Jump to content

Transconductance amplifiers?


Chris A

Recommended Posts

Getting closer

P1020030.jpg

 The 1/4" input jacks needed to by sacrificed to accommodate the current sense resistors. Not planned but that's how it worked out. The 1/4' jacks on the D75 are unbalanced inputs unlike the D45/D75A where they are directly in parallel with the XLR inputs and therefore balanced. I will never have a need to use the unbalanced inputs.

P1020029.jpg

P1020028.jpg

Current sense resistors epoxied to chassis

P1020031.jpg

Dual push switches on left replacing original slide switch. Added 4PDT toggle on right

  • Like 2
Link to comment
Share on other sites

On the negative feedback,

do you have to change the value on the remaining resistor after the 2nd ground is removed ?

 

To maintain the same levels of attenuation on the various frequencies ?

 

Can you jump a variable attenuater around the resistor

so you can vary the input of the Negative feedback ?

Can be compared to the stock channel.....

Link to comment
Share on other sites

34 minutes ago, Bubo said:

On the negative feedback,

do you have to change the value on the remaining resistor after the 2nd ground is removed ?

The books author states the current sense resistor is so low in value that the original feedback resistor will have no effect anymore....we'll see. After all I am trying to mod an existing old Crown amp and it may not be a suitable candidate for this type of operation...again we'll see. I do not understand the question about the second ground.

Have you read the book? Or looked at the author's website?

Link to comment
Share on other sites

Got it all wired up ---you point to point wiring tube guys would be proud but.........

So far it is an EPIC fail:angry2: Appears I have built a fusing blowing oscillator.

Not giving up just yet...it will take all the circuit engineering skills I have to figure this out.

If it takes more than I possess then I will give and change it back.

So I have never used the light bulb as current limiter on the input power before. Do I just wire it in series with the AC hot?

  • Like 1
Link to comment
Share on other sites

Just for fun...

 

Here's a simulation of my Klipschorn bass cabinet, properly loaded by a room corner, no crossover, driven by an amp with a 0.02Ω output impedance vs. a First Watt amp with an 80Ω output impedance.

 

Yellow trace is the voltage across the driver terminals and the Teal trace is the current passing through the driver. If you were to divide voltage by current, you'd get impedance c/o Ohm's Law. Since the two illustrations are very close to constant voltage (left) and constant current (right), you can see the driver/cabinet/corner impedance quite well in the other trace. I love this stuff. : )

 

1238832413_ConstantEvsI.thumb.png.41186b66a20a0423d4215320f4ae8417.png

 

It's obvious that a current source amp for audio application has a lot more to it than just a high output impedance.

 

I need to read the links provided - thanks for the fascinating thread! : )

 

God bless you and your precious family - Langston

Edited by Langston
Link to comment
Share on other sites

Meanwhile...... I need help...any and all thoughts and suggestions are appreciated.

First things first. Build a light bulb current limiter. Then I need to connectorize all the connections from the circuit board in the amp to the various points. Before it was tedious to make and remake the connections. Now with all the added parts it is darn near impossible. Once it is connectorized it will be much easier to teardown to remove or add a part to stabilize this amp(oscillator) and then re assemble.

Link to comment
Share on other sites

  • 2 weeks later...
On 5/15/2021 at 6:05 PM, John Warren said:

99.999999% of audio amplifiers are voltage sources.  If transconductance amplifiers were a better mouse trap, the world would have marched to them immediately.

 

Why do you think it did not?



Likely, development and distribution cost can't fit into what the marketing department is willing to sell.

Link to comment
Share on other sites

  • 3 weeks later...
  • 3 months later...
On 11/30/2016 at 10:35 AM, Chris A said:

Thanks Carl.  This is actually a pretty esoteric subject, apparently.   Many people really don't know what a transconductance amplifier is.  It's a current-source amplifier instead of a voltage-source amplifier. 

 

What most people don't really realize is that their Klipsch loudspeakers operate off of current--not voltage.  Using higher output impedance amplifiers--such as SETs--is really about pushing their amplification paradigm towards current source rather than voltage source amplification.  The problem is that most people are also not prepared to hear the boost in low frequency output that they will get around the woofers' resonant frequency when using a true transconductance amplifier.  So all they have to do is to EQ their current source amplifiers down in the 40-60 Hz region [EDIT 19 Aug. 2018:...and the upper midrange region (2-7 kHz)], depending on exact loudspeaker model you're using.  A lot of people seem to have trouble understanding how this can be true, and abandon their transconductance amplifier trial without first dealing with this phenomenon.

 

Chris

Would the First Watt F8 fall in this category, Current source?

Link to comment
Share on other sites

  • 2 weeks later...
On 6/10/2021 at 5:10 AM, Langston said:

Just for fun...

 

Here's a simulation of my Klipschorn bass cabinet, properly loaded by a room corner, no crossover, driven by an amp with a 0.02Ω output impedance vs. a First Watt amp with an 80Ω output impedance.

 

Yellow trace is the voltage across the driver terminals and the Teal trace is the current passing through the driver. If you were to divide voltage by current, you'd get impedance c/o Ohm's Law. Since the two illustrations are very close to constant voltage (left) and constant current (right), you can see the driver/cabinet/corner impedance quite well in the other trace. I love this stuff. : )

 

1238832413_ConstantEvsI.thumb.png.41186b66a20a0423d4215320f4ae8417.png

 

It's obvious that a current source amp for audio application has a lot more to it than just a high output impedance.

 

I need to read the links provided - thanks for the fascinating thread! : )

 

God bless you and your precious family - Langston

 

 

Current is what moves a speaker not voltage.

 

As you can see with a voltage source the voltage does not change but since the resonance has a very high impedance the current has decreased substantially through that region. With a current source amplifier with a high z out the voltage actually changes along with impedance, so less impedance lower voltage, higher impedance higher voltage to keep the CURRENT the same.

 

How does this add up to sound? A current source or transconductance amplifier will track the impedance curve and distribute the power evenly across all frequencies no matter the frequency/impedance. The thought is 1 amp of current for example will move the motor the same no matter the impedance. Since it is the current and magnetic field that is moving the motor you would want to have a current source. Unfortunately most speakers today are designing their speakers for a voltage source and not a current source things do not work out how they should shown by not only looking at the impedance vs frequency plot but the SPL vs frequency plot. The sound pressure or sensitivity changes so with that high impedance bump in the bass region does not need the same current to reach the output level as other frequencies closer to nominal impedance. So with a current source amp you may have very high increased "one note" bass where the resonance is. 

Edited by captainbeefheart
spelling
Link to comment
Share on other sites

  • 4 weeks later...

Forgive me if this is an ignorant question, but where would the Schitt amps fall into this discussion. They call their amp design Continuity and describe it as:

 

"Technically, Continuity is a way to eliminate transconductance droop outside of the Class A bias region, and extend the benefits of Class A biasing..."

 

 

Link to comment
Share on other sites

@captainbeefheart I am soliciting help. If you have seen any of my previous posts about trying to mod an old Crown amp to be a transconductance amp. So far i have only succeeded in making a fuse blowing oscillator. I need help stabilizing this thing. I am thinking i need to decrease the open loop gain. Any thoughts? Schematics are available if needed.

Link to comment
Share on other sites

17 hours ago, CWelsh said:

Forgive me if this is an ignorant question, but where would the Schitt amps fall into this discussion. They call their amp design Continuity and describe it as:

 

"Technically, Continuity is a way to eliminate transconductance droop outside of the Class A bias region, and extend the benefits of Class A biasing..."

 

 

 

That is different than what we are discussing, they are not saying their amps are transconductance amps.

 

If you ever look at an active device datasheet you will notice that the transconductance is not a constant, it changes throughout different operating points, usually higher current through the device gives higher transconductance. Transconductance is good, it means it changes more current for the same input drive. In class A all devices are conducting the full 360° waveform, Class B the devices are on more than 50% of the 360° waveform but not the full 100% of it, so at one point a device goes into cut-off and shuts off with no current flow. Going back to what I said, the more current through a device the higher the transconductance is so you can see how the total current gain of the output in a Class AB amp is not steady but changing as the operating conditions of the devices change. Schiit is saying they try and keep their current gain of the output stage as stable and steady as possible even during Class of operational changes.

Link to comment
Share on other sites

15 hours ago, babadono said:

@captainbeefheart I am soliciting help. If you have seen any of my previous posts about trying to mod an old Crown amp to be a transconductance amp. So far i have only succeeded in making a fuse blowing oscillator. I need help stabilizing this thing. I am thinking i need to decrease the open loop gain. Any thoughts? Schematics are available if needed.

 

 

I would love to help. Of course the easiest way to turn any amplifier into a transconductance amplifier is to simply place a resistor in series between your amplifier and speaker.

 

I will look for the stock schematic, let me know exactly what you have changed so far.

Link to comment
Share on other sites

5 hours ago, captainbeefheart said:

 

That is different than what we are discussing, they are not saying their amps are transconductance amps.

 

If you ever look at an active device datasheet you will notice that the transconductance is not a constant, it changes throughout different operating points, usually higher current through the device gives higher transconductance. Transconductance is good, it means it changes more current for the same input drive. In class A all devices are conducting the full 360° waveform, Class B the devices are on more than 50% of the 360° waveform but not the full 100% of it, so at one point a device goes into cut-off and shuts off with no current flow. Going back to what I said, the more current through a device the higher the transconductance is so you can see how the total current gain of the output in a Class AB amp is not steady but changing as the operating conditions of the devices change. Schiit is saying they try and keep their current gain of the output stage as stable and steady as possible even during Class of operational changes.

Thank you for this. Now, let's see if I understand...

 

It sounds to me like there is a phenomena of transconductance that occurs in any kind of amp design, but there are amps that are specifically designed to take advantage of that phenomena. What Shitt refers to, then, is a design that attempts to overcome that point where there is a shift from Class A to Class B operation and the current flow ceases (droops).

Link to comment
Share on other sites

3 hours ago, CWelsh said:

let's see if I understand...

 

It sounds to me like there is a phenomena of transconductance that occurs in any kind of amp design, but there are amps that are specifically designed to take advantage of that phenomena. What Shitt refers to, then, is a design that attempts to overcome that point where there is a shift from Class A to Class B operation and the current flow ceases (droops).

 

Since tubes and Mosfets are voltage controlled devices, for a given change of input voltage will have a change in current through the device, this is transconductance. So say 10mA of current change for every 1v change at input. Bipolar junction transistors are a little different, they are not controlled by a voltage but an input current and it's called Beta. So a Beta of 100 means for an input current of 1mA the device will change 100mA through it. This is how us engineers make amplifiers, we take smaller signals/currents and use active devices to enlarge them so to speak, this is called gain. All amplifiers need gain to make small signals/currents larger enough to drive a loudspeaker.

 

All amplifiers take advantage of transconductance or Beta, the ones classified as a "transconductance" amplifiers is nothing more than just one that has a very high output impedance and behaves more like a current source compared to the vast majority of amplifiers that are a voltage source which has a very low output impedance. Schiit is just saying they try and keep the transconductance through the devices they are using as constant as possible, if you have an amp that droops in transconductance it is more or less weaker if you will.

  • Like 2
Link to comment
Share on other sites

On 11/25/2021 at 1:26 AM, captainbeefheart said:

 

 

I would love to help. Of course the easiest way to turn any amplifier into a transconductance amplifier is to simply place a resistor in series between your amplifier and speaker.

 

I will look for the stock schematic, let me know exactly what you have changed so far.

The amp is a D75, not a D75A. If you can't find the scheme I can email a copy. The D75 has two single opamps and one dual. In the A version they combined all four into a quad. I also have a word doc enumerating the changes I made. Since I did 4 of them I needed this just to make sure I did the same to all.

Thanks for your offer to help...Its Thanksgiving weekend....more come Monday when I'm back in the office.

 

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...
On 10/18/2021 at 2:04 AM, VDS said:

Would the First Watt F8 fall in this category, Current source?

 

No, by far, it has an output impedance of 0.2Ω. The F1 and F1J come close (80Ω and 60Ω).

 

In my opinion with a damping factor of lower than 1% you are getting there... More than 800Ω (400Ω) for an 8Ω (4Ω) loudspeaker. I build an amplifier with an output impedance close to 1400Ω.

 

In my opinion a damping factor higher than 100 is voltage-drive (ideal is ∞) between 10% and 10 is power-drive (ideal is 1) and smaller than 1% is current-drive (ideal is 0), the rest are in-betweeners.

Link to comment
Share on other sites

On 5/17/2021 at 11:59 PM, Chris A said:

I use an F3.  As far as I'm concerned, any progress toward using current source amplifiers is welcome in this thread. 

 

I may not be contributing strongly due to the very apparent "growing pains" of putting out this subject for public consumption (and the battle scars that apparently go with that), but anyone that is actually trying something on this subject has highest priority to post here--in order to make this thread more useful.

 

Chris

 

OK, next try... My first post in this thread was on April 28 2018, but no response...

 

In 2017 I designed and build a current-drive system:

- 2-way acoustic suspension loudspeakers, without crossovers and with (sorry, no horns) Scan-Speak drivers,

- a 4-channel current-drive amplifier with an output impedance of 1380 Ω,

- a 2x4 channel DSP/DAC for EQ, crossover and Dirac Live.

 

I used woofers with non-conducting voice coil formers to avoid back-emf eddy currents (the other back-emf that also creates modulated noise).

 

The amplifiers are based on an existing design for a voltage-drive amplifier of which I changed the feedback into negative current feedback and I changed the output stability filters (I would like to build an amplifier with a complementary cascode configuration).

 

In the DSP I used filters and Linkwitz-transformers to create phase coherent crossovers (based on the ideas of Duelund), and to extend the bandwidth to full range. It's a full-range 2-way system (from just below 20 Hz).

 

Since it was operational I gave away my voltage-drive system, no voltage-drive for me anymore...

 

I'm open for questions, about the basic principles behind current-drive and about my system.

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...