Story time:
Professor Leach started his amplifier as a test of Mati Otala's theories.
For the amplifier he constructed he used off the shelf components.
The articles he wrote on the Low Tim were meant for people to be able to build the Leach themselves.
Now, if you take Tony's Barbados Dale's : those were the kind used by Mr d'Agostino and Mr Levinson in their amplifiers.
Not the average type of resistor, you'd have to buy plenty to get them.
On the other side of the world there were amplifiers built too, with very high bandwidth output devices and low inductance meanderband or meanderwire resistors.
Little chance of obtaining these then either, there was an export ban on those devices for years.
In Norway, Per Abrahamsen was a student of Mati Otala.
He started designing and building Low Tim amplifiers the Otala way too, and started his company Electrocompaniet.
His Ampliwire 250 employs high speed ring-emitter output devices, 12 per channel.
Also has low inductance meanderband emitter resistors.
And separate regulated higher voltage rails for the front end.
The AW250 has a high CL bandwidth and slewrate and drives just about anything.
The Leach design is meant to cope with less perfect emitter resistors.
Employing "inductance free" Mills makes it possible to alter the design restrictions, their cost makes it a shame not to.
If you stick to the design by the letter you can suffice with the lowest inductance wirewounds you can find.
I am using Stuart Easson's 10 Watt Panasonic meanderwire resistors, that Stuart was so kind to provide me with.
Professor Leach started his amplifier as a test of Mati Otala's theories.
For the amplifier he constructed he used off the shelf components.
The articles he wrote on the Low Tim were meant for people to be able to build the Leach themselves.
Now, if you take Tony's Barbados Dale's : those were the kind used by Mr d'Agostino and Mr Levinson in their amplifiers.
Not the average type of resistor, you'd have to buy plenty to get them.
On the other side of the world there were amplifiers built too, with very high bandwidth output devices and low inductance meanderband or meanderwire resistors.
Little chance of obtaining these then either, there was an export ban on those devices for years.
In Norway, Per Abrahamsen was a student of Mati Otala.
He started designing and building Low Tim amplifiers the Otala way too, and started his company Electrocompaniet.
His Ampliwire 250 employs high speed ring-emitter output devices, 12 per channel.
Also has low inductance meanderband emitter resistors.
And separate regulated higher voltage rails for the front end.
The AW250 has a high CL bandwidth and slewrate and drives just about anything.
The Leach design is meant to cope with less perfect emitter resistors.
Employing "inductance free" Mills makes it possible to alter the design restrictions, their cost makes it a shame not to.
If you stick to the design by the letter you can suffice with the lowest inductance wirewounds you can find.
I am using Stuart Easson's 10 Watt Panasonic meanderwire resistors, that Stuart was so kind to provide me with.
FWIW, Parts Connexion has a 10% off sale going on, including Mills resistors, Black Gates, etc.
Thanks Jacco.
Would you care to discuss how "coping with less than perfect emitter resistors" affected the design? How could it have been improved if non inductive resistors were available at reasonable prices?
Would you care to discuss how "coping with less than perfect emitter resistors" affected the design? How could it have been improved if non inductive resistors were available at reasonable prices?
jacco vermeulen said:
Well, I guess this is mathematically correct, but I wonder how useful though it is
to represent di as a function of dV. It may be, I just don't see it currently.
As for the inductance, I advise to not worry much about it. I just meant to put
the rationnal answer on this. Personnaly, I won't shake earth to get low inductance resistor. At an audio rate of 15Khz if you put a tone at the input, you will get a small overshoot (300mV) superposed on the fondamental tone at the output (and at the feedback) at 2 A. Well, maybe I'm hard of earing, but I'm not sure I will ear a difference of < 0.5%, a tone is very unlikely anyway. As for harmonics created by this L, it will be filtered by the low pas in the feedback.
BTW, a good way to improve inductance is to keep the leads as short as possible when you solder it. Keep it tight to the board!
JR
Some time ago when I was designing the PCB people wanted the option of using non inductive resistors….. Does anyone know the brand??? Their footprint is smaller than normal wire wound but the resistors should be around 5W or so?
Thanks
\Jens
Thanks
\Jens
resistors
Hi Jens, maybe this one:
http://www.schuro.de/preisl-mpc71.htm
http://www.schuro.de/preisl-manganinwid.htm
German language....
Good luck, Loek
Hi Jens, maybe this one:
http://www.schuro.de/preisl-mpc71.htm
http://www.schuro.de/preisl-manganinwid.htm
German language....
Good luck, Loek
JensRasmussen said:
That was me, Jens.
Those are the meanderband resistors by Fukushima Futuba, Japan. These are not inductance free, but low inductance.
For real low inductance you'd have to step to Mills, Caddocks, NS Dale's and Isabellenhutte.
Tony, if you have 200 Barbados Dale's, you've found a Carribean treasure. These resistors were highly praised by the writing people in those days.
BobEllis said:
If you built the early LT like me, you'll remember using carbon's.
Allan Bradley's for the Boucherot were discussed before.
With rev 4 or 4.5 Professor Leach switched to ceramic wirewounds, likely because production of the big carbons had stopped and were harder to get.
(the ones i have were produced in '66 i believe)
The earlier design had much higher slewrate, above 100V/us i recall.
Overhere we seem to be slewrate and risetime junkies, i remember reading PerAnders post that he is keen on it too.
With lower inductance resistors the bandwidth and slewrate could be raised again, lead compensation altered.
I think that can be read in the combined articles at the homepage of Prof. Dr. Marshall Leach.
I posted the di/dV*dV/dt bit because that seems to show the effect of slewrate and impedance.
dV/dt will be highest at transients, the higher the frequency the higher dV/dt.
With low impedances and large phaseshifts di/dt can be high as well.
So, transients are where the risk of oscillation is high.
imo, that could be another reason why compound output stages, with the resistor on the collector, have added stability.
Low Tim means also little feedback.
Without harsh feedback it becomes harder to keep the thing stable at higher frequencies.
So, high input filter limit, and lead compensation starting at a relatively low frequency.
Thanks Jacco.
Although I built some tube amps in the early 70s (7199 driving 6L6s), I moved on to other hobbies until returning to electronics a few years ago. I bought my Leach boards when version 4.2 was current, but didn't get to building them until 4.5 was released. I remember references to hard to obtain carbon comp resistors.
So, you are one of those DC to blue light bandwidth guys? 😉
Although I built some tube amps in the early 70s (7199 driving 6L6s), I moved on to other hobbies until returning to electronics a few years ago. I bought my Leach boards when version 4.2 was current, but didn't get to building them until 4.5 was released. I remember references to hard to obtain carbon comp resistors.
So, you are one of those DC to blue light bandwidth guys? 😉
BobEllis said:
Afraid so, Bob, and i am a nervous wreck.
In the early 80s i had a thing with the speed devices the German Dieter Burmester made, and 1 Mhz Spectral amplifiers, the famous DMA100.
I switched to tubes after that, from there to class A, the next fall guy was Mr Pass.
What's funny about it, the whole thing started because i just wanted a good, affordable, hifi set.
FYI-I just received my stash of Mills 5w wirewounds for the output emitter resistors today, and they fit very comfortably on the board and in the holes.
Just remember to space the resistors in this position off the board surface to allow air circulation around them. Good to do this for other resistors that carry substantial current, too.
I also want to report my experience with heat sinks in the VAS and predriver stage. My present Leach amp is at the 4.3 level. The VAS and predrivers in my 4.3 version run pretty hot. Leach called them pretty warm, and said you may want to use heat sinks on them, but that it wasn't necessary. (Leach reduced some of the bias currents in the amp since this version to allow some of the transistors to run cooler.)
Well, mine run too hot to keep my finger on them, so I put some heat sinks on them. I found that after I did this, the bias current in the output stage was much more stable, and I turned the pot down to get the proper bias current in the output stage. Likely the high temps in these transistors caused higher gain in them. Further, any air current convection around them, maybe even due to their own heat generation, probably causes a lot more drift in them than when they don't have a heat sink. I recommend heat sinking the VAS and predriver transistors.
It also wouldn't hurt to thermally bond the differential transistors together to keep them from drifting apart, and DC offset to wander. Not a big deal, really. But all you have to do is touch one of the differential transistors to see the DC offset move around.
Just remember to space the resistors in this position off the board surface to allow air circulation around them. Good to do this for other resistors that carry substantial current, too.
I also want to report my experience with heat sinks in the VAS and predriver stage. My present Leach amp is at the 4.3 level. The VAS and predrivers in my 4.3 version run pretty hot. Leach called them pretty warm, and said you may want to use heat sinks on them, but that it wasn't necessary. (Leach reduced some of the bias currents in the amp since this version to allow some of the transistors to run cooler.)
Well, mine run too hot to keep my finger on them, so I put some heat sinks on them. I found that after I did this, the bias current in the output stage was much more stable, and I turned the pot down to get the proper bias current in the output stage. Likely the high temps in these transistors caused higher gain in them. Further, any air current convection around them, maybe even due to their own heat generation, probably causes a lot more drift in them than when they don't have a heat sink. I recommend heat sinking the VAS and predriver transistors.
It also wouldn't hurt to thermally bond the differential transistors together to keep them from drifting apart, and DC offset to wander. Not a big deal, really. But all you have to do is touch one of the differential transistors to see the DC offset move around.
Low Tim is achieved by low feedback and high slewrate, Otala way.
Low feedback is achieved by each stage being very linear, fast and having a large bandwidth, minimised risk of local slewing.
Most of the distortion derives from the output stage, make the output stage more stable and slewrate can go up, and you have a more accurate amplifier.
btw, many a designer regards this as the post-Otala era.
e.g. , NP refers to it in one of his X-series articles.
I use spacer rings for this, looks good and adds damping.
Shame so few commercial producers still do this.
Low feedback is achieved by each stage being very linear, fast and having a large bandwidth, minimised risk of local slewing.
Most of the distortion derives from the output stage, make the output stage more stable and slewrate can go up, and you have a more accurate amplifier.
btw, many a designer regards this as the post-Otala era.
e.g. , NP refers to it in one of his X-series articles.
I use spacer rings for this, looks good and adds damping.
Shame so few commercial producers still do this.
In spite of all the discussion I started about the emitter resistors. I found that regular 5w cement resistors will fit. I am using them with the thought that I can replace them fairly easily if they don't work well. I used these. 0.47R 5w 5%.
They are a slight bit off center but the capacitor wires have plenty of room between them. Just thought I'd let folks know in case they are having a hard time finding the smaller resistors. They are quite a bit cheaper too. I figured if my Krell can sound this good with them, the Low TIM ought to be OK. Time will tell. I'll take a pic of it when I get home tonight and post it if there is interest.
Blessings, Terry
They are a slight bit off center but the capacitor wires have plenty of room between them. Just thought I'd let folks know in case they are having a hard time finding the smaller resistors. They are quite a bit cheaper too. I figured if my Krell can sound this good with them, the Low TIM ought to be OK. Time will tell. I'll take a pic of it when I get home tonight and post it if there is interest.
Blessings, Terry
Terry,
That's what I used in my first Leach amps. They work fine there and in my A75.
I may still go the non inductive route, just to see if I can hear a difference.
That's what I used in my first Leach amps. They work fine there and in my A75.
I may still go the non inductive route, just to see if I can hear a difference.
still4given said:
Actually it is much easier to make a good class A amplifier than a good class AB one, Terry.
http://home.no.net/andiha/articles/audio/classa.htm
Well I've already soldered them in so I guess I'll hear it like this first. If I hear something bad I'll think about changing them.
I do have a question though. Does the value of L1 matter all that much. Will an 8.2R resistor work there instead of a 10R? Isn't the resistor more just for a frame for the coil?
Thanks, Terry
I do have a question though. Does the value of L1 matter all that much. Will an 8.2R resistor work there instead of a 10R? Isn't the resistor more just for a frame for the coil?
Thanks, Terry
- Status
- Not open for further replies.