The Menace

[n_maher]

I like that amp.

I think you'd like it more now.

Overall

Detail of new wiring

Top side trafo enclosure

[cetoole]

Sweet!

[n_maher]

And the listening report is in. Hum gone, bypass cap cured 99% of the noise floor issues. I tried the amp with my Ety ER4Ps, despite having way too much gain it'd be just fine from a noise standpoint. I've had noisier portable amps.

[The Monkey]

Nate, it looks awesome. What's the gain?

[n_maher]

Nate, it looks awesome. What's the gain?

I think Ari figured around 3x but I'd have to defer to him. The pot starts ~7 o'clock and with Grados I don't get above 9 even when listening on the louder end of my comfort zone. The other IEM compliance issue is that the PEC pot has the ususal first few degrees of channel imbalance like the RK27. I was definitely able to find a comfortable listening level with good channel balance with the ER4Ps but there wasn't much in the way of adjustment. One of these days I need to see what the output V of my DAC is for comparison purposes.

[mwl168]

Resurrecting an old thread because I've always wanted to build an OTL dynamic headphone amp and my Google search brought me full circle back to HC.

I have a couple questions reading the amp schematic on page 14:

1. The output capacitors - it shows two 100uf/370vac caps in parallel. Is there a reason other than parts availability/cost to not use a single 200uf caps? Also, what AC potential swing will these output caps see?

2. Any particular reason for the way C4  and C5 are located in the schema? Can C4 be soldered directly on top of C5 like the way C1 and C2 are?

Thanks in advance for your help!   

Edited October 18, 2017 by mwl168

[mypasswordis]

An Egmont style amp makes a lot more sense

[JimL]

On 10/18/2017 at 12:46 PM, mwl168 said:

I have a couple questions reading the amp schematic on page 14:

1. The output capacitors - it shows two 100uf/370vac caps in parallel. Is there a reason other than parts availability/cost to not use a single 200uf caps? Also, what AC potential swing will these output caps see?

2. Any particular reason for the way C4  and C5 are located in the schema? Can C4 be soldered directly on top of C5 like the way C1 and C2 are? 

Tube OTL amps are best used with high impedance cans as their output impedance is relatively high, in the case of the Menace I estimate around 50 ohms.  If we use a Sennheiser HD600 300 ohm as an example, it is specified as 102 dB/volt, so 10 volts will give a deafening 122 dB.  

 

Looking at the pics, I think the output caps C6 and 7 are motor run caps, as is C5 (note that all of these are rated at 370VAC, whereas C4 is likely electrolytic as it is rated in DC volts).

[n_maher]

6 hours ago, JimL said:

Looking at the pics, I think the output caps C6 and 7 are motor run caps, as is C5 (note that all of these are rated at 370VAC, whereas C4 is likely electrolytic as it is rated in DC volts).

Correct, the output coupling capacitors were indeed motor-run caps (about the size of a can of coke) and were paralleled to address your first point, lowering the overall output impedance of the amp to give it some potential to play nicer with lower impedance headphones.

I'll try to take a look at the schematic relative to the second question raised about C4 and C5 later.

[mwl168]

Thanks Jim and Nate!

Did some reading on the motor run caps. Looks like they are often chosen for audio purpose when high capacitance and high voltage are needed in critical positions. They are actually film caps and often oil impregnated polypropylene. No wonder their size.  

[mwl168]

A few more questions:

The 6AS7G and 6080 are both recommended output tubes for this amp. They are both rated 250V for max plate voltage. Would applying the 270V B+ direct to the plates of the output tubes as shown on the schematic shorten the life expectancy of these tubes? 

What other output tubes are recommended to use for this circuit?

I have a few 150uf/350VDC rated BlackGate  caps on hand. Is it a good idea to use them for PS bypass for this amp?

 

Edited October 20, 2017 by mwl168

[dsavitsk]

27 minutes ago, mwl168 said:

The 6AS7G and 6080 are both recommended output tubes for this amp. They are both rated 250V for max plate voltage. Would applying the 270V B+ direct to the plates of the output tubes as shown on the schematic shorten the life expectancy of these tubes?

250V refers to plate to cathode voltage. Because the tube is biased up by the preceding stage, the voltage across the tube is considerably lower than 250V.

28 minutes ago, mwl168 said:

I have a few 150uf/350VDC rated BlackGate  caps on hand. Is it a good idea to use them for PS bypass for this amp?

Even if Blackgates were better than other caps, and even if they are still good, a small regulator using about $4 in parts will sound better. On that note, there is a lot that can be done to improve the circuit with minimal effort/expense.

[mwl168]

Thanks! That makes more sense (re: the plate to cathode voltage).

Maybe I should sell my NOS BlackGate over at eBay or the other site. I've seen them advertised for silly money...

Can you elaborate on what can be done to improve the circuit?

 

[mypasswordis]

37 minutes ago, dsavitsk said:

On that note, there is a lot that can be done to improve the circuit with minimal effort/expense.

This. I know what I would do, but interested in your approach as well. 

[dsavitsk]

In addition to a regulator for the PS, a CCS plate load for the first stage will decrease distortion, a CCS cathode load for the second stage will also decrease distortion, and using a MOSFET instead of the awful 6080 will decrease output impedance, lower noise, and decrease distortion. It will also allow you to eliminate the 6080 heater power supply which is the most likely source of noise, and the source of tons of heat that needs to be dealt with. A first stage biasing scheme that does not involve an RC filter will also do good things.

While we're at it, a 6SN7 gives you too much gain for nearly every headphone out there. That means that you are attenuating and then amplifying, which is a source of both noise and distortion. There are a number of ways to lower gain - the easiest might be simply eliminating the cathode bypass cap (this also increases plate impedance, so whether it can drive HF into the next stage will be an issue, but it is probably fine, but it has the benefit of getting rid of that RC filter). You can also find a lower mu tube, or perhaps a pentode with a low impedance plate load to lower gain. Or you could use a step up transformer on the input.

I've now replaced every part of your amp

 

[mypasswordis]

Hey hey, come on now, the main constraint of this exercise was OTL  give the man what he wants, even if it is worse.

[mwl168]

Yes, I've always wanted to build an OTL amp for my dynamic headphones and that's what lead me to this Menace thread. The Menace's simplistic circuit is also very attractive to me.

With that said, I do like the idea of the CCS and I have experienced what good CCS can do to with JimL's SRX Plus. My plan is to build a simple Menace OTL as close to Nate's published schematic as possible. Once that's working properly I will be able to experiment and add on more sophistications. That, to me, is a big part of the DIY fun! 

[dsavitsk]

I didn't say anything about an output transformer If using a tube in the second position is a must, a higher gm tube instead of the 6080 would be my suggestion. The 6922 has about double the gm. 40 of them will get you close to the Z out of the IRF610 ... 4 of them will get you into the range where Senns seem to shine.

5 minutes ago, mwl168 said:

That, to me, is a big part of the DIY fun! 

It's a great way to learn. One interesting upside of the 6080 is that the mu is so low that if you use a small resistor plate load, the output impedance from the plate is effectively the same as from the cathode. That conceivably opens up some interesting circuit experimentation options unique to the tube.

[kevin gilmore]

6c33 output tube for the win

 

[mypasswordis]

One interesting upside of the 6080 is that the mu is so low that if you use a small resistor plate load, the output impedance from the plate is effectively the same as from the cathode. That conceivably opens up some interesting circuit experimentation options unique to the tube.

Yup, hence my original comment about an Egmont style amp with NFB. Headphone amps only need a gain of around 4 or 5, especially since most sources are balanced these days. 

Edited October 20, 2017 by mypasswordis

[mwl168]

16 minutes ago, kevin gilmore said:

6c33 output tube for the win

 

I actually have 8 or so 6C33C in my house that I bought like 10 years ago, not yet find an application for them. These things are big and requires  ~ 7A filament current. I read getting good tube sockets for these tubes are challenging, most sockets simply deteriorates quickly due to excessive heat.

The Menace's simplistic circuit also lands itself well for a point-to-point construction which also makes circuit experimentation easier.

Edited October 20, 2017 by mwl168

[Emooze]

On 10/20/2017 at 1:25 PM, dsavitsk said:

...a small regulator using about $4 in parts will sound better.

What regulator are you referring to? I'm not having much success right now finding something suitable. I am attempting to modify the transformer output version to use 45 DHTs and so far its a few pages of hastily scribbled schematics.

[Skooby]

I've copied BH zener regulated PS to use for a DTH 45 (ignore the B+ delay/FDA217, I got an unpleasant thump).

 

 

Edited October 22, 2017 by Skooby

[jose]

6c33 output tube for the win

 

 

 

[Emooze]

On 10/22/2017 at 1:51 PM, Skooby said:

I've copied BH zener regulated PS to use for a DTH 45 (ignore the B+ delay/FDA217, I got an unpleasant thump).

Hm hm hm hm, I will have to fool around with this. Many thanks!