Yaqin MS-34D integrade stereo amplifier upgrade

Recently I got a Yaqin MS-34D PP Integrated amp, with tubes upgraded to Russian tubes EL34-EH, and 6J1EB.

Very little review but I have found a Polish website, http://www.popular-hifi.com/projects/nad_stereo/yaqin_ms34d/#6J1P, which gives it a very high regard.

It looks nice, small but heavy, and sound very nice in midrange. However, there are apparently a few issues with it, which are quite normal with Chinese made HiFi gear, as workers were paid very little 10 years ago, and they do stuff randomly.

However, it has a few good stuff such as good exterior design (How ugly does the legendary Marantz 8B look!) solid transformers which add to the weight of 24KG, so I decided to improve it.

So opened it, and drew up the schematic as following:

A typical fixed bias EL34 PP amp, with LTP as gain and phase split stage combined, which makes it very simple and easy to modify.

First thing first, the so called integrated AMP is basically a power amp with motorized remote volume control and input selection.

But the potentiometer used is very bad, in terms of channel balancing and noise. So I have two options to fix it. 1) Scrap the entire "integrated" part and make it a power amp; 2) put a new motorized pot in.

I happen to have a good ALPS motorized pot, so it went in.

Second, the power amp section doesn't have coupling capacitors, so the input select and noise from the pot will be heard from the speakers, which is nasty, so I put two WIMA 2.2uf caps in. Problem solved.

The next part is the most critical part. According to the measurements from the Polish site mentioned before, the <2.5%THD at full power, 0.1% at 15w, is not too bad, (I trust no one would compare these numbers with a SS amp here.) but there is a design defective which can be solved to improve it tremendously.

And this is regarding the phase inverter. The widely used LTP does have a wrong resistor in position, and that is R112, which needs to be slightly lower than the peer, so identical output with same amplitude but with inverted phase will be available for the PP output stage.

(reference: https://www.aikenamps.com/index.php/the-long-tail-pair)

If you have a postive half sine wave considerably smaller or greater than the negative half, the distortion would be horrible as you can imagine.

So it has to be balanced carefully. As a matter of fact, when I check the two outputs, the voltage difference at this stage is almost 3V without load, with load it is even greater. Therefore it has be to fixed.

The idea is to reduce R112 to 75K, or if you don't have 75K, put a 300K 1W resistor in parallel to it. After this mod, the two output is almost with same amplitude. We can imagine how much distortion would be reduced in this alone.

The next step would be dealing with the power rail capacitors. In the old days, high voltage big electrolytic capacitors are expensive. So this amp only has 4x560uf on the main power rails (450V), and the bias power supply has only one 22uf.

No need to say, with bias power supply feed into the cathodes directly, high frequecy and main power noise would be certainly present.

A Russian gentleman on Youtube even put 6800uf for the bias power supply(63V rating ones' price is acceptable), and big 1500uf caps for the high voltage rails.

Also don't forget the high pass small caps to go parallel with these big ones.

If you fancy audio grade MKP capacitors, as in following page, you can also do it easily as the box has plenty of space.

http://www.goldfinch-mod.eu/wzmacniacze-2/yaqin/yaqin-ms34d/

The sound? I haven't got a chance to AB it with another tube amp yet, but it beats my very low distortion SS amp easily.

All these mods, apart from my own the phase inverter fix, are described here in the youtube video, unfortunately however in Russian.

https://www.youtube.com/watch?v=hBryRMPC-RE

And this is truly fun.

TDH measurments updated:

As we can see, when the amp is running at 50% power in TR and UL mode, about 10w and 20w respectively, which means the volumes pod goes to 3pm, and which is an insanely loud, the THD drops down from factory 2.4% to 1.4% and 1.8%.

When in UL mode and the power is 10w, the THD drops evern further to 1.2%.

And drop the power further to 5w (volume pod at 2pm), the THD goes to 0.8%.

We usually listen to music while the volume pod is around 9-10 oclock postion, which means the power is usually about 1w. And the THD drops to 0.5%.

As a matter of fact, these number are maximum. Based on signal source cleanness, (the THD of my signal source is < 0.004%), so the actual THD is considerably less than measurements.