| zega55 |
| Hi guys can you give me some hints on this amp I am using it but with out the second pair of tranzistors and I like it can it be mod. to work nicer |
|
|
| sixtek |
Hi,
- First I would change R12 and R13 to constant current sources.
- 100uF//100nF supply bypass on the board would be beneficial.
- I would change 2N5401/5550 to MJE340/350 in the second stage. 0.5W dissipation is much for a TO-92 device.
- TIP41C/42C are only 100V Vceo devices, not enough for +/-70V rails. MJE15030/15031 is a good driver, but their Ft is 30MHz (TIP's are 3MHz devices), so maybe compensation need to be modified.
- LTP sees 100k in the input and 5k6 in the feedback loop. Make them equal (cap's need to be changed for giving proper rolloff freq).
- I would modify the protection circut to something more elegant. |
|
|
| fotios |
| quote: | Originally posted by zega55
Hi guys can you give me some hints on this amp I am using it but with out the second pair of tranzistors and I like it can it be mod. to work nicer |
I thing it is a very good circuit, a classic BJT amplifier scheme, with all the beneffit that offers it. In these included the high power output without problems. In your case, it is: 70Vp-7V losses approx.=63Vp/1,414=44,5Vrms and Poutmax=44,5X44,5/8=247Wrms theoretical, and the clean output power from 20 to 20000 Hz is about 220Wrms with a good quality power transformer in a logical size of 800 to 1000VA.
This scheme uses a symmetrical input stage with two difference amplifiers and thus practically there is no need for a cascode stage before the individual VAS for each slewing phase(pos. or neg.). Also the input trimmer it is very useful as it balances the base currents in the input transistors and thus the collector currents in each dif. amplifier to cancel the DC offset in output caused by the different beta of each transistor. A verry usefull circuit is the VI limitter included. You can establish very easy the iddle current of output (to get a AB class output stage) by connecting a voltmeter accross the e-c contacts of the Vbe multiplier transistor and turning the trimmer untill you read a value of +/- 1,7Vdc. Now i have to point out some critical items. First, to keep the output devices in SOA, each of the MJ15003/MJ15004 sustain Ic=2A/Vce=70V. For a power of 70VpX70Vp/4Ù = 1225Wp max the total current is Ip=17,5Ap instantly in the worst case. Thus, yet 2 pairs of output transistors are few. In practice for this supply you need at least 5 pairs of output transistors to be safe under each condition. At least 4 pairs needed. Also the drivers TIP41/TIP42 yet if they are of C type (TIP41C=100Vce) touch their limit. Try to change these with the much better MJE15030 - MJE15031 (Vce=150V). It is not so difficult to find these in the market. The remain transistors are good (2N5401/2N5551 Vce=150V). For other queries in your service. Take a look in my web page http://www.eal.gr to get some useful ideas.
Fotios[ |
|
|
| jaycee |
| If you like this circuit but want it to operate reliably at higher voltage, you may as well build The Leach Amp as the topology is pretty similar. |
|
|
| zega55 |
| thanks people I am using it with a transformer that has 400VA and 45 V per rail and it gives about 75 Wrms.The board has hols so I can add the second pair of tranzistors just as it is in the picture I dont need a lot of power 75Wrms is enough for me. |
|
|
| AndrewT |
Hi,
your transformer? Is it 400VA & 45Vac or 45Vdc?
You need two secondaries to make a dual polarity supply.
For 75W into 8ohm, supply rails of about +-45Vdc are about right and a 30-0, 30-0Vac or 30-0-30Vac transformer will do. |
|
|
| zega55 |
| It is 32-0-32 and it gives after greatz 45 Vdc per rail .Is this transformer enough for my use |
|
|
| AndrewT |
Hi,
the 32Vac transformer and one pair of MJ15003/4 will make a very good 80W into 4ohm amplifier. use Re=0r1 and Vre=25mV. this will result in a quiescent dissipation of about 22W/channel into the heatsink so it will need to be big about 0.4 to 0.5C/W for a stereo pair. Smoothing capacitance of +-20mF/chn
If you decide to fit 2pair of 15003/4, the amp will be capable of driving severe 4ohm loads to 150W. I would increase Re=0r15 and this will dissipate 29W/channel, requiring an even bigger heatsink. Try 0.6C/W per channel. smoothing capacitance +-40mF/chn |
|
|
| zega55 |
| Thank you. Would it sound a loot better if I change the tranzistors ? |
|
|
| AndrewT |
Hi,
not trying to be funny but I see only one way to find out.
Build two versions and compare them. I haven't got the money to throw away trying that. |
|
|
| Leolabs |
| Change the input pairs to something low noise like the 2SA970/C2240. |
|
|
|
