D
Deleted member 550749
Hi, I'm new to here, i don't know about engineering stuff, my friend gave me this circuit to try, after comparing Apex fh9 highly doubt about this cirquit, i know only pro engineers can solve the doubt, is this circuit gonna work or needs any changes??
Last edited by a moderator:
looks like a simulated circuit. It may work statically. Due to the lack of any frequency compensation it will probably oscillate wildly. You should simulate it transient response.
It has a gain of about 21 with ac and dc.
The dc gain is undesirable as it will make dc offset unstable.
Get rid of dc gain using 100uf cap between feedback 10k resistor and ground.
I would also make bias current adjustable by making 950R resistor a 2k2 pot.
The dc gain is undesirable as it will make dc offset unstable.
Get rid of dc gain using 100uf cap between feedback 10k resistor and ground.
I would also make bias current adjustable by making 950R resistor a 2k2 pot.
Saturation voltage of output stage will be 6-8V - thus reducing output swing to +-12V approx with +-19V supplies.
Feedback voltage divider way to hi-ohmic.
increases noise and phase drift at higher frequencies - reduced stability
increases noise and phase drift at higher frequencies - reduced stability
D
Deleted member 550749
Yep you r correct, I think 22k is ok but for that I have to change others resistor value like 10 k will be 1 k with 100 to 200uf capasitor, bias resistor ,offset resistor etc 🤔
Looks like huge imbalance on input LTP transistors (1.27mA and 0.356mA). Try simulating with 2.54mA tail current instead for 1.27mA in both transistors.
EDIT: Or increase collector resistance on input transistor until current in input transistor drop to 0.82 mA (approx 1800 ohm)
EDIT: Or increase collector resistance on input transistor until current in input transistor drop to 0.82 mA (approx 1800 ohm)
D
Deleted member 550749
1.As Bucks pointed out, the output saturation wastes about 8 of the 19 Volt power supply. Circuits that avoid this issue are variations on bootstrapping or more likely putting 2x gain in the CFP, ie divider(s) between the driver emitters and the output.
2. The 1K resistor needs a ~100uF cap in series so that the DC input to the LTP is 22K on both sides, not 1K||22K, and DC is not amplified, and a ~10uF series cap on the input to avoid preamp DC.
3. The 100 Ohm in the collector of the negative side of the LTP and the VAS 100 Ohm degen resistor are not useful.
4. The CCS bias probably need a filter cap,. Make the 22.1K two 10K with a ~47uF cap to -19V.
5. A "speed-up" cap across the bias divider can be very important to the stability of the negative output.
6. The VAS needs a compensation cap. Recently "two-pole compensation" is popular, which would be two small caps in series with a resistor to the output.
7. Better THD can be had with a Darlington VAS and a current mirror, but they push the OLG towards stability problems.
2. The 1K resistor needs a ~100uF cap in series so that the DC input to the LTP is 22K on both sides, not 1K||22K, and DC is not amplified, and a ~10uF series cap on the input to avoid preamp DC.
3. The 100 Ohm in the collector of the negative side of the LTP and the VAS 100 Ohm degen resistor are not useful.
4. The CCS bias probably need a filter cap,. Make the 22.1K two 10K with a ~47uF cap to -19V.
5. A "speed-up" cap across the bias divider can be very important to the stability of the negative output.
6. The VAS needs a compensation cap. Recently "two-pole compensation" is popular, which would be two small caps in series with a resistor to the output.
7. Better THD can be had with a Darlington VAS and a current mirror, but they push the OLG towards stability problems.
I guess you use some basic drawing program, which explains the lack of reference designators of the components. I strongly recommend swapping to a simulator like LTSpice for further evaluation. Personally I suspect the Sziklay output configuration will show stability issues and should be replaced by emitter/source followers.
Yep CFP is a disaster
and it is a basically straight forward Differential input.
But pretty horrible topology approach
Unbeatable mosfet design is common Hitachi amplifier
with added thermal tracking if using Vertical type mosfet.
Darlington outputs.
you loose 4 to 5 more volts of the rail to mosfet.
So BJT better at this point for this power level.
and it is a basically straight forward Differential input.
But pretty horrible topology approach
Unbeatable mosfet design is common Hitachi amplifier
with added thermal tracking if using Vertical type mosfet.
Darlington outputs.
you loose 4 to 5 more volts of the rail to mosfet.
So BJT better at this point for this power level.
In a nutshell: it looks like a beginner designed circuit, probably never actually built, only simulated.
Good in a class environment, as a class project, etc. , not bad per se but lacks tweaks and refinements a more experienced designe would add, some of which are mentioned above.
All this said in good faith, trying to help.
If you will actually build an amp, spending time good $$$ on it, I´d suggest go for a tried project instead, built by many and with good documentation.
As a side note, Apex designs in general are good and have been built by many successfully, that´s a good endorsement.
Good in a class environment, as a class project, etc. , not bad per se but lacks tweaks and refinements a more experienced designe would add, some of which are mentioned above.
All this said in good faith, trying to help.
If you will actually build an amp, spending time good $$$ on it, I´d suggest go for a tried project instead, built by many and with good documentation.
As a side note, Apex designs in general are good and have been built by many successfully, that´s a good endorsement.