• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Matisse Fantasy waveform output test

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi All,

Can anyone know my question ?

1) After I finished the Kit Matisse Fantasy, measured its output signal found a small up surging angle at the raising edge of the square wave, it would occur while the input square wave at 1K Hz.

2) When I tested it were using resistor load at 6 ohm. Is it correct to use a fixed value of resistor ? Or instead to use other type of load ?

3) The sound of it present is very well, what the affect of it would be listened ?

CK
 
Hi,

It sounds like overshot and it is not good that it comes even with resitive load as it indicates instability and will get worse with reactive loads. As I assume that the design is basically sound I wonder if you have changed any component values or such when you built it?

With 1kHz square wave I would expect any reasonably designed amplifier to give a nice square wave with a resistive load.

If you have built according to instructions and it is still there yoy can try tweaking the cap in parallell of the feedback resistor that goes from output transformer to some input tube, usually this is effective for reducing this ringing.

Regards Hans
 
Hi Hans,

This preamp is working fine without any noise trouble and hum. But I don't know is it already reach a top of its performance.

I buy a PCB and install it by myself. All components are exactly follow the PCB indicated.

Do you know this circuit or ever play it before ? I didn't have this original circuit to reference.

I can show you this circuit and values for your reference . Also welcome any suggestion to make it better. What value of cap parellel to the feedback path ? There have 2 feedback path, one located on last stage and second located between first and last stage.

Thanks for your help !
CK
 
Hi,

I didn't realised that it was a preamp you where talking about as you said that the load impedance was 6 ohm, is that really correct?

Anyway if the feedback is only taken over one stage I would be very surprised to find any ringing, maybe the feedback path is longer and including several stages?

In order to offer some assistance I think you need to describe the circuit more in detail and ideally to submit a circuit diagram.

Regards Hans
 
I would recommend to load the output with 20K resistor instead for taking measurement. The output impedance of Matissee Fantasy is around 14K. I paralleled a 28pF capacitor to the feedback resistor to stop overshooting or ringing at high frequency. You may try to use a small capacitor and watch the output of the square wave from the oscilloscope.

By the way, I am wondering what is the grid resistor at the 12AT7. Lot of the schematic listed 47K. However, I firmly believe 47K is not a correct value. It should be 470K instead.

I built this line amp myself and experimented different things in there too.

Johnny
 
THanks all of you !

tubetvr -:

Sorry to make you confusing .

kmtang -:

Do you live in HK or China ?

I counter check to the schematic show that 12AU7 input in series with 510 ohm and before it connected with other 47K ohm to ground.

I use the PCB is LS27 which is provided by LITE AUDIO.

CK
 
Hi CK,

I was born in HKG and moved to Vancouver, BC, Canada in 1997.

I believe the 510-ohm is the grid stopper resistor which is used to suppress chance of oscillations within the circuit.

Why don't you try to replace the 47K to 470K and listen to the differences? It may evern better if you could change the coupling capacitor from 0.47uF to 0.1uF.

The last stage should be 12AT7 unless they modify the circuitry for 12AU7 instead.

Do you mind post the schematic?

Best wishes,
Johnny
 
Hello Tang,

When will you visit to HK and going to Disney ?

Sorry I have typo error on "12AU7". You are right it is 12AT7.

Now I am preparing this schematic for your reference. Unfortunately, I did not have a scanner so need to redraw it again.

When I complete the drawing and posted, I want to consult with you about what values of component should be changed.

CK
 
Disney in HK is not interested to me at all. I just want to see my family back there if I have a chance. I missed them a lot.

As I mentioned before you may try to replace the 47K to 470K and listen to the differences? It may evern better if you could change the coupling capacitor from 0.47uF to 0.1uF. I used Riken resistors and Jensen copper paper-in-oil 0.1uF for the coupling capacitor.

By the way, the last bypass capacitor in the Power Supply circuit has big impact on the sound as well. I used 3.3uF Auricap. I have a tube regulated power supply in separate chassis. I can even switch it between tube regulated or simply C-L-C-L-C filtering. I found the simple C-L-C-L-C filtering gives me very nice and smooth sound. I found it's fun playing with the power supply.

Johnny
 
Hi Tang,

Sorry, I don't know just you stay alone in Canada.

Is there only two 47K to ground resistors need to change to 470K ? Exactly, I had saw many to use 47K.

I have two 0.22u AudioNote copper paper-in-oil want to replace them.

Not sure the other value of resistors are correct ?

PSU entirely formed in PCB together with the gain part. Which design is using LM317 working at HV to adjust the DC supply from 380 to 410V. I think it can satify to requirement. Because there have to large amount of backlot, when built a seperate PSU or do modification.

The last by-pass of PSU are using 220u parellel with 1u capacitor.

All resistors mainly I used as RMG gold leg. To read your instructions I want to change all Gird resistor values to 300 ohm (to increase the banwidth) How do you thinking ? Second, the area of 470K , do you think it is important so as to use a best one also ?

Br,
CK wong
 
Hi CK,

The 47K resistor (grid resistor of 12AT7) need to be replaced is the one connected to the 0.47uF capacitor and ground. There are two of them in the amp. One for the L-ch and one for the R-ch. It's better to use 470K RMG resistor to ensure the sound quality is good.

Regarding other resistors value, I see no problem with them at all.

I think 0.22uF coupling capacitor is a little too large but it should be OK in the circuit.

I don't use the grid stopper 500-ohm resistors at all. You may try to bypass these 500-ohm's one by one and check the outptut with an oscilloscope to see if it is still stable and without any oscillations.

Leave the Power Supply as it is for the next step of improvement later.

By the way, I am not an expert but I spent quite some time studying the schematic, and had an A/B comparison with a well made clone Marantz M7 line amp.

The funny thing is that I raised this issue in a Chinese HiFi forum which is specifically for Matissee Line Amp. I had no comment from others and seems no one interested in changing the grid resistor to 470K. As a DIYer, I love to experiment different things.


Johnny
 
Thanks Tang.

Do you know exactly this original design also use 470K ohm ?

May be some people inclined to copy cat but not to evolute .

Another questions :

1) I found many design especially on power AMP they trend to use tude as recitifier. Why they don't use diode on it ? If use diode to recitify, not sure a DC chock is applicable after it ?

2) Can you introduce a circuit for me ? Around 8-20W Class A integrated all tube design, output uses EL34 / 300B / recommended .

CK
 
Hi CK,

I really not sure what value of the 12AT7's grid resistor is? I wish to verify it too. I hope I could open an original one to check as well.

Anyway, it is common sense that the grid resistor of the following stage is at least couple time larger than the plate load resistor. Otherwise, it will overload the first stage which would cause distortion and other problems too. Well, it is up to you to experiment this change and let us know if you could find any improvements at all.

I haven't try switch the power supply from tube rectifier to diode yet. I believe I stay with the tube because it works very well so far. Even with diodes in power supply, you can still use choke (LC) instead of resistor (RC) for filtering. LC filters stores much more energy than RC.

I would recommend you could try the 300B SE amp. Here is the web site of the famous JE Lab 300B amp.

http://members.myactv.net/~je2a3/se300b.htm


By the way, what's the power amp go with the Matissee you are having now?

Johnny
 
Hi Tang,

It is very hard to ask someone check with you about this resistor. Come up this mind not a subjective to me . I think each manufacture were going to test and evaluate many and many before massively produce their products.

Thanks all of you give me advice. Now I completely modified this preamp following Tang's instructions as follows :

1) replace 47K resistor with 470K.

2) change first stage couple CAP to 0.22u.

3) Add 3.3u CAP to the last PSU's bypass area.

4) By-pass 500 ohm at all grids.

5) Use 20k ohm resistive load at output for measurement.

6) Due to the output overshooting still happened. I needed to parallel 33p CAP to the feedback resistors. End up the output looks is well enough.

All things be done even no any impact. The sound can be listened not same as before. For my subjective, doing this really turn better than earlier. But how good you feel need your evaluation to yourself.


Now I use ARC D70 as a power AMP . In the meantime, I am working in the new power AMP with respect to the Pass's A75. Ultimately, I will use this DIY machine.

300B is most popular in HK. Not sure how much per each of this tube ?

CK
 
Hi CK,

Well done. In fact, the improvements won't be too significant but we are getting a step further to perfection. I believe it is rather true because the one I am using is different from yours.

My one employ only one 12AX7 and one 12AT7 for both channels. Yours one with one 12AX7 made paralleled for each channel. The effect on the grid loading for the 12AX7 in case is reduced by half. However, surely it will gives some improvement with suitable loading.

One more thing you could try is to remove the 220uF bypass capcitor from the board and keep the 3.3uF cap in circuit. It should provide a clear sound especially with the bass.

The Electro Harmonix 300B only cost approx US$120 pair. The Tianjin Full Music Mesh 300B priced at approx US$250 per pair.


Johnny
 
You mentioned in your previous message that the PSU employs 220uF capacitor with 0.1uF in parallel.

The Hammond 1627 SE output transformer has good performance with very reasonable price. If you look for a better one, I would say either Lundahl or Plitron toroid output transformer are better choice. If you look for even better one, check out the Perm Alloy output transformer from Tamura.


Johnny
 
Hi Tang,

Do you know the price of these? I can't find their prices.

1) Hammond 1627 SE

2) Lundahl

3) Plitron

4) Tamura

For me I am enough to buy a moderate price and modest quality of output . Which one you suggest to me ? I have not ever listened it to do a comparison. So just rely on your comments. It doesn't matter if you can't help me.

Despite some products are in high prices, however, it does not have its high performance. In contrast, some of it are notorious and are black horse.

CK
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.