Modding a Kenwood L 1000 M


2006-05-01 12:21 am
Good morning to all,
I recently obtained an interesting PowerAmp - Kenwood L-1000 M - manufactured around 1990. Concurrently I was able to locate a service manual as well. I got the Amp since it is a member of Kenwood's famed "L-Series" which was manufactured by Kensonic Labs (also known as Accuphase) for sister company Kenwood.

Specs are as follows:

2x150W @8Ohm - 2x250W @4Ohm
Damping factor : 130
THD 0.006% (20Hz-20KhZ, 130W, 8Ohm)
THD 0.0004% (1kHz, 130W, 8 Ohm)
Frequency response 5Hz-100kHz +0dB -3dB
Signal/Noise Ratio : 115dB (Unbalanced), 125dB (Balanced)

What I'm looking for are ideas how to modify this nice power amplifier to make it even better. It has excellent specs but due to some reason, it didn't receive top grades in German higher end audio journals. (Like Stereoplay, Audio and Hifi-Vision). Maybe because it was too cheap for their preference ? I think the amp originally costed around $3000, that's way below the $10K many good Accuphase poweramps sell for.

It also is full of high quality components. Maybe there might be some "easy tweaks" to unlock the potential of this design, which shouldn't be far too different from Accuphase's own.

The design has some oddities / goodies
1) Full symmetrical layout
2) 2x 270VA transformers
3) 47000µF caps in the power unit
4) Point to point wiring

Aspects that I found questionable (starting points to modding ?)
1) Variable resistors (potis) in the input path, one per channel - is this necessary ? does it introduce noise

2) Maybe try to get caps to tighter tolerances - replace elkos with regular caps where possible

3) Get metal film resistors of smallest tolerances (Holco 0.5%) to replace resistors in audio path, at the extreme: use bulk foil types

4) The "input" board - uses 2 Opamps : 2x Dual Opamp NJM4580 from JRC / New Japan Radio. While these are ok and relatively low noise, there is still room for improvement in noise and especially slew rate (Contemplating AD797 for example)

5) There's a switch from XLR to Chinch Input at the back of the Amp. Well, it would make sense to "migrate" the switch to the front and feed it from 2 seperate preamps: one analog and one digital "cross connect" (with highest quality DA/AD) for example..

Just to give the full picture about my endeavour, it wouldn't be me who would do the modding, a german modding shop would carry out the project. However I'd like to be able to provide useful inputs to them, ultimately in the form of a "checklist" and for me as a a "project supervision" tool. Being familiar with most concepts, I know more than basics in electronics, however I've been at odds with the soldering iron whenever I tried it <g>

any comments and ideas welcome. Even more so because I know this is a rather "strange" and "exotic" amp - manufactured by a company with more than a good reputation..

best rgrds from Vienna/Austria


2008-04-04 12:22 am
This set was the subject of a series of modification articles in the Danish magazine "High Fidelity". The Power amp had only a few upgradable parts and mostly only for when you use it in unbalanced mode: 1) replace the balanced pots with better types or an input attenuator. 2) replace ICs with OP275. 3) Add capacitance to powersupply.

The Control amp is much more fun: there are an amazing number of coupling Capacitors which can all be bypassed. It gives an enormous improvement (and is free)!. In addition all the ICs can be replaced with OP275. Very worthwhile and lifts the unit into another class altogether.


2008-04-04 12:22 am
OP 275 works with no modifications at all. I have also used Analog Devices ICs, but I have forgotten the type.

But the really enormous improvement is to be found in removing the dc-blocking capacitors and possibly also in removing the Inductors from the input (They are there to "stabilise the unit" and were really only there to make sure that an unmodified unit conformed to ALL known electrical requirements across the world).
But the really enormous improvement is to be found in removing the dc-blocking capacitors and possibly also in removing the Inductors from the input (They are there to "stabilise the unit" and were really only there to make sure that an unmodified unit conformed to ALL known electrical requirements across the world).
Accessory Board, C15..18? I'd say the originals are quite generously sized at 22µ / 35 V already and thus unlikely to contribute distortion, and I doubt they used junk quality caps expected to be dead after 20 years. Still, one could install some quality bipolars here for peace of mind (10µ / 35 V will certainly do).

Not so sure about the inductors. How could much of anything improve when bypassing inductors that in the audible range are expected to be in the low double-digit ohms range at most (vs. a ~50 kOhm imput impedance) and thus their effect would be attenuated by about 70 dB to begin with, not to mention that there aren't any significant currents flowing? I'd much rather have RF kept out when the unit is used in unfriendly environments. Why not be glad to see some proper RF filtering for once? But that would require appreciating the work of the EEdiots who constructed the thing... :rolleyes:

As far as choice of opamps is concerned, requirements would seem to be a dual type in DIP-8, unity gain stability, good common-mode linearity and decent load driving. Input impedance imbalance is about 1.2/2.2 kOhms, noise requirements accordingly (~5 nV/sqrt(Hz) will do fine).
The original NJM4580 scores in unity gain stability and load driving, noise is decent; common-mode distortion is an unknown, as is input impedance nonlinearity (though it being a BJT input part, I'd expect average to decent results for the latter). This part must have been fairly new then.
Yesteryear's FOTM part OP275 performs rather worse in all respects (CM distortion, load driving and input impedance nonlinearity in particular).
I might try something from the LM4562 family (good CM distortion, good load driving, decent input impedance nonlinearity, decent noise) or an LT1469 (good CM distortion, good load driving, decent noise, average input impedance nonlinearity).


2008-04-04 12:22 am
THE CAPS THAT I THINK SHOULD BE BYPASSED ARE C51,51,53,54; C39,C40; C31, 32; C35, 37; C17,18; C9,10,11,12; C19,20,21,22; C31, 32. (sorry about the shouting: my CapsLock was stuck). They can be found on the main boar and the auxiliary board.
Thanks for the advice concerning the OpAmps. I am no expert on those and am only relaying the advice given by High Fidelity in around 1994 (this also goes for the Inductors). They suggested mounting all ICs in (gold-plated) sockets so that later substitutions would be easier. I have heard good things about the LM series but have not tried it.


2009-02-21 11:23 am

hi his047,
you mention "should be bypassed" but do not mention "how"
Film caps? if so which values?
Lytic caps? if so which values?

Hi sgrossklass.
Convincent but ambiguous, finally we look for better sound not better specs
in sheet.
You mention LM4562 & LT1469 but why? ,iisimply describing as decent,
good. etc do not describe whatkind of sonic improvement you look for
or expect to get.
Referent to block capacitors I used to bypass in a wonderfull preamp ,the
Pioneer C-21 with tremendous improvements in sonics
I used a array of 8 250k polipropilene bypass and change the capacitor
with ELNA Silmic II

To ALL the members
I really will appreciate if you have "real" results on any amp that can be
extended to the Kenwood L1000M


2008-04-04 12:22 am
for "Bypassed" read "shortcircuited" or "Replaced with wire-links" or "Simply just get rid of them".

What do you mean by the phrase "real Results"?

Bipolars would be disastrous and sound worse than Electrolytics (they are after all, simply two electrolytics back-to-back)
Last edited:


2009-02-21 11:23 am

I understand bypass as adding another cap in parallel to improve
Frankly I can not think that Kenwood/accuphase engineerin department
were so unskilled that supressing so many caps could be a improvement,
maybe the input inductors due that they can produce magnetic fields more
dangerous that the radio filtering they try to help.
In Japan de thousands of radio stations can be a problem we ave not in Europe.
With real results y want mean changes in the sound that will be lately
classided and evaluatedas follow.
1.- Better transparency ?.
2.- Better dynamics?
A change is just a change and should be evaluated ( very individual task)
of course but every one i trying to improve sound in a way that follows
his personal taste, preferences and the way each one understand sound/music.


2008-04-04 12:22 am
I am glad to see that you have such strong faith in the Kenwood/Accuphase design department, but to me it looks like an apprentice designed this preamp.

Take a look at the circuit diagram and think about why it should be necessary to have an Electrolytic Capacitor both in the input and the output of each amplification stage when they are both un-necessary and detrimental to sound. Also ask yourself why one would wish to double the influence of the el-cos by having a configuration where you have "cap+amp+cap+cap+amp+cap+cap+amp+cap": it does not make sense from a sound perspective, especially since the circuit is fully symmetrical and is perfectly stable without it. My L1000C has run for more than 17 years with no probelms and so have the L1000Cs of two friends. Hundreds of people made the modifications in Denmark and I have not heard anyone complain of instability of adverse effects (and believe me they would have written to High Fidelity to complain loudly and persistently).

The modification improves all aspects of the preamp: Stereo perspective, three-dimensionality, dynamics. and best of all, it is free.
The Elcos used by Kanwood are of good quality: Elmas all the way through, but still, the improvement gained by removing them is more than worthwhile.
Last edited:


2009-02-21 11:23 am

I must accept I was wrong.I checked the schematics nad I must agree
that this caps were not only superfluous but also wrong.
The japanese syndrhome agains DC is the only tat can justify such flawed
Have you also info about the preamp?.There are so many caps that
I figure out that maybe most of they should be bypassed as in the amp
Sorry again I should be more cautious before actuating the way I did.
By the way:
I suppose that the improvement with this mods should be terrific
or not?


2008-04-04 12:22 am
Ola Morrosco42,

The caps that I suggested shorting out were all in the pre-amp, L1000C.
The power amp (L1000M) does not have equivalent blocking caps, except in the single-ended to balanced input board. If you run the L1000M in Balanced mode there is no reason to make modifications (except shorting out the input pots or replacing them with a fixed attenuation (for both hot and cold sides of the balanced signal).

I made the mods more than a decade ago and I only remember that they lift an already excellent preamp into an entirely different class.