QSXM2 - BOM questions, etc.

[tk-don]

Just a few questions before I'll finally buy all components.

1) In the BOM for the QSXM2, the Wago part numbers listed for X1, X2 and X3 seem to be fine, but the corresponding Elfa parts number for the 5-pole Wago is a 3-pole connector from Stelvio-Kontek. The comment in the BOM next to the part number suggests fully awareness of the fact that 2 poles are missing.
Or is the intention that I should just buy a 3+2 pole to achieve a 5-pole connector for X1, like the comment for X3 suggests?

2) As you know, I'm going with a SSM2220 and SSM2210 as one of the transistor pairs for the input stage. I was thinking of just going with BC550C/560C for the rest of them. Based on comparisons between BC850C and BC550C data sheets from random manufacturers, I don't see the big difference in specifications. Or is it because the variance of DC current gain is higher between BC5xx devices compared to BC8xx or SMD in general? Would there be much benefit in the BC847BS/857BS when using the SSM2220 and SSM22210 either way, or would that have no real-life impact compared to manually matched BC550/560?

3) Regarding R93: The BOM mentions that I should be aware of something (a note in another document). What should I be aware of - I couldn't find anything about R93 in the circuit description.

4) The amp QSXM2 will be used with a HOMC (2.0mV), so in MM mode, and with about 2 kOhm input impedance instead of 47k. Parallel to this thread: http://sjostromaudio.com/smf/index.php/topic,312.0.html, and keeping your notes in mind - should I also just stay with 15kOhm resistors if I go with BC550/560?

Thanks!

[peranders]

1 Yes

2 SSM2210/20 will only do a good job in the input stage. BC847BS is similar as single BC types but in a differential input stage you will get better performance and takes up less space.

3 R93 determines the gain so if you increase the value, the gain will be lower.

4 If you have one pair SSM2210/20 instead of four pairs with BCxxx you should have the same total collector current. I think I had 200 uA in each BCxxx so that means you should have 800 uA through the SSM's.

[tk-don]

Thanks!
I found out that RS components sells single Wago poles of the 256 series, which turns out as a very cheap option. The end plates are sold in minimum quantity of 25, but they are also very cheap.

Regarding question 4: I'm a bit in doubt here. How do you compute your collector current?
I have the idea that the collector current is one half of the emitter current in each differential pair, since each pair share an emitter resistor. Would the collector current then simply be determined by Ic=((Vcc-Vbe)/Re)/2 ? With Vcc=15V and Vbe=0.6V it makes 480 uA per transistor.
The total collector current you obtain with the collector currents you mentioned would be 1.2 mA, right? The reason you'd prefer a higher collector current with the SSM rather than with the BCxxx, or even equal currents for all differential pairs is to benefit as much as possible from the low noise of the SSMs, or?

[tk-don]

I know you won't be able to respond until next week - no problem. When you have the time:

Maybe I should pay more attention. I thought you'd still want to keep the remaining BCxxx along the matched SSM pair and just let the (manually matched) BCxxx operate with a lower collector current, but I guess it wouldn't make sense and in turn increase the combined noise figure.

In your notes, you write that:

"The noise performance is the best at 100-300 µA and MM-pickup with BC550/560 transistors. 1 mA or more is the best for MC-pickup. If you use MAT02/03 the optimum current is approx. 1-5 mA. If you use MM-pickup it's also recommended to have 1 mA at least. If you choose to use MAT02/03 I recommend 1 mA through each transistor."

In the second sentence, is the 1 mA simply a trade-off between achieving enough gain at the cost of increased noise, and is 1 mA the collector current per BCxxx pair? The gain (hfe) of the SSMs isn't that much higher than a BCxxx - so is one pair really sufficient? So I'm a bit puzzled: Is the main objective just to have closely matched transistors with a matched gain and how do I know how many pairs I need? I think that I want to make the amp future-proof and make it suitable for both LO-MC and for MM/HI-MC, if possible.

[peranders]

About the Wago, if you buy 2-pole you can build any connector you want and you don't have to mess with the end plates.

If you use one pair of SSM's in the preamp, just use one emitter resistor of 15k/4 ohm = 3k3 or something

Mount R16/R17

Do not mount R18-R23

[tk-don]

Thanks P-A.!

The thing that is puzzling me is that you say you want a collector current of 800 uA for one of the SSMs (e.g. corresponding to T5+T23). In the documents you mention the 1.45mA/4 collector current per transistor pair. That would equal 365 uA per transistor pair, right? Or is the 800 uA you mention not a collector current?

How do you know which collector current is right for a given transistor combination, i.e. 1-5 mA for MAT02 but inly 800 uA for SSMs ?


Another thing (sorry for all these questions!): I want to add a switch so I can switch outputs between headphones and external RCA. I want to add a potentiometer to control the volume for the headphones. I've got some BUF634 I can use if all else fails so that the potentiometer sees a high impedance rather than the low impedance headphones. But do you have any suggestion for where in the QSXM02 I could exchange a resistor with the potentiometer to control volume, if possible?

Thanks in advance  :-)

[peranders]

My advice is the easy way out. If you change the collector currents you must recalculate more. 800 uA also work good.

MAT02 is the same device as SSM2210 except for the package.

[tk-don]

Fine by me, 800 uA it is, I was just curious :-)

Regarding my plans of adding a volume control for connecting headphones directly:
I'm aware that R93/E93 should always be present in order not to load the previous stage too severely. Would adding R93/E93 of 680 ohms in series with a 10k potentiometer or something along those line be a way of adding a volume control in the most optimal way, or would something go wrong with this?
Another way would be shunting with the potentiometer, but not sure how that would work out.
Thanks!

[peranders]

I don't recommend that you mess with the gain. A brutal way to fix a volume is to connect 1 k pot at the output. I have used this when I have tested the RIAA amps with headphones.