UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 2nd Nov 2021, 7:04 am   #61
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Golden Ear Laboratory Amplifier Build

Compared to transistors, valves allow you to opt for substantially more gain per stage at audio frequencies. Doing so involves high anode load impedances and this brings down the frequencies of the poles due to stray capacitance, Miller, etc. So the phase lag per stage starts to grow at lower frequencies, the harder you push each stage for more gain.

People see simplicity as a jolly good thing, but they judge simplicity by counting active devices. Valves used to be terribly expensive and were thought so unreliable that though a set might have a year's guarantee, the valves were excluded and got only 12 weeks, with their warranty being from the valve manufacturer. This all shaped human perception and still remains today.

If you aren't tied to using the absolute minimum number of active devices, you can use lower stage gains and find that the lower impedances allow you to reach the same overall gain, but with less overall phase lag at higher frequencies.

Also, you can employ local fedback loops or degeneration to flatten stages within the overall loop. It costs in stage gain and therefore device count, but the trade-offs aren't simple linear ones.

Degeneration is interesting. Essentially it's the use of an un-decoupled cathode resistor. It allows sensing of the anode current and its feedback into the grid circuit and grid voltage. Feedback! .... Now, there are people utterly opposed to using feedback in audio amplifiers. Because they can't see a feedback path on a circuit diagram with an overt line from a late stage going back to an early stage, they think there is no feedback and they are happy. They are protected from seeing things they are scared of, simply by lack of understanding. Otherwise they would see witchcraft everywhere.

So we could do an amplifier as a series of stages with local feedback flattening and linearising each stage to an extent where overall feedback isn't necessary. In this case by stage, I mean there might be more than one valve per stage. The product of the responses of all the stages giving the final performance.

Alternatively, such a string of stages can have overall feedback in addition to local loops. You can have loops within loops (and within loops etc.) but you have to design them carefully.

So, we arrive at a monstrous trade-off, where circuit complexity, flatness, distortion and stability can all be traded off against each other to varying extents.

There is one item that we seem to be stuck with, though.... the output transformer. It isn't hard to design the valve sections of an amplifier so as to leave the output transformer as the dominant cause of phase lags within the loop. essentially, the output transformer is the ultimate controller of the quality of an amplifier.

There is a quirk of transformers though. You can regard them as bandpass devices. They don't transform DC, so there is a low frequency roll-off, affording the delightful prospect of stability issues at the bottom end of the frequency range. This is set by the inability to wind them for infinite impedance. The low frequency end of the response is set by the primary inductance, and the impedance level it is operated at. Operating at lower impedances extends the low frequency response, and the lower turns counts can reduce high frequency strays. Again, there is a compromise and the trade-offs aren't simply linear, but by exploiting lower impedances you can get good transformer operation over more octaves of frequency.

I had the job of designing an impedance measuring bridge. Performance had to be optimised over 10kHz to 18.6MHz. Now that's a lot of octaves. On top of all that, it had to measure balanced impedances, so there had to be transformers in it. To handle the high frequency end, they were transmission line transformers wound on small ferrite toroids with multi-filar wire (6 wires twisted together in one case) The top end precluded the use of the highest permeability ferrites, and dictated the length of the winding. The low frequency end was left out in the cold. I had one dimension of freedom left that I could sacrifice. The detector measuring the bridge balance was ridiculously more sensitive than was needed and so it didn't matter if the bridge was more lossy. So I introduced a minimum-loss pad to reduce the Z presented by the source to the first transformer of the bridge. I didn't need so much primary inductance now, and the same trick worked on the transformer feeding the output to the detector.

So there are similar games which can be played with valve amplifier output transformers. Use lower HT voltages and paralleled, fatter valves and your impedances and turns ratios fall. So does your efficiency, but if you can stomach that, your need for primary inductance is less and you can trade some of this for better top end response.

Ultimately you can bring operating impedances down to the speaker impedance and then it looks like you need no transformer at all, but transformers also do the push-pull thing for us. So we need to change our amplifier output stage topology to handle this, and use something like a totem pole output. This brings us to one design freedom valves lack... you can get complementary pairs of transistors. Where are the complementary valves when you need them? (presumably positron based, but that means anti-matter and all its dangers)

So, we've started with musings about better audio amplifiers and wound up limited by the matter/antimatter asymmetry of our universe. Bit of a b****r.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

Last edited by Radio Wrangler; 2nd Nov 2021 at 9:42 am. Reason: Where not There typo
Radio Wrangler is offline  
Old 2nd Nov 2021, 7:39 am   #62
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Thanks David !!!

AND I mean that. I dont know why I didnt get you as a lecturer or mentor. Your explanations are superb and VERY easy to understand. I hold your opinions in great stead and try to follow and understand them as best I can. So lecture #61 is superb for me.

To my transformers!!. I spent about one year mulling over these traffos. My dad taught me transfomer building when I was about 12, and actually sneaked me into his workshop while he was supposed to be doing overtime !! SSHHH !!! Dad was born in 1906.
ANYWAY, he KNEW with great accuracy what the outcome would be. He taught about insulation values. He taught me about sectioning. He taught me about the different grades of transformer iron.
OHH and varnish, or wire insulation, and even had some dreaded " black enamel" Basically copper wire "painted" with shellac.

My current transformers are the best I could conceive, and I hope ( pray ) they work well.
Another thread at present, What have you bought years ago? or some similar thought !!.
MY wish, I wish I had listened more!..

I will post again when I have done more work.
Thanks again.
your humble student,

Joe
joebog1 is offline  
Old 4th Nov 2021, 1:13 am   #63
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

So after doing a bit of research and testing I will use audiophool cable for my feedback.

In my pic are the three types of audio screened cable I use. I require 400mm length from the output terminals on the back panel to my wafer switch at the front input section.
I measured the capacity of each type using exactly 400 mm. I will earth the screen at the input valve.

The thin cable is 152 pF,
Medium cable is 68 pF,
Audiophool hot pink cable is 30 pF.

I will be using that. They are all made in USA by Soundlink.

Joe
Attached Thumbnails
Click image for larger version

Name:	screened_cable.jpg
Views:	66
Size:	34.3 KB
ID:	244970  
joebog1 is offline  
Old 4th Nov 2021, 1:32 pm   #64
Gabe001
Octode
 
Join Date: Oct 2019
Location: Stockport, Greater Manchester, UK.
Posts: 1,612
Default Re: Golden Ear Laboratory Amplifier Build

Joe, just out of curiosity, can I ask if you've attempted to calculate the gain on this build. It seems to me, with so many cascading steps, that it is going to be a staggering amount. What you are going to use as a signal source?
Gabe001 is offline  
Old 4th Nov 2021, 2:26 pm   #65
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Golden Ear Laboratory Amplifier Build

You can always make more gain than you can use. Nowadays, gain is cheap. In an audio amplifier with overall feedback, phase is a bigger concern. You need to achieve the chosen amount of gain without any more phase lag than is unavoidable.

As frequency goes up, you are racing the roll-off of gain against the build-up of phase lag. You need to let the gain win, in getting to below unity before the phase lag gets above 180 degrees.

If you achieve a chosen amount of gain in the fewest number of stages, you usually wind up with more phase lag than you'd get had the same total gain been artfully spread over more stages. There is an element of swings and roundabouts, but it isn't quite an even balance.

So if you don't mind increased valve count, you can exploit it for your benefit and trade the benefit off across different areas. It becomes a design freedom to spend where you wish.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 4th Nov 2021, 2:48 pm   #66
Gabe001
Octode
 
Join Date: Oct 2019
Location: Stockport, Greater Manchester, UK.
Posts: 1,612
Default Re: Golden Ear Laboratory Amplifier Build

Thank you David. Is phase lag an issue when the triodes are phase split (I don't know if this is the proper technical term) as in this design, or is it always a potential problem even with a simple amp e.g with one half of an ecc81 feeding the other half directly?

Sorry for the many questions, still learning!

Gabriel
Gabe001 is offline  
Old 4th Nov 2021, 6:09 pm   #67
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Golden Ear Laboratory Amplifier Build

Phase lag is always a concern in any amplifier destined to use overall feedback. It sets how much feedback you dare deploy, and how high in frequency you dare let the closed loop gain stay flat.

Yes, even phase-splitter stages. There was some interesting design work done in the phase splitter Dr Bailey did for Arthur Radford. It's gon into in detail in Arthur Bailey's article in Wireless World.

The long-tail pair phase splitter behaves as a cascode from the input grid to the far side anode, but as a simple triode to the near side anode. Thus output impedances, phase delays and bandwidths are not balanced. Bailey rather neatly redressed the balance.

The concertina phase splitter looks like a triode output to the anode output, and a cathode follower to the cathode output. So output impedances are unbalanced. It can be lived with, but you can do better.

If you go over to transistors, these things still get to play, but are nastier, you get the Early effect and pumped varactor diode effects pushing in on the act. So techniques which are helpful in the valve world take on greater importance in solid state circuits.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 4th Nov 2021, 11:07 pm   #68
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Gabe, Yes I have ( probably badly ) calculated the gain. Its huge!!. BUT there are a multitude of feeback paths too, all of which tame the gain down quite a bit in the end.
Cathode resistors are not bypassed.
Common plate feed resistors, all 1% matching both sides.
Voltage regulated input stages.
Cathode follower driver stage.
Finally a LARGE amount of overall feedback, so heres hoping .
I did post the origional article a couple of years back, or you can download the origional magazines from
Radioandbroadcasthistory.com
There were a few notes released after the origional publication too.

You need August and September 1956 for the main article. The circuit is very well explained.

Joe
joebog1 is offline  
Old 4th Nov 2021, 11:43 pm   #69
trobbins
Heptode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 898
Default Re: Golden Ear Laboratory Amplifier Build

The concertina PI is effectively perfectly balanced up to when the following stage grids conduct, so it is best used where the next stage won't be the first stage to overload. For example, Williamson placed the concertina stage behind a driver stage, rather than directly using the concertina stage to drive the output stage where grid current flow may start to become an issue (as well as other reasons to include a driver stage).
trobbins is offline  
Old 5th Nov 2021, 1:17 am   #70
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

I have very ( reasonabley ) sensitive speakers, and my input source is vinyl only. I dont have a radio or a seedie player nor tape player. So overloading stages wont be my problem I hope. My speakers
( origional Tannoy ) are only rated to 30 watts RMS. I have never heard them with 30 watts driving them.
SWMBO chases me with a broom with my seven watts I have at present, at full volume.

The whole exersize is to keep whats left of my brain working, and to use up lots of very exotic components I have. Next up will be the power supply for this amp,
(I have all the bits ready to go) followed by a preamp which is designed and on paper with most layouts done. I have all the components for that too. That is mostly cascode triodes as well.

Thanks to everyone for their comments.

Joe
joebog1 is offline  
Old 5th Nov 2021, 2:49 am   #71
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Golden Ear Laboratory Amplifier Build

Quote:
Originally Posted by trobbins View Post
The concertina PI is effectively perfectly balanced up to when the following stage grids conduct
Sort-of. The following stages will have input capacitance of the simple sort as well as Miller capacitance loading the concertina. Now, as the concertina has different output impedances, the high frequency phase shifts will be different on the two paths and also significant contributors.

The amplifier is wanted to get some benefit from feedback across the audible band to keep it flat, reduce distortion and reduce output impedance to damp the speakers. This means that the forwards path has to have an amount of excess gain remaining at 20kHz and that it will take a few octaves to roll this off to unity. It is right up to the unity gain frequency that we have to take care not to introduce any avoidable lag.

So where you see emitter followers after a phase splitter and before the power valves, they might not be there to support positive grid current, driving the grid capacitance and Miller capacitance can be sufficient reason to include them. In providing buffering, they give a little easement of the problems of the asymmetric impedances of the concertina. Radford and Bailey's trick can help correct the asymmetry of the differential pair type phase splitter.

This is one of those circuits where you have to take care over just about everything if you want the gain at the top end to fade gracefully.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 5th Nov 2021, 5:40 am   #72
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

David and Tim,
Perhaps re-read the origional articles. I understood his arguments perfectly.
ahhmm !! He has written a number of articles dealing with these problems. He starts with a simple 6V6 based amplifier, and he proceeds though several designs over perhaps 4 years to the example I am building. He published an excellent paper entitled, " Extending Amplifier Bandwidth " September 1953 of Radio-Electronics. His 6V6 and 1614 output pairs amplifiers he later modified in a paper entitled "Gilding the Golden Ears". I have searched the web all over for anything he has written and published. I find his discussions and explanations of his circuit techniques to suit my limited brain capacity. I have been working on this "laboratory" version for some years now.
( I told you I am slow !! ) Its my last build for me. Why not a "Rolls Royce" so to speak for my kids to convert to a very poor guitar amp?? ( No distortion ) or sell for sixpence?.

I did prabably 50 designs on output transformers ( all by calculator and in English units ) as I learned about 55 years ago. I suppose it will be they that decide the final result, but its not for lack of trying on my part. 1 1/2" tongue and 3" stack is not very common. I even made prototypes using part of the windings done with Eureka resistance wire to balance the DC resistances. ( It doesnt help anything that I could see ).
So there is an English part of history called "Building Follies" , mostly extravaganza BBQ areas or pagola's, my follie is this amp.

I still like all the comments and suggestions though, One can NEVER know too much.

Thanks again gentlemen and ladies.

Joe
joebog1 is offline  
Old 5th Nov 2021, 5:49 am   #73
trobbins
Heptode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 898
Default Re: Golden Ear Laboratory Amplifier Build

Umm, the concertina circuit with a triode has a balanced output when each output is loaded with the same impedance (such as in the Williamson circuit with a following pp driver). There were certainly some misconceptions that the outputs had differing output impedances that would affect HF response, but that was settled back in the 1950's, including with some measured results as I recall.

However this is off-topic to Joe's cloning effort, although I was sort of chuckling when I read the 1962 MA15 Radford /Bailey articles a few months ago when the topic came up (here) - as the articles started with a lot of hoopla about improving on the Williamson achilles heel of the output transformer, only to end up with an amp that could only use the Radford output transformer as it took almost years to tweak to get right and pretty much no other output transformer could be used without requiring feedback re-design.

Last edited by trobbins; 5th Nov 2021 at 5:57 am.
trobbins is offline  
Old 5th Nov 2021, 9:37 am   #74
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Golden Ear Laboratory Amplifier Build

Emitter followers?..... I meant cathode followers, but same principle.

Yes, the concertina is self balancing in terms of levels, at lower frequencies like audio and provided the loads are equal. Once you go higher in frequency where its own capacitances start to have effects it goes off balance due to output loading even if the loads on its output are equal. With reasonable care, these effects come in well above audible frequencies, but in the zone where the loop is coming down to unity gain, right where phase is important. The concertina, even where its balance is fine is still an amplifier stage, and gives amplitude roll-off as well as phase lag from the stray poles in its response. This happens even if the balance had been perfect.

But the simple balanced pair is known for having different HF responses to is two outputs IF driven single-endedly.

Bailey and Radford knew tht the output transformer was pretty much a dominant factor, but they identified an asymmetric lag which could be reduced. They reduced it. It still left the transformer as the dominant factor, but what they did gave them a bit more margin. They optimised the transformer, Radford's area of expertise and his prime product line. Then they designed a feedback system around it. All told, they were pushing things further than Leak had, a decade or so earlier. The resulting amp design is not buildable with anything other than Radford's transformer design. It is designed into the feedback lop design... meaning not only the reverse path components but also the forwards path components.

The phase splitter can be used in other designs, the idea is not tied to the Radfords, but with a lower bandwidth output transformer, the loop gain cannot be pushed to such high frequencies, and so the benefit of the improved phase splitter becomes less. Not zero, but not as much.

These amplifiers are pretty much at the limits of what can be done with valve-transformer technology. If you want to push things any further, but stick with valves, the transformer's gotta go.

Another nasty is that the lack of a 'PNP valve' means that at high frequencies, or handling transients, stages have different output impedances when slewing in opposite directions. This can lead into asymmetric conditional symmetry where a system which looks ostensibly stable can have an input signal designed which will kick it off into oscillation. Of course, Mr Murphy insists that nature provides these on occasions, just for fun.

Feedback amplifiers may work just fine in their intended bandwidth, but they don't often tail off gracefully above it, they can get quite disgraceful. So, it's wise to have a nice, tame RC lowpass in the input to the amplifier so nothing can get in that it can't handle properly.

These problems also show up in oscilloscopes. No transformer to worry about, but the input is usually single-ended from a probe, and the deflection plates need symmetrical balanced drive. There is usually no overall feedback used, but the whole Y amplifier is a phase splitter. We want 10's or 100's of MHz of bandwidth and good phase response and definitely a lack of ringing. You start learning a lot about the foibles of circuit topologies once you push them to the extreme. There are also some opamps fast enough to make feedback-controlled IF amplifiers at 21.4MHz when you really need extreme flatness and freedom from compression. These are multi-stage amplifiers internally, and you use them with overall feedback. Unity gain is in the hundreds of MHz. There's a bunch of them in the Agilent Noise Figure Analyser, doing just this. There are fields which turn up issues which are a lot less significant at audio, but non-zero.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 5th Nov 2021, 12:48 pm   #75
trobbins
Heptode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 898
Default Re: Golden Ear Laboratory Amplifier Build

Jones in Audio Eng Dec 1951 did some frequency response measurements of a 6J5 cathodyne with large levels of bulk loading capacitance, and derived an estimate of a frequency response of 3-4MHz -3dB for 200pF loading.

Peters in Radio & TV News in 1957 showed meter loading of 50pF on a 6J5 gave some imbalance which could be returned to balance with a padder cap on one output to show -2dB response at 200kHz.

Merlin made measurements of matched response of -3dB to 600kHz with a 12AT7 and bulk loading of 470pF, and also estimated bandwidth well in to the MHz with typical next stage capacitance loading.

Certainly there may be some phase shift improvement to be had with using a trim padder for anyone with the measurement capability, and that would be consistent with maintaining matched symmetry through the forward path in for example the Williamson amp.
trobbins is offline  
Old 5th Nov 2021, 1:43 pm   #76
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: Golden Ear Laboratory Amplifier Build

Those are all measurements of amplitude. The problem is in the overall phase lag from input to output, and not so much in balance between the two outputs.

If something is down 3dB at some frequency, there will be an associated lag of 45 degrees. At lower frequencies the lag diminishes on an arctan law, so the non-linearity is helpful at least.

But in some of these amplifiers the phase margin is not many degrees and a small fraction of 45 degrees is a significant reduction in the overall margin.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 5th Nov 2021, 10:57 pm   #77
trobbins
Heptode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 898
Default Re: Golden Ear Laboratory Amplifier Build

Yes assessing the amplitude corner frequencies is often the first aspect of collating forward path gain/phase contributions to indicate which contributors are likely influencing the response in which frequency regions.

Unlike LF corners from CR networks that are fairly easily calculated due to C and R being dominated by fixed parts, the HF corners from RC networks require more detail to define nominal R and C levels for calculation, and a confirmation measurement is likely deemed too difficult to make. The point is that an estimate can usually be made, and hence particular RC's can be compared for their influence - such as in the relatively simple Williamson circuit where the RC corner frequency from the cathodyne is likely an order of magnitude higher than the RC corners of the other stages, which are nearly an order of magnitude higher than the typical output transformer first resonance. Which is sort of why the effort to pad any measurable unbalance from the PI stage in the Williamson is not a normal action taken by many. But that view can only be made after assessment of all the gain/phase contributors in a particular amp - which Joe may well not have the incentive to pursue for the Golden amp, and as many others do Joe will just aim to tweak a square-wave response by adjustment of the feedback cap and rely on the original designer to get the amp into a region of stable operation.
trobbins is offline  
Old 5th Nov 2021, 11:23 pm   #78
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

You have it in one Tim, thats precisly what I intend to do. If I DO have instability problems you will read about it here. For me at least, this site is "big school" with thousands of lecturers and one student. Me !!.

According to Marshal, he didnt have any problems ( or perhaps the means to measure them/it ).
I have tried to keep all the "strays" out of the build, with as tight as I can true point to point, from valve pin, thru component to valve pin with minimal lead lengths. ( NOT all that easy I might add ).
As Marshal progressed from the simple 6V6 amp to this one he did build at least 4 others that I know of, getting larger power output and higher frequency response with all his additions/modifications. He did not wind his own trasformers however and was stuck with what was available commercially, although I have huge respect for Acrosound and Dynaco transformers. They did seem to be the bees knees back then. This design was based around the Acrosound TO-330, and my hand wound ones I believe to be very close in specs to them.

Anway, time will tell.

Regards

Joe
joebog1 is offline  
Old 9th Nov 2021, 1:48 am   #79
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

So in trying to install my hot pink audiophool cable I found it too large to fit the space I provided.
So I went digging and in my cable box I found some 1.7mm silver wire Teflon coax. Capacity is higher than I would like at 47 pF but it fits so it's soldered in. Another problem I have probably made for myself.

Output stage wiring is complete, but still need some more lacing up.

Joe.
joebog1 is offline  
Old 10th Nov 2021, 12:37 am   #80
trobbins
Heptode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 898
Default Re: Golden Ear Laboratory Amplifier Build

Joe, the feedback path uses quite low resistor values (820R to 82R), so even if the 820R is located at the OPT and the shielded cable shunt capacitance is after that 820R, the RC corner is going to be circa 4MHz, with phase sift contribution above about 400kHz. That may have a slight influence as the bandwidth is indicated as out to 300kHz, but you may be hard pressed to gauge any difference if you located the 820R at the 82R end.

Are you grounding the valve socket spigots?
trobbins is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 8:28 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.