Introducing Tavish Design

[IwanS]

Quite respectable for a hybrid. Solid state amps achieve much lower distortion these days, with much higher feedback. But practically, ~0.1% THD at full power and 20 kHz is clean. I still think you should write an article or white paper on this amp.

Is there a typo in your spec for damping factor? Output Z of 40 mohm -> damping factor of 200 (not 250), is that correct?

Why do we hear from you and not your son?

Iwan

 

[Quip]

We're back from vacation. In response to Quip's request, my son measured the Minotaur distortion levels vs power (from 0.1W up to 140W into 8 ohms), at four different frequencies (25 Hz up to 20 kHz). He also captured some distortion spectra using the Picoscope 4262. I've updated the Minotaur owner's manual, adding an appendix to include these plots.

Minotaur Direct-Coupled Hybrid Integrated Amplifier – Tavish Design

For a hybrid amp with a predominantly vacuum tube signal path and only 24 dB of negative feedback, I think the results are pretty good.

Thanks for asking.
Scott

Wow! Thanks for delivering. Yes, I agree with Iwan. The Minotaur seems well-designed and robust. Are you guys considering international distribution?

 

[TavishDad]

Wow! Thanks for delivering. Yes, I agree with Iwan. The Minotaur seems well-designed and robust. Are you guys considering international distribution?

Hi Iwan, Hi Quip,
Thanks again for the very encouraging positive feedback.

Ok, I’ll plan to write an article on the Minotaur. I certainly don’t consider the design to be secret or anything. But it’s a complicated design (with 7 circuit boards, microcontroller software, etc) and I’m not sure how many DIYers would build it. It takes me and my son 2 weeks to build one, even with all the design work finished. As I mentioned earlier in this thread, I think it might be better for me to design a simpler direct-coupled hybrid that I could offer in kit form.

The damping factor should indeed be 200, not 250. The 40 milliohm output Z is correct. That was some fuzzy thinking on my part. You have a sharp eye! I’ll fix it on my next website update.

I’ve gotten a number of inquiries on international sales, so I’ll look into it. It seems that hybrid amps like this are popular in Europe, based on our emails and website traffic.

The Minotaur would require a couple of custom toroidal power transformers for 240v 50 Hz operation. Getting toroids with dual 120/240v 50/60 Hz primaries is no problem, but getting ones that are quiet when running at 50 Hz might be. Do you have any recommendations on this?

Finally, I have not shipped completed products to Europe before (Quip, I assume you are in Europe?), only PCBs, so I’m not sure what complications may arise with import duties, EU agency approvals, etc. I’ll look into it – if you have any advice let me know.

Scott

 

[jackinnj]

I simulated the overload margin of the phono preamplifier -- comes out to approximate 34dB @1kHz which is very, very good.

edit -- shoulda looked more carefully into the datasheet -- this is almost exactly the number Scott states, but related differently.

 

[awty]

Hi Scott
Forgive my ignorance, but what is the issue with 50hz?

 

[TavishDad]

I simulated the overload margin of the phono preamplifier -- comes out to approximate 34dB @1kHz which is very, very good.

edit -- shoulda looked more carefully into the datasheet -- this is almost exactly the number Scott states, but related differently.

Thanks Jack!!! That’s a great endorsement.

I just remeasured the Classic Phono overload point: referred to the MM input, it is ~190mV RMS at 1kHz, or about 20 V RMS referred to the output, operating into a 22kΩ load, and for a 1% THD level. The way you are specifying it, 190mV RMS is about 32dB above the typical 5mV RMS input reference level.

This simple, classic circuit topology works amazingly well. I’ve attached a distortion spectrum – this is for a ~400mV RMS output at 1kHz into a 22kΩ load, and it shows 0.012% THD, almost entirely 2nd harmonic. Not too bad for a circuit with no feedback (except the cathode degeneration).

I just updated the Assembly and Setup Manual on my website, which includes some component value changes in the schematic and parts list. The changes are intended to maintain accurate RIAA equalization over the wider range of tube-to-tube variations I’ve been seeing. All kits and assembled units are now shipping with the updated circuit. I’ve offered an optional, free update for previous customers, but it is not at all required. Once you’ve got the bug, it can be difficult to stop tweaking something………

Scott

 

[TavishDad]

Hi Scott
Forgive my ignorance, but what is the issue with 50hz?

Hi Paul,

I wasn’t very clear on that. I was referring to possible mechanical noise in the toroidal power transformer. Large power transformers can sometimes “buzz”, and one cure is to use a transformer that is designed with ample flux overhead, that is, with more flux capacity in the core than it needs. In the Minotaur, I am currently using a toroidal transformer from Toroid Corp of Maryland, which operates on a 120V 60Hz line with very low mechanical noise.

But the same transformer might not be as quiet on a 50Hz line, because it would be operating with 20% less flux overhead. So I was concerned about getting a 50/60 Hz power transformer in the US for export to countries that use 50Hz, because I would not know how quietly it would operate. It is a difficult thing for me to test, because the mechanical noise also depends on how electrically noisy the power line is (how much DC offset and even-order harmonics). I would hate to buy a 240V 50Hz transformer in the US, ship it to Europe (or Aus), and then have someone tell me that it made a buzzing noise………… Unhappy customer, unhappy me.

Scott

 

[awty]

Thanks Scott.
Yes noisy transformers are a pain.
If I were to buy just the board, could I use my own transformer and use Salas sshv2 to get the required voltage, or is there another way? http://www.diyaudio.com/forums/group-buys/206033-gb-salas-sshv2-regulator.html

Usually like to separate Power supply and phonostage if I can.

 

[TavishDad]

Thanks Scott.
Yes noisy transformers are a pain.
If I were to buy just the board, could I use my own transformer and use Salas sshv2 to get the required voltage, or is there another way? http://www.diyaudio.com/forums/group-buys/206033-gb-salas-sshv2-regulator.html

Usually like to separate Power supply and phonostage if I can.

Hi Paul,

Oops. I may have still managed to be unclear. The large toroidal transformer I was referring to is the 500VA one in the Minotaur power amp. At present, the Minotaur only runs on 120V 60Hz. I worked hard to find a silent 500VA toroid, but I don't yet have a 50Hz version.

The Classic Phono Stage uses an external 18VAC wall transformer. At present, the wall transformer I supply with the complete kit is 120V 60Hz, but the board will work fine if you provide 18VAC 50Hz. Small transformers, like the ones in the phono stage, seldom have mechanical noise.

Things could get even more confusing when I introduce more products.... ;-)

Scott

 

[TavishDad]

1st Raffle Winner

We’d like your feedback on our products and website, and so we’re holding a raffle, as other vendors on this forum have done in the past. For the every 20 readers who post a question, constructive feedback, or suggestion for a future product on this thread, we’ll give away a free Classic Phono Stage PCB (including free postal shipping up to $10), up to 4 PCBs. The winner will be selected at random from the 20 posts. And we’ll give away at least 1 board per month, so if there is only one post in a month, that’s the winner (no posts = no winner…….).

Previous customers have been entered into their own “raffle” with even better odds (thanks!).

Classic Vacuum Tube Phono Stage (MM / MC) – Tavish Design

Scott

We've selected our first raffle winner, for the month 7/10 - 8/16. The winner gets a free Classic Phono PCB (if they want one.....) and will be notified by PM.

Our next "drawing" will be on or around Sept 15th. To be entered, you have to post a question, constructive comment, or suggestion on this forum. Forum participants who did not win this month are automatically re-entered for next month (winners can only win once).

Scott

 

[IwanS]

Hi Iwan, Hi Quip,
Thanks again for the very encouraging positive feedback.

Ok, I’ll plan to write an article on the Minotaur. I certainly don’t consider the design to be secret or anything. But it’s a complicated design (with 7 circuit boards, microcontroller software, etc) and I’m not sure how many DIYers would build it.

Scott

I'm quite interested in building my own hybrid amp. While we are waiting for your article, could you comment on your use of bipolar transistors in the output? The majority of hybrid designs seem to use MOSFETs in the output (in fact, the majority of present solid-state amps as well).

Thanks, Iwan

 

[awty]

We've selected our first raffle winner, for the month 7/10 - 8/16. The winner gets a free Classic Phono PCB (if they want one.....) and will be notified by PM.

Our next "drawing" will be on or around Sept 15th. To be entered, you have to post a question, constructive comment, or suggestion on this forum. Forum participants who did not win this month are automatically re-entered for next month (winners can only win once).

Scott

Yeah I win!

Thanks Scott, look forwards to building this phono.

 

[TavishDad]

I'm quite interested in building my own hybrid amp. While we are waiting for your article, could you comment on your use of bipolar transistors in the output? The majority of hybrid designs seem to use MOSFETs in the output (in fact, the majority of present solid-state amps as well).

Thanks, Iwan

Hi Iwan,
Sorry for the slow response. I chose bipolar transistors over MOSFETs primarily based on linearity. A bipolar emitter follower tends to be more linear than a MOSFET source follower.

I prototyped several different output stages and measured their standalone distortion performance, and the best results were for the Sziklai configuration (or compound feedback pair) using BJTs. Douglas Self discusses this is his book Audio Power Amplifier Design Handbook – I have the 2nd edition but I assume the latest editions are similar.

I’m not sure why BJTs have become less popular. Their only real disadvantage seems to be the need for safe operating area (SOA) protection. The Minotaur includes dual-slope, frequency-dependent SOA protection that stays completely out of the way during normal operation.

Best,
Scott

 

[TavishDad]

RIAA Curve

Hi everyone,

I was asked by email to clarify which "version" of the RIAA standard we're implementing. It's a good question.

Tavish Design phono stages are designed to implement the traditional RIAA replay curve. We do not implement the 1976 IEC Amendment which rolls off the replay response at 6dB/octave below 20Hz, and we do not implement the occasionally discussed “Neumann pole” at 50 kHz.

We calibrate our phono stages against an inverse RIAA network as described in Lipshitz and Jung, and which I just posted on our "downloads" page. I have verified our network against the theoretical RIAA response to within ±0.08 dB. There is an additional measurement tolerance of ±0.05 dB when measuring the phono stage against the inverse RIAA network, so our RIAA measurements can be considered accurate to within approximately ±0.15 dB.

I discuss my thinking on these versions in more detail on our website. If I'm wrong, I'm certainly willing to be convinced...... ;-)


Downloads – Tavish Design


Scott

 

[awty]

Any suggestion for tubes?
I have a couple of cheap Chinese ones but would rather something nicer.

 

[TavishDad]

Any suggestion for tubes?
I have a couple of cheap Chinese ones but would rather something nicer.

Hi Paul,

I use JJ tubes in the kit and assembled versions, but that is not essential. I do recommend using a JJ 5751 in the first stage, if you can, since the first tube determines the overall noise of the phono preamplifier. The JJ 5751 is among the lowest noise tubes I've measured.

Scott

 

[awty]

Thanks Scott.
I can get a local PCB potted transformer that is 10va with 2 x 120vlt primaries and 15+15vlt secondaries will this be ok?
Also the closest wall plug I can find is 17VAC at1.25amps will that work?

 

[TavishDad]

Thanks Scott.
I can get a local PCB potted transformer that is 10va with 2 x 120vlt primaries and 15+15vlt secondaries will this be ok?
Also the closest wall plug I can find is 17VAC at1.25amps will that work?

Hi Paul,
Yes, I think both will be OK, but it’s difficult to exactly predict the behavior of transformers other than the ones I specified in the parts list. The most important thing is to make sure that neither of the regulated supplies (heater or high voltage) drop out.

For the 18.6v regulated heater supply, you are looking for an unregulated input to the LM1086-ADJ in the range of 20.5 – 22.5 VDC. In my case, the nominal 18 VAC @ 1A wall transformer actually delivers about 18.6 VAC from a 118 VAC 60Hz line, which results in a raw supply of 22.1 VDC with 700 mV peak-to-peak ripple. If instead I apply 18 VAC at 50Hz (using a power amplifier to deliver the 50Hz sine wave), I get 21.2 VDC, and if I apply 17.5 VAC at 50Hz, I get 20.6 VDC. Still OK.

For the 200v regulated high voltage supply, you are looking for an unregulated input to the HV regulator of 210 – 250 VDC. Using the wall transformer to supply 18.6 VAC at 60Hz, I get a 232 VDC raw HV supply. If apply 18VAC at 50Hz, I get 222 VDC, and if I apply 17.5 VAC at 50Hz, I get 215 VDC. All OK.

That’s a long answer, but the short answer is that your proposed transformers should be fine. Just be sure to check the actual unregulated supply voltages you get. And double check that your PCB mounted toroid has the same pin configuration as the one specified, and that it really is a toroid (a conventional E-I core transformer may have much higher magnetic fields.

Have fun,
Scott

 

[awty]

Hi Scott
Board turned up today all in one piece.

Even fits inside the box I started making in anticipation.

Have a bag full of bits from Mouser and some valves on the way.

Will get started some time soon and Im sure there will be a few more dumb questions for you.

Thanks
Paul

 

[IwanS]

Hi Iwan,
Sorry for the slow response. I chose bipolar transistors over MOSFETs primarily based on linearity. A bipolar emitter follower tends to be more linear than a MOSFET source follower.

I prototyped several different output stages and measured their standalone distortion performance, and the best results were for the Sziklai configuration (or compound feedback pair) using BJTs. Douglas Self discusses this is his book Audio Power Amplifier Design Handbook – I have the 2nd edition but I assume the latest editions are similar.

I’m not sure why BJTs have become less popular. Their only real disadvantage seems to be the need for safe operating area (SOA) protection. The Minotaur includes dual-slope, frequency-dependent SOA protection that stays completely out of the way during normal operation.

Best,
Scott

Thanks Scott. I'm still thinking about circuits for hybrid amps, and I hope you don't mind if I ask a few more questions about yours.

Is it really direct-coupled - if a DC voltage is applied to the input, is it transferred to the output? I think I see some Jensen audio transformers in your photo.

If so, how is the DC offset at the output controlled? Is there an offset correction servo? How stable is the output offset?

There are many ICs and transistors on the PCB. What parts of your circuit are tube, and what parts transistors?

Thanks,
Iwan