Hello DIYers
I am going to build this amplifier
A 100-watt amplifier which uses NPN transistors for the two half-cycles. This configuration is known as a Quasi-complementary amplifier. This audio amplifier is a mono version of "la zener". It was given this name because its circuit has a zener diode that stabilizes the differential torque of the input. Its great stability, acceptable power and low harmonic distortion, make it ideal for amplifying the electric guitar preamp, the bass preamp or simply for home use.
If anyone build this amp can please give me some advice about its stability. and how this amp sound, please replay.I am posting web page link (Spanish) of this amp.
construya un amplificador monofonico de 100 watts version 2.0
I am going to build this amplifier
A 100-watt amplifier which uses NPN transistors for the two half-cycles. This configuration is known as a Quasi-complementary amplifier. This audio amplifier is a mono version of "la zener". It was given this name because its circuit has a zener diode that stabilizes the differential torque of the input. Its great stability, acceptable power and low harmonic distortion, make it ideal for amplifying the electric guitar preamp, the bass preamp or simply for home use.
If anyone build this amp can please give me some advice about its stability. and how this amp sound, please replay.I am posting web page link (Spanish) of this amp.
construya un amplificador monofonico de 100 watts version 2.0
Attachments
This has got TIP42 for a driver which I find to be a bit inferior to RCA40904/40905 on high frequencies like top octave piano or tinkly bells. I've got TIP41c/42c in my AX6. I think the 3 Mhz Ft is at fault.
However it sounds, it won't be very stable at high wattage with no temperature sense diodes or Vbe multiplier on the heat sink. The OT idle bias current is not adjustable either. I shoot for about 20 mv across those .47 emitter resistors.
Suggest you build an AX6 or AX8 which do have that feature, which can be found in this thread: A Directory of Apex Audio Amplifiers
Double sense diodes on the AX6 and Vbe multiplier on the AX8. Those features reduce the idle bias current as the output transistors and heat sink heat up.
You can double up on the output transistors on those models if you really want to go 100 W. My AX6 with a single pair of NTE60 output transistors (probably marked up white box MJ15003) was measured at 72 W into 8 ohms with single rail 60v (ic +-30). Use heat sinks on the drivers, use TO220 drivers with 30 mhz Ft like LME15032/33 for good top octave IM.
However it sounds, it won't be very stable at high wattage with no temperature sense diodes or Vbe multiplier on the heat sink. The OT idle bias current is not adjustable either. I shoot for about 20 mv across those .47 emitter resistors.
Suggest you build an AX6 or AX8 which do have that feature, which can be found in this thread: A Directory of Apex Audio Amplifiers
Double sense diodes on the AX6 and Vbe multiplier on the AX8. Those features reduce the idle bias current as the output transistors and heat sink heat up.
You can double up on the output transistors on those models if you really want to go 100 W. My AX6 with a single pair of NTE60 output transistors (probably marked up white box MJ15003) was measured at 72 W into 8 ohms with single rail 60v (ic +-30). Use heat sinks on the drivers, use TO220 drivers with 30 mhz Ft like LME15032/33 for good top octave IM.
Last edited:
As a subwoofer it might work, the Ft of the TIP41/42 driver won't matter if the signal has no high frequencies. The crossover is supposed to filter the high frequencies out of the signal to a subwoofer. I use my AX6 as a broadband amp with woofers and tweeters in the speaker.
The amp appears to be class AB.
If you move the two 1n4007 between the bases of the drivers to locate above the output transistors on the heat sink, you could achieve some bias current control when it heats up. I use a cinch type solder terminal strip at the top of the heat sink, to mount these. The 150 ohm resistor above them could be selected (or replaced with a potentiometer) to control the cold bias current.
With 2SC5200 as output transistors, it is not going to produce 100 W for very long before meltdown. That transistor has a dc soa at 96 v of about 1 A which is light duty. If you reduce the rail voltages to +-36 or 30 2SC5200 might live longer. As I said I'm getting 72 W 8 ohm with +-30 and MJ15003 outputs. If you paralleled two pairs output transistor you might get 100 w for longer, but it would take more heat sink than is in the picture. I have a fan on my heatsink. If you can get real ones, 2SC3263 output transistor has soa of 1.8 a, MJL4281 has about the same. You determine if transistors are real by doing an Iceo test, base open, 12 to 48 v through 1000 ohm on the collector, ground to emitter through a dvm milliamp scale measure the current. Should be low milliamps or less for real parts. Real parts come from real manufacturer authorized distributors like RS, farnell, mouser, digikey. Ordering from an out of country distributor is an invitation for the guys in customs so subtitute fake parts, see many sad stories on the parts thread.
The 1000 pf (.001 uf) capacitor between base & collector of the lower driver is pretty massive, will slow slew rate and may cause cross conduction of the output transistors. I'd try 150-220 pf to start with, and check if the output is oscillating at 1mhz or so. If yes, go up in value on the cap. Use ceramic or mica caps in this position. I do oscillation check with an ANALOG voltmeter, on AC scale, 390 pf capacitor in series with one lead. A voltage shown would be radio frequency. Oscilloscopes are cheaper these days than good analog VOM but if you step on the probe it is a $50 mistake. Also the electrolytic caps in a scope don't last 50 years like an analog voltmeter will.
I suggested faster drivers for a full frequency amp in my previous post. If you're not buying from a legitimate distributor like RS farnell or mouser, anything marked MJE15032/33 would probably really be TIP31/32 anyway. Or something cheaper even.
Have fun.
The amp appears to be class AB.
If you move the two 1n4007 between the bases of the drivers to locate above the output transistors on the heat sink, you could achieve some bias current control when it heats up. I use a cinch type solder terminal strip at the top of the heat sink, to mount these. The 150 ohm resistor above them could be selected (or replaced with a potentiometer) to control the cold bias current.
With 2SC5200 as output transistors, it is not going to produce 100 W for very long before meltdown. That transistor has a dc soa at 96 v of about 1 A which is light duty. If you reduce the rail voltages to +-36 or 30 2SC5200 might live longer. As I said I'm getting 72 W 8 ohm with +-30 and MJ15003 outputs. If you paralleled two pairs output transistor you might get 100 w for longer, but it would take more heat sink than is in the picture. I have a fan on my heatsink. If you can get real ones, 2SC3263 output transistor has soa of 1.8 a, MJL4281 has about the same. You determine if transistors are real by doing an Iceo test, base open, 12 to 48 v through 1000 ohm on the collector, ground to emitter through a dvm milliamp scale measure the current. Should be low milliamps or less for real parts. Real parts come from real manufacturer authorized distributors like RS, farnell, mouser, digikey. Ordering from an out of country distributor is an invitation for the guys in customs so subtitute fake parts, see many sad stories on the parts thread.
The 1000 pf (.001 uf) capacitor between base & collector of the lower driver is pretty massive, will slow slew rate and may cause cross conduction of the output transistors. I'd try 150-220 pf to start with, and check if the output is oscillating at 1mhz or so. If yes, go up in value on the cap. Use ceramic or mica caps in this position. I do oscillation check with an ANALOG voltmeter, on AC scale, 390 pf capacitor in series with one lead. A voltage shown would be radio frequency. Oscilloscopes are cheaper these days than good analog VOM but if you step on the probe it is a $50 mistake. Also the electrolytic caps in a scope don't last 50 years like an analog voltmeter will.
I suggested faster drivers for a full frequency amp in my previous post. If you're not buying from a legitimate distributor like RS farnell or mouser, anything marked MJE15032/33 would probably really be TIP31/32 anyway. Or something cheaper even.
Have fun.
Last edited:
Hello indianjo in AX11 and HX11 which one u suggest for 2.1 home theatre
No knowledge about these. I build single supply 5-7 transistor amps mostly. Look at the index linked to above, find the threads about these, ask people there.
To pick an amp design, know your room size, speaker, desired top and bottom volume, type of music you listen to. 8 or 4 ohms speaker, full range or sub, efficiency. (sound level 1w 1 m) . classical music has 50+ db peaks and valleys, house & adventure movies are pretty compressed, mono volume 10 db range.
![AX11 TO-3 Ver.3 Gray [1600x1200].JPG](/community/data/attachments/606/606408-63546207f0736537edc7640297b2fbe1.jpg)
Attachments
![SDC10566 [1024x768].JPG](/community/data/attachments/606/606428-8ed94c54aa2831c182d92d5b769d5b6f.jpg)
Dear mihailo
Can AX-11 or AX-14 Give me 100Watt on 8ohm load with one Pair of 2sc5200 or 2sa1943 on 50V+/- supply.my posted amp give me 100Watt on one pair.
For about 100 millisec.
Look at the datasheet of 2sc5200 or 2sa1943. DC soa of about .45 A @ 100v. This is the rating Andrew has said predicts output transistor life pretty well. I use datasheets from datasheetcatalog.com but an authorized distributor only guarentees the parts they sell to the datasheet you download from them the day you buy.
(I^2)*R is about 1.6 W. You have to realize currents are higher on inductive speakers than on resistors which have pure resistance.
Your posted amp the videorockola also has a complete fantasy rating. As do most of the amp boards for sale on e-bay. The text of the link you posted suggested you use a different output transistor pair, something from Toshiba (defunct). It also suggested multiple output pairs in the first paragraph. Clone transistors may not have a soa rating at all. Download the datasheet the day you buy. Most of the second line houses like New jersey semi have been dropping the dc soa rating from their substitute parts.
If you use multiple output pairs as I suggested, you have to put heat sinks on the driver transistors. Be aware, with multiple output pairs, each transistor needs its own emitter resistor.
Also be aware the tall package predrivers 2sc2229 in a 1 watt to92 package in general can't be bought. Unless you go to a silk screen supplier that repaints the number of anything you want on the cheapest parts available that don't meet the specs. 99.99999% of real TO92 transistors are 500 mw or 650 mw. I managed to score some 1 W pnp TO92 from On semi as they were shutting down but can't buy npn's.
Look at the datasheet of 2sc5200 or 2sa1943. DC soa of about .45 A @ 100v. This is the rating Andrew has said predicts output transistor life pretty well. I use datasheets from datasheetcatalog.com but an authorized distributor only guarentees the parts they sell to the datasheet you download from them the day you buy.
(I^2)*R is about 1.6 W. You have to realize currents are higher on inductive speakers than on resistors which have pure resistance.
Your posted amp the videorockola also has a complete fantasy rating. As do most of the amp boards for sale on e-bay. The text of the link you posted suggested you use a different output transistor pair, something from Toshiba (defunct). It also suggested multiple output pairs in the first paragraph. Clone transistors may not have a soa rating at all. Download the datasheet the day you buy. Most of the second line houses like New jersey semi have been dropping the dc soa rating from their substitute parts.
If you use multiple output pairs as I suggested, you have to put heat sinks on the driver transistors. Be aware, with multiple output pairs, each transistor needs its own emitter resistor.
Also be aware the tall package predrivers 2sc2229 in a 1 watt to92 package in general can't be bought. Unless you go to a silk screen supplier that repaints the number of anything you want on the cheapest parts available that don't meet the specs. 99.99999% of real TO92 transistors are 500 mw or 650 mw. I managed to score some 1 W pnp TO92 from On semi as they were shutting down but can't buy npn's.
Last edited:
Power transistor
Hello Indianajo
Look at the power transistor I used in this amp. please zoomed to see the numbers printed on the transistor. What you think its original or copy/ fake.
Hello Indianajo
Look at the power transistor I used in this amp. please zoomed to see the numbers printed on the transistor. What you think its original or copy/ fake.
Attachments
Hi Gurpreet
which 2sc5200 ( Toshiba. Fairchild)transistor you going to use in AX-11 if you find original transistor in India. please tell me where you bought that.
Thanks
One can't tell by looking. As I said, clone houses like NJ semi are selling parts labeled 2sc5200 that aren't even tested for SOA. As far as I know Toshiba is out of the business. ST and Fairchild datasheets show an SOA rating.
One quick test if you have a capacitor meter, is measure collector base capacitance. The spec sheet says 150 pf at 1 mhz, I don't have any 2sc5200 to tell you what my meter says. Fake will be smaller.
I don't have any 2SC5200 but I do have some allegro sanken 2SC3263 which are 130 W parts. I got 560 pf b-c on my capacitance meter.n MJ15015 which are 100 W parts read 299. Both of these I bought from Farnell(newark), an authorized distributor
Last edited:
2SC5200 | Bipolar Transistors | TOSHIBA Semiconductor & Storage Products | Asia-Pacific
Seems fine with thousands of devices still in the distribution supply chain.
I guess they never invested enough into marketing in discretes as of late.
Hi Sanbadgujar, Last year got 5 pairs From Chandigarh, that's original. But now i have One pair left. But now original can't anywhere. But i got his replacement TOSHIBA TT5200 & TT1943.First I try to make AX-11 with TTxxxx
hello Gurpreet
Can this amp work in Class AB. Can it gives 100W on 8 Ohm load, And how you make your PCB's like I make my PCB with toner transfer method.I also make a toroidal transformer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.