There's been a fair amount of talk about the new 7591s and 7868s not being as robust as the originals, that on a Fisher 400, for example, some changes in bias voltage, grid resistance and screen stoppers are strongly recommended to avoid issues.
I'll second the idea of the early Heathkit amps. I agree that the Williamson circuits need some tweaking for stability, but Dave Gillespie over at Audiokarma has several very valuable threads offering quite easy changes to improve these amps. Personally I would not alter them to a Mullard circuit. The Williamson topology has a wonderful, rich, natural sound that is not present in the Mullard, for all it's convenience. And just changing to the Mullard circuit doesn't mean you don't need to deal with stability, given the high amount of feedback in both circuits.
The W-2 is the prize, IMO, because of the Peerless output transformer. It's as close to the original "Musician's Amplifier" it was based on as you can get short of sourcing the enormous Peerless S-265-Q output transformers used in that amp. Those are pretty much gone--I have two pair, more about that later. Anyway, these are a bit challenging to rebuild due to the tiny chassis, but what lovely amplifiers! The prettiest 20 watts you could want. They still turn up on eBay and are not terribly pricey, mostly because they always need restoration and they are only 20wpc.
The W-3 has the finicky but amazing Acro TO-300 output tranny, and they make terrific amps but most be modified to be stable and sound their best. Gillespie's mods are excellent.
The W-4 is a workhorse and also benefits from Gillespie's mods. The Stancor outputs are the homeliest of the lot. The advantage is that they are relatively cheap, the chassis are spacious, they can be modded to perform very nicely, and you can even substitute a better OPT. Gillespie also offered a modded version with triode-strapped EL34s that is an excellent 10wpc amp.
The W-5s are the problem children. Beautiful amps when they are working, and they are not difficult to find at all. But they are extremely problematic. The original Peerless outputs were notorious for failing, as were the power trannies. The larger Peerless outputs are more robust and usually fine. But if you go to eBay you'll see that soooooo many of these have replacement power trannies and are otherwise junked up because, as popular as they were, they were also very unreliable. But if you can find a good working pair they are worth updated post-haste for the beautiful sound.
The W-6 was the last Williamson amp Heathkit produced, and I know next to nothing about them. They rarely turn up.
It's true that this is probably more of an Audiokarma thread than a DIY thread, but there's an awful lot to be learned by working on these classic amps. In order to build/restore early Williamson designs, I finally taught myself to use a 'scope, and it's one of the biggest advances in my DIY skills I've ever made. Dave Gillespie's Heathkit restoration threads over there are invaluable reading for anyone who aspires to build a feedback amplifier.
Regarding the Peerless S-265-Q, Heyboer makes a very fine copy of this transformer at a very reasonable price. Anyone who wants to build a truly excellent vintage, American-style Williamson amplifer, triode or "ultralinear," can do so and have a truly magnificent 10-20 watts per channel. I wouldn't swap mine for anything, they are that good.
Respectfully disagree. Integrated amps are a great way to understand signal flow and affordable for newbies and can sound great. In fact often times they used the exact same output transformers in their power amp siblings. Even if you want to use or convert them into pure power amps, their power transformers are beefier than the amp version due to higher current demand. And they allow more real estate for mods.
I've butchered and experimented with so many integrated amps that I picked up from flea markets since I was a teenager. The naive me was wearing a purist hat and always tried to bypass the preamp and kept the amp section. Invariably I ruined the sound that I liked already. The resulted sound of course sounded cleaner with improved transparency but the flavor is gone and tone is dried up and I couldn't listen long term. You thought you were gonna eat flavorful cheese instead you were munching on white chalk!
Unfortunately the added preamp presents its own problem too with increased noise and decreased immediacy! Sometimes too much flavor or too much MSG! Integrated amps provided a great way to find a balance for yourself. Sometimes a simple nonfeedback gain stage in the preamp section is all you need to achieve audio nirvana. Bypassing some extra switching and unnecessary functions helps to bring down the noise level but don't go overboard.
Those tone controls can come in handy for some poor recorded songs...unless you are dedicated to only listening to boring audiophile recordings, including most Sheffield Lab junk. The extra record EQ curves are great for mono records don't exist in most modern RIAA-centric preamps...unless you're, again, dedicated to listening to only audiophile recordings. I want to be able to listen to Earl Hines to Art Tatum to Lennie Tristano using the same knobs.
Any decent condition Scott or Fisher or Eico or HK or Heathkit etc, say, with EL84 PP can compete with many modern production in the same power range including the preamp+amp twofers. When they are well restored or intelligently and tastefully modified they can outright smoked some modern stuff with tacky chrome or gold plated chassis. They're not for the ultra-fi or "The Absolute Sound" crowd that's always looking for perfection or the ultimate neutered, ahem, "neutral" amp. What they offer is something musical, compact, and afford. All they asked for is your attention and some elbow grease.
Happy listening!
Last edited:
And that is why I chose to work on and modify one of my tube amps that was originally in a 1963 Magnavox stereo console.
The somewhat popular push-pull 6BQ5 chassis #93-04-20.
Since the power transformer had to also power a seperate tuner/preamp, it was "beefy enough" to be made a stand-alone unit.
In fact, I had to use the now-unused extra 6.3v (for the preamp tubes) filament winding as a primary bucking winding to lower the secondary voltages a bit, and also compensate for the higher line voltages of today's grid.
As is, the amp sounded great, however changing the OPT's to the clone Dynaco Z565's (ultralinear) brought it several notches up in performance, along with the David Gillespie bias mods.
I chose a selectable feedback arrangement which with my Advent Maestro's is never needed, but handy if ever used with other speakers.
I honestly believe it ranks up with those "high end" amps selling for thousands.
The somewhat popular push-pull 6BQ5 chassis #93-04-20.
Since the power transformer had to also power a seperate tuner/preamp, it was "beefy enough" to be made a stand-alone unit.
In fact, I had to use the now-unused extra 6.3v (for the preamp tubes) filament winding as a primary bucking winding to lower the secondary voltages a bit, and also compensate for the higher line voltages of today's grid.
As is, the amp sounded great, however changing the OPT's to the clone Dynaco Z565's (ultralinear) brought it several notches up in performance, along with the David Gillespie bias mods.
I chose a selectable feedback arrangement which with my Advent Maestro's is never needed, but handy if ever used with other speakers.
I honestly believe it ranks up with those "high end" amps selling for thousands.
Thanks to grovergardner for an overview of various Heath amps.
Finding honest specs for OPTs is notoriously difficult, but here are some hard data from the US Naval Research guys that confirm the worth of the Peerless S-265-Q. Surprisingly, the Acro TO-300 is terrible with a spectacular discrepancy between 1W and full power.
PS: just found these Japanese OPT with very good and detailed spec's for USD 120 a piece + shipping. Anyone tried?
Finding honest specs for OPTs is notoriously difficult, but here are some hard data from the US Naval Research guys that confirm the worth of the Peerless S-265-Q. Surprisingly, the Acro TO-300 is terrible with a spectacular discrepancy between 1W and full power.
PS: just found these Japanese OPT with very good and detailed spec's for USD 120 a piece + shipping. Anyone tried?
Last edited:
I wonder why a government did a test of audio amplifiers and made it "restricted security information".... heh.
Thanks for the pdf, will read it with interest.
Thanks for the pdf, will read it with interest.
That discussion spread into those designs using the original tubes also, iirc-
For instance the 330k grid bias resistors (as the Fisher 400 uses) were believed to cause plate runaway during heavy usage.
I use 100k those locations, like the Pio- and 4.7K grid stoppers. There may be more issues I'm not aware of-
Jim
Attachments
Maybe a good next part for discussion could be "What to Not do" when approaching a newly found/given/purchased Golden Age amplifier. I'll begin with:
1) Don't plug it in to see if it lights up. Old hands already know this, but hopefully this thread will reach folk new to the subject. Gear of all kinds, electronics, vehicle engines, people, etc. take poorly to sitting unused for many years, so need to be woke gently. If you've already done this and maybe blown its fuse, STOP before causing more damage.
2) Don't plug it in at all until you can verify that the primary circuit is safe (ish!). A simple modern DVM and a critical-eyed visual exam is first order of business. Is the power cord itself safe to handle? Are there any capacitors in the primary circuit (they must all be removed or (rarely) replaced) ? Any signs of arcing or charring around fuse holders, switches, mains outlets, etc. ? With the DVM measure resistance of all primary wiring to chassis. If not very, very high, do not procede until that's fixed.
3) The next step is to re-awaken the beast slowly, and for this we need a Variac (ideally) or at least a "lamp limiter", an incandescent lamp of voltage appropriate to your mains (better get some while they're still available) wired in series with the primary circuit. I did this kind of thing half a century ago, as a poor young person, but today anyone can afford, and really needs, a Variac. Wouldn't hurt to have both.
The procedure goes generally: to remove all the valves and to substitute modern 1N4007s for the rectifiers (if you plan to do a lot of this, you can make, or buy, octal socket semi-con rectifiers in the common 5T pinout, or just tack solder 'em on the rectifier sockets' pins). This is to avoid beating up on the valve rectifiers during the procedure, as they're getting rare in some popular types.
Then, connect a DCVM across the first B+ capacitor, and if at all possible, a 'scope. Here we'll observe the electrolytic B+ capacitors "re-forming" as voltage is SLOWLY applied. This could be 24 hours and not be unusual. In many ways the slower the better. With a 'scope (did I mention you need a scope? You need a scope, or at least will in your journey, doesn't have to be fancy, could be a USB plugin for this gig just fine) you'll observe that as you slightly increase mains voltage, ripple (as seen on the scope) first jumps up, then slowly decreases, as the capacitor re-forms (some strange battery-ish chemical process). Without a scope this same slow procedure needs to happen, but it's much harder to see your progress in the decreasing brightness of the limiter lamps. A good selection of wattages helps.
Remember that this process has almost no bleed for the B+, so Watch that DCVM! before any handling, and then only one hand. Live to ride and ride to live.
4) Don't give up. The first part is the hardest, and also the most important. After this initial baptism of fire it's all downhill.
Enough from me. All good fortune,
Chris
1) Don't plug it in to see if it lights up. Old hands already know this, but hopefully this thread will reach folk new to the subject. Gear of all kinds, electronics, vehicle engines, people, etc. take poorly to sitting unused for many years, so need to be woke gently. If you've already done this and maybe blown its fuse, STOP before causing more damage.
2) Don't plug it in at all until you can verify that the primary circuit is safe (ish!). A simple modern DVM and a critical-eyed visual exam is first order of business. Is the power cord itself safe to handle? Are there any capacitors in the primary circuit (they must all be removed or (rarely) replaced) ? Any signs of arcing or charring around fuse holders, switches, mains outlets, etc. ? With the DVM measure resistance of all primary wiring to chassis. If not very, very high, do not procede until that's fixed.
3) The next step is to re-awaken the beast slowly, and for this we need a Variac (ideally) or at least a "lamp limiter", an incandescent lamp of voltage appropriate to your mains (better get some while they're still available) wired in series with the primary circuit. I did this kind of thing half a century ago, as a poor young person, but today anyone can afford, and really needs, a Variac. Wouldn't hurt to have both.
The procedure goes generally: to remove all the valves and to substitute modern 1N4007s for the rectifiers (if you plan to do a lot of this, you can make, or buy, octal socket semi-con rectifiers in the common 5T pinout, or just tack solder 'em on the rectifier sockets' pins). This is to avoid beating up on the valve rectifiers during the procedure, as they're getting rare in some popular types.
Then, connect a DCVM across the first B+ capacitor, and if at all possible, a 'scope. Here we'll observe the electrolytic B+ capacitors "re-forming" as voltage is SLOWLY applied. This could be 24 hours and not be unusual. In many ways the slower the better. With a 'scope (did I mention you need a scope? You need a scope, or at least will in your journey, doesn't have to be fancy, could be a USB plugin for this gig just fine) you'll observe that as you slightly increase mains voltage, ripple (as seen on the scope) first jumps up, then slowly decreases, as the capacitor re-forms (some strange battery-ish chemical process). Without a scope this same slow procedure needs to happen, but it's much harder to see your progress in the decreasing brightness of the limiter lamps. A good selection of wattages helps.
Remember that this process has almost no bleed for the B+, so Watch that DCVM! before any handling, and then only one hand. Live to ride and ride to live.
4) Don't give up. The first part is the hardest, and also the most important. After this initial baptism of fire it's all downhill.
Enough from me. All good fortune,
Chris
Last edited:
An adjustable DC power supply could be an alternative to variac's or lightbulbs. Connecting to the B+ cap and slowly
increasing voltage while watching DC current. When holding voltage steady there should be no current flowing unless bleeder
resistors are incorporated.
increasing voltage while watching DC current. When holding voltage steady there should be no current flowing unless bleeder
resistors are incorporated.
I'd say be aware that the AC line voltage is different now than when these were originally designed. WOTs schematic shows compensation for this. It's given me trouble, when put on top of an amplifier run to the 9's already by design, using tubes that modern equivalents cant match. On certain golden era examples, this will need to be addressed one way or another and it's a pain that it's not simply plug and play...
Many old amps had multiple taps for the input supply, especially here in Europe, so often there is a 245V option, in which case it is slightly reducing the design voltage in the amp. The complication of changing supply voltages mainly seems to impact the American amplifiers, and there it is quite a significant %-age change. One option is adding a high wattage resistor, but in this day and age losing excess electricity as heat does not feel like a sustainable solution.
I have a Japanese amp and use an auto-transformer for that. I suppose a suitable supply has to be part of the baggage, alongside the lamp limiter and variac.
I have a Japanese amp and use an auto-transformer for that. I suppose a suitable supply has to be part of the baggage, alongside the lamp limiter and variac.
Not to forget EL503/EL520, the most recent European (Philips) development of AF power tubes. Extremely unique, scarce and expensive. No replacement possible. Avoid daily use of units featuring them. Also ECC808's, an improved version of ECC83/12AX7A with another pinout. PTP wired units may be rewired for ECC83's, but there's no viable option with PCB's.
Please don't forget that the EAR509 was not a UL design. It rather resembles McIntosh's/Gow's Unity Coupling with a trifilarly wound OT primary. The EL509's are running as true pentodes (= constant screed grid voltage) in class AB or B.
Best regards!
I wouldn't go anywhere near a real EAR 509.
I totally respect Tim de Paravicini (RIP) as a designer, but his "exotic" ideas don't play nice with western production, and his amps had a preference for self-immolation, ever since the TVA days. For some reasons, his previous designs while at Luxman, Japan, seem to stand the test of time. See what Patrick Turner had to say, and check your fire extinguisher.
I totally respect Tim de Paravicini (RIP) as a designer, but his "exotic" ideas don't play nice with western production, and his amps had a preference for self-immolation, ever since the TVA days. For some reasons, his previous designs while at Luxman, Japan, seem to stand the test of time. See what Patrick Turner had to say, and check your fire extinguisher.
Interesting! Until this I literally would have licked my fingers to get hands on a pair of EAR 509's. As Patrick Turner says that the original OT's would get warm, I suppose he ran them cose to or at max. power?
There actually was a four final valve version of this amp, the EAR 549. I found a review of it in an old German HiFi magazine.
Best regards!
There actually was a four final valve version of this amp, the EAR 549. I found a review of it in an old German HiFi magazine.
Best regards!
I just find it amusing.
Turner said: "I had two of these monoblocs brought to me by a customer who probably paid little for them. One had nearly burned down a house, and was badly damaged by smoke and flames under the chassis."
I had a totally opposite experience: I once bought a Quad II for very little money, salvaged from a warehouse that burned down. It had those dark brown chars above the chassis, as if coming from a good BBQ. I cleaned it up and it sang!
Turner said: "I had two of these monoblocs brought to me by a customer who probably paid little for them. One had nearly burned down a house, and was badly damaged by smoke and flames under the chassis."
I had a totally opposite experience: I once bought a Quad II for very little money, salvaged from a warehouse that burned down. It had those dark brown chars above the chassis, as if coming from a good BBQ. I cleaned it up and it sang!
Anybody around here actually retrofit an actual Fisher amp with a comparable sweep tube to avoid 7868 troubles and expense? Or will the power trafo just burn up from the extra amp of heater current? Hot doesn’t begin to describe it under normal operation….. I had one that DID burn up and a new trafo wasn’t an option at the time so it got parted out. Used the output iron 30 years later with vertical output pentodes, which with 125V screen pretty much work as a 7591/7868.
I knew a guy who used to upgrade the 500c with Bendix 6384s. The extra heater current didn't seem to be a problem. They sounded fantastic.
I haven't offered yet but I have a demolished A300 which is no longer worth restoring.
it is complete (power transformers, output transformers, tubes, front panel, volume but has almost no bass treble and switch functions) which makes it a very good candidate for a 21st century modification, if anyone one wants to propose modifications on it, I can implement them, I have more than enough components in stock.
it is complete (power transformers, output transformers, tubes, front panel, volume but has almost no bass treble and switch functions) which makes it a very good candidate for a 21st century modification, if anyone one wants to propose modifications on it, I can implement them, I have more than enough components in stock.