I’ve just repaired and restored this lovely Krell KAV-300i integrated amplifier. Read on to find out what makes the KAV-300i such a superbly engineered amplifier.

Krell still proudly builds amplifiers and other high-end gear, from their base in Orange, Connecticut, USA. Established in 1980, they set about manufacturing a range of really different hi-fi equipment, aimed squarely at the top end of town. The KAV-300i is part of the more affordable range of Krell gear from this period.

All Krell gear that I’ve had the pleasure of working on is superbly well-engineered, well-built, and utilising grade components that you’ll see in consumer-level equipment. This, of course, means it wasn’t (and still isn’t) cheap and that’s where the KAV line came in.

The KAV-300i

Krell released the KAV-300i, as far as I can tell, in around 1995. It has the same distinctive build and materials as seen in amps like the legendary KSA-150, that I owned and restored. That means it’s carved out of blocks of gorgeous anodized aluminium and held together with stainless steel fasteners. I’m not kidding. You couldn’t build it better if you tried.

The difference here is that everything is downsized – capacitors, heat-sinks, and ratings are smaller, as is the 2RU chassis. Compared to my old KSA-150, with roughly the same power output, the KAV-300i is diminutive.

This, of course, means that the KAV-300i cost less to make and was affordable for many more folks. There are still many of them around due to the superb build quality. They really don’t build much like this anymore. This particular KAV-300i matches her power amplifier partner, the KAV-150, both owned by my customer and here for service and repair.

Specifications courtesy of Hi-Fi Engine

Power output: 150 watts per channel into 8Ω
Frequency response: 1Hz to 100kHz
Total harmonic distortion: 0.3% (trust me, wildly pessimistic!)
Signal to noise ratio: 95dB (line)
Dimensions: 475 x 387.5 x 92.5mm
Weight: 11.4kg
Accessories: remote control
Year: 1995 – 2000

Stereophile wrote a great review of the KAV-300i, definitely worth a read if you are looking to get one of these.

The Restoration

OK, let’s get down to it. This KAV-300i was in original condition which is great, but she was misbehaving. The main fault was distorted audio. I found a number of issues with this unit, mostly related to capacitor aging. The unit also exhibited issues with the board-to-board pins and pin header connectors. There were also a few dry joints for good measure.

Krell has been very kind to me over the years, assisting me with service documentation, where they have it and are happy to release it. A big thank you to Patrick @ Krell because, without that support, working on gear like this is more difficult.

Stripping, Cleaning

As always, I start with disassembly, inspection and cleaning.

img 7290
Krell KAV-300i, as she arrived at the workshop.
img 7291
Every control is made of metal. These pushers, for example, are spun out of stainless steel.
img 7293
This is my first dusty look at the interior of the amplifier.

Preamplifier Board

I tend to work through equipment in a logical sequence, so it makes sense to start with the preamplifier board in this amplifier.

img 7294
The preamp board performs preamplifier switching and line-level buffering. All inputs are relay switched and the inputs are shorted when not in use. Nice work Krell.
img 7301
And here it is, free of the chassis. I will rework this board as I consider necessary and replace the small electrolytic capacitors because small caps always age fastest and the owner wants a full restoration.
img 7305
Capacitors replaced with Nichicon FG (fine gold) in this case. I have a large range of premium capacitors in stock at all times. I choose what is most appropriate for the task at hand and based on what I have in stock. If I don’t have what I need, I order it in.

img 7304

Power Amplifier Board

Now to get stuck into the board where all the heavy lifting is done. This amplifier and power supply board is a gorgeous one-piece fibreglass board. Typical of Krell is the absolute minimal use of wiring. The preamp board is connected to the main amp board by pins and pin headers. Power comes up through the board stand-offs, a really nice touch and highly reliable over the long term.

img 7302
Check it out, there are about five wires in the whole amp! Great work Krell.
img 7303
Plan view of the mainboard. Note the two rows of original MJ15024 and MJ15025 output devices. Dual bridge rectifiers sit under the board, middle bottom. To either side, we have two filter capacitors per channel. These are modestly rated and certainly ready to be changed. up through the centre, we have the 5 V and +/- 18 V regulated supplies and associated capacitors. Almost all of these were dead or dying, though they look OK from the top. Another reason why you HAVE TO TEST parts!
img 7307 e1520413463803
Apologies for this blurry naked shot…
img 7309
Main-board out and ready for re-work.
img 7308 1
This is the underside for those playing along at home.
img 7312
Bad capacitor, very bad capacitor. A little on capacitor testing. Replacing every capacitor in anything, without having the ability to test them, is like waterproofing your cat – a complete waste of time. Now, if you can test the parts properly, with an ESR tester that also measures capacitance, you can make informed decisions about how to proceed. This cap was I think meant to be 470uF. It’s now a much smaller cap, with a big resistor inside it, passing DC, which it shouldn’t. This is never good and can cause all sorts of strange circuit misbehaviour.
img 7316
Beautiful, all tested and all changed, replaced with premium laboratory-grade parts that will last a long, long time. Everything I’ve used here is super-low ESR, high temp rated, rated for voltage, high ripple current rated. I use premium parts in refurbs like this so that the unit will give its owner many more years of trouble-free service. More pics below.

Power Supply etc

I replaced literally every electrolytic capacitor in this amplifier, simply because so many were bad. As mentioned, I used only lab-grade components from Nichicon, Panasonic, Nippon Chemicon, CDE, Kemet etc, to guarantee long service life for her owner.

img 7317
Out with the old Panasonic filter capacitors, seen here, and in with some new Cornell Dubilier (CDE) parts.
img 7319
Here are the new CDE filter caps installed, along with the rest on the top of the board. When replacing capacitors, it’s important not only to get the actual values (capacitance) and ratings (temp, voltage) right but also to get the physical dimensions and package type (axial or radial) right. I always spend time making sure that I use parts that physically work on the board in question. My goal is for the finished product look as ‘factory’ as possible. If you look at these parts, and how I’ve installed them, you cannot tell that they are not factory-fitted.
img 7320
Look great don’t they!

img 7321

img 7322
Nice capacitors here. Note I’ve mounted every capacitor with the same spacing, flush to the board, straight. This matters, I do things this way because that’s how I would want the work done.
img 7323 e1520407544147
There are four capacitors on the underside of the mainboard. I replaced them with fresh axials, but despite ordering the slimmest good quality parts I could find, the new ones were a shade too tight with the board mounted back onto the chassis.
img 7325
Temporary new replacement under-board axial. Too tight, unfortunately, so I reused the originals which measured perfectly anyway.
img 7327
Original skinny, but electrically perfect, axial caps
img 9073 e1520407471789
Here I’m preparing to replace the four small electrolytic caps on the front panel board.

Reassembly & De-Bugging

Time to put the KAV-300i back together. Thanks to Krell’s incredible tolerances, everything just fits perfectly.

After reassembly, I spent considerable time testing, adjusting and pulled her apart again, chasing a residual issue. To ease my curiosity, I tested every single semiconductor on both boards, because the amplifier was still occasionally showing signs of distortion, though dramatically less than previously.

I re-cleaned and prepped the board-to-board pins and did a little more re-working of joints and testing of resistors and transistors. Finally, she went back together again and tested perfectly, in fact, way better than spec.

img 9077
Much cleaner now with re-worked boards, new parts and clean chassis.
img 9078
A good view of the KAV-300i reassembled and undergoing testing.
img 9079
A better view of parts and trimmers.

Testing & Measurement

A word or two about testing. I like to quantify things for myself and for my customers. To that end, over the years I’ve invested in the test and measurement instruments needed to characterise this stuff properly.

Here’s a good example, in the final testing and debugging of the KAV-300i. I  feed in signals of known purity and amplitude and measure them again at the output of the amplifier. In a crude sense, this proves that signals pass all the way through the amplifier, but it tells us a lot more than that.

Measuring distortion of the signals output at the speaker terminals tells us a great deal about the health of the amplifier circuitry. In this case, it helped me to diagnose and fix this distortion problem. Having the right equipment takes the guesswork out of debugging and troubleshooting.

The extraordinarily low measured distortion of this KAV-300i is a testament to the design of this great amplifier and to the work I’ve done to restore her.

img 9075
That’s a reading of 0.0023% on my distortion analyser. I feed the amp a very low distortion (0.001%) sine wave input while it drives an 8 Ohm speaker as a load. At the same time, I’m also monitoring the speaker terminals with my analyser. It’s a very precise way of seeing exactly what’s going on.
img 9076
And that’s a reading of 0.0068% distortion, driving a 100 Hz sine wave at about 10 watts RMS. There are absolutely no distortion issues with this amplifier anymore!

So that’s a look at the complete restoration of a Krell KAV-300i. I hope you’ve enjoyed this and found it informative. Don’t forget you can contact me (Mike), right here on my contact page, for assistance with your hi-fi gear.


Discover more from LiQUiD AUDiO

Subscribe to get the latest posts sent to your email.

24 thoughts on “Krell KAV-300i Integrated Amplifier Repair & Restoration”

  1. Fascinating to read your work Michael
    Always of interest to know about the vintage units and to see stage by stage pictures of what, what anf how you fix them
    Keep posting
    I am sure many others feel the same

  2. Most Excellent article. I am recapping my 20 y/o 300i following your guide. Hopefully will do a good job. The inside of my 300i is slightly different in the segment from the power cord to the toroidal transformer.
    I have ordered the capacitors at Mouser Electronics today.

    Andrea Torri
    Northborough, MA
    USA

  3. Mike. Nice article. Could you tell me what caps were sending DC current? Is it a specific set of caps? That is what my 300i is doing. Every time I plug it in I get a surge thru my speakers.

    1. Hi Tim, thanks for your comment. I think you might be asking which caps were dead in this particular repair. Lots of caps measured low for C, high for ESR, and therefore were electrically leaky, meaning they passed some DC. I’ve repaired dozens of pieces since then, so I can’t remember exactly which caps were bad. More importantly though, it shouldn’t make any difference to repairing your amplifier. The only way to approach a repair like this is to test all the capacitors and then replace them, based on any patterns you find. You need the right equipment to test capacitors properly and of course the right soldering and desoldering tools. If in any doubt, take this to an expert. These are really nice amps and can be restored very effectively by the right technician. Good luck getting her sorted!

  4. Excellent tips and very informative article.
    However, I have KAV400xi, now almost 14 years, and the amp start to misbiheave in similar way as mentioned above. Audible distortion and shy bass on the medium-hither level.
    On inspection, everything seems good, but the environment of this little monster run very hot, any component inside is hot as heatisniks. And 80% caps are SMD
    Any experience with KAV400?

    1. Hi Erden, glad you enjoyed this repair! Yes, I have experience with the KAV400xi, it will certainly need an extensive overhaul at this age and the SMD caps will form a big part of that overhaul. It’s tedious work, but it must be done. Leaking electrolyte will likely be an issue and must be dealt with as it is discovered. If you are local, you are welcome to call me to discuss the options.

  5. I appreciate your comment and sharing your experience. Unfortunately, I’m not local but “down under” from your site (Europe), and this is only way to ask/comment 🙁
    I received today most of replacement capacitors, still waiting for small ones. Due to very tight space the choice for appropriate PS electrolytes was limited, so I choose Vishay BC components 105C, since the originals are Nichicon LQ now obsolete 85C lytics, and wanted to go for 105C since the whole amp run very hot, thus 105C makes sense.
    Two questions however arose for me during choice of small capacitors (incl. SMD ones):
    1. the 22uF 6,3V in the power amp input and VAS (C1, C2, C64, C68…etc on the schematics) – all actually decoupling caps.
    The question is – should they all to be replaced? And of what type? There is no clear marks what type they are ( lytic? Polymer?), and will be of very good advantage to replace them with 125C polymers (e.g aluminum or tantalum polymers)
    2. same question is going to SMD (also mainly decoupling capas) in preamp section – 68uF/25V, which also run very hot. marking on the top of the caps says onla E4P, and values. There is stylish “O” also on top, which I do not know are they Oscons (organic caps) or maybe also polymers. On some forums, there a references that in the corresponding preamp KAV-280P, Krell puts everywhere (exept PS) polymer caps, but this can not be confirmed for sure.
    3. Should be changed also the non polars? There are four of them, classic radial 85C Nichicons NP, and from the schematics I can see these are in the feedback loop. Here also planned to go with Panasonic non-polar (SU series), but I must go here with same 85C since there is no 105C bipolars.
    Any clues and help highly appreciated.

    1. Hi Erden, regrettably, I don’t have the time to provide detailed technical assistance of the sort you need here. I don’t have the schematic in front of me and I’m heavily time-limited by my workload. I receive many enquiries each day and they can become a black hole for me in terms of absorbing time, so I try to avoid long technical discussions, outside of those involving customer equipment. Be careful of forums for technical assistance, I suggest using replacement parts of the same type where possible, ie polar for polar, non-polar for non-polar, tantalum for tantalum etc. Triple-check everything and don’t obsess over capacitor types, just carefully repair her using parts at least as good as original and enjoy her again. Best of luck getting her sorted!

    2. Hi Erden, did you ever found 125c of 22uf 6.3v in required size ? 4 yrs on and I am now doing my recap of the KAV-400xi and I can only find for other larger ones using hybrid polymer ones for 125c. Cheers,

      1. Well, I found only polymer type, and that’s good!
        Someone from krell service answered to my request that “they have not in (their) record for failimg any of those particular type of capcitors…” -and usually they did not replace them at all.
        But, in case of need, the can be replaced with rhe same and/os sloghtly bigger voltage (s.g 10V kr 16V).
        Giving this- I just left the originals, replaced the main PS 6800uF and several decoupling.
        The amp works since then as new

    3. Those 68uf 25v ones marked “o” are Nichicon’s logo sometime used and I believe all smd are also Nichicon like UUD clearly marked for 47uF 25V ones as are the thru holes ones, VK, VR and SME(BP) while I still cannot figure out the input coupling NP dark blue ones in the preamp pbca that are marked strangely Пc like some russian capacitors.Cheers,

    4. For the power pcba of stock Nichicon SME BP 85c ones, I replaced them with Nichicon EP BP 105c ones but I want to maintain/not risk some SQ and used the Nichicon ES BP 85c ones in the preamp pcba for input coupling to replace the unknown stock Пc NP ones … figured that I can easily change them out if needed since preamp pcba is more accessible. Cheers,

    1. Hi Erden, no problem at all and no need to apologise, thanks for your comments and I’m glad you are enjoying the site, much more coming. Let me know how the overhaul goes!

  6. Hi! Thank you for this great article.
    Can you tell me(if you remember this), how much does the case of this amplifier heat up when it is on and sounds at a low volume? And does it warm up in standby mode?
    I purchased this unit and was faced with the fact that its body heats up to 40-45 degrees, even in standby mode. And when it is turned on and works at a low volume, the case heats up to 50-60 degrees. And I mean exactly the heating of the case, meaning that the transistors inside the amplifier should be heated even more.
    Thanks!

    1. Hi, my pleasure. This definitely doesn’t sound right. My experience with these is that they need extensive maintenance at this age and many times people have fiddled with them. This one sounds like it has a fault, idle Pd is low enough when correctly adjusted that the sorts of temperatures you mention are not possible. Note my emphasis on correctly adjusted!

      1. Thank you. Mike! I understand correctly that by correctly adjusted you mean the bias adjusting? Or something else?
        And please explain (without abbreviations) what “idle Pd” means? English is not my first language, so I’m not sure if I understood correctly.
        Many thanks!

        1. My pleasure! There are bias (quiescent current) and DC symmetry adjustments, all of which must be made correctly. Idle Pd refers to power dissipation at idle. If you have more questions of a technical nature, you can get in touch via my Contact page.

  7. Mike I am just about done recapping the same as you have here. I need 1 rca jack as the ground caps pulled off. Kind of unusual design. Ground in the middle and back in the 3 pin footprint. Both insulators are black. I have a box full of panel mount dual rca but not this one.
    Thank you

  8. Liked your restoration of the Krell .Got a 300i myself and I am very impressed with it. Brought one for a $ 1000 [it was a trade] and it performs perfectly .Best buy I had in a long time Regards Steve

Leave a comment, like, share and subscribe!

Scroll to Top

Discover more from LiQUiD AUDiO

Subscribe now to keep reading and get access to the full archive.

Continue reading