New Posts New Posts RSS Feed - Valves Seen On Graham Slee's Desk!
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Welcome to the Graham Slee Audio Products Owners Forum

 

Open to all owners plus those contemplating the purchase of a Graham Slee HiFi System Components audio product and wishing to use this forum's loaner program: join here (Rules on posting can be found here)

This website along with trade marks Graham Slee and HiFi System Components are owned by Cadman Enterprises Ltd


Valves Seen On Graham Slee's Desk!

 Post Reply Post Reply Page  <1 121314
Author
Dave Friday View Drop Down
Senior Member
Senior Member


Joined: 07 Apr 2011
Location: Spain
Status: Offline
Points: 173
Post Options Post Options   Thanks (0) Thanks(0)   Quote Dave Friday Quote  Post ReplyReply Direct Link To This Post Posted: 16 Feb 2016 at 10:45pm
Yes Graham sounds like the same books,mine are from when I made the "Three three "
Very interesting to see the distortion figures of those amps!
lp12,oc9mk3,ca610p,krimson40watt pa,kef105.4
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 17 Feb 2016 at 8:10am
I think the reason for the low distortion from this preamp must be because of the 'novel' plate load of the input triode U1.

The 12AU7/ECC82 is 10x higher current than the 12AX7/ECC83 - but is 1/10th the gain supposing the plate load to be a normal resistor.

High current = low noise / low current = low distortion.

The 12AU7 allows a higher current to flow. Not much - it is only around 1 milliamp but sufficiently more than I had using the 12AX7. Therefore the noise is OK, and needs to be otherwise it masks the distortion, and then the distortion can't be read by the meter.

But by 12AU7 standards, the current is quite low, and it has to be at low HT voltage or the grid will draw current. So because of the low current, the distortion is low, and around a tenth of that obtained with the 12AX7.

So, the choice of valve is good, but alone it should not be capable of 0.004% THD according to my train of thought. That's because of the non-linearity of having a resistor as plate load.

But here we have Q1 as plate load. Q1 is a PNP bipolar transistor. It's base is at fixed voltage due to the LED and voltage drop resistor R2, so the current flowing through it and into the plate is determined by R1 - and that current is constant, or fixed.

So U1 (12AU7) plate 'sees' a high impedance load that looks infinite. It's like having something like 1000V HT and a 'long' (really high value) resistor.

That effectively removes the non-linearity of having a simple resistor plate load.

But if you then drive that into any old input, the input resistance would be in parallel with Q1, and would bring back the non-linearity to a degree. So, what the AP was measuring was U1 loaded by the AP which wasn't what U1 had the potential of doing.

U2 (another 12AU7 triode) and associated circuit is a cathode follower. It has near 100% negative feedback so its load can be up to near its own cathode resistance before it starts pushing it non-linear. OK, the resistive load (R7 incidentally) is non-linear, but the load becomes part of R7 and so it's all down to the valve's own output impedance. And this has been brought really low because of the near 100% negative feedback.

U2 grid bias is of the leak variety via R8, and its bias voltage set by R9 in ratio with R7. The junction of R9, R7 follows closely what the grid does so the voltage across R8 hardly changes at all.

R8 is a large value to start with but as the voltage across it hardly ever changes, it becomes several times its value so the output of U1 (first triode) sees a massive resistance in parallel with its constant current load. It's so big as to have near zero effect on U1's linearity.

This is one form of bootstrapping and is called that as it is akin to pulling ones-self up by ones own bootstraps. That's impossible for us to do, but in electronics, it works!

And therefore, U1's linearity is preserved - U2 circuit effectively buffers U1 from the load and we are left with what U1 can do if only it could be isolated - and we have effectively isolated it.

We can only guess at what gain U1 is doing. Its data reckons it has a gain of 10 for a resistive load, but using our "magic" load it should be at least double, and based on what the global negative feedback (R10/R5) suggests I know it's a lot less than 40dB x R10/R5 (250) which would give a gain of R10/R5. It's doing around 3/4 of that so I reckon U1's gain is somewhere between 25 and 100.

So R10/R5 gives us some useful global negative feedback which just falls short of the 40dB negative feedback (NFB) Harold Black of Bell Labs discovered in 1927. So if we guess at 20dB NFB the math suggests distortion is being reduced by a factor of 10, and so the unaided distortion of U1 is around 0.04% or probably a little less - but we'll never see it as that without the rest of the circuit.

The voltage gain of the circuit is somewhere between 8 and 9dB which is just about 2.5 - 2.8.

A properly recorded CD played on a 2V output CD player will be doing around 400mV most of the time, and the output from a phono preamp will be doing about 300mV.

Good power amps have input sensitivities of 400mV to 1V. And those with simple valve amps requiring 1.5 volts? Well, if the input impedance is that of valves, the preamp should do a gain of 2.8, and you'll be 1.2dBW down on max power - close enough!

Therefore this stage stands a good chance of becoming a preamp.

I've been wondering what I could use as a clean gain stage for a preamp for near on five years now. This might be it, and you might also now see the long awaited preamp.

That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 17 Feb 2016 at 5:34pm
Because the load impedance becomes part of the cathode load (I suppose that should be obvious...) then the gain and hence critical measurements will be influenced by the load being driven. Cables too will have some influence but using something sensible (like a CuSat50 or Lautus) won't change things much at all (we use similar for analysis).

So here's the specs...

Frequency response: 3.5Hz - 140kHz (-3dB) or 7Hz - 70kHz (-1dB)
Frequency response (compensated option): 3.5Hz - 60kHz (-3dB) or 7Hz - 30kHz (-1db)
Note: the compensated option might be the one I choose.

Gain: 7.7dB with 22 kilohm load; 8.7dB with 100 kilohm load

S/N: 84dB 20Hz - 20kHz

THD (distortion) at 1kHz: 0.004% into 22 kilohms; 0.014% into 100 kilohms
THD (distortion) at 10kHz: 0.01% into 22 kilohms; 0.02% into 100 kilohms

IMD (intermodulation distortion): 0.012% into 22 kilohms; 0.044% into 100 kilohms

I reckon this is useable!

Here's the FFT plot (spectrum analyser)

Tube Preamp FFT plot

The 2.4kHz peak seen between the 2kHz second harmonic and 3kHz third harmonic is some spurious noise, and being at -90dB should be of no concern (although I'd like to get rid of it).

The 2kHz second harmonic is greater than the 3kHz third harmonic, which is what you'd expect from a tube... but it's doubtful you'll hear its effect at -77dB.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
pcourtney View Drop Down
Regular
Regular
Avatar

Joined: 12 Feb 2015
Location: Stansted
Status: Offline
Points: 59
Post Options Post Options   Thanks (0) Thanks(0)   Quote pcourtney Quote  Post ReplyReply Direct Link To This Post Posted: 02 May 2016 at 3:40pm
that was a great read, and I am now more informed than ever about the Valve/Tube infidelity saga that has so many people (who should know better) saying it sounds better, but in reality they should just be saying "it just sounds a little different to solid state" , and why don't you Mr Punter listen to both pre amps and make your own mind up !  

and to be honest Graham, if you could design a pre-amp that has a valve bypass switch, so that HiFi dealers all over the world could demonstrate the qualities of both, and that it would also work if the valves were completely removed, then some of the hysteria about valve amps may dissipate in time


that could also solve the google SEO for GSP, not only as a new product, but if a "White Paper" could be written by Graham Slee about designing a pre amp that can have valves inserted in the gain path, and what it brings to fidelity ( or not ) 

the GSP loan program (at least for the UK ) would really be a very clever marketing ploy to throw into the mix as well :-)  


Edited by pcourtney - 02 May 2016 at 3:47pm
Back to Top
Paperweight View Drop Down
Regular
Regular
Avatar

Joined: 21 Mar 2011
Location: Lexington NC US
Status: Offline
Points: 84
Post Options Post Options   Thanks (0) Thanks(0)   Quote Paperweight Quote  Post ReplyReply Direct Link To This Post Posted: 12 Feb 2017 at 6:07pm
Good choice on the 12AU7. I have an old military R392 HF mobile receiver that uses 12AU7 tubes in the IF strip with 25-30 volts on the plates. Since their gain is reduced with the low plate voltage, they just tacked on an extra stage or two to make up the difference.
 
12AU7s and its ancestors have very linear plate curves. I bought a bunch of 7 pin 6C4s which are the equivalent of half of a 12AU7. They can drive a speaker at under one watt on 90 volts. Also, at quantity they are fairly inexpensive with military versions often going for $5 or so.
 
Battery tubes have always been fascinating to me. It's amazing they could run the filaments off a 1.5 volt battery and 45-90 volts on the plates.
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 12 Jan 2020 at 3:00am
Resurrecting this old topic, I've had my nephew George staying this weekend, and he's more advanced in guitar playing than my son Ryan. They're about the same age.

We wanted to hear him play the Strat I'd bought Ryan, but all I had for amplification was a Proprius monoblock, which although it can drive the Fane Ascension F70 bass reflex cab Ryan built, it doesn't have the gain, or the sort of sound you'd expect.

Then I remembered the valve preamp discussed in this topic, and dug it out half expecting it not to work, but to my surprise it did.



Plugging the guitar into one input and using just one channel into the Proprius improved the gain and the sound, so I decided to loop it into the other input and take that output to the Proprius.

Not only was there sufficient gain, but the sound took on a signature George really liked. It also went well with the Epiphone Les Paul.

Next a Zoom G2.1u effects unit was inserted and set to "Extreme Distortion" ("Ed" on the display), still on the Epiphone, and things really started buzzin' (not as in buzzing!). We were treated to as many hard rock riffs as George could remember, and it really sounded true to original.

By this time I had to set the Proprius down to roughly 10 o'clock volume setting, because even with all my sound proofing I was starting to get worried.

I got George to guess the power and he reckoned it sounded like a 100 watt amp, and was surprised when I told him it was just 25.

Running on the original 48 volt supply, and although biased for a 72 volt supply, the preamp did a great artistic job. I think it worth building an extra two stages, making a total of 4 stages and stick a tone stack in the middle and some volume controls either end, and to possibly put in a transformer power supply.

After a bit of tube rolling, the RFT ECC82 was chosen to give the best sound, with the JJ Tesla a close second. I think it will make an excellent guitar preamp.


That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
peterb View Drop Down
Senior Member
Senior Member
Avatar

Joined: 02 Feb 2017
Location: Cornwall
Status: Offline
Points: 332
Post Options Post Options   Thanks (0) Thanks(0)   Quote peterb Quote  Post ReplyReply Direct Link To This Post Posted: 12 Jan 2020 at 2:01pm
A proper British made Valve Guitar Head Amp would be something that could fly!
A good Mesa Boogie sells for over 2k!
What it would need is a celeb player to be associated with it then their fans would follow! Wink

Peter
---------------------------------------
Dual 505-1, Cyrus CD T, DIY 80W MosFet amp and PreAmp, 2xKEF 103.2
Back to Top
 Post Reply Post Reply Page  <1 121314
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.01
Copyright ©2001-2018 Web Wiz Ltd.

This page was generated in 0.141 seconds.