Hi,
I'm trying to figure out why after I assembled the board from siliconray (FREE JLH 2005 Class A Amplifier PCB - Boards | Kits | Components | Modules | Tools) the output is -24V.
I've ended up measuring the voltages and on the Q4 i get around 22V for the base, instead of about 0.1V according to the DC voltages table for the 1996 version.
The rail is about 24.5V.
My input isn't connected to anything, am I supposed to connect it before doing these tests?
I'm trying to figure out why after I assembled the board from siliconray (FREE JLH 2005 Class A Amplifier PCB - Boards | Kits | Components | Modules | Tools) the output is -24V.
I've ended up measuring the voltages and on the Q4 i get around 22V for the base, instead of about 0.1V according to the DC voltages table for the 1996 version.
The rail is about 24.5V.
My input isn't connected to anything, am I supposed to connect it before doing these tests?
Thnx, did that. but on the Q4 base I still get about 22V. Is the only possible explanation be that I short circuited something, or is it possible that some voltage leaks from emitter/collector to the base?
p.s. sorry if these are dumb questions, I'm a software engineer, that sometimes tinkers with basic electronics, and this time I might have gotten too far..
It could be any one thing or several things together and so you need a methodical approach. Its no good guessing.
I assume your component references are from you link above
First check and recheck that all the right devices are fitted... no mixed up NPN's and PNP's etc.
Next you need to begin to confirm the operation of the various parts of the circuit. If you measure across the base and emitter junctions of Q5 and Q7 you should see around 0.7 volts. These are the two constant current sources.
Next measure from ground and record the voltages on,
Q8 Emitter
Q8 Base
Q8 Collector
Q7 Emitter
Q7 Base
Q7 Collector
Q6 Emitter
Q6 Base
Q6 Collector
Q5 Emitter
Q5 Base
Q5 Collector
Q4 Emitter
Q4 Base
Q4 Collector
Q3 Emitter
Q3 Base
Q3 Collector
Q2 Emitter
Q2 Base
Q2 Collector
Q1 Emitter
Q1 Base
Q1 Collector
Volt drop across R10
Copy and paste the above and fill in the voltages
Hi Mooly, did that:
Q8 Emitter = 22.7V
Q8 Base = 22V
Q8 Collector = 22.7V
Q7 Emitter = 22.6V
Q7 Base = 22.6V
Q7 Collector = 21.9V
Q6 Emitter = 22V
Q6 Base = 21.5V
Q6 Collector = 11.4V
Q5 Emitter = 22.8V
Q5 Base = 22.2V
Q5 Collector = 21.6V
Q4 Emitter = 11.6V
Q4 Base = 22.3V
Q4 Collector = 22.8V
Q3 Emitter = -24.2V
Q3 Base = 22.6V
Q3 Collector = 22.7V
Q2 Emitter = -24V
Q2 Base = 22.7V
Q2 Collector = 22.7V
Q1 Emitter = -24V
Q1 Base = -24V
Q1 Collector = -24V
Volt drop across R10 = 24mV
The rail voltage = +/- 24.5
There are 20ohm resistors in series on the rail to the amp board, and measured on the amp:
+22.9
-24.3
Also there was a drift(?) of about .2V up and down, so measuring again after a minute would get a bit different values.
Q8 Emitter = 22.7V
Q8 Base = 22V
Q8 Collector = 22.7V
Q7 Emitter = 22.6V
Q7 Base = 22.6V
Q7 Collector = 21.9V
Q6 Emitter = 22V
Q6 Base = 21.5V
Q6 Collector = 11.4V
Q5 Emitter = 22.8V
Q5 Base = 22.2V
Q5 Collector = 21.6V
Q4 Emitter = 11.6V
Q4 Base = 22.3V
Q4 Collector = 22.8V
Q3 Emitter = -24.2V
Q3 Base = 22.6V
Q3 Collector = 22.7V
Q2 Emitter = -24V
Q2 Base = 22.7V
Q2 Collector = 22.7V
Q1 Emitter = -24V
Q1 Base = -24V
Q1 Collector = -24V
Volt drop across R10 = 24mV
The rail voltage = +/- 24.5
There are 20ohm resistors in series on the rail to the amp board, and measured on the amp:
+22.9
-24.3
Also there was a drift(?) of about .2V up and down, so measuring again after a minute would get a bit different values.
The reading that stands out is Q3 with a base-emitter differential of over 40 volts. That's just not possible with a good device (even if you had a PNP fitted by mistake).
I would remove it and just do the basic tests on it. With your meter on "diode" range and the red lead on the base (assuming you have a normal DVM) it should read from base to emitter and from base to collector. The meter should show around .700 on this range give or take. With the black lead on the base you should get no reading (open circuit, meter probably showing 1.999) to either E or C.
Do you see why measuring across junctions is so important for some readings... as you say there is drift and variation also as the mains supply varies
So Q3 has to be the starting point whether or not there are other issues.
I would remove it and just do the basic tests on it. With your meter on "diode" range and the red lead on the base (assuming you have a normal DVM) it should read from base to emitter and from base to collector. The meter should show around .700 on this range give or take. With the black lead on the base you should get no reading (open circuit, meter probably showing 1.999) to either E or C.
Do you see why measuring across junctions is so important for some readings... as you say there is drift and variation also as the mains supply varies
So Q3 has to be the starting point whether or not there are other issues.
Yep I have dealt with Geoff in the year gone by many times he was always very helpfully indeed it would be good if someone can host Geoff's website
Respect
This is my first post here for many a year. I heard earlier today (thanks Rod) that my site no longer existed. It would appear that my ISP (BT) has arbitrarily, and without any advance warning, ceased the free web hosting that formed part of the package when I first signed up with them many years ago. I will be looking into other hosting options but in the meantime the Wayback Internet archive is available.
The last update of my site was on 26/9/2009 (some minor additions to the JLH Articles page) and the most recent Wayback capture was on 20/7/2011 so it is up to date.
The last update of my site was on 26/9/2009 (some minor additions to the JLH Articles page) and the most recent Wayback capture was on 20/7/2011 so it is up to date.
Hi, me again..
Re-soldered all the BD560 transistors, and the voltages have changed, but I get the voltage on the Q3 Collector and Q8 collector (for reference http://web.archive.org/web/20101026224208im_/http://www.tcaas.btinternet.co.uk/jlhupdatefig1.gif) same as on the V+, and even with my limited understanding of how it works, it doesn't seem right. The resistance between this pins and the V+ (when there is no power of course), is 0ohms. Should I continue to investigate why the voltage is same as on the rail, or am I missing something?
thanks in advance, and also to those that helped before, as I was too stressed and might have forgot that part
Re-soldered all the BD560 transistors, and the voltages have changed, but I get the voltage on the Q3 Collector and Q8 collector (for reference http://web.archive.org/web/20101026224208im_/http://www.tcaas.btinternet.co.uk/jlhupdatefig1.gif) same as on the V+, and even with my limited understanding of how it works, it doesn't seem right. The resistance between this pins and the V+ (when there is no power of course), is 0ohms. Should I continue to investigate why the voltage is same as on the rail, or am I missing something?
thanks in advance, and also to those that helped before, as I was too stressed and might have forgot that part
Hi, me again..
Re-soldered all the BD560 transistors, and the voltages have changed, but I get the voltage on the Q3 Collector and Q8 collector (for reference http://web.archive.org/web/20101026224208im_/http://www.tcaas.btinternet.co.uk/jlhupdatefig1.gif) same as on the V+, and even with my limited understanding of how it works, it doesn't seem right. The resistance between this pins and the V+ (when there is no power of course), is 0ohms. Should I continue to investigate why the voltage is same as on the rail, or am I missing something?
thanks in advance, and also to those that helped before, as I was too stressed and might have forgot that part
Update: looking hard into the schematics I figured that the most probable cause was the Q2. Tested all 4 transistors and only one was giving the expected results. Replaced it into Q2 and got about 2V at output. Q1 is fried as well, so I'll have to figure out where to get this ones before moving forward, as none of my usual(read as cheap delivery) suppliers carry them
I'd very much appreciate some alternatives, I currently use 2SC3280.
I have built this amplifier and am listening to it now as I am typing. It is the best sounding amp I have ever built. I recall it was pretty much fool proof when I first applied power to it. I did not use PCB. I used a perf board and wired components myself. I remember I read somewhere those PCBs had an error so even if you put all components accordingly, but they will be wired incorrectly. Here's what I suggest IMHO :-
-Trouble one board at a time
-Desolder all transistors and measure them with a multimeter. I use an old fashion analog multimeter for all my transistors check.
-Trace the tracks on the PCB and compare to the schematic. Do it 3 times to make sure you don't miss anything. This can be time consuming but you ensure the PCB tracks are correctly done to the schematic. Especially making sure the transistor pinouts are marked correctly on the PCB as compare to the schematic. If not, correct the markings.
- When mounting the transistors, make sure the insulation mica or whatever you use between the transistor and the heatsink is NOT shorted. Usually the transistor's collector is on the metal and should not electrically conductive to the heatsink or it will be a short ! Check the transistor's pins, especially "C" for shorts to the heatsink.
- When reinstalling transistors, make sure to verify the pinout. Sometimes, transistors have different pinout even if their external shape and dimension are the same.
- Check your power supply and make sure it is +/- balanced. For e.g., if +/- 15v, you measure both rails and they are the same voltage but different polarity. If this if off (say one side 14v, the other side 15v), it could set a big offset at the output.
- When testing the amplifier first time, apply lower rail voltage if possible to start with, say +/- 10v. Then play around with adjusting the output offset and play some music. When things are going well, raise the voltage and readjust output offset before connecting speakers and try playing music again.
If you have any one transistor broken, you will not be able to set the output offset to zero. When I did my offset adjustment, I didn't need to short the input. I adjust the offset without speaker. When it was zero, I connected a pair of cheapo speakers for testing with my ipod. It amazed me that the cheapo speakers sounded so good with the amp.
I did not use 2sc3280, I used 2n3055 then upgrade to MJ15003. The TO-3 case is more involved to mount but the whole thing is metal and it conducts heat better the heatsink in my humble opinion.
If you are saying Q1/Q2 is/are toasted, they are the output transistors. I do suggest to use TO-3 device if possible. I do not see your location. But if you are in US, you can go to radioshack to get them (although they are pricey there).
Hope this help and let us know how things turn out.
-Trouble one board at a time
-Desolder all transistors and measure them with a multimeter. I use an old fashion analog multimeter for all my transistors check.
-Trace the tracks on the PCB and compare to the schematic. Do it 3 times to make sure you don't miss anything. This can be time consuming but you ensure the PCB tracks are correctly done to the schematic. Especially making sure the transistor pinouts are marked correctly on the PCB as compare to the schematic. If not, correct the markings.
- When mounting the transistors, make sure the insulation mica or whatever you use between the transistor and the heatsink is NOT shorted. Usually the transistor's collector is on the metal and should not electrically conductive to the heatsink or it will be a short ! Check the transistor's pins, especially "C" for shorts to the heatsink.
- When reinstalling transistors, make sure to verify the pinout. Sometimes, transistors have different pinout even if their external shape and dimension are the same.
- Check your power supply and make sure it is +/- balanced. For e.g., if +/- 15v, you measure both rails and they are the same voltage but different polarity. If this if off (say one side 14v, the other side 15v), it could set a big offset at the output.
- When testing the amplifier first time, apply lower rail voltage if possible to start with, say +/- 10v. Then play around with adjusting the output offset and play some music. When things are going well, raise the voltage and readjust output offset before connecting speakers and try playing music again.
If you have any one transistor broken, you will not be able to set the output offset to zero. When I did my offset adjustment, I didn't need to short the input. I adjust the offset without speaker. When it was zero, I connected a pair of cheapo speakers for testing with my ipod. It amazed me that the cheapo speakers sounded so good with the amp.
I did not use 2sc3280, I used 2n3055 then upgrade to MJ15003. The TO-3 case is more involved to mount but the whole thing is metal and it conducts heat better the heatsink in my humble opinion.
If you are saying Q1/Q2 is/are toasted, they are the output transistors. I do suggest to use TO-3 device if possible. I do not see your location. But if you are in US, you can go to radioshack to get them (although they are pricey there).
Hope this help and let us know how things turn out.
Just ordered the 2n3055 and MJ15003. I figured I should start to work with 2n3055 (as they are cheap) and when the circuit is debugged and stable, to upgrade it to MJ15003 as most people recommend it. I've ordered some heat sinks, not very big, but I expect them to be ok at least for the first stage when working with low currents.
I expect to get the parts in about 3 day and get back with updates
I expect to get the parts in about 3 day and get back with updates
I mounted one 12v computer silent fan to blow at the heat sink. My heat sinks are not very big, mostly due to the dimension and mounting options I have with the enclosure. It definitely helps a lot with the heat dissipation. You can find my thread about my JLH amp with pics if you are interested.
