Zero Feedback Impedance Amplifiers - Page 146 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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 10th May 2011, 12:58 PM   #1451
diyAudio Member
 
apelizzo's Avatar
 
Join Date: Sep 2006
Default Matching IXTH20N50D

Quote:
Originally Posted by Susan-Parker View Post
Dear All,

I match IXTH20N50D parts using a grounded Gate and a resistor from the source to ground.

In the first instance I use 10 ohms to do a quick sift selection to get the general bins. This doesn't take too long as I am only interested in ball park readings.

I then retest using a 1 ohm part to do the matching.

Normally +24 volt supply to balance between having a representative drop and the power dissipated. I put on the power and then track the voltage drop until the readings more or less stabilise (takes 200 seconds or more per part).

To get "super" matched pairs I then would retest for closest match allowing groups to be on for up to half an hour. However to do this one would need a large quantity of parts to start with and for this I would use the four windings of one of my transformers cores to make sure that the wire resistances are as identical as possible.

N.B. Parts are clamped to a largish aluminium plate to give some thermal mass to work against (they will get overly hot otherwise).

I have at last been able to get a small allocation from the postultimate IXYS manufacturing run (they are doing a special extra one), and I am expecting delivery in August. It took some doing.

So sadly the IXTH20N50D mosfets are now for all intents and purposes obsolete

Best wishes,
Susan.
Thanks Susan for explaining how to match depletion mode mosfets.
I just received 12 pieces with the hope to find a close enough match.
When you will get yours in August I would be happy to send my mosfets to you so chances to find match pairs will probably increase, if you like to do so.

Anyhow IXIS distributor in Australia Braemac informed me that IXTH16N50D2 will be the closest alternative in production.

Last edited by apelizzo; 10th May 2011 at 01:04 PM.
  Reply With Quote
Old 10th May 2011, 01:48 PM   #1452
miklos is offline miklos  Canada
diyAudio Member
 
miklos's Avatar
 
Join Date: Jul 2004
Location: toronto
Quote:
Originally Posted by AKSA View Post
Stee,

Why the link between the collectors of the 2SC3423s..... interesting circuit.

Hugh
not a link, just a smallish dot
__________________
miklos
  Reply With Quote
Old 10th May 2011, 04:52 PM   #1453
Stee is offline Stee  Italy
diyAudio Member
 
Stee's Avatar
 
Join Date: Jul 2005
Location: milan
Default about the schematic posted before

no, never try to build
if you done, tell me
__________________
Please download my best audio idea https://www.facebook.com/DiyAudioHiEnd
  Reply With Quote
Old 10th May 2011, 07:54 PM   #1454
diyAudio Member
 
Susan_Parker's Avatar
 
Join Date: Oct 2004
Location: London, UK.
Default Closest alternative depletion mosfets

Hi,

Quote:
Originally Posted by apelizzo View Post
Thanks Susan for explaining how to match depletion mode mosfets.
I just received 12 pieces with the hope to find a close enough match.
When you will get yours in August I would be happy to send my mosfets to you so chances to find match pairs will probably increase, if you like to do so.
Yes, sure I would be happy to.

Once I have mine I am hoping there will be enough to get sufficient matched pairs to be able to offer some sets out to others. But can't really commit to any numbers yet.

Quote:
Anyhow IXIS distributor in Australia Braemac informed me that IXTH16N50D2 will be the closest alternative in production.
Unfortunately looking at the current curves between the two "closest" isn't close enough, not to do the auto biasing method using the DC resistance of the output transformer windings.

And as soon as one has to have a separate bias supply, the advantages of the depletion mode negative gate voltage turn on are less attractive since one MUST have the negative bias on first before applying the main power, and keep it up after too while the power rail discharges.

In my mind the HUGE attraction of the current parts is the whole auto bias thing, which really makes to overall circuit so minimalist. Once that is gone the whole simplicity of "one can run it of a couple of truck batteries" goes too.

Anyway, this is how things are. Maybe one day I will be able to fab my own parts, but probably not

Best wishes,
Susan.
  Reply With Quote
Old 10th May 2011, 08:56 PM   #1455
diyAudio Member
 
Join Date: Mar 2008
Quote:
Originally Posted by Susan-Parker View Post
Hi,

Anyway, this is how things are. Maybe one day I will be able to fab my own parts, but probably not

Best wishes,
Susan.
Setting up a manual fabrication line may be possible using used equipments
and depending on what processes is being used. But since it is an hazardous
process it is still going to be very expensive: Semiconductor device fabrication - Wikipedia, the free encyclopedia
  Reply With Quote
Old 11th May 2011, 01:39 PM   #1456
diyAudio Member
 
apelizzo's Avatar
 
Join Date: Sep 2006
Default Matching IXTH20N50D

I tested my 12 mosfets today to get an idea which one could be candidate for a pair.
The test I have done as follow:

Gate to ground
Drain to +12VDC
Source in series with 1R to ground

VGS measured: (ordered by voltage and not by testing sequence)

-0.505
-0.655
-0.720
-0.768
-0.801 **b
-0.813 **b
-0.824 *a
-0.825 *a
-0.838 **c
-0.852 **c
-0.888
-0.991

There is only one close match and two so so.

12 parts are probably sufficient to find 1 match pair.
  Reply With Quote
Old 11th May 2011, 02:47 PM   #1457
diyAudio Member
 
tiefbassuebertr's Avatar
 
Join Date: Jul 2005
Location: D-55629 Schwarzerden
Quote:
Originally Posted by Susan-Parker View Post
Hi,



Yes, sure I would be happy to.

Once I have mine I am hoping there will be enough to get sufficient matched pairs to be able to offer some sets out to others. But can't really commit to any numbers yet.



Unfortunately looking at the current curves between the two "closest" isn't close enough, not to do the auto biasing method using the DC resistance of the output transformer windings.

And as soon as one has to have a separate bias supply, the advantages of the depletion mode negative gate voltage turn on are less attractive since one MUST have the negative bias on first before applying the main power, and keep it up after too while the power rail discharges.

In my mind the HUGE attraction of the current parts is the whole auto bias thing, which really makes to overall circuit so minimalist. Once that is gone the whole simplicity of "one can run it of a couple of truck batteries" goes too.

Anyway, this is how things are. Maybe one day I will be able to fab my own parts, but probably not

Best wishes,
Susan.
it's always annoying, when suitable parts according to own circuit ideas are discontinued (obsolete) after a short offering time. By transistors (MOSFET and BJT) special suited for audio applications often to observe.
I had discover an old description of an old German amplifier brand: Dynacord.
here are three bjt parts in parallel/each half. Naturally there are emitter resistors. But Dynacord uses for each emitter resistor a separate winding part by their special output transformer (six separate winding coils at whole on the primary site). The model was "DEM-286" - go to by post #77 about
2N3055 inside - commercial famous amplifier models, quasi complementary power output
for more informations and several URLs.
This approach could be of interest for you, because then you can use each MOSFET for switching applications, if the value of coil DC resistance is large enough for getting auto bias character. Unfortunately the disadvantage is the more complicated output transformer.
  Reply With Quote
Old 11th May 2011, 02:49 PM   #1458
diyAudio Member
 
apelizzo's Avatar
 
Join Date: Sep 2006
Hi Susan,

do you think the pair marked as **b are acceptable and would not justify buying another set of mosfets and gambling for another match?

Thanks
  Reply With Quote
Old 17th May 2011, 07:42 PM   #1459
diyAudio Member
 
Susan_Parker's Avatar
 
Join Date: Oct 2004
Location: London, UK.
Default Depletion MOSFET pairs

Hi,

Sorry for the delay in replying, only just seen your posts.

I was at the AES UK conference all weekend (Friday to Monday) as well, so still catching up with my correspondence.

Quote:
Originally Posted by apelizzo View Post
Hi Susan,

do you think the pair marked as **b are acceptable and would not justify buying another set of mosfets and gambling for another match?

Thanks
If you are using EI lamination transformers then b and c should be okay to get going with.

There is even a line of thought that says a small imbalance is preferable as it keeps the domains "primed" however I would want to do a bit of specific testing to look at this. A lot of the tube based rule of thumb stuff is not really applicable here as the source impedance and number of turns are so much lower.

If you are using toroids then perhaps not so good, as there is no intrinsic gap unlike where the E and I laminations butt together.

When mine come I can mix and match something better for you.

As to your list, the spread is about 2:1 which is what I would expect from my own experience with a sample group.

Best wishes,
Susan.
  Reply With Quote
Old 17th May 2011, 07:51 PM   #1460
diyAudio Member
 
Susan_Parker's Avatar
 
Join Date: Oct 2004
Location: London, UK.
Default Parallel mosfets with separate windings

Hi,

Thanks for your post.

Quote:
Originally Posted by tiefbassuebertr View Post
it's always annoying, when suitable parts according to own circuit ideas are discontinued (obsolete) after a short offering time. By transistors (MOSFET and BJT) special suited for audio applications often to observe.
I had discover an old description of an old German amplifier brand: Dynacord.
here are three bjt parts in parallel/each half. Naturally there are emitter resistors. But Dynacord uses for each emitter resistor a separate winding part by their special output transformer (six separate winding coils at whole on the primary site). The model was "DEM-286" - go to by post #77 about
2N3055 inside - commercial famous amplifier models, quasi complementary power output
for more informations and several URLs.
This approach could be of interest for you, because then you can use each MOSFET for switching applications, if the value of coil DC resistance is large enough for getting auto bias character. Unfortunately the disadvantage is the more complicated output transformer.
The answer here is Litz windings, which I am experimenting with anyway to look at skin effects between 100kHz and 2MHz.

Except rather than use the Litz bundle as a single conductor, one puts one device per individual wire.

I have tried some variations in the past using pairs of devices, the only small issue is the problem than one needs to provide some resistive cross coupling otherwise the whole set takes off and oscillates!

But yes, as long as the combined gate capacitance does not get one, it should be possible.

Best wishes,
Susan.
  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



New To Site? Need Help?

All times are GMT. The time now is 09:29 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