Low TIM, low distortion hybrid front-end

vzaichenko · 2014-06-22 6:52 am

still4given said:
Hi Valery,

I rarely do 2oz copper. Is there some unusually high current in this design that requires it? Maybe the traces could be widened instead. Looks like there is plenty of room. The cost is almost double for 2oz and since these are usually just for testing and learning I try to keep the cost down when possible.

Let me know.

Thanks, Terry

Terry, please order the way you normally do - I mentioned 2 oz as with my manufacturer it does not really cost any extra. No high currents here, so no problem to go with the thinnest option available.

Cheers,
Valery

MiiB · 2014-06-22 8:51 am

Now that you have this fine design, would it not be possible to use a voltage doubler to increase voltage for the tube LTP. As the current draw is low it should be possible to get good enough smoothing with a few R-C blocks. I like tubes, but the benefit is really biggest if the voltage is high.

thimios · 2014-06-22 8:57 am

MiiB said:
Now that you have this fine design, would it not be possible to use a voltage doubler to increase voltage for the tube LTP. As the current draw is low it should be possible to get good enough smoothing with a few R-C blocks. I like tubes, but the benefit is really biggest if the voltage is high.

vzaichenko · 2014-06-22 9:04 am

Actually, it's agood idea - I will work in this direction

Cheers,
Valery

vzaichenko · 2014-06-22 9:31 am

Bigun said:
I'm looking forward to some listening impressions from you guys on this amplifier. There aren't too many hybrids around to gather listening impressions from.

I remember the Nishiki hybrid amplifier by Wim de Haan which uses an ECC82 with 'singleton' input, buffered by cathode follower to drive a quasi-complementary SS output stage with no global feedback. However, it was, much later, modified for higher power and used an LTP input stage. But no global feedback, the inverting input of the LTP was grounded (I guess it could be used for a balanced input signal). Given that the Nishiki was said to be a good sounding amplifier it might be interesting to include some option with the amplifier of this thread - to have no global feedback. Perhaps a link to switch it from global to local feedback only ?

Yeah, it will be interesting to hear it )

No GNFB - first two stages are fine, but VAS will require some redesign - this one is too high-gain one, I will not degenerate it to appropriate gain levels...

By the way, 12AU7 is an analog of ECC82

Bigun · 2014-06-22 2:25 pm

Reducing the gain maybe something to try since with high gain you have very little signal swing at the anodes of the input stage so that the greater linearity of the tubes c.f. SS is somewhat moot. With the tubes you can afford more swing and perhaps fewer devices (eliminate Q5 and Q6) for overall for less phase shift ?

vzaichenko · 2014-06-22 3:12 pm

Bigun said:
Reducing the gain maybe something to try since with high gain you have very little signal swing at the anodes of the input stage so that the greater linearity of the tubes c.f. SS is somewhat moot. With the tubes you can afford more swing and perhaps fewer devices (eliminate Q5 and Q6) for overall for less phase shift ?

Also a good point to think about.

I will bring this one to life and later on will try to squeeze maximum out of it - increase the tube voltage and try to use its benefits as much as possible. I think it's a good direction for more research - a promising one

still4given · 2014-06-22 4:19 pm

Do I see some changes coming? Should I wait a little longer before ordering boards? We did the CFA-XH boards and before the solder mask was dry we were cutting traces. If you think you are going to make changes to the layout just let me know. I can wait until it is more finalized.

Thanks, Terry

vzaichenko · 2014-06-22 4:22 pm

still4given said:
Do I see some changes coming? Should I wait a little longer before ordering boards? We did the CFA-XH boards and before the solder mask was dry we were cutting traces. If you think you are going to make changes to the layout just let me know. I can wait until it is more finalized.

Thanks, Terry

No, actually it was a bit of a discussion regarding some future development, something for "Tube hybrid 2.0". This current one is final - I don't plan to have any changes here.

thimios · 2014-06-23 12:50 pm

vzaichenko said:
Sorry - made some mistakes. Bl**dy copy/paste It's pretty late here, so...
4 resistors with wrong values, but most important - one compensation cap missing.

Here is a complete new package.
I have carefully checked everything - now it's fine.

Please use this one. My apologies for possible inconvenience

Valery

Hi Valery,please post these PDF in the correct size .

Regards.
Thimios.

vzaichenko · 2014-06-23 7:17 pm

thimios said:
Hi Valery,please post these PDF in the correct size .
Regards.
Thimios.

Hi Thimios, yes - just noticed the previous ones were way bigger the original size. These ones must be ok. Please check.

Cheers,
Valery

thimios · 2014-06-23 9:03 pm

vzaichenko said:
Hi Thimios, yes - just noticed the previous ones were way bigger the original size. These ones must be ok. Please check.

Cheers,
Valery

You didn't post the other side of pcb

vzaichenko · 2014-06-23 9:09 pm

thimios said:
You didn't post the other side of pcb

Agrrrr

Another stupid mistake

Forgot to switch off the parts layer for the first one. Here it is

thimios · 2014-06-23 9:16 pm

Ok now all are posted ,i will go for printing tomorrow again and see what happens

vzaichenko · 2014-06-23 9:21 pm

thimios said:
Ok now all are posted ,i will go for printing tomorrow again and see what happens

Good luck

I have ordered 2 PCBs from the factory - hope to get them by the end of the week :snail:

still4given · 2014-06-24 12:10 am

I sent for a quote from the board factory for this and the Class A you posted in the DX thread. I should hear back sometime tonight. They are all sleeping now in China,

Maybe I should wait to order until one of you prove them.

Bigun · 2014-06-24 2:32 am

vzaichenko said:
- I don't plan to have any changes here.

Sorry, I didn't mean to imply changes were needed. The suggestion was only to incorporate some jumpers in the pcb layout that would allow you to do some experiments

padamiecki · 2014-06-24 8:11 am

vzaichenko said:
Hi All,

Here is my new "baby". Front-end with a vacuum tube IPS.

It is based on one of my previous (prototyped and tested) designs, with certain improvements implemented. Simulation shows very high parameters, promising a good sound.

Anybody interested in further development?

Cheers,
Valery

Hi,

nice try.

I would not like to discourage you, but after swiitching the amp on Q6 and Q10 will see the rail (because of cold heaters) what could destroy them (now I can see the trick with zener regulator)

anyway I like it even if I "see some troubles"

If memory distortion exists Memory Distortion Philosophies - Part 8 : More tests using the tube would almost eliminate them from ips.
Cascoding is another way to eliminate them. Mentioned Q6, 10 are not cascoded but vas stage is another "memdist maker"

light offtopic: this was my way to study them: http://www.diyaudio.com/forums/soli...power-amp-jc-tmc-help-set-up.html#post3771944

bth what kind of ops are you using?

vzaichenko · 2014-06-24 8:52 am

padamiecki said:
Hi,

nice try.

I would not like to discourage you, but after swiitching the amp on Q6 and Q10 will see the rail (because of cold heaters) what could destroy them (now I can see the trick with zener regulator)

anyway I like it even if I "see some troubles"

If memory distortion exists Memory Distortion Philosophies - Part 8 : More tests using the tube would almost eliminate them from ips.
Cascoding is another way to eliminate them. Mentioned Q6, 10 are not cascoded but vas stage is another "memdist maker"
light offtopic: this was my way to study them: http://www.diyaudio.com/forums/soli...power-amp-jc-tmc-help-set-up.html#post3771944

bth what kind of ops are you using?

Hi Pawel,

I have addressed the start-up issue in post #33. Anyway, in the worst case, I'm ready to hit the rail for some time - my intelligent soft-start/protection system will connect the speakers only if it sees "0" at the output. But thank you for pointing it out.

Nice study by the way (memdist power amp). Especially IPS topology is impressive :up:

I use a simple EF3 OPS (number of OP pairs may vary) - see attached.
It is also perfectly fit to Ostripper's Slewmonster OPS, see here:

Ostripper's Slewmonster

Cheers,
Valery

thimios · 2014-06-24 8:20 pm

the first step

Slow progress, photoresist pcb board is old and the result isn't so good

I believe that is necessary Valery uses some bigger width lines and pads,friendly to diy

Low TIM, low distortion hybrid front-end

R.I.P.

Member

Member

R.I.P.

R.I.P.

Member

R.I.P.

Member

R.I.P.

Member

R.I.P.

Attachments

Member

R.I.P.

Attachments

Member

R.I.P.

Member

Member

Member

R.I.P.

Attachments

Member

Attachments