

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf