Woohoo, got up, got my kids a morning drink and immediately came downstairs to check and I I have a +73/-73 volts on both the right and left channel power supplies.
Those bad 2SC1775s and 2SA872s were wrecking havoc. I did install all new zener diodes, new KSC2690 and KSA1220 and replaced all of the 1175s and 872s, where there were only a total of 3 transistors in total in each channel that were in the valley between the four heatsinks.
Im guessing they saw a lot of heat, being between the two sets of heatsinks and next to these large resistors. The original resistors were a little smaller though, but quite heat stressed so I went larger.
Thank you all, now to get the amplifiers back in and see where I’m at.
Dan
Those bad 2SC1775s and 2SA872s were wrecking havoc. I did install all new zener diodes, new KSC2690 and KSA1220 and replaced all of the 1175s and 872s, where there were only a total of 3 transistors in total in each channel that were in the valley between the four heatsinks.
Im guessing they saw a lot of heat, being between the two sets of heatsinks and next to these large resistors. The original resistors were a little smaller though, but quite heat stressed so I went larger.
Thank you all, now to get the amplifiers back in and see where I’m at.
Dan
Good job hope you made the upgrades so that you don't repeat the same cycle soon
Well, there may be another issue. Is there anything in this amplifier that looks like it could reduce voltage with all the boards missing? I went through and checked the voltages before I started putting all the boards back in. I have the +/- 20v and the +/-72v. My rail voltage that goes to the amplifier boards is low now for some reason. I’m not sure if something is dragging it down. It’s not super low, but it’s at +/- 60v.
Of the large bridge rectifiers, two for each channel, B604 and B704 are outputting +/- 75v. B603 and B703 are unfortunately only outputting +/- 59.5v. You can see in the schematic above that the output of B603 goes directly to a 0.025 ohm resistor and then to the connection lugs for the amplifier boards. Both channels are equally low. I measured the voltage at the bridge rectifier. I’m so confused, is it possible that now with both channels circuitry being corrected that it’s pulling it down? I wouldn’t think so. I don’t believe they are being pulled the ground because as soon as you power, the amplifier off those capacitors hold that +/- voltage for a very long time. I don’t remember checking these voltages right after I took all the boards out, but I do know for a fact that when the amplifier was assembled, I did have 74 V on those lugs.
Dan
I would say that was correct.
There are two supplies, a high and a low voltage. B603 delivers the normal 'low' voltage supply which will be what you measure as -/+59v. The high supply from B604 can be superimposed on the lower one when the Thyristors D602 and D603 turn on.
If the ISC trigger line is in an incorrect state then those thyristors may turn on and you measure the higher -/+75 volt rail. Normally the output voltage of the amp determines when they come into play and tbh for normal use that will never happen. A fault in a power amp like a high offset may trigger that line unintentionally.
Isc of at least 1.2V opens the thyristors and the rail voltages go to higher level of ~74V
that's what I meant - looks like some of my powerelectronics habits... 🙂
I've not studied how its done tbh. Usually this type of amp switches the rails when the output reaches say 70% of the max voltage swing on the low rail. Whether there is a problem in that area or not, it should not affect the operation of the basic amp as such beyond increasing dissipation.
With the amplifier being fully assembled I have 74 V rails again. I powered it up this morning for the first time and all I can say is woohoo! it powered up and no signs of issues so far, does not go into protection, the ability to adjust bias on the left channel is back and I can dial both into 8 mV perfectly.
I measured the offset and it’s exactly 1.4 mV on each channel. I have still not measured any type of output, but I want the amplifier to sit for a little bit and let it just run.
I am going to pull the amplifier board back out because I want to make some corrections to it. As I said earlier, roughly half of the output, transistors have had leads that are snapped off, and I’m having to solder little jumper leads from the board to the components, which I’m not super thrilled about.
So in your alls opinion, what would you do, I have plenty of each. Just replace the one channel that has had the board removed to the output leads with new 2SC5200 and 2SA1943, keeping the originals in the right channel? Or replace the outputs in both channels with MJL1302G and MJL3281G. From what I understand the MJL devices are better. I would leave the originals, but since I have to replace one channel anyway…
I checked to see how hard it would be to install new RCA plugs and with how the boards are removed, and I think it would be an impossible task. I checked to see if the sleeves would match the original RCA plugs and that is a no go. I don’t know what these are called, washers with shoulders, stepped washers? Hopefully I can find something that will fit them, but I’ll see how it goes without them. We’ll see.
Dan
I’ve been doing some reading and I guess another option would be NJW0302g and NJW0281g. I guess these are extremely similar to the Toshiba devices, but are superior in a way. I won’t pretend to fully understand it, but I guess it having a smaller die than the MJL1302/MJL3281 will make it superior sonically? I have a few hundred each of the NJW0281/0302 so those are definitely an option since I have more of them than anything. 180w output with 75v rails and 4 pair of outputs, seems okay if you’ll think those would be a better match.
Dan
Dan
Well that all sounds good. The offset is servo controlled (by the opamp) and will be as low as the offset voltage of the opamp used. So that's perfect.
I would be inclined to use original outputs if you have them on the basis that the last thing you want is any unforeseen issue such as bias out of range or some odd stability issue.
On the 74 volt rails... if you mean the output of the regulators you replaced (and marked 71 on the diagram) then it all sounds good.
If you mean the voltage as measure here and marked 74 then I would be wanting to prove it was OK because I think the 74 is the 'high' voltage rails and these should not normally be on.
This is from a different amp but shows what happens with this kind of Glass G design. VH would be your 74 volt line and VL would be the lower 58?? volt supply. When the output approaches the lower value it switches to the higher rail.
I would be inclined to use original outputs if you have them on the basis that the last thing you want is any unforeseen issue such as bias out of range or some odd stability issue.
On the 74 volt rails... if you mean the output of the regulators you replaced (and marked 71 on the diagram) then it all sounds good.
If you mean the voltage as measure here and marked 74 then I would be wanting to prove it was OK because I think the 74 is the 'high' voltage rails and these should not normally be on.
This is from a different amp but shows what happens with this kind of Glass G design. VH would be your 74 volt line and VL would be the lower 58?? volt supply. When the output approaches the lower value it switches to the higher rail.
I still say stick to the originals if you have them. Ideally each set of four NPN's and each set of four PNP's per channel should be matched so that they share current equally.
So where I replaced those reg was for the 71 v at the pin ribbon cables. Once I got all of the bad transistors out of circuit, both channels returned to a +/- 71v, well 72. With all of the boards removed the supply voltage for the amplifier was sitting down at a +/-58v. Now with the boards, and I mean the input boards and amplifier boards it is sitting back at a +/- 74v.
That voltage goes through these bars
That voltage is circled in red here.
So I see what you are saying, you’re saying that the voltage from the rectifier of the high side isn’t supposed to be inserted until a point of drive. I have some other class G amplifiers and understand the different rails kicking in.
So something is causing the higher voltage to kick in? Is there a downside to having it on all the time?
If I stick to original output devices would you say the Onsemi 5200/1943 are close enough to the Toshiba? I only have the Onsemi version. What I read was that the NJW0302/0281 were pretty much an exact copy of Toshiba’s 2SC5200/2SA1943, at least closer than onsemi’s version of the 5200/1943. That could be wrong though.
Dan
Last edited:
What is more likely, something happening in this portion of the circuit?
Or something happening on the amplifier boards, causing the power supply to turn on the high voltage? It’s weird that both channels are doing it, but I guess even with the voltage for the ribbon cable I was having fairly similar issues in both channels.
I hope it’s not the power supply circuitry, I would hate to have to carry the amp completely down again lol. Well, I guess it would be the amplifier boards since it only has 58 V when the boards are removed.
Dan
Or something happening on the amplifier boards, causing the power supply to turn on the high voltage? It’s weird that both channels are doing it, but I guess even with the voltage for the ribbon cable I was having fairly similar issues in both channels.
I hope it’s not the power supply circuitry, I would hate to have to carry the amp completely down again lol. Well, I guess it would be the amplifier boards since it only has 58 V when the boards are removed.
Dan
Its not an easy circuit to follow on a screen. You need to find where the ISC line comes from. All I could find was an ISC2
Check the supply to the opamp but you need to find where ISC (base of Q601) goes. See if there is continuity between ISC and ISC2
Check the supply to the opamp but you need to find where ISC (base of Q601) goes. See if there is continuity between ISC and ISC2
Well it looks like a well implemented class H amplifier, (other implementations are probably not worth it) that defeats the class H when it heats up, Probably there's a faulty part in the chain that keeps it at high rail.
I was discussing with another forum member regarding this amp as he had commented on an old thread I had. He was saying that the Toshiba 2SC5200 and 2SA1943 would definitely not fare well on a four ohm load with 74v rails. He said they could certainly do it on 58v rails.
I found this version of the schematic
The TL082 is the switcher right? This one has voltages posted, woohoo. Pin 5 will have 0V on it, but once a 4 ohm load and 30w is present the voltage goes up to 2.9v. Knowing this, and the fact that this forum member said they couldn’t do 4 ohms on 74, but okay on 58v. Is it at all possible that normally sitting at idle it’s at 74v, but then, once a certain demand on the amplifier is met it then switches to the lower voltage rails for the safety of the outputs? I just checked five of that I see and it is indeed at 0v sitting at idle.
I’m not going to pretend to know how this circuitry works, but it would answer the question as to why I’m seeing the higher voltage and why they were OK with using these outputs into a load if the amplifier switches to lower rails.
That would also explain why the amp is rated at the same power, 180w, in both 8 ohm and 4 ohm. In four ohm the lower rails would equate to lower output power for that load.
Dan
I found this version of the schematic
The TL082 is the switcher right? This one has voltages posted, woohoo. Pin 5 will have 0V on it, but once a 4 ohm load and 30w is present the voltage goes up to 2.9v. Knowing this, and the fact that this forum member said they couldn’t do 4 ohms on 74, but okay on 58v. Is it at all possible that normally sitting at idle it’s at 74v, but then, once a certain demand on the amplifier is met it then switches to the lower voltage rails for the safety of the outputs? I just checked five of that I see and it is indeed at 0v sitting at idle.
I’m not going to pretend to know how this circuitry works, but it would answer the question as to why I’m seeing the higher voltage and why they were OK with using these outputs into a load if the amplifier switches to lower rails.
That would also explain why the amp is rated at the same power, 180w, in both 8 ohm and 4 ohm. In four ohm the lower rails would equate to lower output power for that load.
Dan
It is a 'switcher' or rather it is used as a comparator.
It would be a very a strange mode of operation because normally power levels are just a couple of watts in normal listening. Running at a high voltage all the time increases dissipation significantly.
Did you check if ISC and ISC2 are the same point? It looks like ISC needs to go high to turn on Q608 to turn the HV rails on.