@Cogeniac have you confirmed good results with the MPSA92 and 42 as replacements for the 2sc3478 / 2sa1376? Fairchild also makes one called the KSP42 and 92 and it seems to be identical to the A42/92, except promising higher Hfe. Chart says typical Hfe of 150 @10ma, versus around 90 for the A42/92.
No. I bought some NOS parts from LittleDiode in the UK.
Might make sense to run them on a curve tracer to see how they stack up! It would be really great to have a replacement for those.
I found various equivalents for the non-signal devices, and am OK using them, but I have shied away from using non-originals in the signal path.
I am currently slogging through a cross-match of 32 pieces of both MPSA 13 and 63... I do the 512 measurements one way (waiting for the devices to settle), and then for the matches that are within 10 mV (which is an Ice match of 100 uA)I run them in the reverse positions That cancels out any remaining offset from the test jig (which looks to be about 1-2 mV).
Takes all day while you are doing something else to run them.
Scott
Incidentally, I've type up an FAQ for people who are replacing their boards. I'd love suggestions for improvement.
https://hoppesbrain.com/gfa-565-circuit-board-replacement-notes/
And here's a Mouser BOM with all semiconductor equivalents known to me.
https://hoppesbrain.files.wordpress.com/2017/01/gfa-565-parts-list1.xls
https://hoppesbrain.com/gfa-565-circuit-board-replacement-notes/
And here's a Mouser BOM with all semiconductor equivalents known to me.
https://hoppesbrain.files.wordpress.com/2017/01/gfa-565-parts-list1.xls
Last edited:
This is one hell of a thread. Scott, thanks for the in-depth writeups! I actually read the whole thread! And having Chris involved of course leads to a quality thread with a lot of good and valuable information!
I didn't see mention of one supplier who at least for me and most people that I know is the first we check for semiconductors, particularly rare ones. From all that I know, I am not aware of anyone getting a fake device from them. They are B & D Enterprises (bdent.com). Littlediode always seemed a) high priced, and b) a little sketchy, at least to me. Other places, such as Dalbani I avoid like the plague; lots of examples of fakes from them.
Also, for Chris' semiconductor tester, do you have a BOM for that, or at least a parts listing for the connectors that you used for the DUT? And thanks for the Eagle files... I may add a ZIF socket or two in order to do some SMT packages (SOT-23, etc.) as well, and place an order for boards.
I didn't see mention of one supplier who at least for me and most people that I know is the first we check for semiconductors, particularly rare ones. From all that I know, I am not aware of anyone getting a fake device from them. They are B & D Enterprises (bdent.com). Littlediode always seemed a) high priced, and b) a little sketchy, at least to me. Other places, such as Dalbani I avoid like the plague; lots of examples of fakes from them.
Also, for Chris' semiconductor tester, do you have a BOM for that, or at least a parts listing for the connectors that you used for the DUT? And thanks for the Eagle files... I may add a ZIF socket or two in order to do some SMT packages (SOT-23, etc.) as well, and place an order for boards.
Posts moved to a new thread
Hi all,
There was a discussion started on the transistor Beta matcher and testing the real world results of matched differential pairs.
This thread is here:Matching transistors & measuring the results
Anyone interested in building a "Beta Balancer" should look into this thread.
-Chris
Hi all,
There was a discussion started on the transistor Beta matcher and testing the real world results of matched differential pairs.
This thread is here:Matching transistors & measuring the results
Anyone interested in building a "Beta Balancer" should look into this thread.
-Chris
Well, it's confirmed! I've built up a pair of GFA-565's with the BC550/560, and they work great! The amp measures super low distortion as one would expect, and has survived a battery of stress and torture testing, and listening tests. (Gorgeous, smooth, clean and effortless sound with wicked bass control.) Square wave performance into 2 ohms is clean. This is the first GFA-565s I've done with completely new components throughout. It also has the surface-mounted stabistors and the KSP42/92's in the cascode positions.
So, that is 100% of all parts substitutions for the 565. Here's a a spreadsheet with the full list.
And here it is installed in an amp versus the crusty old board!
amazing!
This has been an excellent thread for me as I bought one of the boards and am about to start my repair journey. My amp was supposedly looked over by a good tech that said no leaky caps but one of the big filter caps is bad. Big caps are fine and the supposedly non leaky caps ( The unholy 3 I am going to call them ) had leaked so bad that the metal had been corroded on the casing and frame of the control board. Thanks to the great work here brethren!
Phil
This has been an excellent thread for me as I bought one of the boards and am about to start my repair journey. My amp was supposedly looked over by a good tech that said no leaky caps but one of the big filter caps is bad. Big caps are fine and the supposedly non leaky caps ( The unholy 3 I am going to call them ) had leaked so bad that the metal had been corroded on the casing and frame of the control board. Thanks to the great work here brethren!
Phil
Well, it's confirmed! I've built up a pair of GFA-565's with the BC550/560, and they work great! The amp measures super low distortion as one would expect, and has survived a battery of stress and torture testing, and listening tests. (Gorgeous, smooth, clean and effortless sound with wicked bass control.) Square wave performance into 2 ohms is clean. This is the first GFA-565s I've done with completely new components throughout. It also has the surface-mounted stabistors and the KSP42/92's in the cascode positions.
So, that is 100% of all parts substitutions for the 565. Here's a a spreadsheet with the full list.
And here it is installed in an amp versus the crusty old board!
Hi Guys!
Been a while! I have just been incredibly busy with work for he past year, and have not had anytime to finish off my replacement boards. I see Phloodpants did that for me! Good job!
When I stopped working on these, I had rehabbed the replacement board to adjust a couple of component positions. The board worked great.
I plan to get back to this after the new year. I'll build up a few of the new boards and then wring out the amps (I have SIX of them...). I'll be selling three at some point, after I assemble my system.
Just bought a set of Magnepan.7s to use these amps with.
Cheers,
Scott
Been a while! I have just been incredibly busy with work for he past year, and have not had anytime to finish off my replacement boards. I see Phloodpants did that for me! Good job!
When I stopped working on these, I had rehabbed the replacement board to adjust a couple of component positions. The board worked great.
I plan to get back to this after the new year. I'll build up a few of the new boards and then wring out the amps (I have SIX of them...). I'll be selling three at some point, after I assemble my system.
Just bought a set of Magnepan.7s to use these amps with.
Cheers,
Scott
Hi Chris;
The board design is done. I built one up, and it worked great, but there was a slight issue with the position of one the of the heat sink pins and some other component placement that I can't recall today. I revised the layout very slightly to fix those issues, and fabbed several boards. And then I got super busy! I can't believe it's been almost 2 years since I posted here last!! Wow, time flies!
I'll try to get back to this in January, and will post details.
I still have the HP distortion meter, but it is pretty slow to use. After I refresh all of the amps, I may rent a modern analyzer and run data on all of the amps. Then I can sell the three with some level of confidence in their performance.
Let me know if you want a board.
I was reminded this thread today when I bought my new Maggies. I mentioned that I was driving them with 565's and the guy mentioned that he thought that Nelson Pass had designed that amp. I am not sure about that. He posted on a 555 thread and said he designed that one.
I must say that figuring out exactly how the 565 worked, I was really impressed with the design. Pure power amplifier..
Scott
The board design is done. I built one up, and it worked great, but there was a slight issue with the position of one the of the heat sink pins and some other component placement that I can't recall today. I revised the layout very slightly to fix those issues, and fabbed several boards. And then I got super busy! I can't believe it's been almost 2 years since I posted here last!! Wow, time flies!
I'll try to get back to this in January, and will post details.
I still have the HP distortion meter, but it is pretty slow to use. After I refresh all of the amps, I may rent a modern analyzer and run data on all of the amps. Then I can sell the three with some level of confidence in their performance.
Let me know if you want a board.
I was reminded this thread today when I bought my new Maggies. I mentioned that I was driving them with 565's and the guy mentioned that he thought that Nelson Pass had designed that amp. I am not sure about that. He posted on a 555 thread and said he designed that one.
I must say that figuring out exactly how the 565 worked, I was really impressed with the design. Pure power amplifier..
Scott
Hi guys. I have two 565's that I'm refurbishing. I purchased new fully populated boards from Chris Hoppe to avoid having to rebuild those.
After replacing key components on the soft start board and caps on the output boards and installing the new input boards.
The first amp came up and worked perfectly, a bit of tweaking the bias settings and it's great.
The 2nd amp. After getting it back together, slowly powering up with the variac, I get to about 20 volts on the rails and the current draw starts to go up abnormally high.
I removed the rail fuses and Powered up, all good.
I installed power fuses one at a time, the problem is on the negative side (orange). So something on the orange side output board is causing a problem.
- if I remove power to the orange side by unscrewing and removing the orange wire on the rail on the output board, I don't have the excessive current draw.
- If I remove the feed (brown wire, connection 4) from the input board to the output board, I don't have the excessive current draw. This is with the power rail reconnected. So full power to the output boards, but no feed from the input board.
- Likewise if I remove the connection to S101, thermostat, I don't get the excessive draw. But that just kills the output on the brown wire, so that's consistent with above.
So something is amiss on the output board(s) on the orange side. Any suggestions about what would cause high current draw there with an input connected but not with it disconnected?
gabo
After replacing key components on the soft start board and caps on the output boards and installing the new input boards.
The first amp came up and worked perfectly, a bit of tweaking the bias settings and it's great.
The 2nd amp. After getting it back together, slowly powering up with the variac, I get to about 20 volts on the rails and the current draw starts to go up abnormally high.
I removed the rail fuses and Powered up, all good.
I installed power fuses one at a time, the problem is on the negative side (orange). So something on the orange side output board is causing a problem.
- if I remove power to the orange side by unscrewing and removing the orange wire on the rail on the output board, I don't have the excessive current draw.
- If I remove the feed (brown wire, connection 4) from the input board to the output board, I don't have the excessive current draw. This is with the power rail reconnected. So full power to the output boards, but no feed from the input board.
- Likewise if I remove the connection to S101, thermostat, I don't get the excessive draw. But that just kills the output on the brown wire, so that's consistent with above.
So something is amiss on the output board(s) on the orange side. Any suggestions about what would cause high current draw there with an input connected but not with it disconnected?
gabo
Thanks Craig.
With the variac supplying enough to put 20vdc on the rails. I get 1.9vdc on pin 3 and -1.9vdc on pin 4.
Sounds high to me, unless that doesn't go up with higher rail voltage. Something definitely "kicks in" when I get the rail voltage to about 30vdc. The current draw from the main power creeps up to .6, .7, .8, .9 amps, but when I hit about 30vdc it starts running and I shut it down before it hits 2 amps. Not sure what the voltage is when I get closer to 30vdc as I was just doing it hand held and don't have enough hands to monitor it and bring up the variac.
gabo
With the variac supplying enough to put 20vdc on the rails. I get 1.9vdc on pin 3 and -1.9vdc on pin 4.
Sounds high to me, unless that doesn't go up with higher rail voltage. Something definitely "kicks in" when I get the rail voltage to about 30vdc. The current draw from the main power creeps up to .6, .7, .8, .9 amps, but when I hit about 30vdc it starts running and I shut it down before it hits 2 amps. Not sure what the voltage is when I get closer to 30vdc as I was just doing it hand held and don't have enough hands to monitor it and bring up the variac.
gabo
I disconnected the brown wire on pin 4 as that allows me to bring up the box to full power without it running away.
With that pin lifted, I get -10vdc on that pin. I also get 6.6vdc on pin3 with it still connected to the other side. So clearly something not right.
Since that board is new and was tested before installation, let me first check that I have it connected up correctly and all the traces look good.
gabo
With that pin lifted, I get -10vdc on that pin. I also get 6.6vdc on pin3 with it still connected to the other side. So clearly something not right.
Since that board is new and was tested before installation, let me first check that I have it connected up correctly and all the traces look good.
gabo
In reading back over this very long thread (seems like I've read it about 5 times). Scott had a similar problem with his amp that was resolved back around post #93 or so by replacing the voltage regulator.
In post #92 he describes a situation where when the rail voltage gets to around 30v he starts to get some transformer hum (current draw) and things go awry.
He puts a scope on the DC rail and observes a low freq sawtooth.
I'll put my scope on the rails tomorrow and see what I get. I'll probably go ahead and order a couple of rectifiers. Looks like a KBPC5010-G would be a nice one.
gabo
In post #92 he describes a situation where when the rail voltage gets to around 30v he starts to get some transformer hum (current draw) and things go awry.
He puts a scope on the DC rail and observes a low freq sawtooth.
I'll put my scope on the rails tomorrow and see what I get. I'll probably go ahead and order a couple of rectifiers. Looks like a KBPC5010-G would be a nice one.
gabo
Hey Gary, a few things come to mind...
There are often blown resistors on the output modules of 565's that have been blown out previously. I test every single component on the output stage before attempting to put them back together. The outputs will appear shorted on your meter's diode test, but if you lift one leg of R201, R202, R301 and R302, you can test that all drivers and outputs will at least pass a diode check on your meter. Given your symptoms, you might have a transistor that only goes haywire above a certain voltage. But generally speaking, transistors that pass the diode test are usually OK.
Bridge rectifier is a possibility, but remote, I think. If you see even voltage on both rails, it's probably fine.
1.9VDC on the outputs of the board, pins 3 and 4, does seem just a wee bit high to me. Without the output stage hooked up, it goes to a maximum of 1.8V with the bias cranked. Seems like it's getting pulled up somehow. What do you get on the working amp?
There are often blown resistors on the output modules of 565's that have been blown out previously. I test every single component on the output stage before attempting to put them back together. The outputs will appear shorted on your meter's diode test, but if you lift one leg of R201, R202, R301 and R302, you can test that all drivers and outputs will at least pass a diode check on your meter. Given your symptoms, you might have a transistor that only goes haywire above a certain voltage. But generally speaking, transistors that pass the diode test are usually OK.
Bridge rectifier is a possibility, but remote, I think. If you see even voltage on both rails, it's probably fine.
1.9VDC on the outputs of the board, pins 3 and 4, does seem just a wee bit high to me. Without the output stage hooked up, it goes to a maximum of 1.8V with the bias cranked. Seems like it's getting pulled up somehow. What do you get on the working amp?
Last edited: