Yes, but it may be ample. On the up-swing the bootstrap cap can boost well above the + rail. On the down swing, the tail of the diff-pair has its own lower rail, -36V re: the -28V fed to the power stage.
It seems awful shy by modern standards, but transistors were expensive then.
Based on the simulation I created and used to guide the entire restoration, I could hit around 70W into 8-ohms or ~+12V and ~-12V at the outputs simultaneously before the boosted voltage would touch the Vceo of the Dual NPN differential pair. That assumes an ideal power supply, however.
You will get higher voltage if you measure the individual output transistors, but not at the outputs.
When the amplifier has an output of +12V at one output, it simultaneously has -12V on the other output which gives you 24V total.
Related thread is here :
https://www.diyaudio.com/community/threads/hadley-622-amplifier-circuit.124757/
https://www.diyaudio.com/community/threads/hadley-622-amplifier-circuit.124757/
The member djk made several posts that referenced the 622. He seemed to know what it was better than anyone except Dawson Hadley himself. It's too bad he's no longer with us.
Bongiorno, on the other hand, seemed to have little to no clue what he was talking about in any of the emails quoted relating to this amplifier in other forums online.
Bongiorno, on the other hand, seemed to have little to no clue what he was talking about in any of the emails quoted relating to this amplifier in other forums online.
I saw a nice ad for a Hadley 622 :
https://reverb.com/item/23108638-vintage-hadley-622-power-amplifier-very-rare-classic
The text is instructive and is shown here with permission of the seller :
The semi-legendary Hadley Laboratories, and its founder and namesake, Dawson Hadley,
made high end amps and pre-amps, in very low numbers, for a few years in the 1960s. They
were in business from 1963 to 1969, and honored their lifetime warranty for an additional 5
years after the company shut its doorsHadley Labs released only 3 products in their short existence,
all 3 in their first 2 years, and all 3 legendary classics of sophisticated audiophile equipment! Audio
Magazine’s annual equipment directory in August, 1963 listed two products from Hadley Labs: the
“Model 601 stereo amplifier…..a dual amplifier designed for the discriminating listener.” The
architecture of the 601 seems similar to that of the Marantz 8 and 8b. The 601 was a tube amp.
The second Hadley product shown in Audio was the “Model 621 Solid-State preamp”, whose
faceplate was somewhat reminiscent of the Marantz 7. The 621 had an unusual feature for the time:
separate tone-controls for each channel.
The March, 1964 issue of Hi-Fi/Stereo Review, the predecessor of Stereo Review,featured an
uncommonly-gushy review of the 601 power amp by Julian Hirsch, known in later years for his
insistence that all “properly designed” amps sound alike. The review begins: “There are a very few
power amplifiers whose performance is so outstanding that they must be put into a special category
for the most discriminating users. The Hadley 601 stereo power amplifier is a recent addition to this
group.”
Hirsch went on to describe how the 601 nearly doubled its rated 40-watt output before exhibiting
significant distortion, and was stable under any load.
Hirsch’s conclusion: “When listening to the Hadley 601, I experienced the same sense of total ease
and almost limitless power reserve I associate with the two or three finest amplifiers I have ever used.”
Audio listed both pieces at $319.50, when the Marantz 7 and 8 were each priced at $264.00—so this was
a rarified realm at the time. Compensating for inflation, each Hadley piece would cost close to $3000 today.
The 601 was the first product by Hadley Labs. The schematic is for Revision C, and is dated Nov 26, 1963.
Sales of the 621 Pre-Amp began in late 1964, with schematics dated in Oct. 1963. The 622 Amp schematics
are dated Oct. 1965. The solid-state 622 amplifier was the final product from Hadley Labs, and that model
is what I have for sale today.
Dawson Hadley made a decision to go with the new, at the time, solid state technology, and the Hadley solid
state gear was known for its warm tube-like sound and superb build quality. The 622 is considered to be one
of the best solid state amps ever made, often favorably compared to the early solid state McIntosh amplifiers
and pre-amps. Words like, “magic”, “irreplaceable”, and “classic” are in the usual lexicon when describing
Hadley audio gear.
Total production run of the 622, over its lifetime, was only around 200 pieces, making this amplifier extremely
rare and hard to find. Hadley ceased selling all their products in 1969. The 622, and all Hadley gear, was
originally sold with a lifetime warranty, which, as I mentioned, the company honored for 5 years after they
went out of business. Dawson Hadley suffered from Parkinson’s Disease, and was unable to continue after that.
The legendary Jim Bongiorno was the final engineer on the 622, who ushered it into production. * See below for
bio of Bongiorno. Truly an audio giant!
This amplifier has been tested and both channels work. However, it may, and probably does, need some routine
maintenance to perform optimally. After all, it is almost half a century old, and has not been used in a pretty long
time, except for today, when I plugged it in and tested it. I don’t know if it has capacitors or resistors or other parts
that routinely wear out over time, but it probably does, and those types of things, while they may not stop an amp
from working, can impact the sound quality, if they need replacement. If you are going to actually use this as an
amplifier, and not just a super collectible extremely rare piece of audio history, then you’ll probably want to have
a qualified technician go over it and tune it up. If I was going to use it, I would do that, but it works now, as it stands,
and can only improve if tweaked. Unfortunately, the guy who really knew these amps, Jim Bongiorno, died a few
years ago, so you may have to do a little research to come up with a qualified person, because I wouldn’t let just
anyone work on this beauty. If you need help finding a competent technician, I can help you find one near your
home.
It’s in pretty good shape, cosmetically, but it does have a few small, stable rust spots on the chassis (see pics).
It’s got 2 on the front, and 1 each on the top and bottom. Also has a small sticker stuck to the top, about 1 inch
square, for the Asbury Park Press (visible in pics).
These amps were 40 watts per channel. Bongiorno used to update them to 80, when he was still working on them.
This one has a very low serial number, 1032. Since they only made 200, I’m guessing they started at 1001, so this
is probably the 32nd 622 that rolled out of the Hadley factory, which at the time, that early on, may have still been
Dawson Hadley’s garage! LOL.
It measures 12x10x5, and is quite heavy, 25 pounds. The inputs and outputs and protective side walls extend the
10 inch side by one inch, making the footprint of the unit 12x11x5.
I’m listing this for local pick up only, but I will ship, within certain parameters. It’s too heavy to ship internationally,
I think. The cost would, undoubtedly, be prohibitive. If you’re willing to pay for it, I’m willing to do it, but it ain’t
gonna be cheap, I don’t think. If you want to ship internationally, send me your address, and I’ll get you a quote.
I will ship within the United States, however, shipping is a little bit complicated. I’ll explain. The amp just fits, in a
USPS large flat rate box. It does not allow the usual 3 inches all around, that I prefer to utilize when shipping a fragile,
heavy item like this, but, it does save a considerable amount of money on shipping. In a flat rate box, I could ship
anywhere in the United States, including Alaska and Hawaii, with insurance, for $30. In a properly sized box, with
proper padding, the rate jumps to $100! If I was buying it, I might spend the extra $70, but that would be a game
day decision. If money was tight, I might gamble on the flat rate. In the flat rate box, there would be a very small
area for padding, but I think I could protect the rear control panel, and there’s room for very thin padding above
and below. The only unprotected area would be the front, but the protection on the rest would not be up to my usual
standards. I would also COMPLETELY reinforce the box with clear packing tape all around, probably in 2 or 3 layers.
This would strengthen the box quite a bit. I am known to be a very thorough packer and shipper, leaving nothing to
chance (Murphy’s Law), and it is solid state, and the only exposed parts that stick out from the chassis are in the rear
control panel, so it might be okay. The choice, and the risk, would be yours. Normally, if an item is damaged in shipment,
I refund the buyer’s money very quickly, and then open a case with the Post Office. That’s the normal protocol, and that’s
what I would do if you paid for the proper shipping, and it was damaged in shipment. In this case, if you choose the flat
rate box, it would be incumbent on you, the buyer, to open the case with the Post Office. It’s a very easy process. I could
walk you through it. I might even be willing to do it myself, under certain circumstances. Let’s hope we don’t have to do
that, but if we do, we’ll talk. USPS is very good about paying for damaged items, so that’s not usually a worry, unless
someone doesn’t follow protocol, like throwing away the packaging, not taking pictures, or something like that.
So, to sum up the shipping in a nutshell - International: send your address for a quote. USA: Flat rate, insured - $30
(you deal with USPS if damaged). Priority non-flat rate, insured (normal conditions apply, if damaged): $100.
Any questions, please ask. I’ll be happy to respond quickly.
Thanks for looking (and reading. LOL).
*James Bongiorno Bio
Tracing roots through Marantz, Dynaco, SAE, SUMO Electric, Great American Sound Company (GAS), Constellation Audio
and his last self-started company, Spread Spectrum Technologies (SST), James Bongiorno's imprint on the audio industry
was as wide as it was deep. Recognized as one of the "pioneers of audio" by The Absolute Sound magazine and one of the
preeminent audio designers of our time, Mr. Bongiorno is credited with designing the first full dual-differential complementary
amplifier topology that is the basis for nearly all modern solid-state amplifiers today. Certainly the name “Ampzilla” is one
many audiophiles around the world fondly remember as one of the most popular amplifiers ever developed. James' circuit
topologies are the stuff of legend, lore and imitation. His deep love for the music drove his desires to create timeless works
of audio art, with great success.
https://reverb.com/item/23108638-vintage-hadley-622-power-amplifier-very-rare-classic
The text is instructive and is shown here with permission of the seller :
The semi-legendary Hadley Laboratories, and its founder and namesake, Dawson Hadley,
made high end amps and pre-amps, in very low numbers, for a few years in the 1960s. They
were in business from 1963 to 1969, and honored their lifetime warranty for an additional 5
years after the company shut its doorsHadley Labs released only 3 products in their short existence,
all 3 in their first 2 years, and all 3 legendary classics of sophisticated audiophile equipment! Audio
Magazine’s annual equipment directory in August, 1963 listed two products from Hadley Labs: the
“Model 601 stereo amplifier…..a dual amplifier designed for the discriminating listener.” The
architecture of the 601 seems similar to that of the Marantz 8 and 8b. The 601 was a tube amp.
The second Hadley product shown in Audio was the “Model 621 Solid-State preamp”, whose
faceplate was somewhat reminiscent of the Marantz 7. The 621 had an unusual feature for the time:
separate tone-controls for each channel.
The March, 1964 issue of Hi-Fi/Stereo Review, the predecessor of Stereo Review,featured an
uncommonly-gushy review of the 601 power amp by Julian Hirsch, known in later years for his
insistence that all “properly designed” amps sound alike. The review begins: “There are a very few
power amplifiers whose performance is so outstanding that they must be put into a special category
for the most discriminating users. The Hadley 601 stereo power amplifier is a recent addition to this
group.”
Hirsch went on to describe how the 601 nearly doubled its rated 40-watt output before exhibiting
significant distortion, and was stable under any load.
Hirsch’s conclusion: “When listening to the Hadley 601, I experienced the same sense of total ease
and almost limitless power reserve I associate with the two or three finest amplifiers I have ever used.”
Audio listed both pieces at $319.50, when the Marantz 7 and 8 were each priced at $264.00—so this was
a rarified realm at the time. Compensating for inflation, each Hadley piece would cost close to $3000 today.
The 601 was the first product by Hadley Labs. The schematic is for Revision C, and is dated Nov 26, 1963.
Sales of the 621 Pre-Amp began in late 1964, with schematics dated in Oct. 1963. The 622 Amp schematics
are dated Oct. 1965. The solid-state 622 amplifier was the final product from Hadley Labs, and that model
is what I have for sale today.
Dawson Hadley made a decision to go with the new, at the time, solid state technology, and the Hadley solid
state gear was known for its warm tube-like sound and superb build quality. The 622 is considered to be one
of the best solid state amps ever made, often favorably compared to the early solid state McIntosh amplifiers
and pre-amps. Words like, “magic”, “irreplaceable”, and “classic” are in the usual lexicon when describing
Hadley audio gear.
Total production run of the 622, over its lifetime, was only around 200 pieces, making this amplifier extremely
rare and hard to find. Hadley ceased selling all their products in 1969. The 622, and all Hadley gear, was
originally sold with a lifetime warranty, which, as I mentioned, the company honored for 5 years after they
went out of business. Dawson Hadley suffered from Parkinson’s Disease, and was unable to continue after that.
The legendary Jim Bongiorno was the final engineer on the 622, who ushered it into production. * See below for
bio of Bongiorno. Truly an audio giant!
This amplifier has been tested and both channels work. However, it may, and probably does, need some routine
maintenance to perform optimally. After all, it is almost half a century old, and has not been used in a pretty long
time, except for today, when I plugged it in and tested it. I don’t know if it has capacitors or resistors or other parts
that routinely wear out over time, but it probably does, and those types of things, while they may not stop an amp
from working, can impact the sound quality, if they need replacement. If you are going to actually use this as an
amplifier, and not just a super collectible extremely rare piece of audio history, then you’ll probably want to have
a qualified technician go over it and tune it up. If I was going to use it, I would do that, but it works now, as it stands,
and can only improve if tweaked. Unfortunately, the guy who really knew these amps, Jim Bongiorno, died a few
years ago, so you may have to do a little research to come up with a qualified person, because I wouldn’t let just
anyone work on this beauty. If you need help finding a competent technician, I can help you find one near your
home.
It’s in pretty good shape, cosmetically, but it does have a few small, stable rust spots on the chassis (see pics).
It’s got 2 on the front, and 1 each on the top and bottom. Also has a small sticker stuck to the top, about 1 inch
square, for the Asbury Park Press (visible in pics).
These amps were 40 watts per channel. Bongiorno used to update them to 80, when he was still working on them.
This one has a very low serial number, 1032. Since they only made 200, I’m guessing they started at 1001, so this
is probably the 32nd 622 that rolled out of the Hadley factory, which at the time, that early on, may have still been
Dawson Hadley’s garage! LOL.
It measures 12x10x5, and is quite heavy, 25 pounds. The inputs and outputs and protective side walls extend the
10 inch side by one inch, making the footprint of the unit 12x11x5.
I’m listing this for local pick up only, but I will ship, within certain parameters. It’s too heavy to ship internationally,
I think. The cost would, undoubtedly, be prohibitive. If you’re willing to pay for it, I’m willing to do it, but it ain’t
gonna be cheap, I don’t think. If you want to ship internationally, send me your address, and I’ll get you a quote.
I will ship within the United States, however, shipping is a little bit complicated. I’ll explain. The amp just fits, in a
USPS large flat rate box. It does not allow the usual 3 inches all around, that I prefer to utilize when shipping a fragile,
heavy item like this, but, it does save a considerable amount of money on shipping. In a flat rate box, I could ship
anywhere in the United States, including Alaska and Hawaii, with insurance, for $30. In a properly sized box, with
proper padding, the rate jumps to $100! If I was buying it, I might spend the extra $70, but that would be a game
day decision. If money was tight, I might gamble on the flat rate. In the flat rate box, there would be a very small
area for padding, but I think I could protect the rear control panel, and there’s room for very thin padding above
and below. The only unprotected area would be the front, but the protection on the rest would not be up to my usual
standards. I would also COMPLETELY reinforce the box with clear packing tape all around, probably in 2 or 3 layers.
This would strengthen the box quite a bit. I am known to be a very thorough packer and shipper, leaving nothing to
chance (Murphy’s Law), and it is solid state, and the only exposed parts that stick out from the chassis are in the rear
control panel, so it might be okay. The choice, and the risk, would be yours. Normally, if an item is damaged in shipment,
I refund the buyer’s money very quickly, and then open a case with the Post Office. That’s the normal protocol, and that’s
what I would do if you paid for the proper shipping, and it was damaged in shipment. In this case, if you choose the flat
rate box, it would be incumbent on you, the buyer, to open the case with the Post Office. It’s a very easy process. I could
walk you through it. I might even be willing to do it myself, under certain circumstances. Let’s hope we don’t have to do
that, but if we do, we’ll talk. USPS is very good about paying for damaged items, so that’s not usually a worry, unless
someone doesn’t follow protocol, like throwing away the packaging, not taking pictures, or something like that.
So, to sum up the shipping in a nutshell - International: send your address for a quote. USA: Flat rate, insured - $30
(you deal with USPS if damaged). Priority non-flat rate, insured (normal conditions apply, if damaged): $100.
Any questions, please ask. I’ll be happy to respond quickly.
Thanks for looking (and reading. LOL).
*James Bongiorno Bio
Tracing roots through Marantz, Dynaco, SAE, SUMO Electric, Great American Sound Company (GAS), Constellation Audio
and his last self-started company, Spread Spectrum Technologies (SST), James Bongiorno's imprint on the audio industry
was as wide as it was deep. Recognized as one of the "pioneers of audio" by The Absolute Sound magazine and one of the
preeminent audio designers of our time, Mr. Bongiorno is credited with designing the first full dual-differential complementary
amplifier topology that is the basis for nearly all modern solid-state amplifiers today. Certainly the name “Ampzilla” is one
many audiophiles around the world fondly remember as one of the most popular amplifiers ever developed. James' circuit
topologies are the stuff of legend, lore and imitation. His deep love for the music drove his desires to create timeless works
of audio art, with great success.
I'm having some issues with the amplifier that I'm working through, but while I was troubleshooting, I made some observations I had not previously made.
Below is one of the boards from my amplifier and a board from a later serial number model. I believe all of the resistors to be original these particular units.
Why would Hadley go from using early film resistors for most everything (except for the 56 and 15k ohm resistors) back to using two additional carbon comp resistors for the 1Ks in the top left and one toward the middle of the board? I've seen carbon comps used for the 10Ks below the 1Ks and the larger 1W 1K resistors the right of the 10Ks. These sets of resistors as well as the transistors at the bottom are both related to the current limiting circuit. Assuming this was not because of cost, would there be any reason to use carbon comps here?
Below is one of the boards from my amplifier and a board from a later serial number model. I believe all of the resistors to be original these particular units.
Why would Hadley go from using early film resistors for most everything (except for the 56 and 15k ohm resistors) back to using two additional carbon comp resistors for the 1Ks in the top left and one toward the middle of the board? I've seen carbon comps used for the 10Ks below the 1Ks and the larger 1W 1K resistors the right of the 10Ks. These sets of resistors as well as the transistors at the bottom are both related to the current limiting circuit. Assuming this was not because of cost, would there be any reason to use carbon comps here?
The carbon comp 1K resistors in the top left are for the DC protection circuit. They're likely carbon comp to soften any sudden rush of current if the protection circuit were to become active.
There were a couple of carbon comps in the power supply that seem to be for a sort of "soft-start" or current limiting at power on.
There were a couple of carbon comps in the power supply that seem to be for a sort of "soft-start" or current limiting at power on.
Below is the abstract for the x-design patent with my notes relating to the 622C and the colored sections in the table below:
AMPLIFIER WITH GAIN STAGES COUPLED FOR DIFFERENTIAL ERROR CORRECTION
Abstract: An amplifier circuit having two identical inverting folded-cascode amplifier stages coupled together through a coupling resistor at the positive terminals of the respective stages. Each stage includes an input gain transistor whose gate forms the negative input terminal of the amplifier stage coupled to receive an input signal through an input resistor, whose source forms the positive input terminal of the amplifier stage, and whose drain is connected to the source of a level-shifting cascode transistor. Both transistors are supplied current from a constant current source also connected to the source of the level-shifting transistor. The drain of the level-shifting transistor forms an output terminal of the amplifier stage. Each stage feeds its amplifier output negatively to the positive input of the other stage, while at the same time feeding its distortion and noise contribution positively, via input gain transistor conduction, to the other stage. Balanced amplified outputs are produced from either balanced or unbalanced inputs, and distortion and noise components are produced in common on both outputs for differential cancellation. Complementary-symmetry and power amplifier versions of the basic single-ended line-level amplifier are also provided.
Below I've redrawn the Hadley 622C schematic into something that clearly shows the beautiful symmetry that I don't think Dawson Hadley wanted to show the world in 1965. One thing is for sure, he knew what he had. Now did he come up with it or did it come from elsewhere? Any tube circuits that look like this?
As one can see, there isn't much left in the abstract not addressed and simply put, the Hadley 622C exhibits the same (if not more) of the "super-symmetry" the x-design does and was capable of doing this 28+ years before the patent.
An example of this behavior: if you inject noise into the inputs, the amplifier will not only cancel the noise and eliminate DC at the outputs completely (to the extent your components are matched and can handle the voltage), but it'll return to its quiescent AC balance as well.
Additional, since this amplifier has one power supply and two of these bridged amplifiers per physical unit, there is a beautiful interaction of the three that takes over at a certain point of matching precision that I can't quite explain.
Inherently, there is no need to match component temperature coefficients to the transistors, but the closer the components match and track each other and the more β & other parameter matched the components are, the better the amplifier will perform up the limitations of the physical construction of the unit as a whole. The circuit is certainly not the limiting factor.
AMPLIFIER WITH GAIN STAGES COUPLED FOR DIFFERENTIAL ERROR CORRECTION
Abstract: An amplifier circuit having two identical inverting folded-cascode amplifier stages coupled together through a coupling resistor at the positive terminals of the respective stages. Each stage includes an input gain transistor whose gate forms the negative input terminal of the amplifier stage coupled to receive an input signal through an input resistor, whose source forms the positive input terminal of the amplifier stage, and whose drain is connected to the source of a level-shifting cascode transistor. Both transistors are supplied current from a constant current source also connected to the source of the level-shifting transistor. The drain of the level-shifting transistor forms an output terminal of the amplifier stage. Each stage feeds its amplifier output negatively to the positive input of the other stage, while at the same time feeding its distortion and noise contribution positively, via input gain transistor conduction, to the other stage. Balanced amplified outputs are produced from either balanced or unbalanced inputs, and distortion and noise components are produced in common on both outputs for differential cancellation. Complementary-symmetry and power amplifier versions of the basic single-ended line-level amplifier are also provided.
| Abstract of X-design patent - 1994 | Hadley 622C in comparison |
| An amplifier circuit having two identical inverting folded-cascode amplifier stages coupled together through a coupling resistor at the positive terminals of the respective stages. | two identical inverting amplifiers that are coupled by the 24Ω resistor in the center (or with your speakers). |
| Each stage includes an input gain transistor whose gate forms the negative input terminal of the amplifier stage coupled to receive an input signal through an input resistor, whose source forms the positive input terminal of the amplifier stage, and whose drain is connected to the source of a level-shifting cascode transistor. | This seems intentionally verbose for no apparent reason beyond attempting to muddy understanding by restating basic concepts. Seems quite basic if you draw it out and substitute in a BJT. |
| Both transistors are supplied current from a constant current source also connected to the source of the level-shifting transistor. The drain of the level-shifting transistor forms an output terminal of the amplifier stage. | The differential pair emitters (or JFET source) of the VAS and output stages are both connected to the same constant current source. The VAS also "forms an output terminal of the amplification stage". |
| Each stage feeds its amplifier output negatively to the positive input of the other stage, while at the same time feeding its distortion and noise contribution positively, via input gain transistor conduction, to the other stage. | there is a differential amplifier that forms the VAS and the emitters of the szikai and darlington outputs are also connected. This is a two stage amplifier with quasi complimentary outputs. no verbose terminology needed in the patent abstract either. |
| Balanced amplified outputs are produced | you bet. |
| distortion and noise components are produced in common on both outputs for differential cancellation | indeed they are. |
Below I've redrawn the Hadley 622C schematic into something that clearly shows the beautiful symmetry that I don't think Dawson Hadley wanted to show the world in 1965. One thing is for sure, he knew what he had. Now did he come up with it or did it come from elsewhere? Any tube circuits that look like this?
As one can see, there isn't much left in the abstract not addressed and simply put, the Hadley 622C exhibits the same (if not more) of the "super-symmetry" the x-design does and was capable of doing this 28+ years before the patent.
An example of this behavior: if you inject noise into the inputs, the amplifier will not only cancel the noise and eliminate DC at the outputs completely (to the extent your components are matched and can handle the voltage), but it'll return to its quiescent AC balance as well.
Additional, since this amplifier has one power supply and two of these bridged amplifiers per physical unit, there is a beautiful interaction of the three that takes over at a certain point of matching precision that I can't quite explain.
Inherently, there is no need to match component temperature coefficients to the transistors, but the closer the components match and track each other and the more β & other parameter matched the components are, the better the amplifier will perform up the limitations of the physical construction of the unit as a whole. The circuit is certainly not the limiting factor.
Hi hadley,
Just what do you mean by "if you inject noise into the inputs, the amplifier will not only cancel the noise and eliminate DC at the outputs completely (to the extent your components are matched and can handle the voltage), but it'll return to its quiescent AC balance as well." ?
The amplifier doesn't know what is noise or a signal. This amplifier is an elegant bridged amplifier design using a quasi-comp output design. It's pretty simple to figure out and shouldn't be difficult for a competent technician to repair properly. It is a minimalist component type design but should work fine.
Back in 1965 (up into the later 1970's in fact), most solid state designs sounded awful. Design techniques were not yet well understood and I remember buying that vintage component back when I was a kid. Testing them later showed they were not very consistent, many even leaky (by later standards). That and everything was a secret. Part numbers, schematics - everything.
I am not taking away from this quality of amplifier, just making observations and recalling what things were like in the early days. But one thing it isn't is magic. I am glad you like it. One thing I do as a business is make meaningful upgrades and measurable improvements to equipment. What I do is not found on the internet, it's all based in design and knowledge of component characteristics.
-Chris
Just what do you mean by "if you inject noise into the inputs, the amplifier will not only cancel the noise and eliminate DC at the outputs completely (to the extent your components are matched and can handle the voltage), but it'll return to its quiescent AC balance as well." ?
The amplifier doesn't know what is noise or a signal. This amplifier is an elegant bridged amplifier design using a quasi-comp output design. It's pretty simple to figure out and shouldn't be difficult for a competent technician to repair properly. It is a minimalist component type design but should work fine.
Back in 1965 (up into the later 1970's in fact), most solid state designs sounded awful. Design techniques were not yet well understood and I remember buying that vintage component back when I was a kid. Testing them later showed they were not very consistent, many even leaky (by later standards). That and everything was a secret. Part numbers, schematics - everything.
I am not taking away from this quality of amplifier, just making observations and recalling what things were like in the early days. But one thing it isn't is magic. I am glad you like it. One thing I do as a business is make meaningful upgrades and measurable improvements to equipment. What I do is not found on the internet, it's all based in design and knowledge of component characteristics.
-Chris
You seem to have turned off the ability to quote, which is odd but I'll manage. It was meant to be DC, not noise. Given the design and to be clear, noise fed in through other parts of the amplifier would indeed be cancelled.
Design techniques were understood just fine in 1965, just not within the solid state audio industry. Components were mostly poor though, but good ones did exist. Dawson Hadley had already stumbled upon the magical dual NPN by then that he used to form the differential pair. Not sure who else was using the dual NPNs by then. Maybe Tek? Never seen one in any of my old HP gear.
The original silicon dual NPNs from 1965 have all tested within 3% per side, far better than most dual NPNs made and available today with the exception of those made by AD. These early units were probably made by Fairchild.
I wouldn't trust it to a tech, any tech, even a competent one. You see, I'm not motivated by money, time, nor do I care how much parts cost. Techs have an inherent conflict of interest relating to these since they actually have to make a living from the work they do.
I also don't need a technician. While I was still a few parts short here in the photo below, I did indeed source the rest of the custom made components (I refuse to use TCC FWIW) and complete the amplifier (two identical boards plus the power supply) to my standards. I extensively used metal foil resistors and I'll freely admit that some of the 1K metal foils were not necessary given their position in the circuit, but I used them anyway because I have a lot of them. My math suggests the CMRR is quite high and the measurements suggest the components exceed the assumed tolerances built into the math given the better than claimed precision of the already spec'd 0.01% or better (on all of the critical parts). The sockets eventually were removed when I was satisfied with the assortment of transistors I had selected. Also tweaked component values to better fit the improved devices (though the values themselves on an absolute basis are not as critical as the matching and tracking between parts).
It performs acceptably, adequately, fine, or whatever, much as a Rolls-Royce did at one point in time.
Design techniques were understood just fine in 1965, just not within the solid state audio industry. Components were mostly poor though, but good ones did exist. Dawson Hadley had already stumbled upon the magical dual NPN by then that he used to form the differential pair. Not sure who else was using the dual NPNs by then. Maybe Tek? Never seen one in any of my old HP gear.
The original silicon dual NPNs from 1965 have all tested within 3% per side, far better than most dual NPNs made and available today with the exception of those made by AD. These early units were probably made by Fairchild.
I wouldn't trust it to a tech, any tech, even a competent one. You see, I'm not motivated by money, time, nor do I care how much parts cost. Techs have an inherent conflict of interest relating to these since they actually have to make a living from the work they do.
I also don't need a technician. While I was still a few parts short here in the photo below, I did indeed source the rest of the custom made components (I refuse to use TCC FWIW) and complete the amplifier (two identical boards plus the power supply) to my standards. I extensively used metal foil resistors and I'll freely admit that some of the 1K metal foils were not necessary given their position in the circuit, but I used them anyway because I have a lot of them. My math suggests the CMRR is quite high and the measurements suggest the components exceed the assumed tolerances built into the math given the better than claimed precision of the already spec'd 0.01% or better (on all of the critical parts). The sockets eventually were removed when I was satisfied with the assortment of transistors I had selected. Also tweaked component values to better fit the improved devices (though the values themselves on an absolute basis are not as critical as the matching and tracking between parts).
It performs acceptably, adequately, fine, or whatever, much as a Rolls-Royce did at one point in time.
Hi hadley,
Well, I come from a test and measurement background and those techniques were well understood. Just not in the audio industry so much. Audio engineers had to transition from tubes to transistor and the thinking is completely different. The diff pair (long tailed pair) existed in the form of tubes, so zero mystery there, and it isn't magical.
I studied parts early, and early transistors were fairly terrible. Todays parts are far, far superior to those used in your amplifier. I commonly match discrete transistors within 1% beta match using a jig I designed and gave to the members here years ago. Matched transistors assemblies have advantages since the best are made from one die, the transistors match very closely and track with temperature better than I can get mine - just the physics of the situation. Engineers through time used the best parts they could for the job at hand. Materials science and manufacturing techniques are so improved over what they could imagine back then it is just silly. Even the purity of the silicon has improved greatly (noise, leakage stability). You can't even begin to compare them - sorry.
The cost of a part often has no bearing on how good it is. It can, but it is often tied to manufacturing volume or desirability in the audio world. Many expensive audio parts I test are inferior to the much less expensive industrial part. Hmmm, well what do you know? lol! Of course I have a lot of equipment dedicated to testing individual components, and jigs for things you can't buy. There are parts that are just plain expensive to produce, so in that you could be right. However you always need to assess when some characteristics no longer make a meaningful difference in performance of the equipment.
You know, every circuit is a collection of both parts and compromises. That is called engineering, and designs are also engineered for manufacturability as well. This design has had the same considerations, I can see them as can many others I'm sure. This isn't a bad thing at all, but you need to have realistic expectations.
I upgrade test equipment as well. Real improvements. That equipment had an eye to performance with price considerations and manufacturing ease also in mind, but nothing close to audio equipment. Those parts and PCB layout are far superior (talking good equipment like HP, Tek and others). But stuff made in the 1980s can be improved simply because we have better parts and even different types of parts that can perform better in certain applications. When you talk mid-60's, wow, much improvement can be made! However often a 0.01% tolerance matters not. Sorry, but it is true. It depends greatly on the circuit and what the part does for a living. So blowing a ton of money for a 0.01% part, you may actually bought the wrong part. I do buy 0.01% parts occasionally when they matter. My jig uses 0.1% parts in certain locations, 5% in others where it honestly doesn't matter. A 0.01% part may have improved things slightly, but the meters most people use are not even 0.1% accurate (not even close on those ranges), so a 0.01% could be good for bragging rights but nothing else.
For your goals, you actually do need a very good technician, or electronics engineer with lots of experience. Whatever. The less you know, the less you know you don't know. There is a lot of truth to that statement. Good technicians are worth their weight in gold, and have no interest in stealing your ideas. Chances are they have seen them before anyway.
-Chris
Well, I come from a test and measurement background and those techniques were well understood. Just not in the audio industry so much. Audio engineers had to transition from tubes to transistor and the thinking is completely different. The diff pair (long tailed pair) existed in the form of tubes, so zero mystery there, and it isn't magical.
I studied parts early, and early transistors were fairly terrible. Todays parts are far, far superior to those used in your amplifier. I commonly match discrete transistors within 1% beta match using a jig I designed and gave to the members here years ago. Matched transistors assemblies have advantages since the best are made from one die, the transistors match very closely and track with temperature better than I can get mine - just the physics of the situation. Engineers through time used the best parts they could for the job at hand. Materials science and manufacturing techniques are so improved over what they could imagine back then it is just silly. Even the purity of the silicon has improved greatly (noise, leakage stability). You can't even begin to compare them - sorry.
Really??? You are kidding me - right? Most techs I know are not interested in stealing ideas or manufacturing their "superior idea". So there is no conflict of interest at all. None. Do you have any idea how many NDA's I have signed? I've never had any interest in stealing a design or idea. Most good techs simply want to do a good job, get paid and move on. If they learn something, even better. That is reality. Besides, most ideas are not great or original. They are good or inspired, and I admire those who come up with them.
The cost of a part often has no bearing on how good it is. It can, but it is often tied to manufacturing volume or desirability in the audio world. Many expensive audio parts I test are inferior to the much less expensive industrial part. Hmmm, well what do you know? lol! Of course I have a lot of equipment dedicated to testing individual components, and jigs for things you can't buy. There are parts that are just plain expensive to produce, so in that you could be right. However you always need to assess when some characteristics no longer make a meaningful difference in performance of the equipment.
Cool, no one said you did. However in your following statements you proved where one could save you a lot of money. Anyone can replace parts. Skill and experience points the way to good parts without wasting money. Even your test equipment and knowing the limitations of that stuff matters a great deal. Most people assume accuracy greatly beyond the ability of their equipment without even doing an error budget (much less a realistic one). Then there is the ability to measure in a meaningful way the performance once you are done. No, listening tests don't even come close to cutting it!
You know, every circuit is a collection of both parts and compromises. That is called engineering, and designs are also engineered for manufacturability as well. This design has had the same considerations, I can see them as can many others I'm sure. This isn't a bad thing at all, but you need to have realistic expectations.
I upgrade test equipment as well. Real improvements. That equipment had an eye to performance with price considerations and manufacturing ease also in mind, but nothing close to audio equipment. Those parts and PCB layout are far superior (talking good equipment like HP, Tek and others). But stuff made in the 1980s can be improved simply because we have better parts and even different types of parts that can perform better in certain applications. When you talk mid-60's, wow, much improvement can be made! However often a 0.01% tolerance matters not. Sorry, but it is true. It depends greatly on the circuit and what the part does for a living. So blowing a ton of money for a 0.01% part, you may actually bought the wrong part. I do buy 0.01% parts occasionally when they matter. My jig uses 0.1% parts in certain locations, 5% in others where it honestly doesn't matter. A 0.01% part may have improved things slightly, but the meters most people use are not even 0.1% accurate (not even close on those ranges), so a 0.01% could be good for bragging rights but nothing else.
For your goals, you actually do need a very good technician, or electronics engineer with lots of experience. Whatever. The less you know, the less you know you don't know. There is a lot of truth to that statement. Good technicians are worth their weight in gold, and have no interest in stealing your ideas. Chances are they have seen them before anyway.
-Chris
For good sounding solid state amplification sure, the thinking is quite different, but I'm sure you've seen the Marantz 7T schematic and compared it to the 7. The thinking was much the same and it did work, but the 7T sounds terrible (unless upgraded).
I hope that you understand that I only called the dual NPN "magical" because you said nothing about the amplifier was "magical" in the previous reply. You've had a problem with this thread since the beginning and your ego got hurt because no one joined you in bashing me and the design of this amplifier. Because that is what this is, the only thing for me to do is to have a bit of fun and double down on these silly things you seem to have a problem with.
Earlier in the thread, there's a photo of a couple of my Analog Devices MAT12AHZ dual NPNs. The β is closer to 0.1% matched than 1% matched and the noise specs are outstanding. Take a look. They had dual NPNs (on one die) when these amplifiers were made because that's what all of mine had in them. I swapped them because I understood why they were there and what was important for this design. The MAT12s are obviously superior to what was used back then, but it's not as if the originals were bad. They're still excellent, but I'm simply not aware of anything approaching the specs of the MAT12s, especially the noise.
I have several early Fairchild catalogs. Actually, they're hard covered books. Beautifully designed, there's art in them, and some very compelling transistor specifications that are on par with what is standard for parts nowadays. So while these may have been top-end in their day, it's not as-if Fairchild didn't have noise or leakage specifications in those days for example. They absolutely did and there were parts that would today still be considered low noise that were marketed as such then.
Our modern 21st century society may have the capability to make amazing transistors, but we don't utilize it unless we're talking about chips with billions of them and we're not. This conversation is about discrete transistors and back then a lot more care was put into how these devices were made generally. Just look at the huge variety of these discrete parts that used to be available.
What you are defining is not a conflict of interest. Intellectual property theft is quite different.
A conflict of interest relates to counterparties who have opposing goals. Using this thread as an example, my original goal evolved into using the best possible parts available installed in the cleanest possible manner to restore an amplifier which I think has a great circuit and a nice history. If I simply asked a tech to restore this amplifier and we agreed upon a price or a ballpark price to do so, inherently the average tech would be motivated to restore the amplifier quickly using parts on hand that maximized their profit on the transaction. No tech that I'm aware of would go to the lengths I have to restore this amplifier. They'd charge me thousands and it's unlikely their workmanship would match my standards. Restoration isn't even a fair term at this point for my work on this circuit. Since our interests do not align/ can be considered to oppose each others, that is a conflict of interest.
"How good it is" can be quantitatively measured in this circuit. It's not about audiophoolery with exotic capacitors and other nonsense. Metal foil resistors simply are the best resistors in the measures that matter when the goal is the most symmetrical tightly matched circuit possible. These metal foil resistors are stable, the tolerances are very tight, they barely drift even after decades (they've been around for nearly 60 years), and they are effective at handling inrush too. I could go on even further.
Where did I prove that? Define what parts I replaced in error beyond the ones I already mentioned did not necessarily need replacement.
The test equipment is all upgraded too. I assure you it is fully capable of measuring what needs to be measured.
Seems quite clear that you were suggesting I needed a tech. Wouldn't have been the first time in this thread. I'll again remind you and anyone else that this forum is a called diyAudio, not techGatekeeping.
There are maybe 200 of these amps that were ever made. I assure you that the layout of the board was not made with ease in mind nor do I think profitability was much of a concern. My expectations were blown away on this project. Do I seem unsatisfied with the performance of the amplifier?
This design is an optimization, not a compromise.
Notice how I put "on all critical parts" above when I mentioned 0.01%?
Sometimes the only option in metal foil is the 0.01% part (or even better) Typically, the "worst" you can do is 0.1% or occasionally 1%. Not a bad problem in my opinion.
I've continued work on the 622C and I've reached the point where DC offset can be measured in μv and a car driving by outside affects the meter reading. Keep in mind that I removed the DC adjustment pot so this is due to the design and the components alone.
According to the HP 3440A voltmeter I use (yes, I use this heavily modified dinosaur for a reason...), the 622C has a DC offset of +/- 50μv. There seems to be no pattern and the reading appears to spend about equal amounts of time positive or negative.
According to the HP 3440A voltmeter I use (yes, I use this heavily modified dinosaur for a reason...), the 622C has a DC offset of +/- 50μv. There seems to be no pattern and the reading appears to spend about equal amounts of time positive or negative.
Last edited:
Thermal drift is cancelled in this circuit.
We are talking about +/- 0.00005 VDC offset here, so there is no problem with the amplifier.
Look at the circuit and tell me what happens to the noise. Now of course, there are limitations like the PCB, but performance is still being limited here by component matching.
This result is why I matched the components in the way that I did to 0.01%. Understand now?
We are talking about +/- 0.00005 VDC offset here, so there is no problem with the amplifier.
Look at the circuit and tell me what happens to the noise. Now of course, there are limitations like the PCB, but performance is still being limited here by component matching.
This result is why I matched the components in the way that I did to 0.01%. Understand now?