|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
166.137.12.55
In Reply to: RE: Eico HF-14 build from scratch posted by banpuku on March 25, 2015 at 06:11:13
...but...did you look over the published specs carefully? Noteworthy are power vs distortion vs frequency ratings and feedback level (see below). Just an opinion but I think your Quads would mercilessly reveal the little EICO's shortcomings.
:
:
:
:
Follow Ups:
Steve,
Elliano, LaFevre, Heyboer, and Purvine all make good "iron". Would Williamson style circuitry using KT88 "finals" and well above average O/P transformers be suitable for mating to the Quad 63 speakers? The OP has stated monoblocks. Otherwise, I'd say acquire and refurb a H/K Cit. 2.
Combining modest cost AnTek power trafos with "fixed" bias seems to be reasonable.
Yeah, yeah, toroids are wide bandwidth. So, put ferrite beads on the primary wire connections to kill crud riding on the AC mains.
Eli D.
My primary concern with the HF 14 is power capability. Additionally, recent experience with a few vintage Williamson designs really points out the demands they place on the OPT and feedback network parameters, mostly resulting from the number of stages contained within the OA feedback loop. Just an opinion but......while I have no doubt that what you propose is possible, I don't believe it would be an easy exercise because there aren't many higher power Williamson amp designs out there to duplicate. IOW, not a great first scratch DIY project due to the number of unknowns involved + need for test equipment nec to "tune" the feedback network.
Again, just an opinion, but if I was in the OP's situation, I'd be considering a clone of one of the Dyna monoblocks such as MKII, MKIII or MKIV. All are proven designs using readily avail. reproduction iron and current production tubes (using the aftermarket boards). The aftermarket circuits can be done p to p if boards aren't acceptable. Another potential advantage of these clones is that the circuits are simple and simple circuits seem to compliment the Quad sound.
Would a very honest 30+ WPC be enough for those Quad 'stats? If so, an "El Cheapo Grande" of 12AT7 splitter/driver and PP 7591 "finals" might be the thing to do. The 2 stage design doesn't have stability issues and it's possible to brute force the phase compensation. So, only a decent multi-meter is needed by the builder. Obviously, a more complete instrumentation setup HELPS.FWIW, Edcor's CXPP60-MS-6.6K O/P "iron" seems quite suitable and modestly priced. Plenty of magnetic headroom here. So, full bass extension can be achieved, without worrying about core saturation from the LF error correction signal.
Eli D.
Edits: 03/25/15
Eli,
I recall mention of the Grande before but I don't recall your posting a schematic. Is there one available?
The Grande idea follows what Fisher, Scott, and Sherwood did. Small signal circuitry that works with "12" W. tubes also works with 7591s. ;> ) Those folks were both good HIFI people and smart businessmen. :> D The Scott 299A, 299B, 299C, and 299D pretty well show how it's done.
Use the original "El Cheapo" schematic and increase the B+ rail voltage.
In another "leg" of this thread, I went into some detail.
"Hit" Frank Philipse site and download the TungSol 7591 data sheet. Pay close attention to the conditions given for a UL mode AB1 PP O/P pair.
Eli D.
30WPC should work nicely given the OP's stated listening conditions/preferences. So an El Cheapo Grande is another possibility.
I recall you (or Jim?) referring to the "Grande" recently but I didn't realize it had been "released to production".
The Grande idea has been out and about for quite some time. Putting together something "official" has not been done.
Compared with the "12" W. tube "El Cheapo", the "taller" B+ rail needed by the 7591 allows the 'T7 plate load to increase to approx. 80 Kohms. 33 K carbon film and 50 K inductive wirewound parts in series could be fine.
Combination biased "finals" would be employed. "Stand" the tied together cathodes of the O/P tube pair on a 470 muF./100 ohm bias network. A single "idle" current set trim pot. is fine and the 100 ohm resistor is a very convenient test point. By making part of the total bias voltage self generated, protection against runaway, which the current production EH 7591 is known to occasionally do, is provided and 330 Kohm grid to ground resistors will be safe. Keeping the 7591 grid to ground resistor value up prevents splitter/driver gain from becoming insufficient.
For monoblocks, AnTek AS-1T300 power trafos should be quite adequate for B+ and heater power.
Give me some time and I'll concoct a scheme for the requisite negative voltages.
Eli D.
Two negative rails are needed: the 1st is "short"/instant on (for bias/C-) and the 2nd is "taller"/slow starting (for LTP power/B-).
Start with an AnTek AN-0240 power trafo.
Phase the 2X 40 VAC windings up and connect them in series. Hybrid bridge rectify the composite using a MBR20200CT common cathode twin Schottky diode as the connection to ground and a 6AL5 forming the connection to the B- supply filter. This bridge is like that in the original "El Cheapo", but the rail voltage is a bit lower (NBD).
Connect the cathode of a MBR1100G Schottky diode to the junction of the 40 VAC windings. Connect the anode of the MBR1100G to the C- supply filter.
Both rails are full wave rectified, with the MBR20200CT providing the path to ground. The "tall" rail is bridge rectified and the "short" rail is FWCT rectified. Zero switching noise too. ;> )
Thank Pete Millett for the 2 rails from 1 CT winding concept.
Eli D.
Not to take anything away from Pete, but isn't this the "economy" supply that appears in technical manuals going back many years? It's a great way to get two different voltages of the same polarity, although it was traditionally employed for B+.
--------------------------
Buy Chinese. Bury freedom.
I also believe the EL Cheapo 7591 is a better amplifier for the application. The 12AT7 tube is a brilliant choice IMO.
At times vintage 7868s are less expensive and the tube insides is about identical to the 7591. 7868s require a different socket from the 7591.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: