Since I've rebuilt five of these amps in the last couple of months, I thought I'd share a bit of what I've learned about these amps.
Built as monoblocks, the L-09M is the most powerful amp ever built by Kenwood. Sold for two years in '79 and '80 (from what I remember), the L-09M is rated at 300W output and retailed for $700 each...so setting up for stereo meant coming up with $1400! This also makes it Kenwood's most expensive amp, by far.
Here's a peek inside:
The L-09M is a fully DC-coupled design, with a FET differential input stage, current-sourced and cascoded. Nothing too unusual there, but this stage feeds a PNP differential VAS, which feeds another NPN differential for the bias stage. The output stage is an unusual cascoded setup, probably chosen to increase the reliability of the output devices (stock devices are Sanken 2SC1116 and 2SA747).
The rail voltages for the output devices are what raise an eyebrow: Schematic calls out +/-106V, and I usually measure about +/-110V with no load! The L-09M does this with a separate transformer for each rail!. I can't recall ever seeing this done before. These transformers feed a bridge rectifier, and then go to two 18,000µf 120V caps, which provide power for the entire unit...there are no extra windings in the power transformer and no regulated voltages in the amp (save for the 28V zeners for the FET input stage). I like this simplicity...regulators are always a trouble spot in vintage gear.
The amp has a front panel switch, which - through a relay - allows two sets of speakers to be connected. You are not allowed to run both simultaneously...choose 'A' or choose 'B', not both. There is also an ususual front panel attenuator, which, through voltage-divider resistors, allows reduction of the incoming signal at -3db, -6db, -9db, -12db, and ∞. Due to its limited range, it is a true attenuator, and not a volume control (thus driving the amp directly from a line-level source without its own output control is not feasable, or recommended...given the power at hand). The meter circuit has three ranges: 1x, .1X, and .01X, which would correspond to 300W output, 30W, and 3W. Very nice, but don't rely on the meters for true readings. While the L-09M meters are better than many, accuracy was not a large design priority.
The L-09M is designed to be powered up either by an external signal voltage, or with a flip of the rear panel switch, powered up normally with the front panel switch (both the L-07C and the L-07CII can provide this voltage through rear panel jacks, so powering up the preamp will power up the amplifiers). Either way, the incoming current surge is carried by an Omron LY-2 relay, so the power switch is unburdened of this heavy power-on transient.
I have found no real weakness with the design, other than the active cooling, and the only reason I call it a 'weakness' is that any time you're moving air, you're moving dirt, and thus an occasional blasting out of the fan and heatsinks with compressed air is a very good idea. Anyway, the heatsink assemblies are not terribly large, but are oriented so that together they form a tunnel with the 100VAC fan mounted at the right rear of the chassis. This fan runs continuously, limping along at a low, nearly noiseless speed under normal operation through a reduced voltage via a large power resistor on the startup/fan board. If run hard enough for long enough, there is a thermal switch mounted on the heatsink designed to trip at 70°C, and this shunts two power resistors on the fan board for increased voltage to the fan, which kicks it into its high-speed mode (which is also admirably quiet). The thermal switch seems to have little hysteresis, and as soon as the temp drops below 70°C, the fan drops back to idle mode. Under normal operation, you will be hard-pressed to get the fan to kick into high-speed mode. Takes about 45 minutes of 50W output to get the amp hot enough to trip this switch.
(Note about the fan: Because Kenwood used a 100V fan, they also sized the voltage limiting resistors on the Startup/Fan Board appropriately. Two of the units here had fans with somewhat worn bearings, and needed replacement. The stock fan is 119mm x 28mm, and the maximim allowable width is perhaps 32mm, so the 'standard' 119mm x 38mm fan is a no-go. I spent days looking at various 119mm fans, and the only suitable long-life replacements were a couple of fans from ebmPapst...a maker of very good, but very expensive fans. I chose the ebmPapst 9906 fan, which is a 119mm x 25mm 115V ball-bearing fan, and cost a bit over $40 each. The stock voltage limiting resistors should be replaced for two reasons: one, Kenwood mounted the stock resistors flush with the PC board, and this is slowly burning up the board. Two, the new 120V fan will spin too slowly with the stock resistor values. The stock resistors are 1.6K @ 10W, and 430 ohms @ 7W. These should be changed to 1.2K @ 12W and 400 ohms @ 8W, and the replacements should be the hollow enamel type, such as the Ohmite B8J and B12J series. These resistors shed heat very quickly, and prevent burning the board and melting their own solder connections. Even if the stock fan is working well, I'd replace the stock resistors with those mentioned above.)
Although Kenwood rated the amp at 300W into 8 ohms, that's an extremely conservative number...of the five amps I've rebuilt, the one that measured the lowest hit 408W into 8 ohms...most were between 410 to 420W. (I have not tested power into 4 ohms, but I will eventually do so and update this thread.) Measured distortion is very very low, and on a properly functioning unit will be around 0.0035% @ 300W (1KHz). Obviously, the L-09M has the power to drive most any speaker you can think of, with the possible exception of something like 1 ohm Apogees. The amp has an 'official' rating of 4 to 8 ohms, but I would not hesitate to drive difficult loads such as the Infinity speakers using Watkins woofers.
One other thing...the distance from center-to-center of the rack cutouts is a nice, standard 4". For my own L-09M's, I've bought some stainless steel handles and attached them to make the amps easier to lug around. And they look cool too. (Edit: the Omnia 8mm dia. stainess steel rounded corner handle is a nice match. Omnia model #9538-102. Many home improvement places online carry Omnia handle pulls)
Overall, I'm very impressed with the build quality and the design of these amps. Judging from what I see, reliability should be very high, and from what I hear, sonics are not compromised by any design shortcuts.
If you're looking for an amp with nearly limitless power, this is one to consider. Not everyone needs 400W of power, but a lot of us want it regardless. The nice thing about it is, you can have all that power without sacrificing finesse. All you need are a pair of these. :thmbsp:
Built as monoblocks, the L-09M is the most powerful amp ever built by Kenwood. Sold for two years in '79 and '80 (from what I remember), the L-09M is rated at 300W output and retailed for $700 each...so setting up for stereo meant coming up with $1400! This also makes it Kenwood's most expensive amp, by far.
Here's a peek inside:
The L-09M is a fully DC-coupled design, with a FET differential input stage, current-sourced and cascoded. Nothing too unusual there, but this stage feeds a PNP differential VAS, which feeds another NPN differential for the bias stage. The output stage is an unusual cascoded setup, probably chosen to increase the reliability of the output devices (stock devices are Sanken 2SC1116 and 2SA747).
The rail voltages for the output devices are what raise an eyebrow: Schematic calls out +/-106V, and I usually measure about +/-110V with no load! The L-09M does this with a separate transformer for each rail!. I can't recall ever seeing this done before. These transformers feed a bridge rectifier, and then go to two 18,000µf 120V caps, which provide power for the entire unit...there are no extra windings in the power transformer and no regulated voltages in the amp (save for the 28V zeners for the FET input stage). I like this simplicity...regulators are always a trouble spot in vintage gear.
The amp has a front panel switch, which - through a relay - allows two sets of speakers to be connected. You are not allowed to run both simultaneously...choose 'A' or choose 'B', not both. There is also an ususual front panel attenuator, which, through voltage-divider resistors, allows reduction of the incoming signal at -3db, -6db, -9db, -12db, and ∞. Due to its limited range, it is a true attenuator, and not a volume control (thus driving the amp directly from a line-level source without its own output control is not feasable, or recommended...given the power at hand). The meter circuit has three ranges: 1x, .1X, and .01X, which would correspond to 300W output, 30W, and 3W. Very nice, but don't rely on the meters for true readings. While the L-09M meters are better than many, accuracy was not a large design priority.
The L-09M is designed to be powered up either by an external signal voltage, or with a flip of the rear panel switch, powered up normally with the front panel switch (both the L-07C and the L-07CII can provide this voltage through rear panel jacks, so powering up the preamp will power up the amplifiers). Either way, the incoming current surge is carried by an Omron LY-2 relay, so the power switch is unburdened of this heavy power-on transient.
I have found no real weakness with the design, other than the active cooling, and the only reason I call it a 'weakness' is that any time you're moving air, you're moving dirt, and thus an occasional blasting out of the fan and heatsinks with compressed air is a very good idea. Anyway, the heatsink assemblies are not terribly large, but are oriented so that together they form a tunnel with the 100VAC fan mounted at the right rear of the chassis. This fan runs continuously, limping along at a low, nearly noiseless speed under normal operation through a reduced voltage via a large power resistor on the startup/fan board. If run hard enough for long enough, there is a thermal switch mounted on the heatsink designed to trip at 70°C, and this shunts two power resistors on the fan board for increased voltage to the fan, which kicks it into its high-speed mode (which is also admirably quiet). The thermal switch seems to have little hysteresis, and as soon as the temp drops below 70°C, the fan drops back to idle mode. Under normal operation, you will be hard-pressed to get the fan to kick into high-speed mode. Takes about 45 minutes of 50W output to get the amp hot enough to trip this switch.
(Note about the fan: Because Kenwood used a 100V fan, they also sized the voltage limiting resistors on the Startup/Fan Board appropriately. Two of the units here had fans with somewhat worn bearings, and needed replacement. The stock fan is 119mm x 28mm, and the maximim allowable width is perhaps 32mm, so the 'standard' 119mm x 38mm fan is a no-go. I spent days looking at various 119mm fans, and the only suitable long-life replacements were a couple of fans from ebmPapst...a maker of very good, but very expensive fans. I chose the ebmPapst 9906 fan, which is a 119mm x 25mm 115V ball-bearing fan, and cost a bit over $40 each. The stock voltage limiting resistors should be replaced for two reasons: one, Kenwood mounted the stock resistors flush with the PC board, and this is slowly burning up the board. Two, the new 120V fan will spin too slowly with the stock resistor values. The stock resistors are 1.6K @ 10W, and 430 ohms @ 7W. These should be changed to 1.2K @ 12W and 400 ohms @ 8W, and the replacements should be the hollow enamel type, such as the Ohmite B8J and B12J series. These resistors shed heat very quickly, and prevent burning the board and melting their own solder connections. Even if the stock fan is working well, I'd replace the stock resistors with those mentioned above.)
Although Kenwood rated the amp at 300W into 8 ohms, that's an extremely conservative number...of the five amps I've rebuilt, the one that measured the lowest hit 408W into 8 ohms...most were between 410 to 420W. (I have not tested power into 4 ohms, but I will eventually do so and update this thread.) Measured distortion is very very low, and on a properly functioning unit will be around 0.0035% @ 300W (1KHz). Obviously, the L-09M has the power to drive most any speaker you can think of, with the possible exception of something like 1 ohm Apogees. The amp has an 'official' rating of 4 to 8 ohms, but I would not hesitate to drive difficult loads such as the Infinity speakers using Watkins woofers.
One other thing...the distance from center-to-center of the rack cutouts is a nice, standard 4". For my own L-09M's, I've bought some stainless steel handles and attached them to make the amps easier to lug around. And they look cool too. (Edit: the Omnia 8mm dia. stainess steel rounded corner handle is a nice match. Omnia model #9538-102. Many home improvement places online carry Omnia handle pulls)
Overall, I'm very impressed with the build quality and the design of these amps. Judging from what I see, reliability should be very high, and from what I hear, sonics are not compromised by any design shortcuts.
If you're looking for an amp with nearly limitless power, this is one to consider. Not everyone needs 400W of power, but a lot of us want it regardless. The nice thing about it is, you can have all that power without sacrificing finesse. All you need are a pair of these. :thmbsp:
Last edited: