Hmm, I don’t know the Leach amp but here are some details for the different devices:
MPS8099/8599: NF, 80V, 0,5A, 0,625W, >150MHz
BC550C/560C: Uni, ra, 50V, 0,1A, 0,5W, 300MHz
A bit risky if you ask me.
Your best bet would be the MPSA06/MPSA56: NF-Tr, 80V, 0,5A, 0,625W, >100MHz
Don’t know if they are easily available.
Other replacements might be:
BC639/640: NF-Tr/E, 100V, 1A, 0,8W, 50MHz
2SC3939/2SA1533: Uni, 80V, 0,5A, 1W, 120MHz
Leach amp gurus could have different opinions
/Hugo
MPS8099/8599: NF, 80V, 0,5A, 0,625W, >150MHz
BC550C/560C: Uni, ra, 50V, 0,1A, 0,5W, 300MHz
A bit risky if you ask me.
Your best bet would be the MPSA06/MPSA56: NF-Tr, 80V, 0,5A, 0,625W, >100MHz
Don’t know if they are easily available.
Other replacements might be:
BC639/640: NF-Tr/E, 100V, 1A, 0,8W, 50MHz
2SC3939/2SA1533: Uni, 80V, 0,5A, 1W, 120MHz
Leach amp gurus could have different opinions
/Hugo
BC550C and BC560C are nice small signal transistors but have only 45V Vce blocking capability. Anyway, Leach Amp uses cascodes on the input stage so full voltage rails are not applied to these transistors
Reliability may be enhanced by properly selecting D13-D16 voltages [ie: 12V+15V instead of 20V+20V] in order to dump half the supply voltage over each BC550C/BC560C. Adjusting R13-R16 values will be required in order to mantain the same bias currents
Q5 and Q6 may suffer voltage stress if supply rails rise too fast [ie: no soft start circuit] since C4 and C5 require some time to charge through R13 and R14 in order to provide the right voltage rail splitting between transistors
Since Q5 and Q6 are cascode transistors, you may use here other devices with higher Vce blocking capability but worse specs, like BC546C/BC556C, 2N5401/2N5551, MPSA06/MPSA56, BD139/BD140, etc...
As an alternative, you may use BC546B/BC556B everywhere since they have better specs than MPSA06/MPSA56 and have still enough Vce blocking capability to work reliably without adittional circuit modifications
Reliability may be enhanced by properly selecting D13-D16 voltages [ie: 12V+15V instead of 20V+20V] in order to dump half the supply voltage over each BC550C/BC560C. Adjusting R13-R16 values will be required in order to mantain the same bias currents
Q5 and Q6 may suffer voltage stress if supply rails rise too fast [ie: no soft start circuit] since C4 and C5 require some time to charge through R13 and R14 in order to provide the right voltage rail splitting between transistors
Since Q5 and Q6 are cascode transistors, you may use here other devices with higher Vce blocking capability but worse specs, like BC546C/BC556C, 2N5401/2N5551, MPSA06/MPSA56, BD139/BD140, etc...
As an alternative, you may use BC546B/BC556B everywhere since they have better specs than MPSA06/MPSA56 and have still enough Vce blocking capability to work reliably without adittional circuit modifications
As stated before, I don't know the Leach amp nor its rail voltage.
Mr. Leach must have had a reason to select the MPS8099/8599 's.
To me it seems that the best replacements would be the MPSA06/56. Just to stay in the SOA. Hfe is probably quite important too. BC550C has a much higher value. 450-900 vs. max. 100 for MPS8099 and max. 50 for MPSA06.
/Hugo
Mr. Leach must have had a reason to select the MPS8099/8599 's.
To me it seems that the best replacements would be the MPSA06/56. Just to stay in the SOA. Hfe is probably quite important too. BC550C has a much higher value. 450-900 vs. max. 100 for MPS8099 and max. 50 for MPSA06.
/Hugo
The Leach Amp uses about 56V supply rails. Mr Leach specifies only american Motorola transistors since they are cheaper and much easier to find on USA than european ones
In contrast, In Europe and Asia, american transistors tend to be very expensive, hard to find, and usually fakes. Some years ago I bought some MPSA06 and MPSA56. A06 turned to be useless fakes due to very bad specs [Hfe = 5 to 25 for Ic=1mA!!!!], A56 were original devices from Motorola but have also poor specs compared to similar european devices. I paid for them about 6 times the money I usually pay for BC series and got mostly unusable crap. Wasted money. I ended using these transistors for LED and relay driving applications...
MPS8098 and MPS8598 are the preferred devices by Mr.Leach and have better specs than A06 and A56, but they are really hard to get in Europe and allways at the risk of paying 5-10 times more money for useless fakes
I think half the low sales-volume or rare transistors sold in Europe are actually fakes
In contrast, In Europe and Asia, american transistors tend to be very expensive, hard to find, and usually fakes. Some years ago I bought some MPSA06 and MPSA56. A06 turned to be useless fakes due to very bad specs [Hfe = 5 to 25 for Ic=1mA!!!!], A56 were original devices from Motorola but have also poor specs compared to similar european devices. I paid for them about 6 times the money I usually pay for BC series and got mostly unusable crap. Wasted money. I ended using these transistors for LED and relay driving applications...
MPS8098 and MPS8598 are the preferred devices by Mr.Leach and have better specs than A06 and A56, but they are really hard to get in Europe and allways at the risk of paying 5-10 times more money for useless fakes
I think half the low sales-volume or rare transistors sold in Europe are actually fakes
The gain of the input devices only affects ouput offset and input impedance [the more gain, the better]. The overall gain of the long tailed pair is limited by emitter resistors so open loop gain is not increased. Placing faster transistors in the LTP will actually increase the phase margin of the circuit
The amplifier has already a pole estabilished by VAS miller capacitance and the next higher pole is introduced by output transistors, so the rest of the poles should be at frequencies high enough where there is no longer enough feedback for oscillation. Using slow transistors in LTP, VAS, etc..., actually increases the risk of unstability since it may place the LTP pole at too low a frequency
Using BC550C/BC560C for LTP and faster devices for VAS and predriver stage may even allow to increase gain bandwith product of the circuit without danger of oscillation. Remember that Leach Amp was designed long time ago and better bipolar transistor are now available
The amplifier has already a pole estabilished by VAS miller capacitance and the next higher pole is introduced by output transistors, so the rest of the poles should be at frequencies high enough where there is no longer enough feedback for oscillation. Using slow transistors in LTP, VAS, etc..., actually increases the risk of unstability since it may place the LTP pole at too low a frequency
Using BC550C/BC560C for LTP and faster devices for VAS and predriver stage may even allow to increase gain bandwith product of the circuit without danger of oscillation. Remember that Leach Amp was designed long time ago and better bipolar transistor are now available
Wagener said:
Eva said:
Excellent post (as usual) from Eva. I totally agree that the cascode transistor bases should be at about 1/2 the supply for best reliability if lower voltage parts are used, so the 30v zener stack makes more sense than 40v.
ON Semiconductor does still make the MPS8X99 parts, so you can purchase them. Leach specified the MPSA06/-56 parts as acceptable alternates in case the 8X99 parts were unavailable.
Since the input pair has very large emitter degeneration resistors (300 ohms), even the input pair does not benefit much from the higher beta of the BC series. But if they are more available or much cheaper, by all means use either BC series as Eva described. Beware that the pinout is reversed for the BC series, so if you use them you have to install them backward from from the silkscreen outline, could be exciting if you forget
Sanyo seem to have some very good ones too. Check out
www.profusionplc.com/product-frame.htm
click Sanyo, then Transistors.
www.profusionplc.com/product-frame.htm
click Sanyo, then Transistors.
I've extracted this info from some Toshiba datasheets I had at hand. Parameters are for PNP since I feel too lazy now to checck NPN datasheets
A1049/C2459 120V 100mA 100Mhz Hfe=400 Cbc=4pF
A1091/C2551 A1432/C3672 300V 100mA 60Mhz Hfe=100
A1145/C2705 A1350/C3425 150V 50mA 200Mhz Hfe=150 Cbc=2.5pF
A1320/C3333 250V 50mA 80Mhz Hfe=160 Cbc=1.8pF
A1349/C3381 Dual 80V 100mA ??Mhz Hfe=500
A1408/C3621 150V 1.5A 50Mhz Hfe=100
A1804/C4689 120V 8A 30Mhz Hfe=100
A1805/C4690 A1941/C5198 140V 10A 30Mhz Hfe=100
A1837/C4793 230V 1A 70Mhz Hfe=150 Cbc=30pF
A1899/C5052 120V 800mA 120Mhz Hfe=160
A1930/C5171 180V 2A 200Mhz Hfe=200 Cbc=26pF
A1942/C5199 160V 12A 30Mhz Hfe=100
A1943/C5200 A1962/C5242 230V 15A 30Mhz Hfe=100 Cbc=360pF
The only thing I could buy was some A1837 and C4793
A1049/C2459 120V 100mA 100Mhz Hfe=400 Cbc=4pF
A1091/C2551 A1432/C3672 300V 100mA 60Mhz Hfe=100
A1145/C2705 A1350/C3425 150V 50mA 200Mhz Hfe=150 Cbc=2.5pF
A1320/C3333 250V 50mA 80Mhz Hfe=160 Cbc=1.8pF
A1349/C3381 Dual 80V 100mA ??Mhz Hfe=500
A1408/C3621 150V 1.5A 50Mhz Hfe=100
A1804/C4689 120V 8A 30Mhz Hfe=100
A1805/C4690 A1941/C5198 140V 10A 30Mhz Hfe=100
A1837/C4793 230V 1A 70Mhz Hfe=150 Cbc=30pF
A1899/C5052 120V 800mA 120Mhz Hfe=160
A1930/C5171 180V 2A 200Mhz Hfe=200 Cbc=26pF
A1942/C5199 160V 12A 30Mhz Hfe=100
A1943/C5200 A1962/C5242 230V 15A 30Mhz Hfe=100 Cbc=360pF
The only thing I could buy was some A1837 and C4793
http://www.ne.jp/asahi/evo/amp/guide.htm
you can see dreamed transistors at this site:fT,Hfe is high and ra
you can see dreamed transistors at this site:fT,Hfe is high and ra
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.