The New-Tron - Page 7

djk · 26th May 2006, 10:12 AM

You may find the resistors to ground may be a much higher value, 500 ohms or so. I think they only need carry the gate current for the outputs.

EUVL · 26th May 2006, 10:26 AM

I think the discussion is not what value they should be (could even be 1k or 10k), but rather how to get rid of them altogether, and perhaps why they should be disadvantageous.

As promised the revised schematics.

Patrick

jh6you · 26th May 2006, 02:05 PM

Quote:

Originally posted by EUVL
jh6you,

How could we address you by name ?

Regards

JH

flg · 26th May 2006, 02:29 PM

Well, there was a question some posts ago and recently regarding why we need those 2 Rs to Gnd... The 2 power supplies that supply the output circuit are floating. They have no Gnd connection. Most discussions of drivers, refrences some drive voltage to ground. Drive voltage refrenced to Gnd will do nothing to the output. The drive needs to be Gate to source. Hence, the Rs allow a gate to source voltage return path while still alowing the output supplies to float.
My circuit has no feedback and it will not work without the Rs. Transformers were the original method to drive this thing. They supply an isolated gate to source drive.
And then ther is me. I'm the one who wants to use some parameters that are most important to make the Rs in my circuit optimum. I've just been putting it off for bench evaluation. I beleive they will be most important on a real circuit with imbalances and offsets, etc. All the perfection of the simulator is not pointing out anything particular to help me understand??? I have simulated small value Rs in 1 source to see what happens. The offsets at the output will be lower with the lower value Rs...
Did I help anyone understand Circlotron, and of coarse New-Tron any better??? Yes it's weird. I don't even think it's one of those agreed upon Class A or AB types of classes...

flg · 26th May 2006, 02:40 PM

Grey, I have a question regarding something I've seen but I just wondered how your prototype would react... What happens when the load increases. Say down to 4 or 2 ohms. Does one side off the amp just keep turning on, in a class AB sort of way???

jh6you · 26th May 2006, 03:08 PM

Quote:

Originally posted by flg

The offsets at the output will be lower with the lower value Rs...

I once again looked into Patrick's . . .
If there are those resistors, I think that the values closer to the speaker impedance would better help to reduce the DC offset . . . but . . .

JH

jh6you · 26th May 2006, 03:33 PM

I was thinking about this . . .

flg · 26th May 2006, 05:15 PM

I don't think that will work jh6you. You need a Drain load don't you. You have a battery??? If I see that right? you are trying to get a common source configuration to work? I have not tried that yet... There are also some inductor loaded follower designs that have gain???

Nelson Pass · 27th May 2006, 07:05 PM

A few thoughts:

Kirchoff's Law dictates that the circuit will not work unless there
is some DC attachment from the input circuit to the output
circuit to allow a difference in drive current to the output devices.

The 2 resistors across the load provide an arbitrary virtual
ground for this, and we see that the currents coming from the
Drains of the input devices as well as feedback currents must
flow through these resistors on their return path to the
input stage. As such, they should be as low a value as
practical, but not so low that they load down the output
undesirably. My own thinking is that 22 ohms would probably
be that lowest practical value.

The performance of this technique will then hinge on the power
supply rejection and common mode rejection of the input stage,
so that cascoding and a high quality CCS will be important.

As to AB performance of this a circuit - forget it. There is not
enough feedback to give low distortion at the crossover point.

EUVL · 27th May 2006, 07:37 PM

Nelson,

Many thanks for the analysis.

> As to AB performance of this a circuit - forget it. There is not
enough feedback to give low distortion at the crossover point.

Meaning pure Class A ?

(Just want to make sure I understand correctly language wise, by no means questioning the statement.)

Would you use a self biasing cascode (like e.g. J174 with high Idss which would give about 5V Vds across the diff pair), or at fixed voltage (e.g. ZVP3310A with gate at -7V wrt. Gnd ? What are the pros & cons ?

Patrick

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Tron Seven	reinhard	Tubes / Valves	2	30th April 2008 12:26 AM
Tron Syren	analog_sa	Tubes / Valves	2	13th April 2005 09:05 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	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

26th May 2006, 10:12 AM	#61
djk diyAudio Member Join Date: Feb 2001 Location: USA	You may find the resistors to ground may be a much higher value, 500 ohms or so. I think they only need carry the gate current for the outputs.

26th May 2006, 02:29 PM	#64
flg diyAudio Member Join Date: Jul 2005 Location: North East	Well, there was a question some posts ago and recently regarding why we need those 2 Rs to Gnd... The 2 power supplies that supply the output circuit are floating. They have no Gnd connection. Most discussions of drivers, refrences some drive voltage to ground. Drive voltage refrenced to Gnd will do nothing to the output. The drive needs to be Gate to source. Hence, the Rs allow a gate to source voltage return path while still alowing the output supplies to float. My circuit has no feedback and it will not work without the Rs. Transformers were the original method to drive this thing. They supply an isolated gate to source drive. And then ther is me. I'm the one who wants to use some parameters that are most important to make the Rs in my circuit optimum. I've just been putting it off for bench evaluation. I beleive they will be most important on a real circuit with imbalances and offsets, etc. All the perfection of the simulator is not pointing out anything particular to help me understand??? I have simulated small value Rs in 1 source to see what happens. The offsets at the output will be lower with the lower value Rs... Did I help anyone understand Circlotron, and of coarse New-Tron any better??? Yes it's weird. I don't even think it's one of those agreed upon Class A or AB types of classes...

26th May 2006, 02:40 PM	#65
flg diyAudio Member Join Date: Jul 2005 Location: North East	Grey, I have a question regarding something I've seen but I just wondered how your prototype would react... What happens when the load increases. Say down to 4 or 2 ohms. Does one side off the amp just keep turning on, in a class AB sort of way???

26th May 2006, 05:15 PM	#68
flg diyAudio Member Join Date: Jul 2005 Location: North East	I don't think that will work jh6you. You need a Drain load don't you. You have a battery??? If I see that right? you are trying to get a common source configuration to work? I have not tried that yet... There are also some inductor loaded follower designs that have gain???

27th May 2006, 07:05 PM	#69
Nelson Pass The one and only Join Date: Mar 2001	A few thoughts: Kirchoff's Law dictates that the circuit will not work unless there is some DC attachment from the input circuit to the output circuit to allow a difference in drive current to the output devices. The 2 resistors across the load provide an arbitrary virtual ground for this, and we see that the currents coming from the Drains of the input devices as well as feedback currents must flow through these resistors on their return path to the input stage. As such, they should be as low a value as practical, but not so low that they load down the output undesirably. My own thinking is that 22 ohms would probably be that lowest practical value. The performance of this technique will then hinge on the power supply rejection and common mode rejection of the input stage, so that cascoding and a high quality CCS will be important. As to AB performance of this a circuit - forget it. There is not enough feedback to give low distortion at the crossover point.

27th May 2006, 07:37 PM	#70
EUVL diyAudio Member Join Date: Oct 2003	Nelson, Many thanks for the analysis. > As to AB performance of this a circuit - forget it. There is not enough feedback to give low distortion at the crossover point. Meaning pure Class A ? (Just want to make sure I understand correctly language wise, by no means questioning the statement.) Would you use a self biasing cascode (like e.g. J174 with high Idss which would give about 5V Vds across the diff pair), or at fixed voltage (e.g. ZVP3310A with gate at -7V wrt. Gnd ? What are the pros & cons ? Patrick