DIY Transistor Amp Using NVA Boards

For the three and more legged things
User avatar
jack
Thermionic Monk Status
Posts: 5493
Joined: Wed Dec 29, 2010 8:58 pm
Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
Contact:

#271 Re: DIY Transistor Amp Using NVA Boards

Post by jack »

izzy wizzy wrote: Mon Jun 06, 2022 1:38 am Cable whys and wherefores are inherently problematic without the full picture. Drive Z and termination Z for starters and also signal level.

No one would use a microphone cable to drive a speaker nor the other way round. It's an extreme example to make a point.

It's interesting to note that studio stuff is 600R drive and termination mostly maximising power transfer. HiFi is mostly medium drive Z and high termination maximising voltage transfer. Studio stuff is mostly balanced, HiFi mainly unbalanced. Studio stuff may also have galvanic isolation where HiFi stuff rarely does.

Unsurprisingly maybe, that HiFi cabling is a mixed bag and studio stuff not so much. And although it does make a difference, for many the effects may only be minor and so claim "there's no difference". It's hard to weigh up "perspective" in a forum environment.

I've noticed over time, reducing drive Z and lowering termination Z removes cable effects to a place I can't really be bothered.
Totally agree. Z-matching the cable (in whatever configuration) should be important if you want (as much as is possible) to remove the cable from the equation. I haven't done any work on that however,

Studio kit is balanced for a host of reasons (as we all know), mostly to remove common mode interference and prevent ground loop issues. Galvanic isolation in order to stop performers being electrocuted, e.g. by being outside in the wet. Standard connectors too so you don't use the wrong cable in the wrong place and all cables that should work in a particular place, do.

There's an engineering agenda in studios - sound quality is excellent but practicality is important too... it's a compromise. We all know that recordings are greatly influenced by the end-to-end signal chain in place when they were made and by the studio engineers & producers' choices when mixing. I'm not a huge fan of messing with the digital signal chain, however the analogue side will have gone through a whole bunch of transducers, active & passive components before maybe hitting an ADC and entering the digital domain. Even the positioning of the source transducers and the acoustic properties of the venue are significant.

At best, we get back to the signal that was recorded (i.e. the start of our signal chain), not the performance itself.
Vivitur ingenio, caetera mortis erunt
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#272 Re: DIY Transistor Amp Using NVA Boards

Post by IslandPink »

A little light reading on skin depth while at work has been interesting.
I need to do more in finding the most relevant pieces before I post more though.

The maths of skin depth shows that there is an exponential decay of current , starting at the surface, then moving inwards towards the centre of the wire. This decay is however a complex function, so the decay is associated with a phase shift. At high frequencies it can result in the current in the centre of the conductor being out of phase with that in the surface.
So, as I suspected, any loss in dB of amplitude , going from LF to HF, will have a corresponding phase delta. I don't have enough to get the numbers yet, however I think that the number of -0.1dB for 20Kz was in comparing twin & earth wire to litz. The amplitude effect there is almost negligible, but based on what Nick, Andrew and I found doing the current-sourced filament supplies, it could be that the phase rotation is significant enough to be audible. That would make sense in terms of what I hear when comparing cables.
An article I'm reading in the context of current-carrying PCB traces, indicates that although the serious effects occur at 100's of MHz or more, they say that subtle effects are already evident at 10kHz.
"Once you find out ... the Circumstances ; then you can go out"
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#273 Re: DIY Transistor Amp Using NVA Boards

Post by IslandPink »

OK, time to backtrack.
This paper, cited by C.W. on the SECA forum , seems to shed much light on the subject.
http://tmr-audio.de/pdf/kabl_cap.pdf
It looks like the main shortcomings of plain heavy-gauge wire or equivalent AWG bunched fine-strand wire , is the inductance, and inductive reactance at high frequencies. Section 3 is the main one that looks at the wires themselves, and there's a comment that for the Levinson wire at 3m length, the skin effect losses would be -0.044dB at 20kHz, but the combiner skin effect and inductive reactance loss would be -0.43dB.
I assume that the inductive reactance will be a -6dB/oct slope, and 0.43dB would therefore represent a considerable phase rotation. Hence the loss of detail and tone in the upper 1 to 1.5 octaves.

So the main advantage of the litz construction is lower inductance.

The later sections from 4 to 7 investigate the additional effects from reactive loudspeaker loads, and amplifier output inductance and resistive Zout , showing that the problems can be compounded with the cable effects.
"Once you find out ... the Circumstances ; then you can go out"
Daniel Quinn
Old Hand
Posts: 859
Joined: Thu Jan 05, 2017 9:19 am

#274 Re: DIY Transistor Amp Using NVA Boards

Post by Daniel Quinn »

You have no means of telling if this article is true or bollocks , and if true audible to the extent you can hear them or Am I wrong ?

Reading these articles is for me akin to reading right wing theory without Marx to help me . That’s why I just listen and conclude at this time,with these parameters it did or didn’t sound better to me .
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#275 Re: DIY Transistor Amp Using NVA Boards

Post by IslandPink »

I have no idea what you're on about, re article being bollocks, Marx etc.
I do make a lot of my decisions based on listening ( nice to see John Curl saying 'Trust your ears' )
Valve power amps are fairly simple and lend themselves to this ; phonos to some extent, although sometimes measurements are necessary. Cables, yes, judged by ear.
Speakers are more trouble - ears used up to a point, but measurements when things go wrong or get confusing.
"Once you find out ... the Circumstances ; then you can go out"
Daniel Quinn
Old Hand
Posts: 859
Joined: Thu Jan 05, 2017 9:19 am

#276 Re: DIY Transistor Amp Using NVA Boards

Post by Daniel Quinn »

You don’t seem to have answered my question .

You have no way of telling wether the article is true or bollocks . If you do , please explain how.
User avatar
rowuk
Old Hand
Posts: 453
Joined: Sun Jun 01, 2014 2:50 pm
Location: Germany

#277 Re: DIY Transistor Amp Using NVA Boards

Post by rowuk »

The bollocks are the continued comparison of repeatable measurements (technical and audible) to the least reliable form of measurement: random listening. I think that everyone here now realizes what is "important" to mr Quinn and that his arguments do not necessarily need constant repeating.

In the case of phase shift and interconnects, ½ dB change in amplitude at 20K will not be audible, but the phase shift could be (for some). Picking the "better" of the two would mean experience and correlation of the technically measured with what we hear. That is a long process before any reliability can be demonstrated.

As our listening capacity is a constantly changing thing based on mood, attitude, what we have learned and above all habit, it is impossible to use "ears alone" with anything that we would want to communicate with others. There is no base for comparison. It is also why I very much disagree with a "claimed" AB or ABX comparison without the exact (and adequate) conditions of test. In the usual "audio sense" not enough care is taken that the differences can be quantified instead of confirming bias.
Whenever I feel blue, I start breathing again.
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#278 Re: DIY Transistor Amp Using NVA Boards

Post by Ray P »

Here's a clue Dennis;

https://www.aes.org/journal/
Sorry, I couldn't resist!
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#279 Re: DIY Transistor Amp Using NVA Boards

Post by IslandPink »

Daniel Quinn wrote: Mon Jun 06, 2022 9:24 pm You don’t seem to have answered my question .
You have no way of telling wether(sic) the article is true or bollocks . If you do , please explain how.
It looks authentically to be a scan of a paper from 1991, published in the AES journal.
As such it would have been scientifically peer-reviewed before publication.
It's linked from several academic portals - which would not be so if there were concerns about its authenticity.
"Once you find out ... the Circumstances ; then you can go out"
Ant
Shed dweller
Posts: 2332
Joined: Mon Jul 31, 2017 6:45 pm
Location: Yorkshire

#280 Re: DIY Transistor Amp Using NVA Boards

Post by Ant »

Took the a20 up to mi dads to have a poke around in it for faults, couldn't find anything wrong with it anywhere. Conclusion is that it hums unless it is connected to something downstream that has a mains earth. Then it doesnt.

While it was open a pair of 39k resistors were paralleled up with the 18k feedback resistors to make the feedback resistor 12 point something k and increase the feedback. This has dropped the gain a little, but it still almost had the windows out with the volume set at 9 o'clock. Still need to dump some gain for it to work with the 15" fanes they are far to efficient for it, with cd the volume is only on about 8 o'clock. Not that it matters, but id prefer abit of volume control travel.

It also looks like a pimple on an arse sat in the place of the f5. Its only about a 6th of the size........

Sounds nice though.
Was interesting to compare the before and after with the altered feedback resistors, it was a clear improvement with the 12 point odd k value, not subtle.
Also starring Rex Hamilton as Abraham Lincoln

www.bte-designs.weebly.com
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#281 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

Yep, the improvement by increasing the feedback is not to be sniffed at.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Ant
Shed dweller
Posts: 2332
Joined: Mon Jul 31, 2017 6:45 pm
Location: Yorkshire

#282 Re: DIY Transistor Amp Using NVA Boards

Post by Ant »

Its far too loud :shock: ive had to back it off further and further until it is just off the bottom stop. Its even worse with these than it was with the 12" fanes for being stupidly loud :D i just dont have the right speakers for it
Should be ok with the records though
Also starring Rex Hamilton as Abraham Lincoln

www.bte-designs.weebly.com
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#283 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

Has that shop down Mexborough still got those Missions in?
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Ant
Shed dweller
Posts: 2332
Joined: Mon Jul 31, 2017 6:45 pm
Location: Yorkshire

#284 Re: DIY Transistor Amp Using NVA Boards

Post by Ant »

No idea, i havent been down that far lately
Also starring Rex Hamilton as Abraham Lincoln

www.bte-designs.weebly.com
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#285 Re: DIY Transistor Amp Using NVA Boards

Post by Cressy Snr »

Cressy Snr wrote: Wed Jun 08, 2022 6:36 pm Yep, the improvement by increasing the feedback is not to be sniffed at.
So much so that I’ve turned up the feedback levels in my own setup a bit more, by paralleling up with 39K to get 10 and a bit K. This was advised ages ago when I had my Fanes by a guy on Paul’s (HFS) forum. He put up measurements to show there would be no problems and quite a few advantages to be gained by increasing the feedback level by 6dB over the ‘standard’ board values. I did it but was terrified of instability even though his measurements showed no such problem existed, so chickened out. Anyway after hearing our Ant’s A20 this afternoon with 12K, I thought sod it and did the mod this afternoon. I already had 15K feedback R as opposed to the 18K standard value and these had cleaned up the sound quite nicely. The 10K has thrown the window wide open. Fan bleedin' tastic. 👍😎

I know Richard was into minimal feedback levels etc etc, and I was a bit loathe to mess with his idea of perfection, but on this evidence I’m afraid that I have to disagree with him. He’s probably up there right now calling me a know-nothing brainless twat. :lol:
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Post Reply