Thanks for reply, I also plan to replace 2SC3421 by TO220 transistor driver, 2SC2238 is one alternative, do you have another suggestion?
That gives 4A of output bias.
The JLH is reputed to put out more than Ibias as maximum ClassA current, some suggest 50% more (cf almost 100% more for push-pull) and that would be around 6Apk for max ClassA output.
6Apk into 4r0 requires supply rails around your target of +-30Vdc.
Expect maximum power to be around 70W into 4ohms. Is that what you intended?
I don't recall anyone reporting a JLH build like this.
The JLH is reputed to put out more than Ibias as maximum ClassA current, some suggest 50% more (cf almost 100% more for push-pull) and that would be around 6Apk for max ClassA output.
6Apk into 4r0 requires supply rails around your target of +-30Vdc.
Expect maximum power to be around 70W into 4ohms. Is that what you intended?
I don't recall anyone reporting a JLH build like this.
I built both 1969 and 1996 versions of this amplifier. In both cases I wish I had gone for larger transformers and heat sinks. With the 1969 I went with an EI transformer of 80 v.a. after a few years use it blew up and I replaced it with a 150 va one that too burned out. These transformers were made locally and perhaps the ratings were overstated. The 1996 one built at that time has a 300 va transformer with a regulated supply and I used heat sinks of 0.3 degrees C per watt. In use these heat up to 60 C in summer and they need a thermostatic switch to operate a fan. The more heat one has the shorter the life of the power transistors. Periodic replacement is not a bad idea.
I am presently doing a rebuild on this amplifier.
If anyone is tempted to build a 70 watt version of the 1969 version, consider the heat sink, power supply requirements, the costs of these and the bulk of the equipment.
Just for fun I am considering Linsley-Hood's 1970 15-20 watt Class AB as the basis of a project. It will hopefully be as good which on the basis of specifications and auditions, Hood said it was, but less a labour of love.
That will be interesting and I hope you go ahead.
Considering the listening comparison results of JLH and his long suffering friends, we may be able to give this full class A deal a bit of a rest. I suspect the sustained interest in the JLH design actually centres on its input stage which, without considering the many alternative CFA designs, is simply "not an LTP" which selectively attenuates even harmonics, particularly the 2nd. In anything but a really high grade amplifier, we can usually pick the difference by preference over time, if not by careful analysis.
That will be interesting and I hope you go ahead.
I suspect the sustained interest in the JLH design actually centres on its input stage which, without considering the many alternative CFA designs, is simply "not an LTP" which selectively attenuates even harmonics, particularly the 2nd. In anything but a really high grade amplifier, we can usually pick the difference by preference over time, if not by careful analysis.
It is on my list for the summer - too cold to venture to the workshop area of my garage at present. Apart from a suitable case I have all the parts I need to hand.
At the cost of one extra transistor the Class AB amplifier there is a possibility for anyone so interested of substituting a new board into a Class A chassis.
There one could choose to run in Class AB mode or increase the quiescent current to allow full Class A running - a possibility mentioned by JLH. Back in the day one of the U.K. kit suppliers offered boards for this amplifier.
The adjustment is via a 100 ohm wire wound potentiometer which could be mounted for external access - switched wire wound resistors could also be used in this way.
Like the Class A circuit I don't see this AB version extending to 70 watts into 4 ohm loads. There are later JLH designs to consider for that use.
Hi,
Just finished 1 channel of the JLH and now realize my heatsink seems a little too small. I can hold my hand on it for as long as I want but if feels hot, I would say around 45-50C.
My bias is stable at 0.9A, power supply is SMPS adjustable from 22-29Vdc, it is now set at 27Vdc.
As anyone set-up this amp at reduced bias and got decent results ? If so and and it still sounded great please share your bias and supply value. I do not mind reducing slightly my output power, my speakers are sensitive.
My transistors are BC560C(T4), 2SC3902T(T3) and NJW21194(T1,T2)
Thanks,
Eric
Just finished 1 channel of the JLH and now realize my heatsink seems a little too small. I can hold my hand on it for as long as I want but if feels hot, I would say around 45-50C.
My bias is stable at 0.9A, power supply is SMPS adjustable from 22-29Vdc, it is now set at 27Vdc.
As anyone set-up this amp at reduced bias and got decent results ? If so and and it still sounded great please share your bias and supply value. I do not mind reducing slightly my output power, my speakers are sensitive.
My transistors are BC560C(T4), 2SC3902T(T3) and NJW21194(T1,T2)
Thanks,
Eric
50°C for the sink is OK.
If the Tc is 60°C then the de-rating Factor for the output transistors can be calculated from
P = 150-Tc / 150-25. DF60=0.72
If you have a 100W device that is de-rated to 72W and your Pq is <50% of that you are probably OK.
If your Pq <1/3rd, then you are definitely OK.
If the Tc is 60°C then the de-rating Factor for the output transistors can be calculated from
P = 150-Tc / 150-25. DF60=0.72
If you have a 100W device that is de-rated to 72W and your Pq is <50% of that you are probably OK.
If your Pq <1/3rd, then you are definitely OK.
Last edited:
The temperatures of 50C are fine, no problem reliability wise. On 27v and 0.9 amp you are looking at around 8 to 9 watts rms output into 8 ohms. If we reduce the bias to around 700ma we see a fall to around 6.5 wrms (all very approximate figures). The heatsink temp would be considerably reduced.
As an experiment I would suggest you try something silly like 200ma bias and just see how loud it takes you before distortion sets in... get an idea of what you need.
Also have a go at this, you might be surprised.
http://www.diyaudio.com/forums/mult...much-voltage-power-do-your-speakers-need.html
As an experiment I would suggest you try something silly like 200ma bias and just see how loud it takes you before distortion sets in... get an idea of what you need.
Also have a go at this, you might be surprised.
http://www.diyaudio.com/forums/mult...much-voltage-power-do-your-speakers-need.html
Hi Andrew,
According to the spec sheet, these are 200W device at Tc of 25C and the spec specify: derate above 25C is 1.6W/C so looks like each output transistor is rated 144W at 60C..is this correct ? Power dissipated in each device is roughly (27V x 0.9)/2 or 12W/transistor so looks like I'm quite OK. Thanks for the explanation.
Thanks,
Eric
Last edited:
Hi Mooly,
I am aware that I don't need much power, my speaker are rated 96dB and I do not blast them, most of my listening is at moderate level. Between 3-5W is enough.
I will leave it as is then with a 50C heatsink temp
. I normally aim for a heatsink temp of 43C so around 50C is a little high for my taste but I will add some thermal switch just in case I forget the amp ON while I'm away.Thanks,
Eric
Help with JLH96
Hello Everyone,
I recently built a JLH96 amp using this schematic:
[Yesterday I posted this message in a different place and someone suggested I post a help request here instead... ]
I built the amp on a breadboard, point-to-point. I have the output transistors mounted on a big heatsink. Everything started out fine... I set my power supply for 0.2 amps and 0 volts. Attached the power supply, turned it on and rolled the voltage up to 2 volts. So far so good. I increased the voltage to 5 volts. All good. Then I went to 10, 15 and finally 22 volts. No smoke!
I have an 8 ohm 35 watt resistor attached to the output terminals. The dc offset was ~76 millivolts so I used VR1 to reduce it to 0 volts. Everything still seems good.
Next I increase the current to 1 amp. After which something must have blown. I didn't hear or smell anything but I could't read any voltage on the output terminals any longer. I turned the power supply off and waited a few minutes. Turned it back on and tried to change the dc offset with VR1 but it had no effect. The output transistors never got warm.
I have no idea where to start or how to begin troubleshooting this thing. Could anyone give me an idea of what to check or where to start??
Thanks much for your help -
Bruce
Hello Everyone,
I recently built a JLH96 amp using this schematic:
[Yesterday I posted this message in a different place and someone suggested I post a help request here instead... ]
I built the amp on a breadboard, point-to-point. I have the output transistors mounted on a big heatsink. Everything started out fine... I set my power supply for 0.2 amps and 0 volts. Attached the power supply, turned it on and rolled the voltage up to 2 volts. So far so good. I increased the voltage to 5 volts. All good. Then I went to 10, 15 and finally 22 volts. No smoke!
I have an 8 ohm 35 watt resistor attached to the output terminals. The dc offset was ~76 millivolts so I used VR1 to reduce it to 0 volts. Everything still seems good.
Next I increase the current to 1 amp. After which something must have blown. I didn't hear or smell anything but I could't read any voltage on the output terminals any longer. I turned the power supply off and waited a few minutes. Turned it back on and tried to change the dc offset with VR1 but it had no effect. The output transistors never got warm.
I have no idea where to start or how to begin troubleshooting this thing. Could anyone give me an idea of what to check or where to start??
Thanks much for your help -
Bruce
Check all your resistor values, and cap polarities. After finding problem, check your transistors for shorts, and re-check exactly as you have described. My guess is one of those three pots had to be a high-ish wattage...you can start there - see if one of those three pots is shorted/open...
I have same amp with two output devices... a very nice amp and worth building.
I have same amp with two output devices... a very nice amp and worth building.
Only when forward biasing the junction (eg B to E or B to C) and the results will only be 'reliable' if you read the devices out of circuit.
Voltage checks are much quicker and should pin the problem down in seconds. Check the two rails and check R10 is OK
Hi
I have a concern with VR1 and VR2 in this circuit. There is no limiting resistor, unlike VR3 which has R11 in series. At 0 ohms Q6 and Q8 might fry, and so then cause Q2 and/or Q1 to expire. Resistor R10 seems to have no purpose other than measuring the current in Q2 (why are the transistors numbered from back to front?).
You may want to unsolder all your transistors, and check them. Do you have a meter with a diode check (or an Avo which you can use on resistance). Suspect some might have INF base to collector!
Next time, I'd put a 47 ohm in series with VR1 and a 22 ohm in series with VR2. You can always reduce these later if necessary. I would also make sure they are MAX resistance to start with, and not rely on the power supply to limit the current when you switch on next time.
I have a concern with VR1 and VR2 in this circuit. There is no limiting resistor, unlike VR3 which has R11 in series. At 0 ohms Q6 and Q8 might fry, and so then cause Q2 and/or Q1 to expire. Resistor R10 seems to have no purpose other than measuring the current in Q2 (why are the transistors numbered from back to front?).
You may want to unsolder all your transistors, and check them. Do you have a meter with a diode check (or an Avo which you can use on resistance). Suspect some might have INF base to collector!
Next time, I'd put a 47 ohm in series with VR1 and a 22 ohm in series with VR2. You can always reduce these later if necessary. I would also make sure they are MAX resistance to start with, and not rely on the power supply to limit the current when you switch on next time.