Carlmart, yes I found a schematic of the M120 and similar topology except for the front end. The apt also uses an OpAmp for DC feedback to trim the output offset which I did not see on the 120. I wonder who copied who? I think the 120 also uses different devices in the output stage. The apt uses the MJ10022/23 devices which I think were quite new when the apt came out.
Here are the two simulations I arrived to with the M120, including one with DC servo. In fact, the original M120 used two IC servos on each channel.
THD gets a bit worst in the sim with the servo.
I never simmed the Apt 1. Perhaps someone did.
As most, if not all, active devices used in the these new "Luxman" versions are different from the original amplifiers, I do not quite agree on calling them "clones". Maybe "architecture" or "inspired". As models for those original parts are mostly unavailable too, we can only watch what happens with the modern parts and listen to the results.
As I have not tools for distortion or noise, auditioning them is the most I can do.
Now I wonder how Baker clamps could be effectively implemented here, in the output drivers I guess. Suggestions accepted to simulate.
Just tried what I think should be a Baker clamp implementation on the Luxman M120 clone. Taken from Bob Cordell's book.
I think D20 and D25 shouldn't be 1N4148. If I remember well from another implementation, that should be a Schottky or a BAV21.
THD was not altered, and the THD curve is practically the same. I'm not sure if there's something somewhere in the sim that will show the clamp effects.
I think D20 and D25 shouldn't be 1N4148. If I remember well from another implementation, that should be a Schottky or a BAV21.
THD was not altered, and the THD curve is practically the same. I'm not sure if there's something somewhere in the sim that will show the clamp effects.
Attachments
Using the better diodes for this particular application (BAV21 and the like vs the ubiquitous 1N4148), i don't think you would see contributions of the clamp until the amp is driven to clipping when the clamp becomes active.
Somewhere around here is some good discussion, simulation and test results showing impact of different diodes used for clamping in.
If I find it again, I come back and post links.
I "think" Syn08 had his fingers in it ...
mlloyd1
Somewhere around here is some good discussion, simulation and test results showing impact of different diodes used for clamping in.
If I find it again, I come back and post links.
I "think" Syn08 had his fingers in it ...
mlloyd1
The 1N4149 is also often a better choice than the better-known 1N4148, since it is similar but with lower capacitance.
Cheers,
Bob
I usually use the 1N4149 parts where lower capacitance matters (or where I suspect it might matter). For example, If I have input stage clamps, I'll probably use them. Where it is a don't care I usually stick to the more common 1N4148.
Cheers,
Bob
I stumbled onto this schematic of the Marantz 510 from 1976; it was very advanced
for the time and I see a lot of similarity to the Apt amp, less the cascodes:
Interesting features are the single diff pair driving a push-pull VAS, output stage
triple, and high class A bias in the drivers for faster turn off of the outputs. Vas
degeneration and base stoppers were not very common back then.
The series output stage was due to the lower Vce devices of the time.
I wonder who designed this amp.
for the time and I see a lot of similarity to the Apt amp, less the cascodes:
Interesting features are the single diff pair driving a push-pull VAS, output stage
triple, and high class A bias in the drivers for faster turn off of the outputs. Vas
degeneration and base stoppers were not very common back then.
The series output stage was due to the lower Vce devices of the time.
I wonder who designed this amp.
Attachments
some years ago, Tom made these comments:
That's good strategy given the output devices available then ...
Also there is Unclejed613's comments that the output devices were selected beta grades in order to mitigate common mode conduction.
And also this comment about driver transistors:
For you folks that have repaired units over the years, what semiconductors did you use to replace these parts and did you test for stability and distortion after repairs to see impacts of the chosen parts differing characteristics?
Looking at the datasheets for the mentioned parts, I don't see fTs anywhere near what Tom mentions.
Just curious ....
mlloyd1
That's good strategy given the output devices available then ...
Also there is Unclejed613's comments that the output devices were selected beta grades in order to mitigate common mode conduction.
And also this comment about driver transistors:
For you folks that have repaired units over the years, what semiconductors did you use to replace these parts and did you test for stability and distortion after repairs to see impacts of the chosen parts differing characteristics?
Looking at the datasheets for the mentioned parts, I don't see fTs anywhere near what Tom mentions.
Just curious ....
mlloyd1
I'd be curious to hear what people use also.
Here's the data sheet for the Sanyo 2SA1011/C2344 parts:
http://www.diyforums.org/MAX/bjt/2SC2344.pdf
They are 1.5A, 100MHz, 23-30 pF Cob parts, we have many similar parts to choose from.
My first choice for drivers would be the 2SC4883A/2SA1859A that have been noted on here.
Here's the data sheet for the Sanyo 2SA1011/C2344 parts:
http://www.diyforums.org/MAX/bjt/2SC2344.pdf
They are 1.5A, 100MHz, 23-30 pF Cob parts, we have many similar parts to choose from.
My first choice for drivers would be the 2SC4883A/2SA1859A that have been noted on here.
Last edited:
I picked up the A92/A42 from digikey and are shown as stocked as of a couple months ago. I used them in place of the D01/D52 original parts. The output devices are still available (MJ15022/15023) although the ones I got had higher beta's than suggested. I think original spec was around 30 and these are around 70. The original 1345 I replaced with an KSC1845. The predrivers on mine are ok, so did not need to figure out a replacement. As mentioned by PB2, I think replacements possible.
Sorry to necromance an older topic, but I have an APT-1 power amp on the bench I just finished servicing and found something to pass along...
DC balance just off enough for the protect to trip (was abou -2.4V) and the DC servo at its max compensation. Balance pot has no effect. DC voltages almost (like off 5%) bang on to service manual, just won't balance.
The input emitter followers, 2SC1345 diode tested OK. On a hunch, put them on the M-Tester and they had a Vbe of 699mV with an hfe of 225. Should be OK, right? Put them back in circuit, same issue.
For fits and giggles, matched up a pair of KSC1845, 633mV, 455 hfe. Put those in for the C1345.
Voila! Fixed!
Sometimes you just have to use "Vacuum tube policy" - nevermind the test readings, just change it
DC balance just off enough for the protect to trip (was abou -2.4V) and the DC servo at its max compensation. Balance pot has no effect. DC voltages almost (like off 5%) bang on to service manual, just won't balance.
The input emitter followers, 2SC1345 diode tested OK. On a hunch, put them on the M-Tester and they had a Vbe of 699mV with an hfe of 225. Should be OK, right? Put them back in circuit, same issue.
For fits and giggles, matched up a pair of KSC1845, 633mV, 455 hfe. Put those in for the C1345.
Voila! Fixed!
Sometimes you just have to use "Vacuum tube policy" - nevermind the test readings, just change it
You are wright, those 2SC1345 measures fine and produce DC offset when they are internally defective, I repaired a couple few days ago and changed it with 2SK1845 and problem solved, Q1 is the most common problem, also if the relay last more than 5 seconds change caps C205 and C214 of 47 uf 16v on the power supply boardSorry to necromance an older topic, but I have an APT-1 power amp on the bench I just finished servicing and found something to pass along...
DC balance just off enough for the protect to trip (was abou -2.4V) and the DC servo at its max compensation. Balance pot has no effect. DC voltages almost (like off 5%) bang on to service manual, just won't balance.
The input emitter followers, 2SC1345 diode tested OK. On a hunch, put them on the M-Tester and they had a Vbe of 699mV with an hfe of 225. Should be OK, right? Put them back in circuit, same issue.
For fits and giggles, matched up a pair of KSC1845, 633mV, 455 hfe. Put those in for the C1345.
Voila! Fixed!
Sometimes you just have to use "Vacuum tube policy" - nevermind the test readings, just change it
thanks for sharing the info!
i wonder if a redesign might be better in the long run. seems a little risky to have a protection circuit design so dependent on BJT hFE and VBE range. those parameters vary a lot with temperature, too.
especially with mfrs discontinuing parts at the rate they're going.
i wonder if a redesign might be better in the long run. seems a little risky to have a protection circuit design so dependent on BJT hFE and VBE range. those parameters vary a lot with temperature, too.
especially with mfrs discontinuing parts at the rate they're going.