Yaqin MC 100B Chinese Tube Amp Mods

[-G]

I purchased a Yaqin MC 100B Chinese tube amp from Hong Kong off ebay and it's a really solid amp that seems to use a traditional tube amp circuit design with 12AX7 SRPP input stage, 6SN7 schmitt trigger, cathode follower and KT-88 output tubes.

After careful review of the Yaqin MC 100B amp I found it to be a little too rich in THD or harmonic content but a very solid amp that will give many others a run for it's money. I would like to use this thread as an investigation into the Yaqin MC 100B mods.

There is a kind fellow over at the Elektor.fr forums who has suggested a THD reduction mod, he states that 6SN7 has cold bias and not working very linear thus causing artificial harmonics for the Yaqin MC 100B tube amp.

I will translate it from french to english here... from bablefish... if someone can help to refine the translation for technical terms it would be greatly appreciated. It is located here:

http://www.elektor.fr/forum/forum/realisations-publiees-par-elektor/tubes-audio-anciens-recents/yaqin-mc100-b.936195.lynkx

This yaqin mc-100B tube amp is a very good model, which built a serious reputation.
However, the audiophiles which sought to further go, by changing tubes, by modifying certain components, did not succeed in obtaining better than the assembly of origin, and it is not normal!

My analysis of yaqin mc 100 tube amp diagram (pdf file) revealed characteristics:

- first stage amplifier of voltage with a ECC83 assembled in SRPP with resistances of 1k. It functions with an high voltage about 330V and under a current of 0,65mA approximately, which is completely quite selected. Nothing to announce on this stage, if it is not the well-known problem of the SRPP, related to the leakage current filament cathode, in the triode the top whose cathode is carried to 164V approximately.
Certain tubes ECC83 do not appreciate this voltage, a positive polarization of the filament could improve that, and I will give suggestions for solution later (it is not the most serious problem…)

- second stage: phase-converter of Schmitt with a double triode 6SN7, loads of plates of 47k and common resistance of cathode of 68k.
High voltage of 354V approximately.
Direct connection with the preceding stage, the grids are thus with 164V, and cathodes with 170V (2,5mA in the R of 68k) thus polarization of -6V.

This stage works under very weak current of 1,25mA by triode, therefore into full in the round-off with the curves…
It produces a great quantity (too much) of even harmonic. 2nd order harmonics are useful, but nevertheless!

- third stage: driver, in cathode follower with load of cathode of 100k to attack the grids of the KT88.
High voltage of 414V, quiescent current of 3mA per triode.
Direct connection with the second stage, therefore grid voltage of 296V approximately (it is the tension plates on the phase-converter).
With 3mA in Rk of 100k one puts cathode at +300V what the filament should not appreciate of the whole…
Polarization is of -4V for this stage.

- stage of power: Push-pull of KT88, assembly UL or triode with the choice, adjustable negative polarization, high quiescent current of 60mA approximately with negative bias towards -57V. Very well done, nothing to say.

It is thus the second stage which I propose to re-examine:

- to work under 3mA approximately, which is a minimum, and product already full of 2nd order harmonics. (To be linear and not to distort, the 6SN7 should work under 9-10mA…)

- not to modify the high voltages of the other stages, which go very well like that.

that gives this:

- R6 and R7: the loads of plates of the two triodes of the phase-converter pass from 47k to 18k

- R9: Rk resistor changes from 68k to 27k

- R11: resistance series on the alignment always passes from 20k to 9,5k (10k should be appropriate) in 3W. This change makes it possible to preserve the identical tensions of alim, although the consumed currents are higher.

- R14+R15: two R of 5,1k in series what gives 10,2K, pass to 7,5k for the same reasons as above.

With all that, the points of operation of stages 1,3 and 4 should not move, only second is modified and maintaining each triode 3,15mA sees approximately, and its negative bias passes from -6V to -4V approximately.

And if you find that does not go better, it is enough to return to the old values…

Note: the voltages which I indicate are those obtained with sector of 220V (envisaged for this amplifier). Under 230V have must find a little more, and the flows will be also hardly higher.
The heating filament will be also stronger under 230V, which is not good, but then not good of the whole for the tubes…

The second stage of yaqin mc 100b tube amp could be cold bias where biascurrent is under 3ma stock and increasing biascurrent on 6SN7 could make it work more linear.

Updates and refinements on Yaqin MC 100B tube amp THD reduction mod will follow.


tag cloud: yaqin mc 100, chinese tube amp, yaqin mods, canadian hifi, kt-88 tube amp, sylvania 6sn7, 5751 preamp tubes, tube amp

[-G]

translated from homecinema-fr.com:
http://www.cinetson.org/phpBB3/vintage-tubes-f22/ampli-yakin-100b-t27290-135.html#p440580

Hello, I speak about this Yaqin MC 100b tube amp Schematic diagram:

- Far too much gain before the phase-converter, moreover, there is a switch of sensiblity to increase (further! ) the negative feedback.
- Very bad choice of tube for the phase-converter, not enough current, can only be terribly dissymmetrical, 3 to 4% without doubts.
- Not enough current (+-2mA? ) in the following 6SN7 of cathode (of which Vfk max is probably not respected).
- Commutation triode/UL sub optimal, the required load is not the same one.

Under these conditions, the 6SN7 are polarized in the zone where the linearity is bad, it is not surprising that the differences between tubes are exacerbated.

Yves.
http://www.dissident-audio.com

--

Hello with all,

of agreement with Yves on this assembly:
- too much gain
- 6SN7 used under too weak current
- even impedance of load for the two assemblies UL and triode, which is not ideal, but that passes

Concerning the gain, with direct connections between the SRPP and Schmitt, and between Schmitt and the driver, one cannot just change the flow of a stage without all to recompute…

At the rigour, one can put one 5751 in the preamp: the gain which was about 33 with 38dB (according to the HT of alignment that I am unaware of…) with the ECC83, will lose 2 with 3dB, it is already that.
With one 5755 one passes to 28-30dB of gain, it is still better.

I have the impression that the originator with put as much gain in closed loop (with CR) quite simply because it could not put more CR!!!
(the circuit becomes unstable)
With a gain reduced on the first floor, one must be able to fix a rebouclé profit a little less extremely: it is enough to change a R, but I do not arrive at reading his value nor his number on the diagram put on line…

--

Problems of gain in the Yaqin MC-100B tube amplifier:

- the first stage gives gain: 38dB in the case of the SRPP of ECC83 with uncoupled cathode
- the second, phase-converter of Schmitt, in additions still a little: +23dB approximately
- the driver, to exit cathode, does not give any, ouf…
- the stage of power still adds some: I do not have any opinion then I am of agreement. - 19dB

And the transformer of exit lowers the voltage, in the report/ratio of its rollings up: 3,8k: 8ohm that submits to us a report/ratio of impedance of 475, therefore 1/21 for the tensions = -27dB

Thus in open loop, there is approximately 53dB gain! (I do not have the diagram under the eyes, I give approximate values, on the principle that does not change large-thing)

As the CR loop fixes the gain (by the ratio of two resistances) at approximately 30dB on the entry 600mV and 38dB on entry 250mV, there are thus a CR rate of 23dB in the first case, and of 15dB in the second.

BUT: in open loop the high cut is given by the transformer of exit, which forms with the impedance of the tubes a second-order high-pass filter, because of the stray capacity between rollings up on the one hand, of the coil of escape of the primary education on the other hand.

In other words, the transformer behaves like a circuit LLC (filter of order two, with resonance)
The coil of escape gives a first order slope starting from 25-30kHz approximately, and after the stray capacity intervenes, which creates a small bump of resonance in the answer, towards 100-150kHz, followed by a second-order slope.

and it is all the problem there: with a second-order slope, when one applies CR the signal returned to the entry is (out of HF) in opposition of phase with that of entry.
As the stage of entry withdraws the two signals, the dephasing of 180° on the CR signal amounts adding both, therefore the profit tends towards the infinite one, the amplifier oscillates…

With a transformer of exit, one cannot establish a CR rate higher than 12dB great maximum, in general. With weaker CR, the cut is on the first order slope, and there is no instability, but with CR stronger, the cut is rejected at a higher frequency, where the cut is of order two: instability.

Here are (ouf) why one should not give too much gain!
And it is as for that as one should not lower the gain stupidly by increasing CR: assured instability!

In your case it would be good to remove some a little, by putting a tube with µ low than the ECC83.
The 5751 with µ of 70 would be rather well, the 5755 still better but its stitching (pinout?) is different.

One can also remove the capacitor decoupling of cathode, on the triode of the bottom of the first 12AX7 SRPP stage, that will decrease a little the gain. **

Good, it was the MP.
Now, to decrease the total gain of the amplifier, it should initially be decreased in open loop, then to increase CR.
I propose to test one 5751 in first stage, with not uncoupled cathode: the gain of the stage changes from 37-38dB to 31-32dB approximately, therefore -6dB.

Afterwards, one increases the CR rate to bring the gain rebouclé around 28dB (there is 31dB currently) by changing R10 of 12k by a value of 8.3k obtained by simply putting in parallel on R10 a R of 27k.

If -3dB is still wanted, one approximately changes R10 to 5.5k, by putting a 10k in parallel.

A simple idea nevertheless to reduce the gain without changing the tubes nor to recompute the points of operations:
To separate two cathodes from the phase-converter and to insert a resistance in each one towards R9. Probable values, from 100 to 1000 Ohms… to test and listen

[-G]

I found this important warning on www.hifi-forum.de about the Yaqin MC 100B tube amp and translated it to english. My particular Yaqin MC 100B has plastic on the tops of the caps. The metal is not exposed as seen in the photos below:

Warning: Owners OF the Yaqin MC-100b.

One copy OF this amp have passed for my workbench and I have remarked that it which lethal for the operator.

The power supply electrolytic capacitors, situated outside the case, acres without isolation RK the top and in two OF them there which the full HT rail (~ 500 V).

Please inspect with A voltmeter, the VOL-meets between the chassis and the top OF the capacitors and if you measure the same VOL-meets, make some insulation it ton the top OF the glue in plastic on top of the capacitors.


Yaqin MC 100-B owner, please consider.

ATTENTION MORTAL DANGER!!!!!!!

With confusion I had just now to determine from the Chinese Yaqin MC 100-B has acute mortal danger because DC voltage-prominent on capacitor parts are not contact-voltage-proof. From the power supply capacitors danger is possible because the metal top (negative pole of the Elkos?) on half anode voltage potential is through series connection. With the Yaqin MC 100-B that is approx. 260 V of DC voltage with affects will be very probably be deadly.

See Photos; shows the naked metal above MC 100B capacitors and HT voltage.

[-G]

Bias alignment mod for Q2 6SN7 stage of Yaqin MC 100B chinese tube amplifier:
(Thanks to John from Champ Electronics http://www.chambonino.com for these instructions)

Change R6 & R7 from 47k down to 18k.
 
Change R9 from 68k to 27k.
 
Change R11 from 20k to 10k.......3 watt.
 
Change R14 & R15 to one resistor of 7.5k.......also 3 watt.
 
I can see the guy's reasoning for doing this & would agree it should be better. Trust your ears & if you don't like it you can always put it back to original!

Bias alignment for Q3 6SN7 stage of Yaqin MC 100B chinese tube amplifier:

Refering to R12 & R13, the 3 watt, 100k cathode resistors on the 6SN7. First put in series with each one a 1%, 1 Ohm .6 watt resistor (it doesn't matter which side of the 100k's you do this). Now with your meter set on DC Volts......meter across each 1 ohm resistor. The m/v reading you will get is equal to the m/a that each half of the tube is drawing. For example, if you get a reading of 3 m/v across one of the 1 Ohm resistors.......this is equal to 3m/a of tube current.
 
You simply now lower the 3 watt, 100k's until you get the current draw of what you are looking for. Try 3 watt, 82k's first & see what current you have. You could even put 5 watt multi-turn pots inplace of the 100k's (still with the 1 Ohm in series thoughfor your meter reading). Like that you would be able to "tune" the currents to what you want & match them on each half too, plus with a change of tube (6SN7) you can resett the current. 5 watt multi-turn pots are expensive though!
 
I have heard these amps before & I think they are OK for the price.

[-G]

Yaqin MC-100B Bass Enhancement Mod

I can recommend you the following mods.

- change tubes, for more and firmer bass bass try 4x EH KT90 or JJ Electronic KT-88, 4 EH 6SN7 and 2 x EH 5751 Goldpin.

Try also JAN "chrome dome" type 6SN7. I have compared these with RCA 60's 6SN7GTB and they have increased low end detail.

- change coupling capacitors C6 and C7 for Clarity Caps SA 0.47uF or Auricaps. You can also try 1uf caps here however it may overdrive output transformer? Mcintosh use .47uf in their classic MC-30 / MC-40 monoblock power amps.

[-G]

A friend emailed me some additional information about this amplifier... so I will post it; in reference to the schematic he says:

I am not a specialist on high-end tube amplifiers, but I can tell you what I know from electronic magazines and books about tube amplifiers and my own experiences in making tube amps (simple ones) and power transistor amps (all kinds of).
 
A. Input resistance. The original input resistor (grid-ground) was 39 K. This is raised to 510 K or 470 K. Raising this value makes the tube more sensitive. It can also change the way the volume control works, because there is a direct connection to the 50 K volume control potentiometer. I dont see problems in raising the input resistance of the 12AX7. The input resistance is always a compromise between sensitivity, damping, keeping the signal between the right levels to prevent distortion. But: there is a relation between the negative grid voltage that the 12AX7 needs and gets via R4 and R2 and the value of the 39 K (or 510 K or 470 K input resistor), because R5 is part of a voltage divider with R4 and R2 (see B).    
 
B. I presume  the 0,47 uF capacitor that is added to the circuit is a series-capacitor to the steering grid of the 12AX7 (pin 7). If it were mounted from grid (pin 7) to ground it would damp the high frequencies of the tube amplifier severely. It is no problem I think to use a series capacitor of 0,47 uF to the grid of the 12AX7 (common and often seen in tube amps). But when to pin 7 a negative grid voltage must be added to let the tube operate properly, the DC path may not be blocked by such a series capacitor. I am almost sure a negative grid voltage is applied to the 12AX7 because of R4 and R2. This negative grid voltage is in most tube amplifiers realised by R4 (common way) and thus the value of R4 (1K2) and in this case combined with R2 (330 Ohm). But no worries, when the amplifier at this moment works properly (good sound) there is no need at all to do changes (the negative grid voltage is critical, if it works good, let it be).
 
R10 (12K) and C10 (50 P) build a frequency dependent circuit. In Dutch we call this the "tegenkoppelcircuit", litt. translated the "backcoupling circuit". The circuit feeds the output signal from the loudspeaker/transformer back to the amplifier in the right phase (if the phase were not good the amplifier would start to oscillate). The circuit is important for the frequency response (and the amount of distortion) of the amplifier, and as far as I know from theory, the value of C10 and R10 have a great effect on the sound of the amplifier. So you can do slight changes to the values of R10 and C10. Start with very small changes, for instance 100 PF instead of 50 PF and 20 K instead of 12 K listen to the produced sound (do not lower R10 too much, I assume that can damage the circuit). Because this is the backcoupling circuit, big effects can be expected by very small changes in the components (R10 and C10). R10 and C10 in parallel build an audio filter that is sensitive for a certain part of the audio spectrum (20 Hz-20 KHz), the frequency band to which this filter is sensitive will be amplified most or more (study it with an oscilloscope and sine wave at the input). That is the reason for the very low value of C10. When you do tests, it is advised to connect the oscilloscope parallel to the loudspeaker (check first whether the scope input can handle the output voltage of the tube amp, set all to low values) and study the waveform. Also very important that the amplifier may not start to oscillate when the value of C10 and R 10 is changed.

I never touch audio tube amp circuits with my fingers, only the input grid of a tube amplifier can be touched with a finger (big hum). And I even don't do that for safety reasons. If you want to avoid risks in touching that input grid you can take a capacitor or resistor first between your fingers and touch the input grid with the other end of that capacitor. But in reality you always come with your hand close to (perhaps) high voltage parts of the tube amp, so it is not my preferred way. The other electrodes of the tube cannot be touched, and of course you know the anode cannot be touched at all, especially in your case  (high end audio amplifier with hundreds of volts on the anode). The "hold one hand in your pocket" rule is a good rule, it prevents that voltage from one finger/hand can travel to the other finger/hand through the heart area. 
 
When I do such alignings I work with one hand and always use a potentiometer with a plastic (nylon) shaft, or an insulated screwdriver to turn the potentiometer (types without shaft). 
 
As far as I know from audio theory the harmonics of musical instruments and voices give the "typical" sound of the instrument, thats the reason why the same note on a flute sounds different on a violin or trumpet. The harmonics from instruments can indeed reach high audio frequencies, up to 20 KHz or (perhaps, have to study this further) more. Generally spoken the human ear can hear sounds up to 18 KHz or so, as far as I know it is individual and depends for a great deal on age. When you watch my audio video's on You Tube I mostly show the amplification of an amplifier on the oscilloscope (sine wave) on different frequencies from 50 Hertz up to 20 KHz or so. In some video's you can see (oscilloscope views) that the amplifier circuit does not amplify equally strong on all frequencies, sometimes the strength of the signal weakens above 14 KHz. I show this in this way for people interested in audio, when they want to copy the circuit they know the properties of amplification. I dont want to exaggurate the qualities of the circuit. The fact that the amplification weakens (sometimes only a little bit, but sometimes more) above 14 KHz or so, does not mean that the signal is not reproduced or amplified. The loudspeaker box also does a great deal and it is possible that a box with a well chosen sound characteristic compensates the weakening of high frequencies in the audio amplifier (14 KHz is for me the absolute minimum, it is possible to hear it, better is 18 KHz). But of course when the weakening of high frequencies is too big, the box cannot compensate it. In modern high-fi terms the amplification of an audio amplifier must be equally strong for all frequencies from 20Hz up to 20 KHz, some Hi Fi amplifiers even go higher, the reason is (theory) that some harmonics of instruments reach very high requencies, so they determine the typical sound of an instrument. I have good information about harmonics of instruments (how high they come) in a technical manual of the fifties. I am a little bit conservative, in my book "Schematics 2" (yet to come) I explain my ideas about sound. My approach is that sound is a question of personal taste (of course within certain technical margins in terms of distortion). When an amplifier is suitable and has enough technical qualities, I agree that testing (goal: undistorted amplification of a sine wave) from 20 Hz up to 20 KHz is the ideal situation. You can be sure that your loudspeaker boxes can be driven along the complete audio spectrum.                       
 
Indeed the backcoupling circuit is also called negative feedback or so. It is a way to "correct" the frequency bandwith and (sometimes) distortion of an amplifier (tube or transistor). It was common in tube circuits.