Post by Owen Y on May 29, 2017 23:42:09 GMT 12
So, continuing that possibility, the black wires probably head to those 2 round diodes I mentioned above - these could comprise a simple 'full wave' rectifier to produce raw 6.3vdc for the 2 axial smoothing caps, which need to be (over)rated only 20-25v or so (as printed on the bd).
|
Post by Graham on May 30, 2017 9:45:12 GMT 12
Keep up the good work Owen/Sherlock. Another brain teaser question seeing as how you are on a roll. The speaker terminals on the amp have a choice of 8 ohm or 4 ohm, while most speakers are rated at 8 ohms but measure closer to 6 ohms. So -- what difference would it make using either on my '8 ohm' speakers ? Is one technically preferred to suit the amp and/or speakers, and will there be any audible difference ? Cheers Grasshopper.
|
Post by Owen Y on May 30, 2017 10:46:41 GMT 12
"Patience Grasshopper, work hard, persevere." OPT connections: The so-called output 'taps' refer to the connections available on the output (secondary) winding of the OPT - but you already know this As discussed earier, OPTs are necessary in valve amps because valves have very high impedance/plate resistance (1000s of ohms) & would not like being connected directly to a low 8 ohms load. Yes, you choose say 4 or 8 ohms to suit your lspkr specif, but: - it's worth trying all the options with your lspkrs, let the ear be the judge. - measuring the lspkr with an ohms meter gives you the DC resistance, but not the AC impedance (just another way of referrinng to the resistance under an AC signal) - it's normal for an 8 ohm imped lspkr to measure a bit lower 6-7 ohms DCR. - arguably most important is the impedance curve of the lspkr in the bass & mid freqs, because the bass in particular is most power-demanding of your amp. - theoretically, it's fine to put a higher imped spkr load (say 8 ohms) onto a lower imped tap (eg. 4 ohms), than the other way around - which would be more demanding loading on the amp output stage. There may be some loss in power output (may be unimportant) & there may be some difference in sound (possibly lower distortion?).
|
Post by Owen Y on May 30, 2017 12:14:22 GMT 12
|
Post by Graham on May 30, 2017 14:04:15 GMT 12
Yes they are MX6 Silver Series. OK, must try 4 ohms again. I remember experimenting with this when I first got them but couldn't really detect any difference.
|
Post by michaelw on May 30, 2017 14:10:54 GMT 12
back when he was reviewing for hifi news, martin colloms was fond of running speakers off the 4 ohm taps of tube amps.
he found noticeable sonic improvements at the cost of less ultimate level.
my old quicksilver monos had taps for 1, 4 and 8 ohms.
4 ohms was the sweet spot.
|
Post by Owen Y on May 30, 2017 22:17:46 GMT 12
Graham - You can see that the PSU side of an valve amp is significant. Those numerous chokes (4) & big high voltage caps (6) looks like: A 'choke pi-filter' (DC smoothing) circuit: This is a traditional config using chokes instead of resistors - accepting the increased size, weight & cost, because chokes sound better, bandwidth & dynamically under load demands. The numerous caps are probably because they are 'stacked' - because the KT88s can handle over 500vdc & high-voltage caps are rated normally only 500vdc, they can be 'stacked' in pairs, in order to double the voltage capability, ie 1000vdc per pair. Something like this schematic below: (Except that 2 x valve rectifiers are used, not 2 x SS diodes shown.) This shows a 'pi' filter - so-called because the Cap-Choke-Cap (C-L-C) config resembles the Greek symbol 'pi'. In the Yarland, possibly there is a double C-L-C-L-C config. ('L' denotes an Inductor/choke.) Probably part of the reason why it's a good amp sonically. Haven't fallen asleep yet?
|
Post by Owen Y on May 30, 2017 22:41:08 GMT 12
Graham - feel free to butt in with Qs, change the direction of this. Not sure what's of specific interest to you. PS. Did you see all those audiophile (black) Clarity Cap SA film caps in the front end?
|
Post by Graham on May 31, 2017 8:46:12 GMT 12
Hi Owen
My specific interest is understanding how it all works, and you are doing a great job of explaining this. I'm not sure if anyone other than me and Ben are following this thread but I am certainly finding it interesting and informative. My initial impression after looking inside was that this amp is particularly well designed and constructed. I have seen quite a lot of forums bagging Chinese products, and yes they do make some rubbish if that's what the client wants, but they are capable of good quality also. I'm surprised it doesn't have more of a following. I've never seen it on a 'recommended valve amp' list, but it has really good performance and value compare to the absurd prices charged by the more recognised brands. Bit like the snobbery that exists in the car industry where certain brands popularity is not necessarily reflected in the quality or reliability.
|
Post by guitardude on May 31, 2017 14:01:06 GMT 12
I'm still here too, although admittedly a lot of this is going straight through to the keeper !
|
Pundit
Post by Gryffles on May 31, 2017 14:06:48 GMT 12
I'm completely interested. Keep going
|
Post by Owen Y on May 31, 2017 16:04:58 GMT 12
Oh good, hopefully not blind leading the blind Graham - how would you describe the sound of this amp? (Compared, if you like, with what you had before.) Have you got it running again (with a full or spare set of KT88s)? Or is it idle presently? (You may need to go under the bonnet again ) PS. Are the Bias points & adjustment, also on the chassis side? (I see most Yarland pics show this on the chassis top.)
|
Pundit
Post by papahemi on May 31, 2017 18:22:36 GMT 12
Me too, but the keeper is really busy, still trying to get the occasional edge.
We are what we pretend to be, so we must be careful about what we pretend to be.
|
Post by Graham on May 31, 2017 21:35:03 GMT 12
Oh good, hopefully not blind leading the blind Graham - how would you describe the sound of this amp? (Compared, if you like, with what you had before.) Have you got it running again (with a full or spare set of KT88s)? Or is it idle presently? (You may need to go under the bonnet again ) PS. Are the Bias points & adjustment, also on the chassis side? (I see most Yarland pics show this on the chassis top.) Yep, the amp is up and running again with the original Shuguang Electron Tube KT88-98s and sounding very nice indeed. I would make a hopeless audio reviewer as I don't understand references to 'air around notes' or PRAT, I'm just a nuts and bolts engineer! However what I listen for is realism. By that I mean do the instruments or singers sound like the real thing as if they were performing in the room somewhere between the speakers, with good individual placement and depth. The Yarland does this very well, no doubt helped by the acoustics of my large cave, with nice deep but controlled bass, lovely mid range, and excellent high frequency which is demonstrated on some of my violin concertos. It is never tiring to listen to which is a good test of a hi fi system.
The bias points are half way down each side which proved to be a PITA to get at so I manufactured a plug in extension harness to the 3 sockets to bring the meter access points to the front when biasing. I also had to make up a very short screwdriver to tweak the adjusters. The 3 good Gold Lions KT88s are sitting here probably with a bit of life still in them, but I am still deciding what to do about buying a matched quad for a future back up or upgrade. Cheers
image uploader
Numerous world patents pending.
|
Post by Owen Y on Jun 1, 2017 12:01:10 GMT 12
Awesome! Don't you love seeing DIY? PS. What are the insructions? Usually you measure a voltage - which is taken across a resistor within the amp.
|
Post by Graham on Jun 1, 2017 12:32:49 GMT 12
Correct. Bias is adjusted to 0.4 volts for KT88-98, or 0.45 volts for KT90 according to the manual. Info on some forums indicate that with KT88s anywhere from 0.38 - 0.42 is fine for a balance of valve life and performance.
|
Post by Owen Y on Jun 1, 2017 13:05:12 GMT 12
OK - this brings us to a key part of understanding the amplifier design - the valves' operating configuration, set-up by the designer. Of particular interest are of course, the KT88 output valves. First, we always look at the KT88 datasheets:
|
Post by Owen Y on Jun 2, 2017 9:54:11 GMT 12
I've highlighted the Power output of the 3 main operating modes. As I recall, Yarland specified power output is 40wpc (?), which concurs with the Genalex data for 'Push-pull Ultra-linear with Cathode Bias'!
|
Pundit
Post by Gryffles on Jun 2, 2017 22:30:59 GMT 12
@owen Y
Would you mind listing what each of those abbreviations stand for? In lay-mans terms of course!
eg
Va(b) Va, g2
etc
|
Post by Owen Y on Jun 3, 2017 12:58:55 GMT 12
Hi Gryffles - good question! Maybe the best place to start is to look at the basic operation of a valve & the simplest valve is a 2-element Diode: If a +ve voltage is applied to the Anode (plate), electrons (-ve) generated from the heated (oxide-coated) Cathode are attracted across the vacumm gap to the Anode (+ve):
|
Post by Owen Y on Jun 3, 2017 13:05:01 GMT 12
For amplification, the simplest valve is a Triode - requires an additional element, a 'grid' is inserted between Cathode & Anode:
|
Post by Owen Y on Jun 3, 2017 15:25:13 GMT 12
This grid (of wires) is used to control the flow of electrons (like a tap as mentioned by RdM earlier), so that when say a +ve signal appears at the grid, more electrons are attracted from the cathode & electron flow is increased. OTOH when the grid becomes more -ve, electron flow is reduced. Thus this grid is often called the Control Grid. Small signals at the grid are able in this way to create larger signals through the valve. In Graham's terms, it's something like airflow through a carburettor, accelerated by the venturi & controlled by the throttle flap
|
Post by colinf on Jun 3, 2017 18:34:59 GMT 12
The typical audio triode doesn't operate with positive grid voltage as the diagram above shows. It uses negative grid voltage. Gryffles, Va means the voltage applied to the anode relative to the cathode. Vg2 is the voltage applied to the second grid in a pentode, which has 3 grids. A triode's grid is called g1 or control grid.
AMR-iFi R&D
|
Post by Owen Y on Jun 3, 2017 19:24:40 GMT 12
colinf is quite correct, the Physics & Radio Electronics drawing is misleading - I have amended it. Normally a valve Control Grid is 'biased' -ve in relation to the Cathode - and sufficiently so that incoming +/- signal 'swings' do not cause it to drop to anywhere near 0 volts (otherwise it will start to suck up electrons which we want to go to the Anode). (Also amended, Current flow direction is, by convention, opposite to electron flow direction.)
|
Post by Owen Y on Jun 3, 2017 19:47:36 GMT 12
Moving on to Tetrodes - which have 4 electrodes - an additional Screen Grid (g2) which has the desirable effect (due to shielding, increasing electron flow) of greatly increasing amplification:
|
Post by Owen Y on Jun 3, 2017 20:35:15 GMT 12
Then you get Pentodes, which as you can see above, have a 5th electrode a Suppressor Grid (g3) - which helps 'suppress' the 'secondary emission' of electrons from the Anode when it is being bombarded by electrons from the Cathode. The overall effect is that Pentodes can handle greater voltage 'swings', produce greater power output. Examples are the EL34 & EL84 - by Philips. This brings us to the KT88 which is a 'Kinkless Tetrode' (KT) or Beam Tetrode: This power tube was an invention of GEC(UK) & is a tetrode with additional 'Beam Plates' as an alternative solution to the problem of 'secondary emission' mentioned above. Other Beam Tetrodes are KT66, 6550, 6L6. Sorry, lomg winded but that's a simple outline of valve types - which will help us understand the valve characteristic abbreviations that Gryffles asked about. Next.....
|
Pundit
Post by Gryffles on Jun 3, 2017 20:39:34 GMT 12
@owen Y
All very useful and informative. Keep up the good work please!
|
Post by Owen Y on Jun 3, 2017 20:42:56 GMT 12
( Gryffles - you have to type @ a d m i n (no spaces) for the tag to appear. The reason is that some members have been registered under different names than their display name - mine was default, sorry.)
|
Post by Graham on Jun 3, 2017 21:04:00 GMT 12
I'm managing to keep up ( just) with your excellent explanation Owen. Keep the info coming, we'll let you know when we reach information overload !
|
Post by colinf on Jun 4, 2017 9:54:02 GMT 12
The Yarland uses a 12ax7 as an input amplifier stage. It's called a twin triode, which means it has two triodes inside the one glass bulb. Have a look at one closely while I talk about it. It operates typically like this: the filament heaters need to be warm first by applying 6.3 volts at 0.3 amps to make them glow red hot, the bit you can see glowing. It takes about 11 seconds to heat them to red hot from cold. The heater filament is inside the small diameter vertical tube that is called the cathode. The heater heats the cathode to red hot also. The cathode then boils off electrons. The grey box surrounding the heater-cathode construction is the anode. The control grid has been inserted between the cathode and the anode and is hard to see from the outside without pulling one apart. You can see two anodes in a 12ax7, one for each separate triode. You can have both left and right channels going through one 12ax7 twin triode. There is a triode for each channel. The voltages needed to operate a 12ax7 would be (it varies from manufacturer to manufacturer): if the cathode is at 0v or grounded; applying negative 1.1 volt to the control grid and positive 150 volts to the anode results in a current of 1 milliamp (mA) (0.001 amp) flowing from the cathode to the anode. Making the control grid more negative (like -1.3 volts, a difference of 0.2 volts) reduces the cathode current to about 0.6 mA and conversely increasing it to -0.9 volts makes the cathode current about 1.4mA. Almost no current flows into the control grid. It is electrostatic in operation. Let me know if this explanation was understandable! I'll explain what a 6sn7 twin triode is soon.
AMR-iFi R&D
|