QRV-08 Building and test

[peranders]

Glad to hear it. Someone else listened to the amp and thought the bass was too thin! The thing was that he used my headphones and not his. My HD545 is not very "disco", very neutral and yes not accentuated bass!

[galouche]

I listen with a akg 701.
  No bass excess ( I like neutral as i'm a musician ) I hate bass boost :). when I talk about bass it's about the bass instrument and its caharcteristics.( 100hz to 200 hz ) with fingers picking ( 1000/3000 hz) and its caharcteristics like ( fret noises, slides, chord vibration, etc ... ). The voice is very natural too ... I've just listened some of my favorites songs ...

[wasp]

Finally I built QRV-08! But when I checked idle current of output stage I've measured only 3 mV across R20. It gives I=3/2.35=1.28 mA.
I'd check all resistor values by DMM, all is OK.
MicroCap 9 sim also shows around 4 mA of idle current.
Please help me to find a problem.

[peranders]

What is your supply voltage?

Voltages across and values:
R16
R17
R18
R19


R3 = 22k?

Voltage across?

[wasp]

What is your supply voltage?

Voltages across and values:
R16
R17
R18
R19


R3 = 22k?

Voltage across?

Vsup=+13.92 V & -13.9 V
R16=220R V=168 mV
R17=220R V=168 mV
R18=10R V=7.4 mV
R19=10R V= 7.5 mV
R3=22k (measured 19.6k) V=24 V

[peranders]

This seems not to be unnormal but could you also measure the Vbe of the transistors involved?

[wasp]

All seems like OK - all transiskor have 0.61..0.63 Vbe. Only T24 Vbe=0.6 V.
My BJT's is in SOT-323 case from NXP, but I think this is not a root cause.

[peranders]

Those currents you have right now will still get you incredible low distortion but if you want to increase the idle current a bit, lower the value of R16, R17.

[wasp]

Those currents you have right now will still get you incredible low distortion but if you want to increase the idle current a bit, lower the value of R16, R17.

This amp will work on 25 ohms headphones, so I want to increase idle current to have class A operation as much as possible.
I've tried R16(17)=R18(19)=100R, but it gives around 60 mA of idle. This current overheets output BJT's.

[peranders]

You can't think like that I'm afraid. There is no linear relationship with this resistor value and the current in the output transistors. Try to use half of the used value first. (Solder an another resistor on top of the original one.

[qusp]

Hi, i'm about half way through building the first of 2 QRV08, I am following the BOM (as close as possible) for the most part.

a few differences, please let me know if this is OK

I could not get the right size PPS caps from rifa from mouser and although I have some on the way, I decided to build it up with EVOX/RIFA PEN (is this OK?),

I am building 2 of these to use as an output stage of a balanced dac , serving as a pre for my active monitors, is it as simple as feeding +/-L to one and +/-R to the other?, or is it better to feed L/R + to one and L/R-  on the other?

I would also like to use LM329 for the reference as its good and I have some on hand. what changes need to be made to the reg gain to accommodate this? I have some hole mounted 431 as well, but would prefer 329.,

I am using some nichicon 4700u for the main reservoirs, should I change the value of the multi-fuses? or is inrush not enough to worry about?

regarding the servo opamps, do the chips used have an effect on the sound signature of the amp? I was thinking I might try OPA827, but I do have the 134 on hand, I just feel if we are talking SQ, I find them a bit boring personally, if they are not influencing the sound quality so much and do the job as best you know how, I will simply stick to them


can the input film cap be omitted if DC offset is at a known low level, fed from a dac inside the same enclosure? and of course CMMR

the instructions say the trim pot is optional, I have installed it, but do I really need it if the servo is doing its job properly?

thanks it for now, sorry for all the Q's, its coming together quite nicely and it really does lok cool too, especially since I have a couple of brands of resistors, so its multicoloured ;)

thanks for your time

[peranders]

Cap: it's not critical as long as it fits proberly.

Balanced mode: Use one pcb to each channel.

Reference: I'll recommend a SMD part but you can of course use a holemounted part but it's a positive thing to have the parts close to each each other.

Fuse: Will probably work alright with the original values. These fuses are rather slow.

Servo: Yes, they are a part of the audio chain for frequencies up to 1 kHz but the part of the audio signal is very low. It's onlý below 10 Hz the have a major impact and at those frequencies even the the old 741 is good.

They should be decent at least. OPA134 is much more than you'll need so you won't gain much to have a really good opamp unless you already have them and don't have to pay for them.

Input cap: You should not exclude unless you have a low impedance connection to your signal source. The trimpot is a must if you have a high impedance DC-path (which you will have if you'll use an input cap) in order to inject base current let the DC-servo work less.

[qusp]

Cap: it's not critical as long as it fits proberly.

thanks for your reply, yep, the ones I bought JUST fit, I only have about 0.5mm of pad on either side, but enough with a bit of persuasion, I just soldered one pad first and fired it up while I slid the cap over the pad, then just flowed solder onto the tip on the other side until surface tension gave way and flowed onto the cap and pad, they are the 7.3 size

Balanced mode: Use one pcb to each channel.

OK, that makes it easy

Reference: I'll recommend a SMD part but you can of course use a holemounted part but it's a positive thing to have the parts close to each each other.

more important than being a higher quality reference? not sure if 329 is available in SMD, I will check; otherwise I will grab some other SMD ref (suitable). If I do end up using 329 though, can you please advise of the appropriate adjustmnt to the gain of the reg? or just let me know the position and formulae for setting it, that would be fine too

Fuse: Will probably work alright with the original values. These fuses are rather slow.

ah OK, I have other fuses in this build anyway, there is a wanky silver fuse on the inlet :D

Servo: Yes, they are a part of the audio chain for frequencies up to 1 kHz but the part of the audio signal is very low. It's onlý below 10 Hz the have a major impact and at those frequencies even the the old 741 is good.

right, so there IS an incremental improvement, I have the OPA827 here, so will use them; they are great opamps and not being put to any use at the moment, so probably should

They should be decent at least. OPA134 is much more than you'll need so you won't gain much to have a really good opamp unless you already have them and don't have to pay for them.

see above

Input cap: You should not exclude unless you have a low impedance connection to your signal source. The trimpot is a must if you have a high impedance DC-path (which you will have if you'll use an input cap) in order to inject base current let the DC-servo work less.

OK, well that does sound like I could do without both then, the output impedance of the DAC is around 195R per pole and the connection will consist of about 5cm of 16AWG (equivalent) silver foil; this is appropriate yes?

i'm getting quite excited now actually, the DAC its for is coming together well, the hexateq case is gorgeous!, my custom titanium volume knobs and selectors have arrived today and look great (made to my CAD file design) and i'm keen to hear this amp in dual mono balanced mode. I may end up using my own potted transformers and your RFB03 and bypass the onboard rectifier, but for starters will build up as 'standard' there is no pot to account for with the input impedance which is cool and thinking I will wire up the gain jumpers offboard with a switch to choose for headphone use (gain of 2) and unity gain for preamp duties. will make sure to take some pics for the forum, as although I have seen a couple of people mention building 2 up and using them balanced, nobody seems to have gone through with it. fair bit of work to do yet. OH and the LEDs have to be 25mcd yes? or oter alterations need doing?

[qusp]

with a low impedance load of >30R (well actually more like 15R since i'm running balanced) how can I tweak the amp so it never falls out of class A into AB?

any way I cn ft the LM129 on this board? its only 2 pin and I gather is the smd LM329, or can you suggest another upgraded reference perhaps? i'm getting the tweakers itch and i've been quite well behaved on this build, only upgraded the reservoirs, servo opamp and gain setting R's

oh yeah and built a whole second amp to run dual mono ;)

[peranders]

Class A:
I'll leave that up to you. I don't support it. You can calculate how high volume the class A current corresponds to. I think the sound level is way over 130-140 dB in you ears!

The reference:
LM329 doesn't come in SMD at all. You have good references from Analog Devices. You can look there.

I must also point out that the design is a 20 MHz design with a 40 MHz power supply. It can be hard to tweak and I don't support it. I support my tested design. I'll hope you'll understand that.