How to modify a class D amp to work below 20hz?

nate san

Member
2016-05-23 9:32 pm
I am using an audio amp and speakers for a very unconventional method. I want to compress a fluid with a speaker to mimic a heart beat. That means a sub 1Hz response is required.

However, most amps aren't rated below 20Hz. As I understand it, I can remove the output capacitors on a class D amp to remove it's low pass filter. In this case it's output range goes from DC to it's rated maximum. I would just need to ensure that my input power is clean to reduce noise.

Would this be a viable option?
 

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
I'm not aware that Class D amps have 'output capacitors'. That term normally relates to AC coupled conventional designs.

Most Class D amps use a bridge configuration with the speaker directly coupled between two outputs stages that are driven in antiphase with respect to each other, however there is normally AC coupling at the input which would roll the response off. Any caps on the output of a Class D amp are there for filtering all the noise and hash normally produced by such output stages. The speaker current does not flow in these caps.

Do you really need Class D for this ? I'm thinking a simple 'chip amp' would suffice here.
 

nate san

Member
2016-05-23 9:32 pm
I'm not aware that Class D amps have 'output capacitors'. That term normally relates to AC coupled conventional designs.

Most Class D amps use a bridge configuration with the speaker directly coupled between two outputs stages that are driven in antiphase with respect to each other, however there is normally AC coupling at the input which would roll the response off. Any caps on the output of a Class D amp are there for filtering all the noise and hash normally produced by such output stages. The speaker current does not flow in these caps.

Do you really need Class D for this ? I'm thinking a simple 'chip amp' would suffice here.

I may have confused what someone told me about class D and Class A amps.

Well I'm driving about 160 W. I'm not sure I could get something with that wattage on a chip.
 

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
OK, well Class A amps are usually smaller low powered designs designed for out and out fidelity. They are spectacularly inefficient and run very hot. Class D is the other extreme, very efficient but not the ultimate in fidelity.

160 watts is a lot of power (assuming you mean the proper text book definition of it being rms or root mean square) and would normally be done by a conventional class B type design... and they cover most audio type amplifiers. That kind of power still means pretty big power supplies and heatsinks.

'Chip amps' are commonly available up to 50+ watts rms, some go higher.

Perhaps if you could give an idea of the type of speaker you are trying to drive then we might have a better idea what would be best.

(to give you an idea of 'power', most audio listening on a good hifi system is done at only a couple of watts or so)
 

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
Interesting (speaker design isn't my specialty and I haven't seen anything like that before).

I think the bottom line comes down to how hard you actually need to drive these to get the results you need. I couldn't begin to guess at that... and you might find its a lot less than you think. So much will depend on how you couple these to the fluid... I'm guessing you might have some flexible membrane that the speaker works into.

My advice, if you are starting from zero on this, would be to try a small DC coupled chip amp design and see if that gets you in the right ballpark on this. You would at least get a feel for what is possible or not.

I don't know what your construction abilities are... would you be making the amp yourself or would you be looking to modify an existing offering such as a kit.

I'm thinking something like a TDA2050 based design could work here. Its simple and has a basic maximum current drive of around 5 amps before it limits. At such low frequencies you are really looking at the basic properties of ohms and amps rather than complex impedance curves over a wide frequency range and so on.

Interesting, I'll look in again later.
 

29285

Disabled Account
2010-01-17 9:20 pm
You could use an array of 8 el cheapo class-d-modules, each one firing a single speaker.
To extend to sub-audio-frequencies, you have to change input coupling caps, and the bootstrap caps as well. As a rule of thumb, you might increase all capacitances by a factor of ten.
 

nate san

Member
2016-05-23 9:32 pm
You could use an array of 8 el cheapo class-d-modules, each one firing a single speaker.
To extend to sub-audio-frequencies, you have to change input coupling caps, and the bootstrap caps as well. As a rule of thumb, you might increase all capacitances by a factor of ten.

That is an interesting solution.

So you're saying i should short the input and and bootstrap caps, and then increase all other caps on the board by a factor of ten?
 

gpapag

Moderator
Paid Member
2002-11-17 6:42 pm
Athens-Greece
I am using an audio amp and speakers for a very unconventional method. I want to compress a fluid with a speaker to mimic a heart beat. That means a sub 1Hz response is required.
However, most amps aren't rated below 20Hz.

You don’t have to use a sub Hertz sinusoidal signal for that.

You can get away from the low frequency limitation of the amplifiers by driving the amplifier -any audio amplifier- with a pulse repeating every x seconds, where x is the time period of the heart pulse you want to imitate (e.g 1.5s is the time period of 40 beats per minute, 0.7s is the time period of 80 beats per minute)

So, you don't need to modify an amplifier for doing your experiments. (what you suggested as a mod in your first post, won't work anyway)
George
 
I am using an audio amp and speakers for a very unconventional method. I want to compress a fluid with a speaker to mimic a heart beat. That means a sub 1Hz response is required.

However, most amps aren't rated below 20Hz. As I understand it, I can remove the output capacitors on a class D amp to remove it's low pass filter. In this case it's output range goes from DC to it's rated maximum. I would just need to ensure that my input power is clean to reduce noise.

Would this be a viable option?

No. Input capacitors must be omitted in a full bridge, symmetrical supplied amp.
 

Pano

Administrator
Paid Member
2004-10-07 6:05 am
Panama
No need to worry about the output capacitors. They are part of a LOW pass filter. That filter passes frequencies that are below 50Khz or whatever the switching rate of the Class-D amp is. They won't do a thing to the frequencies you need.

It's the INPUT caps you need to worries about. They "may" be limiting the incoming signal. You can replace them with a larger value cap. Just be aware that on some amps those input caps need to charge up when you turn the power on. That can cause a startup thump in the speakers. The larger the cap, the bigger the thump. But thump is what you are looking for anyway. :D
 
I have done something similar using an amp and a speaker driver to pulse puffs of air in a small chamber for fluid dynamics experiments. I needed absolute DC control of the voice coil position. You basically need a complementary output class AB amp with DC coupling. You can build one of just buy a Burr Brown (now TI) OPA541 High power DC capable op amp. You can go down to DC so sub Hz is possible. It's capable of 40v rails and 5v continuous current. Handy little voice coil driver. Probably would make a decent audio amp.

OPA541 | Precision Amplifier | Operational Amplifier (Op Amp) | Description & parametrics

The TO220 package is a bit friendlier and this AP version goes to 70v. $22.

http://www.newark.com/texas-instruments/opa541ap/op-amp-1-6mhz-10v-us-10mv-to-220/dp/72K9048
 
Last edited:
I know you said driver choice is convoluted and fixed, not subject to change, but I would feel bad if I didn't point out (may already be known to you) that the Dayton exciters you linked to are wholly inappropriate for the task and likely severely inadequate. Their resonant frequency is 365 Hz, they are smallish and lightweight, meant for full range sound reproduction, and likely have horrid low frequency response.

Their ability to drive anything, much less a mass of fluid, at a frequency of 1Hz is probably just about nil. So IMO, it doesn't matter at all what you do for amplification, it isn't going to work.

If it is at all possible to change drivers and not waste your time, do so. If you want to stick to pistonic linear motion speaker like devices, you should be looking for bass-oriented tactile transducers. Search for "bass shakers" and you will come up with some options. Crowson, Aurasound, and Buttkicker come to mind.

Another reasonable alternative is a rotational servo motor, connected like a locomotive drive wheel, similar to some of the old Seaton servo subwoofers.

Then, you can start asking about appropriate amplification. Buttkicker and Crowson happen to sell packages with both amp and transducer, so it may be worth checking out their frequency response limits.
 

BWRX

diyAudio Moderator Emeritus
2005-01-17 5:29 am
Pennsylvania
Nate San, did you ever accomplish what you were out to achieve? Just curious if you came up with a solution and what you ended up using. Currently, I have an interest in doing something similar to what you had originally mentioned.

Nobody mentioned a linear actuator, which could provide more force than a tactile transducer of similar size and cost.

You don’t have to use a sub Hertz sinusoidal signal for that.

You can get away from the low frequency limitation of the amplifiers by driving the amplifier -any audio amplifier- with a pulse repeating every x seconds, where x is the time period of the heart pulse you want to imitate (e.g 1.5s is the time period of 40 beats per minute, 0.7s is the time period of 80 beats per minute)

If you need to use a tactile transducer, this was the best suggestion when the thread was active.
 
Last edited: