montesquieu
02-03-2012, 13:17
Just posted this on the 'Wam, which is the main Hifi board I tend to post on having arrived there some time before finding AoS, but I thought my friends here might also be interested especially given the profile of Tannoys on the board .... hope the double post is OK. I'll try and get some pics together soon.
Anyone who has had big Tannoys, be they Monitor series, HPDs or even the later ones, knows how wonderful they are at big scale, filling the room with marvellous music of amazing weight and substance. They are fabulous at big 3D soundscapes, and have a wonderful full-bodied way with timbre. In horn enclosures like the GRF they also have a lot of dynamic punch, though they can be a bit flatter and more tubby sounding in small volume standard Lancaster boxes, the up side being deeper bass of course.
They do have their critics though and for some people, a slight coarseness detectable in the treble (with Golds and HPDs at least) is enough for them to pass on the whole Tannoy thing. I spotted this issue too (especially coming after the wonderful transparency of Quad ESLs), and put it down to the standard crossover on the Golds, which though being reasonably sophisticated, also uses some fairly cheap mass-produced components (mine were 45 years old to boot) On top of that, Golds aren't always the most intricately-detailed of speakers - musical yes, room-filing, yes, but detail-meisters, they aren't. Not in standard guise anyway. They also aren't the easiest to drive and get the best out of, needing a high damping factor to keep those 15in cones under control and not all boomy and soft, something that in my experience means a good PP amp with strong current delivery. I've never had much joy with SET despite trying half a dozen at least.
I've owned my GRFs for over two years now, picked up from Coco when he was fund-raising for his Voyd Reference. After a year or so, I had new crossovers done, with £400 worth of premium bits, which took away (as I had hoped) most of the slight throatiness, as well increasing overall refinement and detail, though enough of a gap between what I was getting and what I thought the potential was remained to keep me wishing for more.
Last autumn I took the cabs up to Russ Collinson in Barnsley for a full refurb. The cabs were of reasonable design but quite old and were built of thinner ply than would be used today, also there was little bracing on them so they did audibly resonate in places, creating a bit of bloom mainly in the upper bass. Russ went to town in the bracing department, putting multiple solid posts in to reinforce the folded horn section and bonded on sheets of plywood to kill panel resonance. He also got rid of the mahogany stain they were coated in and refinished in natural wood, and as a finishing touch made new grilles with vintage cloth. Now, the old girls look stunning.
http://i1181.photobucket.com/albums/x440/montesquieu61/newtannoy1.jpg
At the same time he also made up new back plates with speaker posts attached hat enabled the two drivers to be directly wired. Suddenly bi-wiring, or even bi-amping (passive or active) was a possibility. Russ also did some lovely wood boxes to take the custom passive crossovers external. A wonderful job, as usual from Russ.
http://i1181.photobucket.com/albums/x440/montesquieu61/newtannoy2.jpg
Ever since coming across this http://www.hilberink.nl/codehans/tannoy86.htm in the Tannoyista site I've fancied a crack at taking them active. A try with a Behringer DCX 2496 was a bit abortive, mainly because of gain and sensitivity issues (in the GRF cabs they are something like 97db). I managed to pick up (from 'gjm' on the Wam) a 2A3 integrated that would be perfect to experiment with as second amp, while the Audion 300b PP amps (back in my posession after four years doing the rounds ...) have volume controls for variable sensitivity - ideal for level matching.
A conversation with Nick Gorham 'lurcher' resulted in him agreeing to build the circuit above - and last week it duly arrived. Very nice work - he had the boards run up after laying out the circuit for op amps, and added buffer stages to handle any impedance nasties with valve amps at input and output.
http://i1181.photobucket.com/albums/x440/montesquieu61/activexover1.jpg
My idea had been to use PP Audion 300bs for decent bass from the 15in cones, and gjm's sweet little 2A3 for the 2in compression driver. People use a mix of drivers in horn systems all the time - why not in a GRF which is a mid-horn and treble horn in the one drive unit?
First listen, though, was disappointing. Bit of fiddling with level matching got the crossover element into the right ball park, but I felt the overall sound was a little course and sub-optimal. Just, not as musical as it should be, not gelling, while detail seemed to have gone a bit blurry and the lovely delicate sound from my valve pre and power amps had gone. It just sounded a bit - well - cheap and solid state. My first thought was that it was an issue with using two different amps - I went looking for a matching pair to try (I'm still hoping to do this at some point!).
I had a discussion with Nick and it transpired that in my conversations with him I'd described this as 'an experiment' - so he'd done his best to keep the cost minimal, including specifying some fairly cheap opamps. He proposed replacing these with OPA604AP. At more than £2 each (times 12) this was four times the cost of the originals, though still hardly a fortune. They duly arrived from RS components and I slotted them in (a new DIY experience for me but I managed ok quite chuffed with myself. 845 SET build here I come :) ). Instant revelation - much much clearer, and the music had returned. Not only that, I was getting the benefits of active coming through loud and clear, less background hash, more low-level detail, better layering in terms of soundstage, more punch and dynamics. It also became obvious that the tweeters were out of phase - duly fixed, this was a big improvement again. I had been worried about putting a bunch of opamps in between my valve pre and power amps, but I needn't have been concerned, this is now a pretty transparent piece of kit and there seems to be no detectable hit on the subtleties of my AN DAC, Nick-designed phono stage or SJS preamp.
I've left it on now for several days to warm up and burn in and it's been getting better and better. As it has settled in, more and more of the valve character has come back, all that lovely ambience and sweetness, without losing the punch and 3D layering. The soundstage has improved over what I had with the passives, as has low-level detail. The only quibble is a very slight comparative loss of focus around the point of the crossover frequency .. something I'm sure that was there before, but was covered up by other hash. And it is only slight, you have to be looking for it. The acid test - listening with the wife - was passed with flying colours. I got the 'now don't you dare go and change anything' lecture that I get occasionally when I do something very right.
Additional refinements remain possible, which I hope might just fix that slight haze around 1khz - improvements to PSU, or (according to Google) I could even spend £25 a pop (times 12 ...) on even fancier opamps. I will definitely go ahead with improvements to the regulation (as recommended by Nick), the opamp option remains on the cards but prohibitive in the short-term I think, I'd need to be really sure any gains would be worth it.
Also the 2A3 is used at max volume, but the Audions level match at around 11 o'clock on the dial - combined with the lowish output on my DAC that means I'm a bit short of gain on the DAC input (no probs on vinyl). So I may still experiment with a higher-sensitivity amp for the 2in driver. And I really ought to twist the arm of JP 'diceman' or Coco to come over with their measuring gear and make sure I've properly matched the levels (by ear it sounds good but measurement would take away any lingering doubt that I have it right).
Of course a valve version of the active crossover circuit would be intriguing as well - now we know it works accurately - and Nick has promised to take a look and see how it might be done ....
But for the moment I'm a happy bunny. I feel a springtime classical music day coming on to canvas the views of the usual suspects as well as some newcomers - will keep the board informed.
http://i1181.photobucket.com/albums/x440/montesquieu61/newtannoy3.jpg
Anyone who has had big Tannoys, be they Monitor series, HPDs or even the later ones, knows how wonderful they are at big scale, filling the room with marvellous music of amazing weight and substance. They are fabulous at big 3D soundscapes, and have a wonderful full-bodied way with timbre. In horn enclosures like the GRF they also have a lot of dynamic punch, though they can be a bit flatter and more tubby sounding in small volume standard Lancaster boxes, the up side being deeper bass of course.
They do have their critics though and for some people, a slight coarseness detectable in the treble (with Golds and HPDs at least) is enough for them to pass on the whole Tannoy thing. I spotted this issue too (especially coming after the wonderful transparency of Quad ESLs), and put it down to the standard crossover on the Golds, which though being reasonably sophisticated, also uses some fairly cheap mass-produced components (mine were 45 years old to boot) On top of that, Golds aren't always the most intricately-detailed of speakers - musical yes, room-filing, yes, but detail-meisters, they aren't. Not in standard guise anyway. They also aren't the easiest to drive and get the best out of, needing a high damping factor to keep those 15in cones under control and not all boomy and soft, something that in my experience means a good PP amp with strong current delivery. I've never had much joy with SET despite trying half a dozen at least.
I've owned my GRFs for over two years now, picked up from Coco when he was fund-raising for his Voyd Reference. After a year or so, I had new crossovers done, with £400 worth of premium bits, which took away (as I had hoped) most of the slight throatiness, as well increasing overall refinement and detail, though enough of a gap between what I was getting and what I thought the potential was remained to keep me wishing for more.
Last autumn I took the cabs up to Russ Collinson in Barnsley for a full refurb. The cabs were of reasonable design but quite old and were built of thinner ply than would be used today, also there was little bracing on them so they did audibly resonate in places, creating a bit of bloom mainly in the upper bass. Russ went to town in the bracing department, putting multiple solid posts in to reinforce the folded horn section and bonded on sheets of plywood to kill panel resonance. He also got rid of the mahogany stain they were coated in and refinished in natural wood, and as a finishing touch made new grilles with vintage cloth. Now, the old girls look stunning.
http://i1181.photobucket.com/albums/x440/montesquieu61/newtannoy1.jpg
At the same time he also made up new back plates with speaker posts attached hat enabled the two drivers to be directly wired. Suddenly bi-wiring, or even bi-amping (passive or active) was a possibility. Russ also did some lovely wood boxes to take the custom passive crossovers external. A wonderful job, as usual from Russ.
http://i1181.photobucket.com/albums/x440/montesquieu61/newtannoy2.jpg
Ever since coming across this http://www.hilberink.nl/codehans/tannoy86.htm in the Tannoyista site I've fancied a crack at taking them active. A try with a Behringer DCX 2496 was a bit abortive, mainly because of gain and sensitivity issues (in the GRF cabs they are something like 97db). I managed to pick up (from 'gjm' on the Wam) a 2A3 integrated that would be perfect to experiment with as second amp, while the Audion 300b PP amps (back in my posession after four years doing the rounds ...) have volume controls for variable sensitivity - ideal for level matching.
A conversation with Nick Gorham 'lurcher' resulted in him agreeing to build the circuit above - and last week it duly arrived. Very nice work - he had the boards run up after laying out the circuit for op amps, and added buffer stages to handle any impedance nasties with valve amps at input and output.
http://i1181.photobucket.com/albums/x440/montesquieu61/activexover1.jpg
My idea had been to use PP Audion 300bs for decent bass from the 15in cones, and gjm's sweet little 2A3 for the 2in compression driver. People use a mix of drivers in horn systems all the time - why not in a GRF which is a mid-horn and treble horn in the one drive unit?
First listen, though, was disappointing. Bit of fiddling with level matching got the crossover element into the right ball park, but I felt the overall sound was a little course and sub-optimal. Just, not as musical as it should be, not gelling, while detail seemed to have gone a bit blurry and the lovely delicate sound from my valve pre and power amps had gone. It just sounded a bit - well - cheap and solid state. My first thought was that it was an issue with using two different amps - I went looking for a matching pair to try (I'm still hoping to do this at some point!).
I had a discussion with Nick and it transpired that in my conversations with him I'd described this as 'an experiment' - so he'd done his best to keep the cost minimal, including specifying some fairly cheap opamps. He proposed replacing these with OPA604AP. At more than £2 each (times 12) this was four times the cost of the originals, though still hardly a fortune. They duly arrived from RS components and I slotted them in (a new DIY experience for me but I managed ok quite chuffed with myself. 845 SET build here I come :) ). Instant revelation - much much clearer, and the music had returned. Not only that, I was getting the benefits of active coming through loud and clear, less background hash, more low-level detail, better layering in terms of soundstage, more punch and dynamics. It also became obvious that the tweeters were out of phase - duly fixed, this was a big improvement again. I had been worried about putting a bunch of opamps in between my valve pre and power amps, but I needn't have been concerned, this is now a pretty transparent piece of kit and there seems to be no detectable hit on the subtleties of my AN DAC, Nick-designed phono stage or SJS preamp.
I've left it on now for several days to warm up and burn in and it's been getting better and better. As it has settled in, more and more of the valve character has come back, all that lovely ambience and sweetness, without losing the punch and 3D layering. The soundstage has improved over what I had with the passives, as has low-level detail. The only quibble is a very slight comparative loss of focus around the point of the crossover frequency .. something I'm sure that was there before, but was covered up by other hash. And it is only slight, you have to be looking for it. The acid test - listening with the wife - was passed with flying colours. I got the 'now don't you dare go and change anything' lecture that I get occasionally when I do something very right.
Additional refinements remain possible, which I hope might just fix that slight haze around 1khz - improvements to PSU, or (according to Google) I could even spend £25 a pop (times 12 ...) on even fancier opamps. I will definitely go ahead with improvements to the regulation (as recommended by Nick), the opamp option remains on the cards but prohibitive in the short-term I think, I'd need to be really sure any gains would be worth it.
Also the 2A3 is used at max volume, but the Audions level match at around 11 o'clock on the dial - combined with the lowish output on my DAC that means I'm a bit short of gain on the DAC input (no probs on vinyl). So I may still experiment with a higher-sensitivity amp for the 2in driver. And I really ought to twist the arm of JP 'diceman' or Coco to come over with their measuring gear and make sure I've properly matched the levels (by ear it sounds good but measurement would take away any lingering doubt that I have it right).
Of course a valve version of the active crossover circuit would be intriguing as well - now we know it works accurately - and Nick has promised to take a look and see how it might be done ....
But for the moment I'm a happy bunny. I feel a springtime classical music day coming on to canvas the views of the usual suspects as well as some newcomers - will keep the board informed.
http://i1181.photobucket.com/albums/x440/montesquieu61/newtannoy3.jpg