Hi Mikelm
A good question. The original Mullard spec for these gave fT=250 MHz typ. The Philips 1999 spec stated 190 MHz (at 5V/50 mA). I don't know what ST's spec. is, nor if NXP (taken over from Philips) continue to state the fT.
I was concerned at one time and measured the hfe to check a sample was not an "epibase" - no drop at all at 1MHz so think it was OK.
John
A good question. The original Mullard spec for these gave fT=250 MHz typ. The Philips 1999 spec stated 190 MHz (at 5V/50 mA). I don't know what ST's spec. is, nor if NXP (taken over from Philips) continue to state the fT.
I was concerned at one time and measured the hfe to check a sample was not an "epibase" - no drop at all at 1MHz so think it was OK.
John
They say, for this part, only use Philips. But I think I have seen/owned black colored (as well as the common grey) Philips.
I used 2SD669AC for the phase spliter TR3, it worked fine for me. And BC556 is fine for TR4 as well as 2N5401 in my JLH amp. These transistors are easier to find.
Jaz
Jaz
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Not sure how "destroyed" your amp sounds. But normally bad sound is due to the distortion, that is caused by unmatched output pair and/or malfunction of any of of the two output transistors.If you build a two channel(stereo) amp and both channel sounds bad, and if you amp sound ok when playing in low volume, then you may have to rise the quescent current.
Regards.
Jaz
Hi
Have you checked the current flow in the output transistors? Can your power supply handle the current drain? The output transistor current was stated to be 1.2A from about a 30V supply for an 8 ohm load. It can be set by adjusting the load resistors in the driver stage.
There might be some AC ripple appearing at the output if you have not got a good earthing scheme. It could be that the amplifier oscillates, causing spiky sounds or a hot loudspeaker. One simple approach to stabilising an oscillating JLH is to add a small capacitor across the feedback resistor (in the region of 47-100pF) not to add all the capacitors which JLH proposed, which IMHO would seriously degrade the sound.
If it is bass notes that sound weak, the output capacitor needs to be increased. The low frequency response is pretty poor with 2,200 uF (2.2 mF) even though many amplifiers used this typical value in the 70's with reasonable performance for small speakers and average listening.
If you are getting noise the input bias arrangements are certainly not optimum. The base of the input transistor (I would have called this Tr1 but JLH labelled his transistors from the back!) is biased with high impedance resistors which give a high noise level at low frequencies which needs a good low preamp impedance to overcome. YOU might also find that a small capacitor from the input transistor base to ground might stop any oscillation (e.g 100pF).
My preference for biasing an input stage for single-ended (AC coupled) amplifiers is to form a voltage divider using two relatively low value resistors (e.g. 10k+12k) across the supply, with the mid point decoupled with a capacitor (which can also be a lower voltage rating) and take a single bias resistor from the tap to the base (also 10 to 12k) for better LF noise performance.
John
Have you checked the current flow in the output transistors? Can your power supply handle the current drain? The output transistor current was stated to be 1.2A from about a 30V supply for an 8 ohm load. It can be set by adjusting the load resistors in the driver stage.
There might be some AC ripple appearing at the output if you have not got a good earthing scheme. It could be that the amplifier oscillates, causing spiky sounds or a hot loudspeaker. One simple approach to stabilising an oscillating JLH is to add a small capacitor across the feedback resistor (in the region of 47-100pF) not to add all the capacitors which JLH proposed, which IMHO would seriously degrade the sound.
If it is bass notes that sound weak, the output capacitor needs to be increased. The low frequency response is pretty poor with 2,200 uF (2.2 mF) even though many amplifiers used this typical value in the 70's with reasonable performance for small speakers and average listening.
If you are getting noise the input bias arrangements are certainly not optimum. The base of the input transistor (I would have called this Tr1 but JLH labelled his transistors from the back!) is biased with high impedance resistors which give a high noise level at low frequencies which needs a good low preamp impedance to overcome. YOU might also find that a small capacitor from the input transistor base to ground might stop any oscillation (e.g 100pF).
My preference for biasing an input stage for single-ended (AC coupled) amplifiers is to form a voltage divider using two relatively low value resistors (e.g. 10k+12k) across the supply, with the mid point decoupled with a capacitor (which can also be a lower voltage rating) and take a single bias resistor from the tap to the base (also 10 to 12k) for better LF noise performance.
John
Hi
Some more thoughts - do both your output transistors work? This push pull design needs both to work to work correctly. ALso the bootstrap capacitor needs to be moderately large- at least 220 uF for 8 ohms, but if you increase the output capacitor you should also increase the bootstrap too.
Last comment for now - I don't know what transistors you have in the output stage but if they are 2N3055's (ON semi or ST) these should have fT's in the region of 3-6 MHz. That is OK and the gains might be around 50 at 1A. So my guess at the driver load resistors (as a starting point) would be 100 ohms for the one connected to the supply (which should be at least 2W rating) and the bootstrapped one can be about 220 ohms (but you may want to start with 470 ohms and check the output currents first). The poewer rating of these needs to be at least 1W too, and higher for 220 ohms (might dissipate over 2W but if made up with parallel 2W resistors that will not be a problem. To simplify checking you could put 0.1 ohm resistors in series with the emitters of the output transistors which you could remove or bypass later.
John
Some more thoughts - do both your output transistors work? This push pull design needs both to work to work correctly. ALso the bootstrap capacitor needs to be moderately large- at least 220 uF for 8 ohms, but if you increase the output capacitor you should also increase the bootstrap too.
Last comment for now - I don't know what transistors you have in the output stage but if they are 2N3055's (ON semi or ST) these should have fT's in the region of 3-6 MHz. That is OK and the gains might be around 50 at 1A. So my guess at the driver load resistors (as a starting point) would be 100 ohms for the one connected to the supply (which should be at least 2W rating) and the bootstrapped one can be about 220 ohms (but you may want to start with 470 ohms and check the output currents first). The poewer rating of these needs to be at least 1W too, and higher for 220 ohms (might dissipate over 2W but if made up with parallel 2W resistors that will not be a problem. To simplify checking you could put 0.1 ohm resistors in series with the emitters of the output transistors which you could remove or bypass later.
John
I suggest to measure the bias current. It is simple to do. If it is about arounf 1.3 amp and your speakers are 8 ohms nominally and quite efficient then there is no reason for the sound to distort.
However, if you are using four ohm speakers then the output power would be half or the speakers are very inefficient and you may need to turn the volume up to the point that the amplifier severely clips because the current is limiting.
The amplifier does not enter class AB mode as one would expect as with over biased class B. This amp sounds sounds great or when running out of steam it sounds terrible, there is no middle way.
However, if you are using four ohm speakers then the output power would be half or the speakers are very inefficient and you may need to turn the volume up to the point that the amplifier severely clips because the current is limiting.
The amplifier does not enter class AB mode as one would expect as with over biased class B. This amp sounds sounds great or when running out of steam it sounds terrible, there is no middle way.
Hi Nico
A good point. I would also point out that in this JLH amplifier, if it is overloading, then the base-emitter junction of the upper output transistor could become reverse biased. This might be due to a fast transient or a low impedance or short circuited load.
If the base emitter voltage exceeds about 10V it may breakdown and then the transistor could conduct and that might lead to a catastrophic breakdown, so it could be that the amplifier has overloaded and one of the output transistors has expired.
Simulations also reveal in this amplifier that the bootstrap drive is not "fast". The transient response of the amplifier is limited by the speed of the output transistors, and in fact is worse than Class (A)B amplifiers which have driver transistors even if they also have a bootstrap.
One solution for this is possible if a poor transient performance is thought to be the problem is to use a pair of high frequency transistors such as 2SC5200's or MJL3281A's.
These certainly oscillated in a test circuit I built which needed 47pF feedback compensation capacitors (in parallel with the feedback resistor, not a Miller!!) to stabilise, but after that gave audibly improved performance.
John
A good point. I would also point out that in this JLH amplifier, if it is overloading, then the base-emitter junction of the upper output transistor could become reverse biased. This might be due to a fast transient or a low impedance or short circuited load.
If the base emitter voltage exceeds about 10V it may breakdown and then the transistor could conduct and that might lead to a catastrophic breakdown, so it could be that the amplifier has overloaded and one of the output transistors has expired.
Simulations also reveal in this amplifier that the bootstrap drive is not "fast". The transient response of the amplifier is limited by the speed of the output transistors, and in fact is worse than Class (A)B amplifiers which have driver transistors even if they also have a bootstrap.
One solution for this is possible if a poor transient performance is thought to be the problem is to use a pair of high frequency transistors such as 2SC5200's or MJL3281A's.
These certainly oscillated in a test circuit I built which needed 47pF feedback compensation capacitors (in parallel with the feedback resistor, not a Miller!!) to stabilise, but after that gave audibly improved performance.
John
Hi there all,
I am going to build the JLH for Quad ESL - I have a stacked pair. I have got hold off all the major components (MJ802s, BD139-16, Caddock resistors) but can't find a 2w trimpot for the bias adjust in SA; nor Mouser amongst a few others - any suggestions? Also heatsinks - where could I source 0.5degC/W units?
I am going to build the JLH for Quad ESL - I have a stacked pair. I have got hold off all the major components (MJ802s, BD139-16, Caddock resistors) but can't find a 2w trimpot for the bias adjust in SA; nor Mouser amongst a few others - any suggestions? Also heatsinks - where could I source 0.5degC/W units?
Hi Ferret,
You can also try Mantech Electronics they are in Johannesburg. You will not get a 2W trim pot easily but you could use a 2 watt multi-turn potentiometer. Why do you want to use a 2 watt pot?
FAX : (011) 493-9319
sales@mantech.co.za
Mantech Electronics | Leader in Electronic and Industrial Components, Tools, Test and Measurement
You can also try Mantech Electronics they are in Johannesburg. You will not get a 2W trim pot easily but you could use a 2 watt multi-turn potentiometer. Why do you want to use a 2 watt pot?
FAX : (011) 493-9319
sales@mantech.co.za
Mantech Electronics | Leader in Electronic and Industrial Components, Tools, Test and Measurement
Just happen to come back to this thread today. I have built a dual rail version with 2n3055 and love it. Now I want to build the 2005 version with 2xCCS for better output offset adj and Q current adj. Also will be using better transistors for the new build.
I think this has probably been asked before but I seem to be unable to find the answer - Can I use a regular small 50R blue plastic trim pot for VR2 ? This pot seems to be carrying some current (Q current / hFE) and I'm afraid a regular small trim pot will get overheat. Please refer to the following schematic.
The Class-A Amplifier Site - JLH Class-A Update
(Really enjoying my JLH amp. Currently pushing a pair of Tannoy MKII-611 with it and love it every moment)
Regards,
-AC
I think this has probably been asked before but I seem to be unable to find the answer - Can I use a regular small 50R blue plastic trim pot for VR2 ? This pot seems to be carrying some current (Q current / hFE) and I'm afraid a regular small trim pot will get overheat. Please refer to the following schematic.
The Class-A Amplifier Site - JLH Class-A Update
(Really enjoying my JLH amp. Currently pushing a pair of Tannoy MKII-611 with it and love it every moment)
Regards,
-AC
Hi ferret,
Re; heatsinks. They are a perrenial problem and expensive. Have you thought of fabricating your own? Down here I got the idea from a local mag' who suggested buying sheets of Al, 2-3mm thick and then cutting them up in to smaller pieces, say 6''x6", bending them appropriately (so as to form the necessary pathways for air to move) and bolting them together in stack of half a dozen. Needs a bit of time but not a lot of skill.
Edit: not happy with "perrenial" how do you spell it pls?
Cheers, Jonathan
Re; heatsinks. They are a perrenial problem and expensive. Have you thought of fabricating your own? Down here I got the idea from a local mag' who suggested buying sheets of Al, 2-3mm thick and then cutting them up in to smaller pieces, say 6''x6", bending them appropriately (so as to form the necessary pathways for air to move) and bolting them together in stack of half a dozen. Needs a bit of time but not a lot of skill.
Edit: not happy with "perrenial" how do you spell it pls?
Cheers, Jonathan
