Thanks for posting. Squares are 100 kHz. OK, so 1,2,3 has 7 pF without Cdom. Will try this in VSSA since this is on the bench. 100 kHz resistor loaded squares in VSSA are the same.
Hm... seems my problem is troublesome, it is not consistent. I was going to try input filter, then everything went to nice even without filter. Will try more to see whats the problem.
Anyway, I notice something strange. I trimmed the pot to make output zero without any source, but as I connect to my netbook, the output immediately turn to 0.17mV. Is this source-related problem ?
Anyway, I notice something strange. I trimmed the pot to make output zero without any source, but as I connect to my netbook, the output immediately turn to 0.17mV. Is this source-related problem ?
Is it 0,17 mV or 0,17 V?
If it is 0,17 mV variation, that's is nothing as we speak about output DC offset.
Input bias current makes voltage drop on 10 k resistor, and when source is connected this is changed, causing different output DC offset. Solution is input DC offset compensation but for 0,17 mV difference it is absolutely not necessary.
Is it 0,17 mV or 0,17 V?
If it is 0,17 mV variation, that's is nothing as we speak about output DC offset.
Input bias current makes voltage drop on 10 k resistor, and when source is connected this is changed, causing different output DC offset. Solution is input DC offset compensation but for 0,17 mV difference it is absolutely not necessary.
haha, srry, typing error. It is 0.17V, which is significant. But will it be detrimental ? extra 7mW across 4ohm speaker for nothing. Anyway, is there any other way to make the amplifier input more RF-resistant ? ( other than RC filter )
Low pass filter at input will not decrease the listening quality, on the contrary, it reduce any high speed signal, (Including high order distortion from the source) thus reduce induced IM distortion. It provide better immunity to RF (same remark). I use to set the best value for square waves, as said, then try to lower the corner frequency with careful listening.
It is more important with BJTs in input, as they are less RF immunity than FETs, may-be the reason why some prefer FET's sound for their input devices.
Ps: Sorry for the typo, yes, it was 100Khz square waves. Caution, the amp simulated is not *the* SSA, but one i'm playing with to experiment some alternatives: don't use my caps values.
It is more important with BJTs in input, as they are less RF immunity than FETs, may-be the reason why some prefer FET's sound for their input devices.
Ps: Sorry for the typo, yes, it was 100Khz square waves. Caution, the amp simulated is not *the* SSA, but one i'm playing with to experiment some alternatives: don't use my caps values.
Last edited:
I try to keep any cap on the input below 47pF, so the amp can be used with a volume pot. Otherwise you'll have an LP filter that changes with volume. I rather put LP filtering before the pot, where source impedance is much more predictable. My amps typically have a simple input buffer which makes them much more tolerant to a variety of source impedances. This solution is less expensive than creating an amp without an input buffer and then having to buffer the volume pot/attenuator.
Another option is to employ a fancy bootstrapped input LP filter to cancel input impedance, but this must be done carefully or you'll have an oscillator.
Another option is to employ a fancy bootstrapped input LP filter to cancel input impedance, but this must be done carefully or you'll have an oscillator.
Is oscilloscope available to you?
Usually there are 50/60 Hz plus harmonics, some local devices interferences like nearby SMPS, LCD, lamps etc in the spectrum. But there shouldn't be clear and strong MHz components present in the output signal when touching live input pin.
Without scope you're blind, electrically speaking.
Is oscilloscope available to you?
Usually there are 50/60 Hz plus harmonics, some local devices interferences like nearby SMPS, LCD, lamps etc in the spectrum. But there shouldn't be clear and strong MHz components present in the output signal when touching live input pin.
Without scope you're blind, electrically speaking.
I believe I'm blind now ^^ I'm a blindman touching/feeling an elephant
Will proceed to TSSA when i'm free, and will wait for bad thing happens now. (they are rather obedient now)
Completely omitting input cap is out of an option at CFB, since any HF interference pick-up with input/cable can make output stage to saturate in max HF current, even to oscillate. As you've said from min 20-30 pF directly from input bases to GND is a must.
Good input buffer is also a must to connect higher impedance sources like tube preamps etc. With VSSA I'll try to analyse amp's response and stability if vary input impedance from 10k to 100k.
Hm... So it is not wise to direct connect source to input bases of input pair ? Recommendation ? (including volume pot)
I believe I'm blind now ^^ I'm a blindman touching/feeling an elephant
Will proceed to TSSA when i'm free, and will wait for bad thing happens now. (they are rather obedient now)
I apologize for any inconvenience, but scope is a must when DIY fast AC circuits.
???
Please, use 1K in serial and 50 or 100pf to the ground. Are your speaker's cables long ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.