I have a few other things im going to try.
1. Im going to try different output transistors. MJL4281A (NPN)
MJL4302A (PNP) they have a higher FT and higher beta.
2. Different base stopper resistors on the output stage. RSMF1JT2R20CT-ND
3. Different emitter resistors on the output stage A137272CT-ND
Testing one thing at a time of course.
But other than that.
Things maybe as good as its going to get due to the limitations of this design and possibly the pcb layout which in my opinion is not ideal.
It was originally designed to be a true diy pcb. Single sided with jumpers.
Ideally a new optimised double sized layout would probably make some improvements especially the grounding and signal ground interaction.
Also the current mirror could be improved and the 2EF could be upgraded to a 3EF so less load is placed on the VAS at high current conditions when the current low power output transistors start to suffer from beta droop.
1. Im going to try different output transistors. MJL4281A (NPN)
MJL4302A (PNP) they have a higher FT and higher beta.
2. Different base stopper resistors on the output stage. RSMF1JT2R20CT-ND
3. Different emitter resistors on the output stage A137272CT-ND
Testing one thing at a time of course.
But other than that.
Things maybe as good as its going to get due to the limitations of this design and possibly the pcb layout which in my opinion is not ideal.
It was originally designed to be a true diy pcb. Single sided with jumpers.
Ideally a new optimised double sized layout would probably make some improvements especially the grounding and signal ground interaction.
Also the current mirror could be improved and the 2EF could be upgraded to a 3EF so less load is placed on the VAS at high current conditions when the current low power output transistors start to suffer from beta droop.
What limits this design? A high beta EF2 is quite adequate for 8R. This design was discussed to the limit.
The PCB IS ideal for what this amp was intended for. Grounding is star based, signal also goes to that star.
Yes , you could make improvements .... maybe 5db hum/noise or a few
PPM THD - so what ?
As jwilhelm mentions , there is the slewmaster. But ... one would find it hard to distinguish a badger from a wolverine.
Compared to this (below) - a "natural sound" Yamaha , the Badger is quite advanced. Yamaha uses the -9V aux and a resistor as the current source (ha).
They based 15 years of receivers and amps on this simple design. Sure sounds nice !! You would not believe the jumpers on the yamaha , they
even run the right channel LTP all the way across a 9" board (across the left channel circuit).
PS - A optimized design WAS discussed , but DIYA wanted this PCB to be compatable with the heatsinks and cases in the store.
OS
Attachments
Hi OS,
Thanks for your comments and welcome back to the forum.
In no means was my comments meant to be critical of the HB design.
I am very happy with it.
I guess as I own a QA-401 and enjoying tinkering I wanted to investigate making some minor adjustments to this amplifier.
One thing is clear from my testing. Is has very low distortion at 1K but the distortion increase as the frequency goes up due to reduction in loop gain and other factors. As a designer I'm sure you know all about this.
I am seeing about 0.00245% THD at 1khz and 0.22% at 16khz
So I guess that I was enjoying learning about amplifier and doing some testing along the way.
Here is a link to my build album if your interested in taking a look.
First Amplifier build - Google Photos
There you will see many plots and also a loop gain measuring setup that I made to measure the phase and gain margins.
historically the tripple darlington by Lochanti, and the series connected output devices by Leach, Meyer, Bonjiorno etc, were brought about by the lack of highish hFE at the time, and the lack of high Vceo output devices in the mid 70's...
today, we are not limited to such choices present in those times....
if i were to upgrade this circuit, i will use jfets as input differentials, mosfet VAS and mosfet big CCS or even depletion mosfets for small and big ccs....
so then we will be moving towards Nelson Pass......
today, we are not limited to such choices present in those times....
if i were to upgrade this circuit, i will use jfets as input differentials, mosfet VAS and mosfet big CCS or even depletion mosfets for small and big ccs....
so then we will be moving towards Nelson Pass......
Of course, the design is just fine. I was just saying that if anyone wanted to try and lower the distortion even more than what I have shown can be achieved.
One of the things holding them back would be the pcb. It was not meant to be critical it was just stating a fact.
I am going to continue to test and tune and will report back once I have some more news.
One of the things holding them back would be the pcb. It was not meant to be critical it was just stating a fact.
I am going to continue to test and tune and will report back once I have some more news.
... but with larger holes for resistors
a reasonable wish....
i am fine with large solder pad holes, but i wished the traces for low current signals be made thinner.....
the honeybadger is just about the best example of a single ended input stage and VAS implementation and is quite different to the "push pull design all the way" by Leach and others, i have the Leach Super amps high up there in my diy amps, and having two honey badgers built, i can assure builders that they will surely be amazed no end...
HF distortion depends on a number of factors including cabling and enclosure layout but the best first step is to add regulated PS for LTP and VAS 3-5V higher than power rails voltage.
cheers,
PS you may also increase R14 from 2k2 to 10k and reduce mirror resistors a little and see what happens. The first mod gives a tiny reduction in HF in simulation. Of course removing anti-clipping diode will substantially reduce HF distortion but at a known price.
I believe that Janusz is referring to the
Diode D10.
Which is the Baker Clamp option.
I do not have this installed
It seems Badger 2.0 with an EF3 is due.
I have been playing around with my parts and simulations , I can't believe
I once was a believer in EF2's.
I have (have not) noticed much difference in audio quality between the 2 output stages. I repair amps , I have seen and listened to dozens of them
on my polk's recently. The triples seem to dominate in the "bass world".
Just about all the higher class AB sub amps are Ef3 or class D. The 250w+
Dayton plate amps are EF3.
I like the fact of that you can run 3mA VAS's with the EF3. Use the NJW
TO-3p's and parallel 6+ pairs of outputs with no beta droop.
PS - I would of originally designed the Badger with a EF3 , but I was scared
of injecting instability into the forum.
Badger 2.0 would be a work of art . PS onboard , just hook up the trafo ...
plenty of empty space on the present Badger PCB. Also , IPS cap multipliers,
DC servo and basic DC protection.
OS
I have been playing around with my parts and simulations , I can't believe
I once was a believer in EF2's.
I have (have not) noticed much difference in audio quality between the 2 output stages. I repair amps , I have seen and listened to dozens of them
on my polk's recently. The triples seem to dominate in the "bass world".
Just about all the higher class AB sub amps are Ef3 or class D. The 250w+
Dayton plate amps are EF3.
I like the fact of that you can run 3mA VAS's
with the EF3. Use the NJWTO-3p's and parallel 6+ pairs of outputs with no beta droop.
PS - I would of originally designed the Badger with a EF3 , but I was scared
of injecting instability into the forum.
Badger 2.0 would be a work of art . PS onboard , just hook up the trafo ...
plenty of empty space on the present Badger PCB. Also , IPS cap multipliers,
DC servo and basic DC protection.
OS