| Posted by Zekk | April 12, 2015 - 09:50 pm |
| Hi Dave, I find no words to describe adequately how delighted I'm, after following your modded schematic. Have gone thru 4 builts with the same massive potted A-431 but none came close! However, the bass was a bit heavy. Wired the feedback to a switch and have found no effect. I believes with minor tweating, it'll sounds fabulous! I'm using a #26 linestage . Nevertheless, I'm very, very happy with it . Dave, thank you very much for contributing this HF-60 MOD to the communityof diyers. With Best Regards. Zekk |
| Posted by Dave | December 05, 2013 - 04:42 pm |
| Hi Al -- There is no doubt that the EF86 could be triode connected, the feedback reduced, and a reasonable input sensitivity then still maintained. As for the feedback level though, it is always a trade off. If you reduce it to target certain performance aspects, then others will be degraded. If you increase it to address degraded aspects, then others will be compromised, and so on. I believe that a definite point of feedback level can be reached where upon any further feedback provides little if any useful return, and can act to degrade sound quality. But I also believe that too little feedback is harmful to performance and sound quality as well. The correct level, applied correctly, will always yield a maximum of benefits, with a minimum of collateral damage in the process. Regarding the production of TIM, I think any effort spent trying to isolate any real effects from this form a distortion in classic vacuum tube amplifier designs would be all but impossible -- if it is even produced at all. My experience is that it takes feedback levels far in excess of the typically maximum 30 db or so used in classic designs for TIM to become of any significant concern. Ultimately, the intent of this article never was to address the HF-60 as a complete redesign, but simply to address properly operating the amplifier with 6550 tubes, rather than the original EL34 tubes. Simply changing the tubes and modify the bias appropriately will actually degrade performance, where as done properly, performance is superior to that of the original design. Thanks for your interest! Dave |
| Posted by Al | December 05, 2013 - 09:20 am |
| Do you have some comment about my post dated December 4, 2013 ? Regards |
| Posted by Al | December 04, 2013 - 08:38 am |
| I think it is possible to use a EF86 triode connected and get an imput drive level approach 1 VAC RMS reducing the NFB from 20dB to 12dB. Besides less NFB improve the TIM. Best regards |
| Posted by Dave | August 01, 2013 - 10:46 pm |
| A triode connected EF86 can in fact be used, but as my previous post explained, it will drive the input requirements of the design up significantly. The primary intention of this effort was in providing a notable improvement in output stage performance, while maintaining the original input sensitivity specification. Thanks for the interest! Dave |
| Posted by Al | August 01, 2013 - 03:00 pm |
| why do not you set an tiode conected EF86 as imput stage? Best regards |