I just wanted to update everyone with some tests that I have been working on for a long time now. Trying to test this amplifier at the performance level it is capable of is no small task, both in time and equipment. The Wolverine team spent a lot of effort on not just making 1k low distortion but also 20k, issues abound though when actually trying to measure 20k and harmonics cleanly and accurately.
To be able to actually test at this level I had to purchase one of Viktor's oscillators tuned to 20kHz, build a suitable low noise and ripple free power supply for it, and then figure out the best way to actually measure the damn thing. I have a Cosmos APU that has notches at 1k and 10k, so I decided to modify the 10k section to turn it into a 20k notch, swapped out the needed resistors and capacitors and I was in business. It was calibrated to the output frequency of the oscillator which works out to 20,040Hz once at temperature.
Using REW I was able to create a correction curve for the filter, and using the D10S with REW's signal generator, which allows you to inject a known amount of distortion into the signal, I was able to verify that the calibration curve was correct and accurate to +/-0.5dB.
First run at it turned out well, but there was power supply/mains noise that had to be cleaned up. Stuart came up with the idea of powering the voltage regulators with lithium batteries, which did the trick!
Attached are two FFT's that were taken today @ 20kHz, one of them is full bandwidth 192kHz and the other is limited to a more reasonable 91kHz. These were taken with a Wolverine EF3-4 using Sanken 2SC3264 and 2SA1295 outputs on +/- 71 volt rails. The measurements were taken at 25Vrms into 8ohms and the amplifier was at stable operating temperature.
I am not sure of many amps that measure better then -120dB @ 20k... If you know of any please let me know!
To be able to actually test at this level I had to purchase one of Viktor's oscillators tuned to 20kHz, build a suitable low noise and ripple free power supply for it, and then figure out the best way to actually measure the damn thing. I have a Cosmos APU that has notches at 1k and 10k, so I decided to modify the 10k section to turn it into a 20k notch, swapped out the needed resistors and capacitors and I was in business. It was calibrated to the output frequency of the oscillator which works out to 20,040Hz once at temperature.
Using REW I was able to create a correction curve for the filter, and using the D10S with REW's signal generator, which allows you to inject a known amount of distortion into the signal, I was able to verify that the calibration curve was correct and accurate to +/-0.5dB.
First run at it turned out well, but there was power supply/mains noise that had to be cleaned up. Stuart came up with the idea of powering the voltage regulators with lithium batteries, which did the trick!
Attached are two FFT's that were taken today @ 20kHz, one of them is full bandwidth 192kHz and the other is limited to a more reasonable 91kHz. These were taken with a Wolverine EF3-4 using Sanken 2SC3264 and 2SA1295 outputs on +/- 71 volt rails. The measurements were taken at 25Vrms into 8ohms and the amplifier was at stable operating temperature.
I am not sure of many amps that measure better then -120dB @ 20k... If you know of any please let me know!
The time and attention that @fireanimal has put into actual being able to measure this amplifier is something that we should all be very grateful for. I would hate to put a number to the amount of hours he has spent researching, testing, setting up and refining his measurement process. Prior to these measurements all we have really been doing is measuring the measuring equipment and not the Wolverine it's, because as it turns out the Wolverine is actually better than the most measuring equipment.
We need an oscillator with a loopback distortion of -140db or 0.00001% at 20k to be able to confidently know that we are now measuring the Wolverine.
Looking at the final numbers -120.4db is 0.000095% THD tested at 20khz and driving 80 watts into a 8 ohm load.
A truly Amazing result and inline with what the simulator predicted. That's one of the reasons we kept refining our measurement technique.
Well done Andy. 🙏🙏🙏
We need an oscillator with a loopback distortion of -140db or 0.00001% at 20k to be able to confidently know that we are now measuring the Wolverine.
Looking at the final numbers -120.4db is 0.000095% THD tested at 20khz and driving 80 watts into a 8 ohm load.
A truly Amazing result and inline with what the simulator predicted. That's one of the reasons we kept refining our measurement technique.
Well done Andy. 🙏🙏🙏
My musician wife has little appreciation of the engineering finesse and effort put into Wolverine BUT she does think the circuit diagrams, the pcb artworks, the finished board photos and now the noise spectrum analysis charts should all be sold as framed "modern" artworks.
You may want to read the development thread in full before you make remarks like that...all I'm saying 🤔
Speaking of this amp's history...
The introduction in the build guide is a great starting point for understanding the contextual background of this amp and the motivations of everyone who pitched in here (awesome work to achieve these kinds of numbers!)
I was wondering if there were any other blurbs about how this project came to be. I caught ostripper's first posts in the discussion thread, but I guess I'm looking for more of the "marketing hype." Ironic, in this world where marketing is shoved down our throats, right? lol
Maybe these and more are answered in the other, discussion thread... I'll have to spend some more time flipping through that.
The introduction in the build guide is a great starting point for understanding the contextual background of this amp and the motivations of everyone who pitched in here (awesome work to achieve these kinds of numbers!)
I was wondering if there were any other blurbs about how this project came to be. I caught ostripper's first posts in the discussion thread, but I guess I'm looking for more of the "marketing hype." Ironic, in this world where marketing is shoved down our throats, right? lol
- Why was the EF3 topology employed?
- I did some searching on the topology and came across Locanthi's white paper. There's some cool history there and with EF3 in general, but I wish I had more context on why EF3 was chosen over other topologies. Or, is it as simple as "it sounds good, and EF3 > EF2 (Badger)"?
- Is the Wolverine's schtick its modular IPS design? Or, are there even more impressive feats within its design?
- Was a lot of attention brought to the distortion spectra? In particular, was the goal from the outset to have a 2nd order dominant spectrum?
- Seems like this tends to be tough with BJTs, but I'm generalizing here.
Maybe these and more are answered in the other, discussion thread... I'll have to spend some more time flipping through that.
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger.192431/post-6541345
Hi Omed,
Please see above link.
Respectfully and politely I will say that the driving force behind the development of the Wolverine was not Ostripper. He posted some initial Schematics however the development to get it to where it is now was carried out by a group of people. Posting the sort of thing you did above just gets tiring for the people who have invested alot of time and effort to get it working at this level.
I post this to avoid an argument in this thread again. When it should be about building the Wolverine amplifier.
Best regards
Daniel
Hi Omed,
Please see above link.
Respectfully and politely I will say that the driving force behind the development of the Wolverine was not Ostripper. He posted some initial Schematics however the development to get it to where it is now was carried out by a group of people. Posting the sort of thing you did above just gets tiring for the people who have invested alot of time and effort to get it working at this level.
I post this to avoid an argument in this thread again. When it should be about building the Wolverine amplifier.
Best regards
Daniel
The whole of the events that ultimately led to the creation of this amplifier is, in my opinion, too complex to summarize in a single message. For some in depth research you could read the following threads, and from there from your own opinion.
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger.192431/
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger-build-thread.211905/
https://www.diyaudio.com/community/threads/slewmaster-builds.260268/
https://www.diyaudio.com/community/threads/slewmaster-cfa-vs-vfa-rumble.248105/
https://www.diyaudio.com/community/threads/slewmaster-project-boards.257057/
https://www.diyaudio.com/community/threads/diya-store-wolverine-son-of-badger-suggestions.369758/
https://www.diyaudio.com/community/threads/diy-class-a-b-amp-the-wolverine-build-thread.385920/
I might have missed some, it's late.
EDIT: Interesting milestone. Wolverine IPS 2014-02-26
https://www.diyaudio.com/community/threads/slewmaster-cfa-vs-vfa-rumble.248105/post-3835079
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger.192431/
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger-build-thread.211905/
https://www.diyaudio.com/community/threads/slewmaster-builds.260268/
https://www.diyaudio.com/community/threads/slewmaster-cfa-vs-vfa-rumble.248105/
https://www.diyaudio.com/community/threads/slewmaster-project-boards.257057/
https://www.diyaudio.com/community/threads/diya-store-wolverine-son-of-badger-suggestions.369758/
https://www.diyaudio.com/community/threads/diy-class-a-b-amp-the-wolverine-build-thread.385920/
I might have missed some, it's late.
EDIT: Interesting milestone. Wolverine IPS 2014-02-26
https://www.diyaudio.com/community/threads/slewmaster-cfa-vs-vfa-rumble.248105/post-3835079
Last edited:
Hi @sp33ls
Thanks for your post.
The idea of the wolverine was birthed out of post that I made on the Honey Badger forum back in February 2020.
I had been working with Keantoken for about 6 - 12 months on and off trying to improve my honey badger that I had recently finished. I was ultimately trying to reduce the 20khz distortion and I was keen to learn from any information that Keantoken could provide. Finally we go to the point where it appeared that we had gone as far as we could and the limitations of the board layout and topology choices limited what we were going to be able to achieve.
I announced it to the forum members in that thread. OS who had not been seen for many many months read it and must have got a little upset by my commets but after a few days he realised it was actually the truth.
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger.192431/post-6541345
Please read post 2198 to 2220
This was when the idea of the wolverine was born. There was a few posts in the honey badger thread about it before OS started a new thread.
I can't speak for all the wolverine team members but I do know that right from the beginning of the Wolverine design my intention was to support the process and try and design the best amplifier possible no mater the time it took. For OS it appeared that near enough was good enough and my efforts and suggestions including those of the other forum members which included the future wolverine team members were not alway given the time or attention they derived.
When OS suddenly disappeared. Possibly from a COVID issue, myself and Jeremy wanted to keep going. So we created a chat group on WhatsApp and began to move the project in a new direction were everyone's input was evaluated and tested to see if it would improve the distortion. We call upon the knowledge of as many people and resources as we could to help our fellow DiyAudio members in the quest for the best amplifier.
This took many many months. Months of developing accurate simulation models. An accurate and valid test environment with LTspice and 100 and 100's of hours looking at datasheets and reviewing and revising the pcb. Over and over we meticulously went over very detail. Even moving traces fractions of a mm were discussed.
Our ultimate goal was to produce an amplifier with the lowest distortion possible. For those who have boards the wolverine team members initials are listed on the boards. Without the help of all those guys plus a few others who didn't want to be mentioned this would not have been possible.
So for someone to say. Thanks to one particular person these results were possible is a massive injustice to the people that actually make this happen. Please note that the wolverine design team followed through with it right to the end and we are still supporting you guys.
I wish you all the best and hope your builds are progressing well.
Thanks for your post.
The idea of the wolverine was birthed out of post that I made on the Honey Badger forum back in February 2020.
I had been working with Keantoken for about 6 - 12 months on and off trying to improve my honey badger that I had recently finished. I was ultimately trying to reduce the 20khz distortion and I was keen to learn from any information that Keantoken could provide. Finally we go to the point where it appeared that we had gone as far as we could and the limitations of the board layout and topology choices limited what we were going to be able to achieve.
I announced it to the forum members in that thread. OS who had not been seen for many many months read it and must have got a little upset by my commets but after a few days he realised it was actually the truth.
https://www.diyaudio.com/community/threads/diyab-amp-the-honey-badger.192431/post-6541345
Please read post 2198 to 2220
This was when the idea of the wolverine was born. There was a few posts in the honey badger thread about it before OS started a new thread.
I can't speak for all the wolverine team members but I do know that right from the beginning of the Wolverine design my intention was to support the process and try and design the best amplifier possible no mater the time it took. For OS it appeared that near enough was good enough and my efforts and suggestions including those of the other forum members which included the future wolverine team members were not alway given the time or attention they derived.
When OS suddenly disappeared. Possibly from a COVID issue, myself and Jeremy wanted to keep going. So we created a chat group on WhatsApp and began to move the project in a new direction were everyone's input was evaluated and tested to see if it would improve the distortion. We call upon the knowledge of as many people and resources as we could to help our fellow DiyAudio members in the quest for the best amplifier.
This took many many months. Months of developing accurate simulation models. An accurate and valid test environment with LTspice and 100 and 100's of hours looking at datasheets and reviewing and revising the pcb. Over and over we meticulously went over very detail. Even moving traces fractions of a mm were discussed.
Our ultimate goal was to produce an amplifier with the lowest distortion possible. For those who have boards the wolverine team members initials are listed on the boards. Without the help of all those guys plus a few others who didn't want to be mentioned this would not have been possible.
So for someone to say. Thanks to one particular person these results were possible is a massive injustice to the people that actually make this happen. Please note that the wolverine design team followed through with it right to the end and we are still supporting you guys.
I wish you all the best and hope your builds are progressing well.
Appreciate the detailed reply, Stuart.
This is definitely a badass story, fitting for such a badass amplifier. I'm excited to get building, thanks to the support of a few fellow DIYA folks for advising and supplying some key components (ahem @fireanimal)
The kind of stuff that makes me glad to have found this community...
Nice work, Wolverine team!
This is definitely a badass story, fitting for such a badass amplifier. I'm excited to get building, thanks to the support of a few fellow DIYA folks for advising and supplying some key components (ahem @fireanimal)
The kind of stuff that makes me glad to have found this community...
Nice work, Wolverine team!
Hi Andy, first of all many thanks for your great efforts! You did an really impressive work! I have a question for you. Given the enormous number of Wolverine boards that you stuffed and the multitude of possible component selections you tested, do you think that, looking at the measuring results you got in the last tests, these ones could be applied on the 1st group buy boards too? I'm just curious, I won't build the Wolverine again on 2nd GB boards due to time and money lack. 1st GB one I built sounds awesome as it is...
Thank you,
Gaetano.
Wow, thank you very much @fireanimal; I know how hard it is to measure stuff at -120 dB.
Btw. it is quite common to use batteries as power sources when measuring such things. Even "floating" oscilloscopes can introduce too much noise via ground loops (since they are never 100% floating).
Btw. it is quite common to use batteries as power sources when measuring such things. Even "floating" oscilloscopes can introduce too much noise via ground loops (since they are never 100% floating).