Joe Rasmussen "Trans-Amp" - 40 Watt Transconductance "Current Amplifier" - diyAudio
Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 17th July 2013, 02:53 AM   #1
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
Default Joe Rasmussen "Trans-Amp" - 40 Watt Transconductance "Current Amplifier"

.
PLEASE BUILD THIS WITH DUE CAUTION!

[Click on it to enlarge]

Click the image to open in full size.

There are good reasons why there are no "Current Amplifiers" on the market today, with exception of the Nelson Pass F1 and F1J amplifiers. This DIY project is about building a low cost 40 Watt Transconductance Amplifier using National Semiconductor's LM3875T.

WHY BUILD ONE?

Before going ahead: READ THIS BY NELSON PASS: Current Source Amplifiers / Full Range Drivers

Conventional amplifiers are of cause "Voltage" types, where the amplifier outputs Voltage similarly to what we see on an oscilloscope with X-Y axis' representing Voltage (vertically) versus Time (laterally). The current is dependant on the load.

With a "Current" type amplifier, this is far less intuitive to understand, but basically it is the opposite, here the voltage is dependant on the load.

It becomes then clear that the Voltage Amplifier needs to have a low output impedance, so that the amplifier can control the Volts without any sagging. Effectively, the Voltage is 'regulated.'

What is not so clear is that a Current Output Amplifier needs to have a VERY HIGH OUTPUT IMPEDANCE!

So a Voltage Amplifier is defined by its Low Impedance.

A Current Amplifier is defines by its High Impedance.

The notion that there is any such thing as "Damping Factor" - with a Current Amplifier - forget it!

Indeed Damping Factor is largely a myth, what really happens, as Richard Small told me way back in 1975 (he is the S in T-S), that any series impedance/resistance with the Voice Coil erodes and if high enough, destroys the speakers Bass Alignment - so that the Electrical Damping within the speaker is non-existent.

If the DCR of the Voice Coil is 6 Ohm and Qes = 0.5, then an additional series of 6 Ohm will double the Qes = 1.0... and so on. Your Bass Alignment no longer works.

SO THIS AMPLIFIER CAN DAMAGE THE WRONG SPEAKERS!

Also, the Crossover (High Pass) to the Tweeter could also damage it in a very short time.

Yes, the use of Current Output Amplifiers are suitable for Full Range Drivers and also speaker systems that are designed to work with current.

Such as the latest Elsinore Mk5 Loudspeaker Project.


Cheers, Joe R.
.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY

Last edited by Joe Rasmussen; 17th July 2013 at 12:15 PM.
  Reply With Quote
Old 17th July 2013, 03:29 AM   #2
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
.
A FEW NOTES:

What is a Transconductance Amplifier?

Basically, this is an amplifier that has a Voltage input and outputs an equivalent defined Current output.

In this case, this amplifier is designed for 100mV signal input, to produce 0.25 Amp - or 250mA output. Hence 200mV input will produce 500mW input and 400mV input produces 1 Amp, and so on.

Voltage in and a specific Current out.

Look at the column to the right of the Schematic, called STRUCTURE.

If 100mV input and we have a 4 Ohm Load, then 250mA through that 4 Ohm Load should gives us 1V exact. We come very close indeed at 1.031V measured.

I an 8 Ohm Load is used, then we should see 2V and we get 2.033V measured.

We can now see that this is indeed behaving as a transconductance device, and the small error means we can calculate the Output Impedance as near 270 Ohm.

Note too, that the Voltage Gain is dependant on the Load. Hence the Gain is 10 (20dB) with 4 Ohm and 20 (26dB) with 8 Ohm, and so on.

PLEASE NOTE THAT THE SCHEMATIC IS ALSO A WIRING DIAGRAM.

Also:

USE ENTIRELY DUAL-MONO CONSTRUCTION.

So you will need TWO Power Transformers, min 160VA.

You could also use a single Power Transformer, but will double the number of Secondaries. If you don't get the wiring right, expect a fair amount of hum. It seems that the LM3875 is internally signal referenced to the Negative Rail +35V. So it is important to get it right.

When constructed, you will need to null the output. Use a 10R Resistor, then to fine-tune I would use a 100R value. Keep in mind that DC on the output also changes with Load. Ideally, with no Load connected, the 470R should see no DC.

Note the Resistor 0R33 - I used three 1 Ohm Resistors in parallel. These must be low inductance and the 6.5W rating is minimum.

This 0.33R is the "Current Sense" Resistor. The LM3875 input voltage on Pin 7 "+" input, also appears at Pin 8 "-" input. Don the maths and with 100mV on both pins, the current through our "Current Sense" Resistor will be 303mA. This does not seem to match our target of 250mA? But it does match after the input loss of the Bandpass Filter on the input of the RCA, so that 100mV on the RCA input will give us close to 250mA output.

About that Bandpass Filter, it is worth experimenting with the sonic influence of 1n5, so tweak this, reduce it to brighten the balance or find the sweet spot you like, and increase it as well. So tweak away, but don't entirely remove it. There are slew-rate reasons to include it as it inter-acts with the internal "Comp-Cap" of the LM3875 and helps define reduced feedback above 100MHz.

Next... some interesting measurements...

Maybe refrain from adding comment until they are posted, and I am finished.


Cheers, Joe R.
.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY

Last edited by Joe Rasmussen; 17th July 2013 at 03:33 AM.
  Reply With Quote
Old 17th July 2013, 03:40 AM   #3
frugal-phile(tm)
diyAudio Moderator
 
planet10's Avatar
 
Join Date: Oct 2001
Location: Victoria, BC, NA, Sol III
Blog Entries: 5
Good to see you finally get this going Joe.

We have had a somewhat similar transimpedance amplifier kicking around for the last few years (ours has variable output impedance so that you can dial in the output impedance that works best with your speakers).

Buried in what Joe said, and that is important to remember, is that the system response is the FR of the speaker (measured being driven with a voltage amp) convolved with the speaker's impedance.

Since a speaker is controlled by current not voltage there are some significant gains in distortion reduction to be had with a current amp -- with a voltage amp the speakers impedance acts as an imperfect V/I converter.

A good reference NEW: Book on Current-Drive of Loudspeakers Published. His book is recommended.

dave
__________________
community sites t-linespeakers.org, frugal-horn.com, frugal-phile.com ........ commercial site planet10-HiFi
p10-hifi forum here at diyA
  Reply With Quote
Old 17th July 2013, 04:00 AM   #4
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
.
1 WATT DISTORTION MEASUREMENTS:

The "First Watt" is always the most important. All below measurements were made into an 8 Ohm Dummy Load.

Click the image to open in full size.

1 Watt THD 0.023% & THD+N 0.051%

Note that Odd Orders are well suppressed. Small amount of 1/3rd and 1/5th. All the others are Even-Orders and of less concern, even if elevated relative to Odd-Orders. Higher power measurements shows an interesting trend, the Odd-Orders are on their own plateau than the Even-Oders on a higher plateau. So please continue...


Click the image to open in full size.

1 Watt IMD 0.026%

The clusters form around Even-Order frequencies. Quite pleased.

Next, higher 10 Watt measurement.

Cheers, Joe R.
.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY
  Reply With Quote
Old 17th July 2013, 04:12 AM   #5
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
.
10 WATT DISTORTION MEASUREMENTS:

Click the image to open in full size.

10 Watt THD 0.023% & THD+N 0.028%

Here we see that interesting trend again. All Odd-Orders are suppressed relatively to the Even-Orders immediately above and below - on a lower plateau.

Click the image to open in full size.

10 Watt IMD 0.024%

Quite pleased.

Cheers, Joe R.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY
  Reply With Quote
Old 17th July 2013, 04:26 AM   #6
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
.

And interesting trend when looking at 10 Watt THD:

Click the image to open in full size.

I would ignore that 16KHz peak, I seem to get this quite often when using 96K sampling frequency with other DUTs (devices under testing). So it is an aberration of some kind.

DO YOU SEE THE TREND?

The Black lines join Even-Order Harmonic Distortions and the Blue the Even-Orders. Note they are on different levels or plateaus?

I see a similar charactersitics in the 10 Watt and 20 Watt IMD measurements, the IMD clusters are visible for the Evens, but noticeable suppressed for the Odds.

This may well seem a characteristic for transconductance amplifiers.

I can say that the amplifier is very nice to listen to, IF you have compatible loudspeakers. Not your ordinary Chip Amp.

Cheers, Joe R.

PS: I also have 40 Watt THD measurement, showing that it clips well above 40 Watt. Also have IMD measurements up to 20 Watt (with current setup limited to 20W IMD). May post these later.
.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY

Last edited by Joe Rasmussen; 17th July 2013 at 04:28 AM.
  Reply With Quote
Old 17th July 2013, 04:39 AM   #7
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
Quote:
Originally Posted by planet10 View Post
Good to see you finally get this going Joe.
Thanks, Dave. The last two months has been as busy as I have ever experienced, rather tired at the moment.

I have that book and have had a few exchanges with him. The point is really this, for those who are reading this and not in the know, all 'motors' are current devices and it is current that makes them move. So there are significant reduction in phase noise (increased clarity), that comes from current driving.

But what do you think of those measurements, this is not the way a regular chip amp would measure, that's for sure. So into a non-reactive 8 Ohm Load, they look rather good and not exactly the norm.

I wish I had a straight chip amp here that I could measure and compare.

Off to the Post Office...

Cheers, Joe

PS: The dial-in output impedance is a neat idea - not so easily implemented with this design, as I see it.

BUT... Take a 15 Ohm large Wattage Resistor and put in parallel with the speaker (so that the amp also sees it), and that 270R output Z will come down - vary that resistor, but don't drop below, say 10R, with 4 Ohm speakers. So 10R to 100R will do a similar job - albeit drawing a bit of power/loss.

.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY

Last edited by Joe Rasmussen; 17th July 2013 at 04:42 AM.
  Reply With Quote
Old 17th July 2013, 12:13 PM   #8
diyAudio Member
 
Joe Rasmussen's Avatar
 
Join Date: Jan 2003
Location: Sydney, Australia
Send a message via MSN to Joe Rasmussen
This is the one I built. I uses only a single 300VA transformer, but I don't recommend it. It was very tricky indeed. Best to go with separate transformers, or be prepared to sort out some issues.

Note that the Heatsink needs to be a good size.

Note the use of Jantzen Superior Red input coupling capacitor. I highly recommend it - as this needs to be of high quality.

Click the image to open in full size.
__________________
The "Elsinore Project" DIY Speaker System
Custom Analogue Audio - we also support and promote non-profit DIY
  Reply With Quote
Old 17th July 2013, 02:26 PM   #9
gadut is offline gadut  Indonesia
diyAudio Member
 
gadut's Avatar
 
Join Date: Dec 2008
Location: Jakarta, Indonesia
nice project, you always come with a fresh design. how does it sound compare to a standard LM3875?

Regards
  Reply With Quote
Old 17th July 2013, 05:26 PM   #10
diyAudio Member
 
Loudthud's Avatar
 
Join Date: Nov 2007
Location: Near Dallas Texas USA
Is the DC Trim pot shown correctly on the schematic? Is there any reason that other power amp ICs won't work or not be suitable for this design?
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
"WTB" "The Wire AMP" Class A/AB Power Amplifier based on the LME49830 with Lateral enantra Swap Meet 3 23rd December 2012 06:36 PM
Building "The Wire AMP" Class A/AB Power Amplifier" wirewiggler Solid State 12 23rd September 2012 03:40 PM
That curious extra knob, or: "Trans-Amp" Duo Solid State 111 24th November 2011 11:57 PM
What makes an amplifier "bright", "warm", or "neutral"? JohnS Solid State 51 13th December 2009 06:42 PM
WTB: Focal & Audiom & Volts 11", 12", 13", 15" woofer gengis Swap Meet 0 17th March 2005 02:29 AM


New To Site? Need Help?

All times are GMT. The time now is 12:48 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2