Measurements of vintage Kenwood KA-5010 amplifier clone - Now with SMPS!

[pma]

When I first saw schematics of vintage Kenwood KA-5010 amplifier, 2 years ago, I was immediately attracted and decided to build a circuit copy, better say a copy of the circuit topology with some newer solid state parts used. There were 2 points leading me to this decision - frequency compensation used and 3EF output stage, that grossly reduces loading of the VAS stage. I used lower supply voltage, like 2x25V - just what I had at the moment. The clone schematics looks like this:

This was built as a complete 2-channel stereo in a case with side heatsinks.

The result was much better than I expected, so please let me share some measurements with you. First, THD and THD+N at 1kHz, 5W/4ohm

Kenwood KA-5010 distortion 1kHz 5W/4ohm

Next, THD vs. amplitude at 1kHz, 4ohm load

Kenwood KA-5010 clone THD vs. amplitude at 1kHz and 4ohm load

Next, THD vs. frequency at 5W/4ohm

Kenwood KA-5010 clone THD vs. frequency at 5W/4ohm

Last, 19+20kHz CCIF IMD measurement

Kenwood KA-5010 clone CCIF 19+20kHz IMD at 12.5Wpeak/4ohm

These are very good results and the result of 19+20kHz IMD is simply great by any means. It shows that the amplifier has excellent HF linearity and that it keeps low distortion at high frequencies, as also seen from THD vs. frequency plot.
If there is an original somewhere, it would be nice, IMO, if Amir had an opportunity to make a measurement and compare.

 

[graz_lag]

What a super nice work you did!!!

 

[pma]

Thank you very much for your kind words! It is my hobby and passion.

 

[maty]

https://www.hifiengine.com/manual_library/kenwood/ka-5010.shtml (1989-1990)

 

[maty]

I LOVE the graph. Dominant H2 and monotonously decreasing harmonic profile.

http://www.sengpielaudio.com/calculator-thd.htm

THD 0.0014% -> H2 at -97 dB

 

[maty]

[PDF] https://www.hifiengine.com/hfe_downloads/index.php?kenwood/kenwood_ka-5010.pdf you need to be registered, free

 

[maty]

http://www.hifi-review.com/140018-kenwood-ka-5010.html

The internal design is essentially that of a lower-powered KA-7010 clone, featuring a star-earth system and common disc, line, tone and protection circuitry, but with a single mains transformer and just two 10,000μF reservoir capacitors. The output stage uses two pairs of Sanken bipolar transistors; compared to the 7010 these are lower-rated 2SA1215/2SC2921 devices...

http://www.hifi-review.com/140017-kenwood-ka-7010.html

Kenwood's 'Pure Signal Ground Line' star-earthing topology helps minimise interference from circulating earth currents while the main output stage is mechanically decoupled from the mother PCB using a strip of circuit track and flying leads. Paralleling the cheaper KA-5010, the '7010 features an indentical FET headamp for the MC disc input, followed by series-feedback equalisation around a dual-channel NJM4560D op-amp. Such commonality must surely help keep production costs down...

 

[restorer-john]

Very interesting. But you do need to show us some actual pics of the project.

Interesting you'd use the MJ21194 (4MHz) when the original Sankens (50MHz) are still floating around in great numbers even though they are obsolete too. Why that choice? You are in Europe/UK by the component choices by the look of it.

Tell us some details on the PSU you used and what modifications you've made to the original design. Clearly, you've deleted the current limiters and gone with a much simpler output zobel with no inductor. Have you done any reactive load testing on it?

 

[pma]

There are some photos left from the construction. It was built just because of my curiosity and not meant as my main amplifier, that's why I did not care about lower power, used the PSU I just had and reduced output pairs to one. MJL21193/94 are my favourite parts because of their robustness, I never happened to burn a single one of the plenty of them I used. Stability is OK up to 22nF at least, and as it is for my purposes only, I do not need more and that's the reason why I omit output inductor.

 

[maty]

Resized and optimized

 

[graz_lag]

I'am curious abt. what the pile on the upper-left corner is ... it looks like to be either unpaid invoices or recycled 1/4 cuts out from A4/Letter size sheets ...
I do the same to recycle my one-side-only printed sheets ...

 

[amirm]

That is an incredibly clean build! Well done.

 

[restorer-john]

Very nice, very neat build. That's a really lovely case too.

Well done.

I do however question using a two pin, unearthed IEC with no chassis earthing. Yes, your wiring appears to be excellent and you've stuck with double insulated mains 3 core wiring, but the main power PCB is sitting just above open grilles on the base panel. There needs to be elephant-hide or similar below the PCB IMO.

 

[pma]

I do however question using a two pin, unearthed IEC with no chassis earthing. Yes, your wiring appears to be excellent and you've stuck with double insulated mains 3 core wiring, but the main power PCB is sitting just above open grilles on the base panel. There needs to be elephant-hide or similar below the PCB IMO.

You are of course right and in case that the amp was intended for sale or to be used by another person than me, it would definitely be provided with a 3pin IEC socket and chassis connected to PE. As shown, it was just an evaluation sample for me, to measure the circuit parameters and to make some listening tests. At the moment, the same case is used for another project sample and again for evaluation only. The KA-5010 clone was moved to another case, now with 3pin IEC socket. In any case, thank you for mentioning the safety concern.
To conclude, I found this end-of-the-eighties topology very good and competitive even now and I will probably build a higher power (like 2x200W) amp based on this topology in the future. Re sound - it sounds very very transparent, without any trace of distortion and can compete to best classA amplifiers.

Again on chassis PE protection, below is the link which shows an amplifier that I built and sold in small series 7 years ago, it has a 3pin IEC socket and chassis connected to PE:
http://pmacura.cz/pa4_amp2.jpg

 

[pma]

That is an incredibly clean build! Well done.

Thank you Amir.

 

[pma]

Cannot resist to post multitone distortion test, 4ohm load . LF noise rise comes from my test gear, not from the amp.

 

[pma]

I have decided to make the same output force torture test to measure output impedance, as I did here:
https://www.audiosciencereview.com/...ss-d-amplifiers-measurements.9940/post-272799

with a small difference, resistor that serves to control the current was now only 4ohm and no output filter is used.

Force voltage amplitude was 8V, so the forced current was 2A (peak for sine, amplitude for square)

First, 2A square forced to KA-5010 clone output

We can see that after settling, which takes about 100us, Zout = 0.01 ohm approx.

Now, Zout for sine 20kHz

Zout is somewhat below 0.03 ohm

We can see that amplifiers have hard times with this torture outforce test. As they are feedback systems, it take some time for the whole thing to recover, and, probably more important, the output impedance is complex, not purely resistive. It can be best seen from the huge overshoot as a response to square force current. This is unlikely to happen as a back EMF from the speaker, but it is definitely interesting, IMO. Every amplifier behaves differently. These are measurable differences even between very low THD amplifiers.

 

[pma]

A couple of measurements in time domain

1. Step response into 2.5ohm resistive load. Rise time Tr(10%-90%) 1.16us, from which we may calculate corner frequency where amplitude response falls of -3dB

F(-3dB) = 0.35/Tr(10%-90%) = 301.7kHz
So bandwidth of this amplifier is > 300kHz. This is important to keep flat phase response in the audio band.

The next measurement shows amplifier 1kHz square response into 6.8uF capacitive load, to simulate ESL panel.

We can see that after very short time (50us approx.) the small ringing that is a result of resonance of the 6.8uF capacitor with inductance of about 1m of connecting wires disappears and the response is settled and stable.

 

[pma]

Coming soon ....

 

[pma]

Back to KA-5010 clone, made some more measurements . Not so bad from this 1989-1990 circuit topology, right? Especially compared to some latest @amirm AVR reviews.

THD vs. frequency

Twin-tone CCIF IMD 13+14kHz